Public entities that operate fixed route transportation services for the general public are required by the U.S. Department of Transportation (USDOT) regulations implementing the Americans with Disabilities Act of 1990 (ADA) to provide ADA Complementary Paratransit service for persons who, because of their disability, are unable to use the fixed route system. These regulations (49 CFR Parts 27, 37, and 38) include six service criteria, which must be met by ADA Complementary Paratransit service programs. Section 37.135(d) of the regulations requires that ADA Complementary Paratransit services meet these criteria by January 26, 1997.

The Federal Transit Administration (FTA) is responsible for ensuring compliance with the ADA and the USDOT regulations. As part of its compliance efforts, FTA, through its Office of Civil Rights, conducts periodic reviews of fixed route transit and ADA Complementary Paratransit services operated by grantees.

The purpose of these reviews is to assist the transit agency and the FTA in determining whether capacity constraints exist in ADA Complementary Paratransit services. The reviews examine policies and standards related to service capacity constraints such as those measured by on-time performance, on-board travel time, telephone hold times, trip denials, and any other trip-limiting factors. The reviews consider whether there are patterns or practices of a substantial number of trip limits, trip denials, early or late pick-ups or arrivals after desired arrival (or appointment) times, long trips, or long telephone hold times as defined by established standards (or typical practices if standards do not exist). The examination of patterns or practices includes looking not just at service statistics, but also at basic service records and operating documents, and observing service to determine whether records and documents appear to reflect true levels of service delivery. Input also is gathered from local disability organizations and customers. Guidance is provided to assist the transit operator in monitoring service for capacity constraints.

An on-site compliance review of ADA Complementary Paratransit service provided by New York City Transit (NYCT) of New York, New York was conducted September 10-18, 2003. Planners Collaborative, Inc., located in Boston, Massachusetts, and Multisystems, Inc., located in Cambridge, Massachusetts, conducted the review for the FTA Office of Civil Rights. The review focused on compliance of NYCT's ADA Complementary Paratransit service with one specific regulatory service criterion: the "capacity constraints" criterion. Section 37.131(f) of the regulations requires that ADA Complementary Paratransit services be operated without capacity constraints.

This report summarizes the observations and findings of the on-site review of NYCT's ADA Complementary Paratransit service. First, a description of the approach and methodology used to conduct the review is provided. Then, a description of key features of transit services provided by NYCT - fixed route, demand responsive, and ADA Complementary Paratransit service - is provided. All of the findings of the review are summarized in Section IV. Observations and findings related to each element of the capacity constraint criteria are then presented in Sections V through IX. Recommendations for addressing some of the findings are also provided.

NYCT was provided with a draft copy of the report for review and response. A copy of the correspondence received from NYCT on June 28, 2004, documenting their response to the draft report

This review focused on compliance with the ADA Complementary Paratransit capacity constraints requirements of the DOT ADA regulations. Several possible types of capacity constraints are identified by the regulations. These include "wait listing" trips, having caps on the number of trips provided, or recurring patterns or practices that result in a substantial number of trip denials or missed trips, untimely pick-ups, or significantly long trips. Capacity constraints also include other operating policies or practices that tend to significantly limit the amount of service to persons who are ADA Complementary Paratransit eligible.

To assess each of these potential types of capacity constraints, the review focused on observations and findings regarding:

• Trip denials and "wait listing" of trips,
• Trip caps,
• On-time performance, and
• Travel times.

Observations and findings related to two other policies and practices that can affect ADA Complementary Paratransit use also are provided, including:

• Determinations of ADA Complementary Paratransit eligibility, and
• Telephone capacity.

ADA Complementary Paratransit eligibility determinations were assessed to ensure that access to service was not adversely impacted by inappropriate denials of eligibility for the service or unreasonable delays in the eligibility process. Telephone capacity was assessed because access to reservations and customer service staff is critical to the effective use of any ADA Complementary Paratransit service.

The review first involved the collection and examination of key service information prior to the on-site visit. This information included:

• A description of how the ADA Complementary Paratransit service is structured;
• Public information describing the ADA Complementary Paratransit service; and
• A description of the New York City Transit's (NYCT) standards for on-time performance, trip denials, travel times, and telephone service.

NYCT was requested to make additional information available during the on-site visit. This information included:

• Copies of completed driver manifests for recent months;
• Six months of service data, including the number of trips requested, scheduled, denied, canceled, no-shows, missed trips, and trips provided by NYCT;
• A breakdown of trips requested, scheduled, and provided;
• Detailed information about trips denied in the last six months including origin and destination information, day and time information, and customer information;
• Detailed information about trips identified in the last six months with excessively long travel times;
• Telephone call management records; and
• Records of recent customer comments and complaints related to capacity issues (trip denials, on-time performance, travel time, and telephone access).

An on-site review of the service was conducted from September 10-18, 2003. The on-site review began with an opening conference, held at 1:00 PM on Wednesday, September 10, 2003. NYCT representatives attending the meeting included: Lawrence Reuter, President; Millard Seay, Senior Vice President; Barbara Spencer, Executive Vice President; Mary McCorry, Legal Department and from NYCT Paratransit Division: Howard Ende, Vice President Paratransit Division; Brian Altschul, Director Telecommunications & Information Services; Michael Cosgrove, Contract Management Officer; Dennis Erkus, Command Center Officer; John Gaito; Aldo Martelli, Standards and Compliance Officer; Beverly Morris, Eligibility & Planning Officer, Maritza Troncoso, Financial Services Officer; and Carol Zwick, Customer Relations Officer. Marc Albrecht, Grants Management and Noreen Din, Civil Rights, from the New York Metropolitan Transportation Authority also attended. Don Kidston and David Chia of Planners Collaborative, and Russell Thatcher and Rosemary Mathias of TranSystems, Inc. represented the FTA review team. Cheryl Hershey, David Knight and David Schneider of the FTA's Office of Civil Rights in Washington, D.C. also participated in the opening conference via telephone.

Cheryl Hershey opened the meeting by thanking NYCT for their cooperation in the review. She described the purpose of the review and emphasized that it was intended to assist NYCT in providing effective ADA Complementary Paratransit service and acknowledged the service challenges that NYCT faces in such a large urbanized area. Ms. Hershey outlined the steps in the review process:

• Preliminary findings and an opportunity to respond would be provided at a closing meeting on Thursday, September 18;
• A draft report would be provided to the NYCT for review and comment; and
• NYCT's comments would be incorporated into a final report, which would then become a public document.

Don Kidston indicated the objective of the review to identify significant impediments, if any, to people with disabilities receiving the service that they are entitled to under the ADA, and to assist NYCT in improving service if warranted. He described the scope of the review as including review of policies, procedures, practices and performance that can affect availability of effective service. The areas to be addressed include eligibility, telephone access, reservations and scheduling, performance, operating procedures and practices and adequacy of resources. Mr. Kidston went on to present the schedule for the on-site review, including the parts of the operation that would be observed by day.

Mr. Reuter noted strong support for Access-A-Ride (AAR) by NYCT and MTA as evidenced by substantial increases in financial support for the program in recent years.

Following the opening conference, the review team toured the AAR offices at 2 Broadway in Manhattan and met with senior AAR staff to discuss the service structure and standards and the information available on-site. The review team also accepted reports and information provided by NYCT in response to FTA's earlier request. Members of the review team also observed call takers in the reservations section of the call center. Team members sat with selected reservationists, listened in on calls from riders and recorded the handling of trip requests.

During the morning of Thursday, September 11, 2003, three team members continued observations of the trip reservations and scheduling process. One team member observed call takers handling customer calls related to same day service issues. During the afternoon review team members began review and analysis of telephone performance reports, the trip scheduling process and reports on disposition of trip requests and performance in serving completed trips. Review team members also continued to observe call taker handling of trip reservations and scheduling and same-day trip inquiries.

On Friday, September 12 and Monday September 15 the review team split into teams of two and visited each of the eight primary service providers. One team visited Star Cruiser Transportation Inc. and the other Transit Facility Management Corp. on Friday morning; and MV Transportation, Inc. and RJR Paratransit Corp. on Friday afternoon. The teams visited Maggie's Paratransit Corp. and PTM Management Corp. on Monday morning and Atlantic Paratransit of NYC, Inc. and American Transit, Inc. on Monday afternoon. At each operator's site review team members observed dispatchers and interviewed drivers and managers regarding operating procedures and practices. Team members also collected information on personnel and equipment, as related to their capacity to provide effective service.

On Tuesday review team members returned to AAR offices at 2 Broadway and continued to interview managers and collect and review reports and information. One team member continued to review disposition of trip requests and performance in serving completed trips, focusing on one sample day. Team members also initiated review of the eligibility process, complaints received by NYCT, service design parameters, and the AAR budget process.

On Wednesday review team members continued review of disposition of trip requests and performance in serving completed trips and eligibility. One team member visited an NYCT AAR eligibility contractor. Another team member spent much of the day at NYCT's Headquarters at 370 Jay Street in Brooklyn analyzing duration of paratransit trips compared to fixed route trip duration for the same trip origin and destination. Team members also summarized the results of driver interviews and initiated analysis of operator resources. The review team leader briefed FTA on progress of the review.

Due to hurricane Isabel, which was forecast to impact New York City on Thursday, September 18, NYCT requested that the Closing Conference be rescheduled from 1:00 PM on Thursday to 8:00 AM. As a result the review team spent Wednesday evening continuing its analysis and preparing for the closing conference.

An exit conference was then held at 8:00 AM on Thursday, September 18, 2003. Attending the exit conference for NYCT were: Lawrence Reuter, President; Millard Seay, Senior Vice President; Barbara Spencer, Executive Vice President; Robin Bergstrom and Mary McCorry, Legal Department and from NYCT Paratransit Division: Howard Ende, Vice President Paratransit Division; Michael Cosgrove, Contract Management Officer; Dennis Erkus, Command Center Officer; John Gaito; Aldo Martelli, Standards and Compliance Officer; Beverly Morris, Eligibility & Planning Officer, and Carol Zwick, Customer Relations Officer. Attending from FTA's Office of Civil Rights were Michael Winter, Director; John Prince, Jr. Civil Rights Officer, Region II; and David Knight. Attending for the review team were Don Kidston and David Chia of Planners Collaborative, and Russell Thatcher and Rosemary Mathias of TranSystems.

Mr. Winter opened the exit conference by thanking the NYCT staff for their cooperation in the review.

The review team members then presented an overview of the assessment and initial findings in each of the following areas:

• Consumer input,
• Service design parameters,
• Eligibility determinations,
• Telephone access,
• Handling of trip requests and trip denials,
• Trip scheduling,
• On-time performance and service delivery,
• Trip duration,
• Operations, and
• Resources (vehicles, manpower, and financial resources).

There was discussion throughout presentation of the preliminary findings. Following the presentation, Mr. Winter noted that NYCT has made progress in recent years in improving AAR service. He expressed concern about the number of trips classified as no shows and the number of late trips. NYCT agreed to perform a more detailed analysis of these trips. Mr. Winter expressed concern that NYCT's suspension policy, were it fully enforced, could be overly restrictive and an impediment to use of the service by people who have disabilities. Mr. Winter also encouraged NYCT to increase public information on and availability of travel training to promote use of fixed route services by people with disabilities.

Mr. Winter and the review team again thanked NYCT for their cooperation during the field review. NYCT officials thanked FTA and the review team for their input.

New York City Transit (NYCT) is a New York State Public Benefit Corporation and public authority and is an affiliate of the Metropolitan Transportation Authority (MTA). NYCT is the principal transit operator in New York City, providing a rapid rail and bus transit service throughout all five boroughs that comprise the city.

General - The train fleet consists of 5,758 subway cars and the bus fleet consists of 4,489 buses (Source: 2000 National Transit Database). NYCT operates 25 subway lines, 204 local and 31 express bus routes.

NYCT operates the nation's largest subway system with 228 miles of right-of-way linking three boroughs (Bronx, Brooklyn and Queens) with Manhattan and its central business district. NYCT also operates the Staten Island Railway (SIR), a single line, 14-mile rapid rail facility owned by the Staten Island Railway, another MTA subsidiary. The Staten Island Ferry links the SIR to Manhattan.

The NYCT rapid transit system consists of 731 track miles, of which 70 are on elevated structure. The deepest station is 181 feet below street level; the highest station is 87.5 feet above street level. The system picks up and drops off passengers at 468 stations and from 1077 train platforms. Average weekday ridership is approximately 4.3 million passengers with 1.7 billion unlinked passenger trips served annually.1 In addition NYCT serves approximately 822 million unlinked bus passenger trips each year. (Source: 2000 National Transit Database)

As of May 17, 2004, NYCT has 55 stations (40 accessible key stations and 15 other accessible stations) that are accessible individuals including people who use wheelchairs.

The NYCT rapid transit system is actually the consolidation of three historically separate systems known as the IRT, BMT and IND. These systems were constructed and operated by private and public entities in the early part of the 20th century. A major task of NYCT since its creation in 1953 has been the integration of the three systems into a single operating authority.

In addition to the core rapid rail system, NYCT is responsible for operation of service on the Staten Island Railway (SIR), another subsidiary of MTA. SIR is responsible for the railroad facilities, with the City of New York responsible for funding SIR. SIR was once owned and operated by the Baltimore and Ohio Railroad as a passenger/freight railroad. In 1971, following years of increasing deficits, the SIR was acquired by the City of New York and later conveyed to the MTA as a subsidiary corporation.

SIR operates a single route over a 14.3-mile double track, grade separated right-of-way between the Staten Island communities of St. George and Tottenville. SIR uses 64 rail cars, which were overhauled in 1990 and 1991 by NYCT to serve 22 stations. St. George, the terminal for the Staten Island Ferry service, has the highest volume passengers of stations on the route. The Staten Island Ferry service, linking Staten Island with Manhattan, is operated by the City of New York. All SIR service is coordinated with the ferry service. In 1999 SIR served five million annual passenger trips. The heaviest concentration of passenger trip ends is at the mid sector of the route near the communities of Bay Terrace, Great Kills, Eltingville, Annadale and Huguenot.

In 2001 NYCT reported operating 3,840 buses during peak periods out of a fleet of approximately 4,500 buses.

Service Area - The NYCT service area is the City of New York including all five boroughs, the Bronx, Brooklyn, Manhattan, Queens and Staten Island. The area covers 322 square miles and had a 2000 population of 8,008,278. New York City is the center of the New York - Northern New Jersey Urbanized Area, which covers 2,967 square miles and had a 2000 population of 21,200,000. (Source: United States Census)

NYCT buses operate throughout the five boroughs and in several locations serve areas (within three-fourths of a mile of the bus route) beyond the city limits. A number of NYCT bus routes operate less than three-fourths of a mile from the Bronx-Westchester County Border and the Queens-Nassau County border with routes 5 and 85 extending less than a mile into Nassau County.

Service description - The NYCT A Division consists of the former IRT system. This is the oldest NYCT Division with most facilities built between 1904 and 1928. The A Division includes the Lexington Avenue Line (Lines 4, 5 and 6), the Broadway-Seventh Avenue Line (Lines 1/9, 2 and 3), the Flushing Line (Line 7), and the 42nd Street to Grand Central Terminal Shuttle (S). The Lexington Avenue and Seventh Avenue Lines link Manhattan with the Boroughs of Brooklyn and the Bronx. The Lexington Avenue Line is the second most crowded line in the system. The cars used on the A Division are shorter and narrower than those of the other Divisions.

NYCT's B Division consists of the former IND and BMT systems. Both the IND and BMT systems shared identical car dimensions and tunnel clearances. NYCT Subdivision B-1 is the former BMT system. This is the second oldest NYCT Division, largely constructed between 1915 and 1931. Subdivision B-2, formerly the IND system, is the newest Division, largely constructed between 1925 and 1948. The B Division lines include the Eighth Avenue Line (A, C and E), the Sixth Avenue Line (B, D, F and Q), the Broadway Line (N and R), the Nassau Street Line (J/Z and M), the 14th Street Line (L), the Cross-town Line (G) and the Franklin Avenue Shuttle (S). Many of these lines also feature branches that extend into the Bronx, Brooklyn and Queens. The Queens Boulevard corridor line (E, F, R and G) is the most crowded line of the rapid transit system.

Service Hours - Citywide service is available on a 24-hour basis. Train connections with the ferry service are scheduled 24 hours a day seven days a week. Many of NYCT's bus routes also operate 24 hours a day 7 days a week.

Fares - The base fare (not including discounts) throughout the NYCT and SIR systems is $2.00 regardless of distances traveled. Free transfers are provided between vehicles. Fares on express buses are $4.00 and also include free transfers to other NYCT services.

On the SIR, fares are collected as passengers enter and exit the system at the St. George Terminal. No fares are collected at any other SIR station.

On the NYCT system, fares are collected at fare gates (turnstiles) as patrons enter the system. MetroCards are predominately used as the system-wide fare medium. Tokens are still accepted.

MetroCards may be purchased at staffed fare booths at the stations and from MetroCard vending machines. MetroCard rates are:

• 1-day Fun Pass - $7 for unlimited use of NYCT subway and local service for one day;
  
  
• 7-Day - $21 ($10.50 reduced fare) for unlimited rides on NYCT subway and local services;
  
  
• 30- Day Unlimited Ride - $70 ($35 reduced fare - for unlimited rides on NYCT subway and local services;
  
  
• Pay-Per-Ride - A $4 - $80 debit card for use on NYCT with free transfers within a two-hour period. Pay-Per-Ride offers an additional 20% in value on the debit card for refills of more than $10 in value.

When traveling via subway, people with disabilities and senior citizens upon presenting Medicare card or other specifically approved ID, such as a Paratransit ID, can purchase a $2 MetroCard from the station agent and receive a return-trip ticket. Reduced-fare MetroCards are also available by special application to NYCT.

General - NYCT assumed responsibility for operation of Access-A-Ride (AAR), its ADA Complementary Paratransit service in 1993. Previously the service had been operated by the New York City Department of Transportation. The Division of Paratransit in the NYCT Department of Buses is now responsible for AAR service. AAR has a 15 member advisory committee with new (replacement) members selected by consensus of the committee. The committee controls meeting agendas.

Trip requests are made to a central reservation center located at the offices of the NYCT Paratransit Division at 2 Broadway in New York City. Scheduling of trips is also performed in these offices. NYCT Paratransit managers oversee the reservation center operation and scheduling function, which are performed by a contractor, First Transit. Dispatch and operation of service is performed by each of eight contracted carriers at their operating facilities. The eight carriers and their locations are:

AMERICAN TRANSIT, INC.
Yonkers, N.Y. 10703

ATLANTIC PARATRANS OF NYC, INC.
Maspeth, N.Y. 11378

MAGGIES PARATRANSIT CORP.
Brooklyn, N.Y. 11208

MV TRANSPORTATION, INC.
Brooklyn, N.Y. 11233

PTM MANAGEMENT CORP.
Jamaica, N.Y. 11435

RJR PARATRANSIT CORP.
Staten Island, N.Y. 10306

STAR CRUISER TRANSPORTATION, INC.
Brooklyn, N.Y. 11234

TRANSIT FACILITY MANAGEMENT CORP.
Brooklyn, N.Y. 11208

A more complete description of operator locations appears in Appendix C. Although each of the carriers operates throughout the entire service area, each carrier is assigned trips based on 9 geographic zones. The assignment results in each carrier operating in a primary service area. The service zones are discussed further elsewhere in this report. Each of the eight operators is providing service under the terms of a five-year contract executed during August of 2001.

Customers direct concerns on the service day to the NYCT Paratransit Division's service center located at 2 Broadway. The service center addresses service issues with the appropriate carrier. The service center is also operated through a contract with First Transit under the supervision of NYCT managers.

Service Area - The DOT ADA regulations require that ADA Complementary Paratransit service be available within three-fourths of a mile of all bus routes, and within three-fourths of a mile of all rail stations (49 CFR §37.131(a)). NYCT bus and subway service, with a few exceptions, covers all of New York City. AAR provides service to all points within New York City.

There are a number of NYCT local bus routes in Queens and the Bronx that are within three-fourths of a mile of Nassau and Westchester Counties, respectively (in addition, there are bus routes in Queens operated by private carriers under contract to New York City that are within three-fourths of a mile of Nassau County). Table III.1 lists some of the NYCT bus routes that are within three-fourths of a mile of Nassau or Westchester Counties.

The areas in Nassau and Westchester Counties within three-fourths of a mile of an NYCT bus route are part of the mandated ADA Complementary Paratransit service area. With the exception of the areas associated with routes Q5 and Q85, AAR does not provide service beyond the borders of New York City.

NYCT coordinates cross-jurisdictional trips for ADA Complementary Paratransit service with three adjoining fixed route providers:

• Long Island Bus (ABLE Ride)
• Westchester County DOT (B Line)
• New Jersey Transit (Access Link)

For AAR riders who want to travel into Nassau County, AAR and ABLE Ride have three designated transfer points near the Queens/Nassau border. For AAR riders who want to travel into Westchester County, AAR and B Line have five transfer points near the Bronx/Westchester border. For AAR riders who want to travel into New Jersey, the transfer point is the Port Authority Bus Terminal in Manhattan. For any cross-jurisdictional travel, a rider must be a registered client of both services. The rider must make separate reservations with each service, be responsible for coordinating the transfer between the two services, and pay separate fares. AAR does not track the number of cross-jurisdictional trips that its clients make.

Service Hours - The USDOT ADA regulations require that ADA Complementary Paratransit service be available during the same hours and days as fixed route service (49 CFR §37.131(e)). NYCT's subways operate 24 hours a day, seven days a week. Many of NYCT's fixed bus routes also run around the clock. AAR also provides service 24 hours a day, seven days a week.

Fares - The DOT ADA regulations require that the fare for ADA Complementary Paratransit service be no more than two times the base fare for a comparable fixed route trip (49 CFR §37.131(c)). NYCT's base fare for the bus and subway is $2.00. The fare for AAR is also $2.00 for a one-way trip.

• The telephone answering standard is that the reservationist must respond to an average of 95% of all incoming calls within a maximum of 2 minutes after the call has been answered by the automatic call distribution ("ACD") system.

• NYCT has a goal of 0% denials.

• If the vehicle arrives at the pick-up location between five minutes before the scheduled pick-up time and 25 minutes after the scheduled pick up time the vehicle is considered to have arrived on time. The NYCT goal is that vehicles arrive on-time 95% of the time.

• NYCT's goal is to schedule 100% of the trips using the following travel time standards:

  Miles	Minutes
  > 0 and < 3	50
  > 3 and < 6	65
  > 6 and < 9	95
  > 9 and < 12	115
  > 12 and < 14	135
  > 14	155

Prior to and during the field portion of the compliance review the review team gathered input from the perspective of consumers to assist the reviewers in identifying compliance issues of concern to consumers. Input was collected from review of recent litigation, complaints on-file with FTA, consumer interviews, and review of customer complaints on file with AAR.

Three individuals and Disabled in Action of Metropolitan New York, the Brooklyn Center for the Independence of the Disabled, Inc., Bronx Independent Living Services, The Center for the Independence of the Disabled in New York, Inc. and Queens Independent Living Center, Inc. filed a lawsuit against the Metropolitan Transportation Authority and New York City Transit in November of 1998.

The complaint was settled on September 30, 1999. Among the agreements reached were:

• The denial rate will not exceed 1% of requested trips for the period December 1999 through the end date of the agreement, December 31, 2001.

• Late pick-ups will not exceed 10% monthly. The agreement defines a late pick-up as a pick-up past the pick-up window. Lateness caused by forces beyond NYCT's control is not included in this calculation.

• Carrier no-shows (missed trips) will not exceed 0.3% per month from July 2000 through the end date of the agreement. Carrier no-shows are defined as a trip in which the carrier's vehicle fails to appear within 45 minutes after the scheduled pick-up time.

• Negotiation between the parties of a methodology for comparing paratransit trip duration with the door-to-door travel time of the same trip using fixed route bus service. For inter-borough trips, use of subway service would also be included.

• NYCT must provide monthly reports that include:
  • Demand citywide, inter-borough and intra-borough;
  • Trips scheduled;
  • Trips completed;
  • Carrier no-shows (missed trips) by carrier;
  • Late pick-ups by carrier;
  • Vehicles in operation by carrier;
  • Vehicles ordered by NYCT for each carrier;
  • Report on complaints of long trips;
  • Number of taxi reimbursements;
  • Registered riders by borough and wheelchair use;
  • Number of passenger no-shows, early and late cancellations -city-wide, by borough, inter-borough and intra-borough;
  • Number of subscription trips.

NYCT reported that the U.S. District Court dismissed this action with prejudice in May 2002. According to NYCT the court found that NYCT had met all of the terms of the agreement except one and that it had met its good faith effort on that one term.

There were 2 recent complaints relating to AAR service on file with FTA. Both of those complaints, which were filed between April 25, 2000 and May 10, 2000, were closed.

Issues raised in the complaints were:

• Denied eligibility for Personal Care Attendant (PCA);
• No-show suspension policy;
• Poor scheduling;
• More than one wheelchair user scheduled to use vehicle, whose capacity is limited to one wheelchair at one time;
• Late pick-ups; and
• Malfunctioning lifts.

In August one of the reviewers met with a number of AAR customers and people who represented customers. Among those interviewed were three AAR customers and three consumer representatives. The purpose of these meetings was to identify concerns of consumers prior to the review as a means of focusing the review effort on potential problems. In addition a reviewer interviewed a customer representative.

Among the concerns identified were:

• Service Design - One consumer representative advocated greater use of NYCT fixed route services by people with disabilities and more emphasis on AAR as a "complementary" service. Other consumers indicated limitations in the accessibility and reliability of the fixed route system. Particular concern focused on NYCT contracted bus services. Options suggested for fixed route service were use of paratransit as a feeder service, and provision of strong customer support - both in providing strong customer support to address problems during trips and through more extensive fixed route travel training for people with disabilities.

• Eligibility - Consumers indicated that the eligibility process appears to be lenient, particularly with respect to limiting people to conditional eligibility. Consumers also described the process as long and without a clear definition of when the process began (when the application was considered complete). Consumers expressed concern that they had to deal with different eligibility staff rather than having one person assigned to their application and that this approach contributed to confusion, inefficiency. Consumers also indicated that eligibility staff was sometimes rude.

• Telephone Access - Between 4:00PM and 5:00 PM on weekends the call center will often stop taking calls for reservations. One customer said that when they got through at the end of a weekend day that they were told to call back later. Consumers indicated that callers using TTD were disconnected. Calls can take a long time when call takers are scheduling trips. Customers also indicated that it takes a long time to confirm ride status when vehicles are late.

• Reservations and Scheduling - Customers noted that the number of trip denials has declined in recent months. One customer indicated that times offered are usually good but on occasion are too early or too late. Customers said that subscription trips are changed by AAR without notifying the customer. Pick-up times are changed from that agreed upon at the time of the trip request resulting in drivers and customers having two different times for the same trip. Schedules were also described as illogical. Service was also described as variable. It appears that subscription routes are well designed but when changed result in poor performance which is corrected after a period of time. Customers said that rides scheduled with AAR appear to be lost in the system. On the day of the trip AAR would have no record of the requested trip. Customers questioned the turnover rate and training of call takers as possible contributing problems. One customer felt that schedules are not adjusted to traffic conditions in autumn resulting in longer trips than in summer.

• Transportation - Vehicles sometimes arrive early and leave before the pick-up time. One customer cited a vehicle arriving at 10:30 and departing at 10:37 for an 11:30 pick-up. Customers do not have the name of the operator making it difficult for them to identify the vehicle. Visually impaired customers are often stranded because the driver doesn't approach the customer and the customer is unaware that the vehicle has arrived. People have been dropped off by as much as 1.5 hours early resulting in customers discontinuing use of the service.

There is no effort to contact customer no-shows with the exception of one operator, RJR, which does it sometimes. Customers are categorized as no-shows when the vehicle arrives late and the customer is no longer waiting for the ride. A consumer provided an audiotape, which purports to demonstrate dispatchers directing drivers to classify missed trips as customer no-shows. Verification of what is said on the tape has not been possible because of the poor quality of the audio.

Missed trips from vehicles going to the wrong place are a problem. AAR has been addressing this problem by establishing standard pick-up locations.

Pick-ups are late in the evening. When a second vehicle is dispatched for a late pick-up it often arrives 1.5 hours late. AAR provides vouchers for Yellow Cab service for ambulatory customers who are stranded (missed trip). The vouchers are usually authorized after the trip is significantly late (1 hour). The number of vouchers available is limited by budget. Since Yellow cabs are not accessible to wheelchairs, the vouchers are not available to wheelchair users.

Travel times are long in the evenings causing some customers to avoid use of the service. Trips have been as long as 3 hours, exceeding NYCT's standard of 155 minutes (2 hours and 35 minutes) for trip length.

One consumer representative indicated that NYCT oversight and coordination with operators appeared limited.

• Resources - Consumers expressed concern about inconsistent vehicle maintenance and high driver turnover rate that they attributed to low pay rates. One indicated that drivers were insensitive.

• Customer Service - Almost all of the consumers and consumer representatives cited complaint tracking and resolution as a problem. NYCT has a goal of responding to complaints within 60 business days. A tracking system has been in place since November but according to consumers, performance in responding to complaints does not appear to be measured. Customers making complaints often are not informed as to the disposition of the complaint. They noted that calls to the resolution hot line were not treated by NYCT as complaints. They noted that the hot line was used for ride status or "where's my ride" calls.

Disabled in Action of Metropolitan New York also provided FTA with an extensive report on problems with AAR Services. Concerns raised include:

• Limitations of conditional eligibility
• Long eligibility process
• Eligibility of children
• Re-certification required of people with permanent or deteriorating conditions
• Policy changes not provided in alternate format for people with low vision
• Suspension policies too severe - suspension after 7 late cancellations and/or no-shows in 6 months, suspension for 15 days rather than 14 days in policy, 30 day appeal begins on date letter issued, not received. Subscription riders suspended from subscription service for 25% cancellations in 30 days.
• Not responding to telephone calls in queue at 5:00 PM
• Long time to make reservations
• Denial of next day trip requests
• Use of standby list for reservations
• Trip reservations lost in the system - AAR sometimes has no record of rides scheduled with customers
• Trips scheduled to the wrong destination
• Limits on schedule for return trips - 1.5 hour delay within borough, 2.5 hours between boroughs
• Unresponsive trip offers - Offers more than 1 hour from request or before customer is able to leave his/her activity
• Changed pick-up times - Subscription trip schedules are changed by AAR without notifying the passenger.
• Pressure to accept early pick-ups
• Late Cancellation and subscription missed trip policy
• Late pick-ups - Particularly during the afternoon; drivers appear to be reporting passenger pick-ups before they occur.
• Unreliable service
• Long hold times for ride checks
• Vehicles at wrong pick-up location
• No Show Policy - Passengers are incorrectly classified as no shows with no effort to confirm by AAR
• No effort to communicate vehicle arrival to people with low vision
• Stranded riders
• Inadequate back-up service - long wait for approval, pre-payment of taxi fare, delayed reimbursement, limited to ambulatory customers
• Long rides
• Indirect routing - Trips transporting customers in a zigzag pattern
• Inexperienced drivers - low driver pay rates and limited training, poor knowledge of routes
• Inaccessible sedans
• Poor van maintenance - high entry step, lift failures
• Unresponsiveness to complaints

Access-A-Ride (AAR) receives rider comments on its service by phone, e-mail, fax, and in writing. The telephone number for rider comments appears in the AAR brochure. The telephone number is also listed on the AAR web site, as well as in each rider newsletter. Telephone comments for all NYCT services are directed to this number.

AAR records and tracks comments concerning ADA Complementary Paratransit service through an email based system ("Right Now") that is used by all NYCT departments. A member of AAR's Customer Service Unit reviews the comments related to ADA Complementary Paratransit service. Comments related to carrier activities are forwarded to the appropriate carrier. The carrier is responsible for replying to these complaints (with approval from AAR). For comments concerning in-house activities, policies, or procedures, the Customer Service Unit develops the response, with help from other units (e.g., eligibility determination, contracts) as needed. A majority of replies are made in writing. Replies over the phone tend to be for complaints initially received by phone. All replies and other actions related to developing the reply are tracked in Right Now.

The goal of AAR is to reply to each complaint within 60 days. However, at the time of the review team's visit, AAR was not tracking its performance in meeting this goal.

For the first 7 months of 2003, AAR had a total of 6,902 complaints recorded in Right Now (AAR had also received 394 commendations during that period). Table III.2 presents the top categories (at least one percent) of complaints for that period. AAR's Customer Service Unit developed these categories.

The 6,092 complaints yield an average of 986 per month. In 2002, AAR recorded 15,660 complaints, or 1,305 per month. In 2001, AAR recorded 12,012 complaints, or 1,001 per month.

The Customer Service Unit staff said that they use the tracking system to "proactively determine problems." Based on the trends in number and types of complaints, they alert other service units for areas requiring focus.

The Customer Service Unit has not analyzed complaints by individual carriers. Until 2003, they had not consistently recorded that information in Right Now.

As can be seen in Table III.2, principal customer complaints were for missed trips (26.3%), late pick-ups (15.8%) and long ride times (8.5%).

1. NYCT provides AAR service 24 hours a day, seven days a week.

2. NYCT's fare of $2.00 for AAR is the same as NYCT's base fare for bus and subway service. This fare complies with the regulations for ADA Complementary Paratransit service.

3. AAR provides service to all points within New York City. However, AAR does not provide service in portions of Nassau or Westchester Counties that are within three-fourths-of-a-mile of NYCT local bus routes. Service to customers in these areas is provided through carriers operating in these areas requiring customers in the AAR service area to transfer to AAR services. The requirement for customers outside of New York City within three-fourths-of-a-mile of NYCT fixed route service to reserve separate trips from both the operator in the county of trip origin and AAR, to transfer, and pay two fares does not appear to meet the three-fourths-of-a-mile obligation.

• NYCT should consider adjusting the service area for AAR to include all addresses within three-fourths-of-a-mile of any NYCT bus route, including addresses in Nassau and Westchester Counties. Alternatively, AAR should consider reserving both legs of trips of a one-way trip for the customer with one call. This would permit the customer to reserve the trip with one call as is done with other AAR customers and permit AAR reservationists to schedule the entire trip with the adjacent county to assure transfers do not result in significantly long trip times. Similarly, AAR could consider free transfers for such trips so that these customers are paying the same fares as AAR customers.

The following summarizes the findings made as a result of the review. The findings are observations of policies, procedures, practices and performance that relate to delivery of service as required by USDOT ADA Regulations at the time of the review. Findings may be positive, neutral, or observations of opportunities to improve service delivery. The bases for these findings are presented in other sections of this report. Findings of opportunities to improve service should be used to identify corrective actions proposed by NYCT. Recommendations are also included in the body of the report for NYCT's consideration in developing corrective actions.

1. Service Parameters
   
   

   1. NYCT provides AAR service 24 hours a day, seven days a week.

   2. NYCT's fare of $2.00 for AAR is the same as NYCT's base fare for bus and subway service. This fare complies with the regulations for ADA Complementary Paratransit service.

   3. AAR provides service to all points within New York City. However, AAR does not provide service in portions of Nassau or Westchester Counties that are within three-fourths-of-a-mile of NYCT local bus routes. Service to customers in these areas is provided through carriers operating in these areas requiring customers in the AAR service area to transfer to AAR services. The requirement for customers outside of New York City within three-fourths-of-a-mile of NYCT fixed route service to reserve separate trips from both the operator in the county of trip origin and AAR, to transfer, and pay two fares does not appear to meet the three-fourths-of-a-mile obligation.

   go to top of page

2. ADA Complementary Paratransit Eligibility Determinations
   
   

   1. NYCT has established a thorough process for determining ADA Complementary Paratransit eligibility. In addition to a paper application, documentation/verification of disability is collected when appropriate. In-person interviews and functional assessments are also conducted on an as-needed basis. Professionals, with experience in assessing functional abilities, conduct in-person assessments.

   2. A review of 41 recent determinations of eligibility found that denials of eligibility appear to be appropriate. The practice of conducting an in-person interview and/or functional assessment before eligibility is denied (in most cases) is also a good way to ensure that there were no misunderstandings and miscommunications in the paper application process.

   3. A review of 23 determinations where "conditional eligibility" was granted indicated that applicants did appear to be able to use fixed route service some of the time. However, in two instances it was noted that the specific types of conditional eligibility granted might not have covered all circumstances under which applicants may not be able to use fixed route service. In one case, the type of conditional eligibility granted appeared to be specific to one trip that the applicant reported making. In the second case, an applicant who could walk limited distances and was unable to use the subway system was granted eligibility for trips involving travel of "more than 1-2 blocks," but not "inter-borough" eligibility. As a result, in the rare instance that the customer wanted to schedule a trip from within a block of a subway station the trip might not be accepted. In making determinations of conditional eligibility, NYCT should ensure that the types of conditional eligibility granted cover all situations that prevent travel by fixed route.

   4. Conditional eligibility based on path-of-travel barriers (the "navigational" type of conditional eligibility) does not appear to be used very often - even in cases where it appeared that applicants likely would be affected by such barriers. It was also noted that even if this type of conditional eligibility were to be granted, the specific barriers that would affect the person would not be included in the determination (e.g., uncontrolled intersections, lack of curb-cuts or accessible sidewalks, etc.). Without more specific information about the barriers that prevent travel, making appropriate decisions about trip eligibility for individuals affected by path-of-travel barriers would be difficult, if not impossible.

   5. Use of the "trip-by-trip" conditional eligibility category, which recognizes travel limitations from day-to-day due to episodic/variable disabilities or health conditions, was reportedly not being utilized at the time of the on-site visit due to a lack of compatibility with the subscription function of the reservations and scheduling software system.

   6. Based on an interview of staff at the Gramercy assessment site, it appears that contractors consider only assistance needed to use fixed route service when deciding whether a PCA is needed. The contractor in making the eligibility recommendation did not consider PCA assistance needed by the traveler at his or her trip destination.

   7. If NYCT requires additional information of applicants or requires applicants to participate in an in-person interview/assessment, it suspends the application processing time. As a result the time required to process applications for applicants requiring an interview for eligibility from the time that the written application was first submitted can be considerably longer than 21 days. Such applicants, who may ultimately be determined eligible for service, are not offered service during the period beyond 21 days from receipt of the initial completed application. This procedure appears to prevent ADA Eligible people from using the service during this period.

   8. About 36% of determinations that do not involve an in-person interview and/or assessment take more than 21 days and 2% take more than 35 days; about 58% of determinations that involve an in-person interview and/or assessment take more than 21 days and 14% take more than 35 days. NYCT has excellent policies and practices in-place, though, to enable applicants to use the service when determinations take more than 21 days to complete. The practice of sending a confirmation letter stating the date the application was received and indicating that the service can be used 21 days after that date, and the practice of immediately entering rider information into the ADEPT system with an "eligible date" 21 days after the date of receipt of the completed application, is particularly proactive and helpful.

   9. NYCT's current practice of identifying obvious instances of excessive no-shows and then carefully evaluating possible circumstances beyond the rider's control that may have caused the no-shows appears to meet the intent of the regulations.

   10. While the current practice appears to be appropriate, the formal policy, which considers seven or more no-shows or late cancellations in a six-month period to be an abuse of the service could unreasonably limit service to ADA eligible customers and does not appear to be consistent with the intent of the regulations. Appendix D of 49 CFR Part 37 indicates that, "suspensions of eligibility for no-shows are intended to prevent a pattern or practice of no-shows." It is further noted, "a pattern or practice involves intentional, repeated or regular actions, not isolated, accidental or singular incidents." Seven no-shows or late cancellations in a six-month period, particularly when considering a customer who uses the service frequently, may not rise to the level of a pattern or practice as intended by the regulations and described in the associated appendix.

   11. The regulations allow transit systems to suspend service for a reasonable period for riders who abuse the system by regularly "no-showing" for scheduled trips. While transit agencies have in recent years also considered "late cancellations" to be an abuse of the system and have considered this in their suspension policies, the effects of a late cancellation should be operationally equivalent to a no-show in terms of the negative impact on the service. Cancellations made several hours in advance of the scheduled pick-up time would still seem to allow the system's dispatchers to use the open vehicle time to respond to same-day operating issues. Systems, which operate without "floater" vehicles or with limited "floater" capacity, often rely on same-day cancellations to be able to operate reliably and on-time. NYCT should revise its policy of suspending persons who do not cancel by 5:00 PM the day before service and should ensure that its definition of a "late cancellation" is operationally equivalent to a no-show in terms of its impact on the service.

   go to top of page

3. Telephone Access
   
   

   1. AAR's phone system capacity appears to be adequate to handle calls in a timely manner. Average hold times and call times for calls to reservations are under twenty seconds. An exception to prompt response times is occasional 30-minute periods with long average hold times. These infrequent periods of poor performance seems to result from unforeseen shortage of reservationists staffing. Overall the system performance appears to be excellent and not an impediment to using AAR service.

   2. AAR's call management system does not provide data to explicitly present AAR's performance relative to its standard. The system generates data by 30-minute period and by day for average and maximum telephone hold times and call times, but not by AAR's telephone answering performance standard of an average of 95 percent of calls answered within two minutes after the call is transferred to the reservationists.

   go to top of page

4. Reservations
   
   

   1. AAR does not accept next day trip requests for the period 12:00 Midnight to 4:00 AM. AAR defines "next day" to begin at 4:00 AM rather than 12:00 Midnight. For a trip between Midnight and 4:00 AM, a caller must call two days in advance (e.g., Monday for a trip on Wednesday at 1:00 AM).

   2. NYCT offers trips between 61 and 65 minutes after the pick-up time requested by the customer. One occurrence of this practice was identified while observing reservations. A scheduled time was offered that was 61 minutes after the requested time. The review team also looked at all trip offers for requested pick-up times on May 15. Of 5,278 requests for pick-up times, 108 (2%) were offered trips that were between 61 and 65 minutes after the requested trip times. Of those 108 trip offers, 24 were classified as customer refusals. This was 8.7% of 277 customer refusals on May 15. This practice does not comply with 49 CFR §37.131(b) and these offers of pick-up times more than one hour from the requested time should be classified as denials of service. This practice apparently results from using the pick-up window in the scheduling software in a way that considers the beginning of the pick-up window (5 minutes before the scheduled time) rather than the scheduled time, when offering trips to customers. NYCT indicated that they are in the process of reprogramming the software to discontinue this practice.

   3. The practice of limiting the times riders can book return trips to more than 90 minutes after the going trip pick-up or appointment time for intra-borough trips and 2.5 hours after the going trip pick-up or appointment for inter-borough trips, does not appear to provide service that is comparable to fixed route service as required by 49 CFR §37.121.

   4. Other than as noted above, no trip denials were noted during observations of reservations. A total of 151 trip requests were observed and scheduling options existed and trips were offered in all cases. In reviewing NYCT data for May 15, no denials, other than as addressed in Finding D.1, were identified. NYCT reported 2 denials during the month of May and none between June 1 and August 31.

   5. Waiting lists and trip caps also do not appear to be used. Some trips are placed on standby, but in these cases riders are told that the trip will be provided and that a scheduler will get back to them with an exact pick-up time.

   6. Reviewers observed a thirteen percent (13%) trip refusal rate by riders. This observation appears to be significantly higher than the percentage of refused trips reported by NYCT. It is recommended that NYCT examine the way that trip refusals are captured to be sure that all refusals are appropriately reported.

   7. A detailed training program and call handling script have been prepared and reservationists appear to be well trained and were observed to follow the established script in most cases.

   8. Reservationists were observed to diligently repeat back and confirm key trip reservation information with callers. Reservationists also consistently conducted a final verification of the trips scheduled at the end of each call.

   9. In some cases, reservationists repeated back and verified trip requests information very rapidly and it was not clear in some cases that callers were focusing on all of the information being repeated.

   10. The current practice of offering only one possible pick-up time in response to a trip request and not searching for alternate times within the one-hour before or after the requested time may not be consistent with the concept of "negotiat[ing]" times as detailed in 49 CFR §37.131(b)(2) of the USDOT's ADA regulations.

   11. Some reservationists made limited use of the "pick-up time with earliest departure" in responding to customer needs in scheduling trips. While some reservationists often used this option, others indicated that they only used this scheduling option if callers specifically said they could not leave earlier than a certain time. Also, procedures for using the "pick-up with earliest departure" scheduling option did not appear to be clearly addressed in the Reservations Manual. Use of this option for all trips to which it applies could result in trip offers that are more responsive to customer needs and avoid seeking additional trip times when the customer can't use the offer.

   12. Some reservationists appeared to continue to place trip requests from riders who could not be served by the supplemental taxi program (i.e., required an accessible vehicle) and which were received after 2:30 PM the day before service, in "standby" mode until 4:00 PM. Others appeared to offer "standby" status up to 5:00 PM. Procedures for handling these situations do not appear to be addressed in the "Reservations Manual."

   13. Procedures for when to search for options to the one time given by the system did not appear to be clearly addressed in the Reservations Manual.

   14. There appear to be enough reservationists on-duty throughout the week to handle calls from riders in a timely way.

   go to top of page

5. Scheduling
   
   

   1. The scheduling process appears to be well structured. NYCT and First Transit staff appeared highly knowledgeable of the scheduling process and software. Staff appears to be pro-active in reviewing schedules, addressing scheduling problems, and seeking out improvements to the scheduling process and resultant schedules.

   2. Computer software modifications made over the past year seem to have significantly reduced earlier problems encountered during the transition from the PASS to the ADEPT software system.

   3. Use of the beginning of the pick-up window (5 minutes before the scheduled time) in the scheduling parameters rather than the scheduled time appears to have resulted in offers of schedule times to customers that are more than 60 minutes from the requested time. NYCT staff indicated that they are modifying the software parameters to prevent trip offers of more than 60 minutes from the requested time.

   4. Subscription trips are handled by a separate unit and appear to be adequately staffed for the current volume. Changes to the subscription program (allowing subscriptions for one trip per week instead of limiting subscriptions to three per week) could increase the burden on this department as they process and monitor subscription requests.

   5. Schedulers' use of The Point Coding List Report, Multi-Boro Report, standby list, and other reports appears to be effective in refining schedules.

   6. By working with the operators, NYCT appears to effectively address scheduling issues that affect service performance.

   7. Active efforts to improve schedules through the application of the affinity factor, on-street travel time checks, and the use of "Peter Pan" trips appear to have had a positive effect on the quality of the schedules.

   8. Scheduler's do not appear to review actual service performance to identify significantly late or long trips but instead rely upon customer complaints and the Multi-Boro Report and operator feedback to identify such trips and "ghost runs." Review of actual performance of significantly late, long and missed trips should be done to assist schedulers in identifying the cause of such trips, such as travel time assumptions, geo-coding errors, etc., and further reduce the number of significantly long, late and missed trips.

   9. Assigning trips with uncertain pick-up times (such as return trips from dialysis) to the black car voucher program appears to provide better service for the passengers and removes potential problem trips from scheduled routes.

   10. There appear to be adequate First Transit schedulers on duty throughout the week to handle trip-scheduling responsibilities, including customer callbacks, as needed.

   go to top of page

6. Transportation Operations
   
   
   1. Dispatch
      
      

      1. Dispatch procedures and operations generally appear to be effective with good coordination between the First Transit Control Center and the Operator's dispatchers.

      2. Routing of customers to different departments for each service issue can consume more staff time and discourage customers from resolving their concerns and using AAR service than addressing as many issues as practicable with one staff person. Customers who call the control center for reasons other than a same day service issue are transferred to another AAR department. In one instance a caller was rerouted to reservations to make a next day trip change after making a same day cancellation. In another instance a customer who was provided incorrect or incomplete information was routed back and forth between the control center and eligibility before resolving the customer's problem.

      3. Observed calls from customers inquiring as to the status of their ride appeared to be addressed effectively, with responses in less than four minutes. However, operator dispatchers indicated that the control center would often call to check on the status of rides that were operating within the "on-time" window. This practice can unnecessarily consume staff time and distract dispatchers from more important activities.

      4. The wording of dispatcher calls to customers announcing driver arrivals before the pick-up window could be construed to encourage, or require, customers to board the vehicle before the beginning of the pick-up window. In some instances when drivers arrive at pick-up locations before the pick-up window, dispatchers were observed telling customers that their "vehicle is waiting for them."

      5. In general, operators were observed applying procedures for customer no-shows effectively. In one instance a driver classified a customer as a no-show without dispatcher approval. In several other observations the dispatcher called the customer before the beginning of the pick-up window but not during the pick-up window. Since the customer may not be expecting the vehicle until the pick-up window begins, the early call could result in a missed connection with the customer.

      6. Same day changes in customer schedules were generally handled very well. Reviewers observed control center dispatchers making a positive effort to accommodate real time requests by customers for schedule changes.

      7. The practice of leaving a scheduled trip unchanged, when customers indicate that they will be late, until one hour before the scheduled pick-up time, limits utilization of the slack time created by the rescheduled trip.

      8. Some dispatchers appeared unclear that the scheduled time used in the software is intended to guide the driver in performing their trip. Conversely, dispatchers all seemed sensitive to the importance of the promised time in the schedules when monitoring service.

      9. Most dispatchers appeared to very effectively track performance of their routes and move trips as needed. Dispatchers poll drivers for location rather than drivers calling in their locations as indicated in AAR procedures. Polling drivers permits dispatchers to communicate with drivers at a time of their choosing, thereby avoiding interruptions from drivers when dispatchers are busy addressing late runs or no-shows.

      10. Reviewers observed some dispatchers at MV as much as 60-90 minutes late in polling drivers and devoting little time to monitoring late runs and taking corrective action.

      11. It appeared that dispatchers did not consistently report incorrect customer or trip information to the responsible department sufficiently for appropriate corrective action or that the responsible department was not always following through to correct incorrect information. Reviewers observed a number of instances of incorrect information causing service problems. This included incorrect customer IDs, addresses, phone numbers and GIS locations. In one instance an applicant for eligibility could not schedule service because of incorrect information on a form letter. Some of these problems with misinformation were recurring.

      12. The time required for operators to obtain authorization from the control center for recovery routes (route extensions or additions to address service problems) may contribute to late and missed trips. The use of route extensions and additions appears to be an effective way of addressing real time service problems. However, the time from identification of a late route by an operator's dispatcher to the route appearing late in the scheduling/dispatch system to request and authorization for recovery could be extensive. Most operators indicated that the time is excessive and they dispatch vehicles in advance of control center authorization to minimize delays to customers.

      13. The number and assignment of dispatchers appeared to be adequate for each of the operators. Most operators maintain a ratio of no more than about 20 runs per dispatcher most of the time. MV Transportation appears to have dispatchers handle up to 28 runs each on Monday and Friday afternoons. Managing 28 runs at one time could reduce the dispatcher's effectiveness in maintaining schedule control of the runs.

      14. The skill of the dispatchers appeared to be good but varied among operators and among dispatchers, which is to be expected in a large operation with personnel having differing amounts of experience. While some operators, particularly those with more experienced dispatchers, appeared to be very effective at monitoring their runs and quickly taking corrective actions when delays occurred and working cooperatively, others, particularly at MV, appeared to have difficulty tracking their runs and identifying and correcting problems.

      15. Slowdown in computer operation was reported to occur frequently by Atlantic Paratrans dispatchers and was observed on Monday afternoon, September 15, 2003. During this time, dispatchers were not able to use the system for other functions.

      16. Communications equipment appeared to be adequate to support dispatch operations.

      go to top of page

   2. Transportation
      
      

      1. Operator's practice of reviewing and revising schedules before and during the service day appears to positively contribute to service delivery.

      2. Not all drivers understood the difference between the estimated time and the negotiated pick-up times listed on the manifest. This could cause confusion in running their route.

      3. Not all drivers knew the correct pick-up window of -5 minutes/+25 minutes. This could contribute to drivers encouraging passengers to leave early, late pick-ups and miscommunication with customers.

      4. Some drivers made changes to their routes and told the dispatcher after the fact or not at all. This failure to obtain prior dispatcher approval could cause routes to run late, by either inefficient changes or dispatcher having incorrect information on the status of the run.

      5. Operators appear to have sufficient equipment to provide service.

      6. MV Transportation appears to have a high percentage of its passenger equipment assigned to maintenance and repair. In August of 2003, 14 to 29% of MV Transportation's fleet was assigned to maintenance and repair and insufficient vehicles were assigned to meet peak service requirements. Since that time the MV fleet has increased from 100 to 110 vehicles.

      7. Operators appear to have an adequate force of drivers to perform assigned routes.

      go to top of page

   3. Performance
      
      

      1. For a sample day of May 15, 2003, 94.7% of scheduled trips were completed, 3.3% were Customer No-Shows, 1.2% were categorized as No-Fault No-Shows, 0.4% were Carrier No-Shows and 0.4% were other.

      2. Taxi and black car vouchers were issued for 167 (1.7%) of the scheduled trips for May 15, 2003. Voucher redemptions accounted for 0.3% of scheduled trips for the month of April 2003. Based upon this redemption rate it appears that approximately 20% of the vouchers were used. Because of lag time in redeeming vouchers it is not possible to accurately estimate the utilization rate. Based upon these numbers, 1.4% of trips otherwise classified could be missed trips. These are trips that were not completed by the customer and the vehicle did not arrive within the pick-up window agreed to with the customer.

      3. It appears that more than 50% of the trips categorized as No-Fault No-Shows or 0.2% of scheduled trips may be missed trips. In total missed trips for May 15, 2003, are estimated at 0.9% of scheduled trips including Carrier No-Shows (0.4%), No-Fault No-Shows (0.2%), and unused vouchers 0.3%

      4. Taxi vouchers appear to effectively provide a backup service to prevent customers from being stranded. However, the apparent low utilization of the vouchers by customers (estimated at 20%) and authorization of some vouchers for trips that are already late may indicate that the vouchers have limited effectiveness in reducing the number of missed and late trips. The late authorization for customer use of vouchers, the initial customer outlay of money for the taxi fare and time and effort required for reimbursement may limit the attractiveness of vouchers to customers.

      5. For the sample day of May 15, 2003, 91.8% of pick-ups were on-time or early. This is consistent with on-time performance of 91.5 to 93.8% as reported by NYCT. For May 15 0.6% of pick-ups were more than 35 minutes after the pick-up window or 1 hour after the pick-up time negotiated with the customer. 0.1% of pick-ups were more than 1 hour after the pick-up window or 1 hour and 25 minutes after the negotiated time.

      6. On-time performance internally tracked by MV Transportation indicated some performance issues on weekends. Managers at this provider indicated that because there were fewer vehicles on the road on weekend evenings, the opportunity to reassign trips between runs if in-service problems developed was more limited than on weekdays.

      7. NYCT does not have a performance standard nor does it track performance for on-time drop-offs. On-time drop-off performance is often more critical a concern to consumers than on-time pick-up performance, particularly for trips to appointments for work, school, medical services, etc.

      8. NYCT's goal is to schedule 100% of trips using distance-based travel time standards. The distance parameters overlap potentially causing confusion. NYCT does not have a travel time policy for actual performance of trips and does not measure performance based upon AAR trip travel times.

      9. Based on reconciled trip information for May 15, 2003, the average travel time for all trips was 42 minutes; the overall average trip distance (based on odometer readings) was 9.3 miles. 68% of all trips were 50 minutes or less. Of those, the average travel time was 36 minutes and the average trip distance was 5.8 miles. 81% of all trips were 65 minutes or less and 95% of trips were 95 minutes or less. 23 trips (less than 1%) were more than the maximum travel time standard for any trip set by NYCT (155 minutes or 2 hours and 35 minutes).

      10. The average travel time for sample of 70 long trips from May 15, 2003, was 105 minutes and the average travel distance was 23 miles. The average trip mileage calculated by MapQuest for the sample trips was 13 miles.

      11. The analysis identified approximately six trips from the sample of 70 long trips that appeared to be significantly long considering the trip distance, AAR travel time and comparable fixed route travel times. An additional 16 long AAR trips from the sample that took more than 30 minutes longer than fixed trips by the "fastest route" could also be considered significantly long.

      go to top of page

7. Resources
   
   

   1. Financing of AAR services appears to have strong support from MTA and public officials as evidenced by an increase in the operating budget from $88,860 million for the year 2000 to $153,889 million for 2003. Funding does not appear to limit delivery of ADA Complementary Paratransit services.

   2. The NYCT budget process appears to effectively address the anticipated market demand for service through the use of a demographic based model and trends in use of AAR service.

   3. The present fleet appears adequate in size to serve current demand with a cushion for increased demand.

   4. Staffing of the call center and scheduling appears to be adequate and not limit service. Staff involved in management of AAR service at all levels appeared to be professional, highly experienced and dedicated to provision of public service.

   5. MV dispatch staff appeared to be taxed managing assignments of 28 runs per dispatcher. All other dispatch functions appear to be adequately staffed.

   6. The availability of drivers for some of the newer carriers may limit service. The stable labor force of long-term carriers indicates that current limitations of labor may be corrected by the passage of time.

The process used to determine ADA Complementary Paratransit eligibility was reviewed to ensure that determinations are being made in accordance with the regulatory criteria and in a way that accurately reflects the functional ability of applicants. The timeliness of the processing of requests for eligibility was also assessed. Finally, the process used to identify and suspend riders for excessive no-shows was reviewed. The review of these issues was completed as follows:

• Input about the eligibility determination process was obtained through meetings with riders and advocates, a review of rider comments on file at NYCT, and a review of formal complaints on file at FTA.

• An understanding of the eligibility determination process was developed through a review of current eligibility materials, interviews of eligibility determinations staff, and an on-site visit to a selected in-person assessment contractor site.

• Eligibility decisions were reviewed for the period from January 2003 through August 2003.

• The application files of 35 recent applicants who had been granted conditional eligibility or who had been denied ADA Complementary Paratransit eligibility were reviewed.

• The application files of 6 recent applicants who had been denied eligibility and who then appealed the initial decision were also reviewed.

• The processing times for 2,990 applications processed in August 2003 were reviewed.

• No-show suspension records between January and September 2003 were reviewed and the no-show tracking and suspension process was discussed with NYCT staff.

As noted in Section III of this report, a meeting was held with advocates and riders in advance of the on-site visit to discuss rider experiences with the paratransit service. A report prepared by Disabled in Action of Metropolitan New York was also received and reviewed prior to the on-site visit. This meeting and report identified the following eligibility determination issues:

• Some advocates/riders indicated that the process can take a long time and that there was no clear indication of when applications were considered complete.

• It was stated that applicants have to deal with different eligibility staff throughout the determination process rather than having a single person review their eligibility. It was suggested that these multiple contacts contribute to confusion about the process and appeared to be inefficient.

• Some concern was expressed about the limitations placed on applicants determined conditionally eligible.

• The approach used to determine eligibility of children was identified as a concern.

• It was noted that riders with permanent disabilities whose travel abilities were not expected to improve were still required to reapply and be recertified periodically.

• Some advocates also expressed concern that the eligibility process appeared to be too lenient and that individuals who might be able to used fixed route service some of the time were still using paratransit services for these trips.

Two formal ADA complaints regarding AAR were filed with FTA in 2000. One complaint claimed that authorization to travel with a personal attendant had been inappropriately denied. The process used to suspend riders who had accumulated no-shows was also challenged.

A relatively small number of rider comments received directly by NYCT also cited issues with various aspects of the eligibility determination process. In calendar year 2002, 145 of the 15,660 rider comments received by NYCT (about 1%) concerned the eligibility process. Between January and early September 2003, 51 of 6,902 total complaints (0.7%) concerned the eligibility process.

NYCT managers indicated that prospective clients are made aware of AAR and the eligibility process through out-reach efforts, bulk mailings. Individuals interested in applying for ADA Comprehensive Paratransit eligibility are referred to the Eligibility Determination Unit (EDU) of NYCT's Paratransit Division. Six to eight staff persons in the EDU (depending on the day and time) answer phones, provide general information about the service, and application process, and send "ADA Transportation Assessment Application" forms to callers who request them. The application form is available in large print as well as standard print and is also available in Spanish as well as English.

The standard print application form is eight-pages long (Attachment D). The first page provides instructions for completing the form/process and provides contact information should applicants need assistance. The instructions request that two photographs be submitted with the completed application form. One photo is kept on file and one is used to prepare an ID card for persons determined eligible.

The second page of the application form is a statement that must be signed by applicants. The statement notes that the applicant understands that: the form will be returned if incomplete; an evaluation by a professional certifier selected by NYCT might be required as part of the process; the information provided in the form is true to the best of the applicant's knowledge; misrepresentation of information could result in a revocation of eligibility; a failure to cooperate with requests for additional information to verify information provided in the application could be grounds for a suspension or revocation of eligibility; and failure to adhere to service policies and procedures could be grounds for suspension or revocation. The statement also indicates that applicants agree to notify NYCT if paratransit is no longer needed.

The remainder of the application form requests: general information (name, address, phone, etc.); information about current travel and use of MTA services and programs (fixed route bus, subway and reduced fare services); information about the applicant's disability and mobility aids used; information about travel abilities (i.e., ability to travel to the nearest bus stop and subway station, ability to board and ride buses and enter and use subways, obtain and understand information needed to use the service, etc.). Open ended questions are included that ask applicants to identify how their disability prevents them from using bus or subway services and to provide any other information they feel is important. One question also lists typical issues that might prevent travel by fixed route (i.e., distances to/from stops, inclement weather, hills, path-of-travel barriers, etc.) and asks applicants to identify all that might prevent their travel by bus or subway. A question about the need for a personal attendant is also included and applicants are asked to describe what assistance an attendant provides. Finally, information about the applicant's three most frequent destinations is requested.

When application forms are received, they are reviewed for completeness. Two EDU staff members perform these reviews. If significant information is not provided (e.g., the up-front statement is not signed or questions are left completely blank) the application is returned with a letter requesting that the application be sent back once the missing sections are completed.

Once applications are determined to be complete, the information contained in the application is entered into the ADEPT system. The date that a completed application was received is also noted and a confirmation letter indicating that the application has been received and is being reviewed is sent to the applicant. The confirmation letter provides the date the completed application was received and informs applicant that they can call and use the service if an eligibility determination has not been made within 21 days of that date. The telephone number to call to request service is also provided. To facilitate use of the service by persons whose applications are not processed in 21 days, the EDU staff members also then enter the applicant's name and the date they should be considered eligible (21 days after the application receipt date) into the ADEPT system.

Once information is entered into the ADEPT system, the application forms are forwarded to one of five EDU Certifiers. Each of the five Certifiers handles determination for one of the boroughs. The Certifiers review the information provided in the application form. If an eligibility decision can be made based on the information provided, they will prepare an appropriate determination letter. If the information in the application form is not sufficient to make a final determination, Certifiers might request more information from applicants (through a follow-up phone call). In some cases, this might simply be a phone interview to ask for clarification about information provided in the application. In other cases, it might be a request for documentation/verification of the disability indicated by the applicant. Documentation/verification of disability might be requested, for example, if the applicant indicates a "hidden disabilities" such as a health condition or a psychiatric condition. If requested, it is then up to the applicant to obtain and forward some form of documentation.

If documentation/verification of disability is requested, applicants are given 10 days to provide this information. If the documentation is not received within 10 days, the determination is placed in a "closed" or "suspended" category. No further action is taken on the application until documentation is provided. If provided at a later date, the file is then reopened and the process continues.

If applicants call and indicate that they do not have any documentation beyond what was provided in the application form, NYCT staff indicated that they typically would ask the person to then participate in an in-person assessment. Also, if the information in the application and follow-up information is otherwise still not sufficient to make a final determination, Certifiers can make a decision to ask applicants to participate in in-person interviews and/or functional assessments. NYCT has contracts with three health care/medical centers, which perform the interviews/assessments at five locations and then provide a summary of observations to the Certifiers. One contractor has interview/assessment sites in the Bronx, Manhattan and Staten Island. A second is located in Brooklyn and the third is located in Queens. These contractors and the boroughs they serve are:

NYCT provides annual ADA Complementary Paratransit eligibility training to assessment site contractor staff. A review of the assessment process is also conducted on-site once each quarter by NYCT.

If a NYCT Certifier decides that an interview and/or assessment are needed, the applicant is contacted and a date and time for an in-person interview/appointment are set. Applicants are given ten days advance notice of an appointment, and provided a voucher for transport to the center by a car service. A confirmation letter giving the day, time and location of the assessment is then sent. NYCT staff members create the schedule of assessments for each contractor and provide the schedules with applicant files in advance of upcoming appointments.

NYCT Certifiers specify for each in-person assessment whether they would like the contractor to only perform an interview or to do both an interview and an assessment. Professionals at each site generally follow these requests, but they have the leeway to do an assessment if, based on the interview, they believe an objective observation of abilities is needed.

Most assessments begin with a short interview. Staff at each assessment center asks each applicant a standard set of questions, contained in a "Certifier In-Depth Review Report." Applicants are asked if they are able to get to and use the subway, get to and use buses, and for information about the need for personal attendant services when traveling.

Each assessment center is also equipped to allow a functional assessment to be performed if requested by NYCT (or if determined appropriate by the assessment center staff). Each assessment center has a mock-up of a bus step-well and entrance and a wheelchair-lift. Physical therapists, occupational therapists, or other similar professionals at each site observe applicants' abilities to climb bus steps, use a bus lift, use a fare box, get to and use a seat, hold onto overhead support rails, use the stop request system, and recognize destination signs. Assessors also observe general mobility as applicants travel the equivalent of about two blocks throughout the assessment center. Gait, balance and endurance are assessed and assessors indicate on the "Certifier In-Depth Review Report" the distance that they believe applicants are able to travel to get to and from fixed route stops or stations.

If applicants fail to appear for a scheduled interview/assessment, they are sent a letter noting that the appointment was missed and that the application has been "withdrawn" as a result. The letter then indicates to applicants that if they wish to be reconsidered for eligibility they should call the EDU. If applicants call, a new interview/assessment appointment is arranged.

Completed "Certifier In-Depth Review Reports" are then sent back to the NYCT Certifiers. This information is then used, along with information provided in the application and any follow-up information obtained, to make a final determination of eligibility. The decision on eligibility is made by the NYCT Certifiers. Applicants can be found either fully eligible for paratransit service, conditionally eligible, temporarily eligible, or not eligible. If temporary eligibility is granted for less than the standard period (which is three years), applicants are made either fully or conditionally eligible for this temporary period.

A limited number of specific categories of conditional eligibility are granted. Applicants can receive conditional eligibility under one or more of the following categories:

• Extreme Cold or Inclement Weather: Individuals given this type of conditional eligibility are able to use paratransit service from December through March but not at other times of the year.

• Extreme Heat: Individuals given this type of conditional eligibility are able to use paratransit service during the months of July and August but not at other times of the year

• Inter-borough Trips: Applicants who can use bus service to travel within a borough but who are unable to use the subway system to travel between boroughs are able to use paratransit service only for trips between boroughs.

• Navigational/Unfamiliar Locations: This category of conditional eligibility is granted if applicants cannot travel to and from certain locations due to path-of-travel issues (e.g., steep terrain, lack of curb-cuts) or are only able to travel to places with which they are familiar. Applicants who have cognitive disabilities or vision disabilities and who can make some trips to places they have learned but who cannot travel to unfamiliar locations might receive this type of conditional eligibility.

• Distance: If applicants cannot travel more than a certain distance to get to and from fixed route stops/stations, they are given conditional eligibility when travel is more than this distance. Within this category of conditional eligibility there are subcategories that specify maximum distances. Applicants are given eligibility for trips that involve walking: more than 1-2 blocks, more than 3-4 blocks, or more than 5 blocks.

• Trip-by-Trip: Applicants whose disability is episodic or variable and who might be able to use fixed routes on "good days" but not on "bad days" might receive this type of conditional eligibility. NYCT staff indicated, however, that this category of eligibility is not currently being used because of software/policy issues regarding the provision of subscription service for persons with episodic/variable disabilities. They indicated that use of this category has been suspended until these software/policy issues are resolved.

Once a final determination has been made, applicants are sent a letter indicating whether they have been found eligible and the type of eligibility being granted. An "Access-A-Ride Paratransit Service User's Guide" is sent to all persons determined eligible. Letters indicating a denial of eligibility or any other limitation on eligibility (i.e., conditional or temporary eligibility) also indicate that applicants can appeal the decision and include a copy of an "Access-A-Ride Eligibility Appeal Form." The appeal form gives applicants the choice of appealing in-person or having the decision reviewed based on the written record. The form encourages applicants to provide a statements indicating why they disagree with the determination and to provide additional supporting documentation if they choose not to appear in-person.

NYCT staff also indicated that in most cases, a decision to deny eligibility is not made without the benefit of an in-person interview/assessment. This allows the NYCT Certifier to have information and observations by the assessment contractor before a decision is made to deny eligibility. In some instances, though, it may be clear from the application form and/or follow-up contacts with an applicant that the individual misunderstood eligible for AAR and it was clear that the person was not eligible. In these cases, eligibility might be denied without an in-person assessment.

If applicants request an appeal based on the written record, the application file along with any additional documentation provided is forwarded to the NYCT Appeals Officer. The Appeals Officer and a staff physician review the record. A decision is then made and the appellant is sent a letter indicating the outcome of the appeal. Because two individuals decide appeals, all decisions are made by consensus. NYCT has a goal of making appeal decisions, based on the written record, within 30 days of the receipt of a request from an applicant.

If an in-person appeal is requested, a date and time is set and the appellant is informed by telephone. In-person appeals are also heard and decided by the Appeals Officer and the staff physician. It is NYCT's goal to hear and decide on in-person appeals within 45 days of the receipt of a request.

Occasionally, applicants who appeal and who did not participate in an in-person interview/assessment as part of the original determination might be asked to participate in an assessment before the appeal is decided. The Appeal Officer and the staff physician will then consider the results of the assessment.

It is also NYCT's policy to not reduce an appellant's level of eligibility as the result of an appeal. If any changes are made to the initial decision, more eligibility rather than less is granted.

All riders are required to be recertified when their eligibility expires (three years from the date eligibility is granted unless temporary eligibility is provided). Riders are sent a notice with a copy of a new application form 60 days before their eligibility is scheduled to expire.

According to the "Eligibility Determination and Appeals Monthly Report" provided by NYCT staff for the period from January-August 2003, a total of 77,961 individuals were on record as ADA Complementary Paratransit eligible. An average of about 3,200 applications were received each month during this period. This included about 2,010 new applications and about 1,010 requests for recertification each month.

The report also indicated that about 48% of all applicants are referred for an in-person interview and/or assessment. For the eight-month report period, 11,572 applicants were referred for an in-person interview or assessment. Some applicants who are referred decide not to pursue eligibility. During the same period, 9,261 interviews/assessments were performed, which suggests that about 2,311 applicants referred (20%) do not continue with the process once asked to participate in an assessment.

For the eight-month period from January 1 through August 31, 2003, a total of 23,423 eligibility determinations were made. About 54% of these determinations granted full eligibility to applicants. Conditional eligibility was granted 40% of the time. Six percent (6%) of applicants were found not eligible.

From January 1 through August 31, 2003, NYCT received 767 requests for appeals. Twenty-seven (27) appellants had not yet participated in an in-person assessment and were asked to do so as part of the appeal process. In 520 cases (68%), the original determination decision was upheld. About 18% of the time the original decision for conditional eligibility was changed to full eligibility. And in about 14% of the appeals the conditions of eligibility originally granted were modified (in most cases applicants were given additional conditions under which they would also be eligible).

As part of the on-site assessment, 36 recent determinations where reviewed with NYCT staff. This included 13 determinations that applicants were not eligible and 23 determinations where applicants were found conditionally eligible. Determinations made by each of the five Certifiers were selected for review. In each case, the information in the file - the paper application and the interview/assessment summary - was reviewed. The reasons for the final determination were then discussed with the Certifier who made the decision.

The review found that all 13 of the "not eligible" determinations appeared to be appropriate. In-person assessments were performed in all 13 cases, and the observations made by the contractor staff suggested that the applicants could travel to and from and use fixed route transit services.

The review of selected determinations also indicated that in most of the 23 cases where applicants were given "conditional eligibility" the decisions appeared to be appropriate. There was evidence in each file that applicants could use fixed route service in some instances and required paratransit services only under certain conditions.

In a few instances, however, there was some question about the appropriateness of the specific categories of conditional eligibility granted and whether these categories covered all circumstances under which the applicants would not be able to use fixed route transit. In one case, the applicant reported severe arthritis and use of a support cane. An in-person assessment was performed and the report suggested that the applicant could only be expected to travel about one block to get to or from transit stops/stations. The applicant was only granted eligibility for "inter-borough" trips, however. The Certifier explained that eligibility based on "distance to and from stops" was not granted because the applicant only indicated a need for AAR services to get to and from her doctor, who was located in another borough. The decision appeared to focus on a specific trip need rather than on the broader conditions under which fixed route service could not be used.

In a second case, the applicant reported using a cane or walker and sometimes a wheelchair. Use of inaccessible subway stations appeared to be an issue. Conditional eligibility was only granted for trips "more than 1-2 blocks" and for travel during times of "extreme heat" and "extreme cold." Eligibility for inter-borough travel (when the subway system would need to be used) was not one of the conditions granted. The Certifier explained that giving eligibility for trips involving travel of more than 1-2 blocks should cover inter-borough travel since the subway stations were likely to be more than 1-2 blocks away. While this might be the case for most trips, it is possible that stations could be closer at both the origin and destination for some trips that the applicant may request in the future.

NYCT Certifiers noted that some applicants report having variable disabilities and might reasonably be expected to travel by bus or subway on some days but not on days that their disability or health condition caused severe limitations. They noted that the "trip-by-trip" category of conditional eligibility had been created to address these instances. They noted, however, that the granting of "trip-by-trip" conditional eligibility had been discontinued because of incompatibility with the ADEPT reservations and scheduling system. Apparently, granting subscription service for individuals granted conditional eligibility based on a variable condition was a problem. As a result, that type of conditional eligibility was discontinued.

The handling of applications from children with disabilities was also discussed, since this was an issue specifically raised by advocates and riders contacted prior to the assessment. Certifiers noted that if applicants were less than eight (8) years of age, their ability to use the fixed route system with an accompanying adult is considered. Independent travel ability is only considered when applicants are eight years of age or older. Staff explained that this policy was established to mirror the fixed route policy, which required children less than 48 inches tall to be accompanied by an adult.

The granting of eligibility for a personal care attendant (PCA) was also discussed. Staff at Gramercy Physical Therapy, who were interviewed as part of the assessment indicated that they consider whether or not an applicant might require the assistance of a PCA to perform the tasks needed to use public transit. They then make their recommendation regarding the need for a PCA on this specific need rather than on a broader need for assistance at a destination. Discussions with NYCT Certifiers also indicated that they often would rely on the contractor's observations to make a decision on the need for a PCA. As a result, these decisions may not always consider the need for attendant services at destinations rather than just for actual travel.

Finally, it was observed that eligibility determinations did not appear to always note travel limitations caused by path-of-travel barriers. The "navigational" category of conditional eligibility did not appear to be used very often. It was not used in any of the 23 determinations reviewed even though it appeared that some applicants would be affected by the lack of curb cuts or an accessible path-of-travel (e.g., applicants who indicated they used wheelchairs).

Six recent appeal cases were also reviewed with the Appeal Officer. The information in each file was examined and the appeal decision was discussed. In all six cases, the appeal decisions appeared to be appropriate.

The USDOT regulations implementing the ADA state that applicants must be treated as eligible if a determination of eligibility cannot be made within 21 days of the receipt of a completed application (49 CFR Part 37.125(c)). Information about the time required to process requests for ADA paratransit eligibility and the handling of cases where the processing time exceeds 21 days was therefore collected.

As noted above, when paper applications are received, NYCT staff review them for completeness. If applications are found to be incomplete they are returned with a letter asking the applicant to provide the missing information and then return the completed application. When applications are determined to be complete, the date they were received is entered into a computer database. At the same time, the applicant is entered into the ADEPT reservations and scheduling database and an "eligible" date that is 21 days from the date of the receipt of the completed application is entered. This ensures that if the processing of the application takes more than 21 days, the applicant will be able to call and request trips.

A confirmation letter letting the applicant know that the application has been received and is in process is also sent to the applicant (see Attachment E for a sample letter). The letter indicates the date the application was received and lets the applicant know that if a decision is not made within 21 days he/she can use the service until a decision is made. The phone number to call to request rides is also provided.

If an applicant is required to participate in an in-person interview/assessment, the processing time is "suspended" once the person is notified and the count of days does not resume until the person has appeared for the interview/assessment. Further, if an applicant schedules an appointment for an interview/assessment and is a no-show, the application is considered "closed" and the applicant is notified that the processing of the application will not resume until an interview/assessment is conducted.

Similarly, if an NYCT Certifier determines that documentation/verification of disability is needed, the applicant will be notified that documentation is needed and the processing time is "suspended" until the documentation is received. If documentation is not received within 10 days, the application is considered "closed" and the applicant is notified that processing will not resume until the required documentation is received.

NYCT maintains a computer database that includes the key dates in the process. This includes:

• The date of receipt of completed applications;
• The date applicants are notified of the need for an in-person interview/assessment (if applicable);
• The date the interview/assessment is completed;
• The date documentation/verification of disability is requested (if applicable);
• The date documentation of disability is received; and
• The date a final determination is made.

Two printouts listing the processing dates for 2,990 determinations completed in August 2003 were generated and reviewed. One printout was for 1,201 determinations that were made "in-house" without in-person interviews/assessments. The second list was for 1,789 determinations that included an in-person interview/assessment. Table V.1 below shows the reported application processing times for determinations done in-house without in-person interviews or assessments. Table V.2 presents reported processing times when in-person interviews and/or assessments were required.

As shown in Table V.1, 64% of all determinations that are made "in-house" (without an in-person interview/assessment) are made within 21 days of the receipt of a completed application. Another 18% take between 22 and 28 days to complete. Sixteen percent (16%) take between 29 and 35 days, and about 2% take 36 days or more. The longest processing time noted for this sample of 1,201 determinations was 53 days.

A review of the sample of 1,789 determinations that did involve an in-person interview/assessment showed that processing times were somewhat longer. As shown in Table V.2 below, about 42% of all determinations were made in 21 days when an in-person interview/assessment was required. Twenty-seven percent of these determinations were made in 22-28 days, 17% took up to 35 days, and 14% took 36 days or more. The longest determination time in this sample took 80 days.

It should also be noted, though, that the reported processing times in Table V.2 do not include the time required to arrange for and conduct the in-person interview/assessment. As noted above, once it is determined that an in-person visit is required, the processing time is "suspended" until the interview/assessment is completed. Essentially, NYCT considers that application "incomplete" at that point and only restarts the processing time when the information from the assessment is obtained.

To estimate how much longer it typically takes to arrange for an in-person interview/assessment, the processing time records for a sample of 133 determinations made in August 2003 that involved an interview/assessment were reviewed. The total processing times for these determinations (from dates the applications were recorded as complete to the dates of the final determinations) were compared to the processing times reported by NYCT (that did not include the time required to arrange for and conduct the assessments). This analysis indicated that, on average, an additional 15 days was required. In most cases it appeared that interviews/assessments were arranged and conducted in 7-14 days, but occasionally a longer time was required - probably due to scheduling difficulties on the applicants' end. If the time required to conduct in-person assessments was included in the processing time, it appears that probably only about 10% of determinations would be made in 21 days or less.

49 CFR §37.125(h) of the USDOT's ADA regulations states that transit agencies "... may establish an administrative process to suspend, for a reasonable period of time, the provision of complementary paratransit service to ADA eligible individuals who establish a pattern or practice of missing scheduled trips." Related language in Appendix D of the regulations states that "It is very important to note that sanctions could be imposed only for a 'pattern or practice' of missed trips. A pattern or practice involves intentional, repeated or regular actions, not isolated, accidental, or singular incidents."

A copy of NYCT's "No-Show/Late Cancellation Policy" is provided as Attachment F. As shown, the policy considers cancellations made after 5:00 PM on the day before service to be "late cancellations." Riders can be suspended from AAR service for two weeks for the first time they accumulate seven or more no-shows or late cancellations in any rolling six-month period. If riders subsequently accumulate seven or more no-shows or late cancellations in a second six-month period they can be suspended from service for three weeks. And a four-week suspension is called for if there is a third occurrence. Subscription service can also be canceled if a rider accumulates seven of more no-shows in six months and riders must wait at least four months after the date of the suspension to reapply for subscription privileges.

If a suspension is proposed, NYCT sends the rider a letter which details the policy, lists the no-shows and late cancellations by date and trip addresses, and indicates the start and end date of the proposed suspension. The letter also informs riders that they have a right to appeal the proposed suspension. Riders must complete and submit a "Notice of Intention to Appeal Suspension" form and can either appeal in writing or in person. A "Statement of Appeal" form is provided to assist rider in making written appeals (although letters stating the reasons for the appeal are also accepted).

As noted in the description in Attachment F, the current policy was adopted on May 15, 2000. NYCT staff noted, however, that it was not enforced for a period of time in 2001 and 2002 because of general service performance issues. The identification of riders with excessive no-shows and late cancellations resumed in March of 2003 and suspensions began again in April of 2003.

While the policy calls for suspensions when there are seven or more no-shows in six months, the actual practice since March of 2003 has been to only propose and enforce suspensions of riders with the worst no-show/late cancellation records. Staff indicated that since February of 2003 a list of the 10 riders with the greatest number of no-shows/late cancellations in the last two months has been generated each month. These riders usually have no-shows or late cancellations on at least 20 separate days in a two-month period. It was also noted that when generating this list staff consider "days of no-shows/late cancellations" rather than individual occurrences. So, for example, it is possible that these riders no-showed or late cancelled 20 or more round trips (40 or more total one-way trips) during this period.

Staff also indicated that before proposing a suspension, the record of no-shows will be checked against trip records and dispatcher notes to ensure that service performance issues did not cause or contribute to the no-show or late cancellation. It was also noted that staff would sometimes make follow-up calls to riders or professionals who might be working with riders to determine if no-shows or late cancellations were beyond the riders' control. Information collected from these investigations or follow-up contacts will also be used to correct "system" problems that might be causing no-shows and late cancellations. In general, staff indicated that they will "look for any reason not to have to carry out a proposed suspension" and that enforcement actions to date have really focused on what they consider to be clear cases by the worst offenders in the system.

Records indicated that a total of 81 suspensions have been enforced between March 1, 2003 and the date of the assessment (September 10, 2003). About 70 individual riders have been suspended (a few riders have been suspended multiple times).

1. NYCT has established a thorough process for determining ADA Complementary Paratransit eligibility. In addition to a paper application, documentation/verification of disability is collected when appropriate. In-person interviews and functional assessments are also conducted on an as-needed basis. Professionals, with experience in assessing functional abilities, conduct in-person assessments.

2. A review of 41 recent determinations of eligibility found that denials of eligibility appear to be appropriate. The practice of conducting an in-person interview and/or functional assessment before eligibility is denied (in most cases) is also a good way to ensure that there were no misunderstandings and miscommunications in the paper application process.

3. A review of 23 determinations where "conditional eligibility" was granted indicated that applicants did appear to be able to use fixed route service some of the time. However, in two instances it was noted that the specific types of conditional eligibility granted might not have covered all circumstances under which applicants may not be able to use fixed route service. In one case, the type of conditional eligibility granted appeared to be specific to one trip that the applicant reported making. In the second case, an applicant who could walk limited distances and was unable to use the subway system was granted eligibility for trips involving travel of "more than 1-2 blocks," but not "inter-borough" eligibility. As a result, in the rare instance that the customer wanted to schedule a trip from within a block of a subway station the trip might not be accepted. In making determinations of conditional eligibility, NYCT should ensure that the types of conditional eligibility granted cover all situations that prevent travel by fixed route.

4. Conditional eligibility based on path-of-travel barriers (the "navigational" type of conditional eligibility) does not appear to be used very often - even in cases where it appeared that applicants likely would be affected by such barriers. It was also noted that even if this type of conditional eligibility were to be granted, the specific barriers that would affect the person would not be included in the determination (e.g., uncontrolled intersections, lack of curb-cuts or accessible sidewalks, etc.). Without more specific information about the barriers that prevent travel, making appropriate decisions about trip eligibility for individuals affected by path-of-travel barriers would be difficult, if not impossible.

5. Use of the "trip-by-trip" conditional eligibility category, which recognizes travel limitations from day-to-day due to episodic/variable disabilities or health conditions, was reportedly not being utilized at the time of the on-site visit due to a lack of compatibility with the subscription function of the reservations and scheduling software system.

6. Based on an interview of staff at the Gramercy assessment site, it appears that contractors consider only assistance needed to use fixed route service when deciding whether a PCA is needed. The contractor in making the eligibility recommendation did not consider PCA assistance needed by the traveler at his or her trip destination.

7. If NYCT requires additional information of applicants or requires applicants to participate in an in-person interview/assessment, it suspends the application processing time. As a result the time required to process applications for applicants requiring an interview for eligibility from the time that the written application was first submitted can be considerably longer than 21 days. Such applicants, who may ultimately be determined eligible for service, are not offered service during the period beyond 21 days from receipt of the initial completed application. This procedure appears to prevent ADA Eligible people from using the service during this period.

8. About 36% of determinations that do not involve an in-person interview and/or assessment take more than 21 days and 2% take more than 35 days; about 58% of determinations that involve an in-person interview and/or assessment take more than 21 days and 14% take more than 35 days. NYCT has excellent policies and practices in-place, though, to enable applicants to use the service when determinations take more than 21 days to complete. The practice of sending a confirmation letter stating the date the application was received and indicating that the service can be used 21 days after that date, and the practice of immediately entering rider information into the ADEPT system with an "eligible date" 21 days after the date of receipt of the completed application, is particularly proactive and helpful.

9. NYCT's current practice of identifying obvious instances of excessive no-shows and then carefully evaluating possible circumstances beyond the rider's control that may have caused the no-shows appears to meet the intent of the regulations.

10. While the current practice appears to be appropriate, the formal policy, which considers seven or more no-shows or late cancellations in a six-month period to be an abuse of the service could unreasonably limit service to ADA eligible customers and does not appear to be consistent with the intent of the regulations. Appendix D of 49 CFR Part 37 indicates that, "suspensions of eligibility for no-shows are intended to prevent a 'pattern or practice of 'no-shows.'" It is further noted, "a pattern or practice involves intentional, repeated or regular actions, not isolated, accidental or singular incidents." Seven no-shows or late cancellations in a six-month period, particularly when considering a customer who uses the service frequently, may not rise to the level of a pattern or practice as intended by the regulations and described in the associated appendix.

11. The regulations allow transit systems to suspend service for a reasonable period for riders who abuse the system by regularly "no-showing" for scheduled trips. While transit agencies have in recent years also considered "late cancellations" to be an abuse of the system and have considered this in their suspension policies, the effects of a late cancellation should be operationally equivalent to a no-show in terms of the negative impact on the service. Cancellations made several hours in advance of the scheduled pick-up time would still seem to allow the system's dispatchers to use the open vehicle time to respond to same-day operating issues. Systems, which operate without "floater" vehicles or with limited "floater" capacity, often rely on same-day cancellations to be able to operate reliably and on-time. NYCT should revise its policy of suspending persons who do not cancel by 5:00 PM the day before service and should ensure that its definition of a "late cancellation" is operationally equivalent to a no-show in terms of its impact on the service.

1. When making determinations of conditional eligibility, it is recommended that NYCT tailor these decisions to the specific travel limitations of each applicant rather than fitting them to categories. It is also recommended that path-of-travel barriers be more fully explored in the in-person functional assessments and final eligibility determinations. Explaining conditional eligibility in a more specific way rather than through general categories may also be more helpful to riders. For example, finding someone conditionally eligible "when travel to or from a bus stop or subway station is prevented due to a lack of curb-cuts or the lack of a sidewalk" may be more understandable than "navigational" eligibility.

2. NYCT should work to eliminate impediments to use of the "trip-by-trip" conditional eligibility category so that this type of conditional eligibility can again be used. In the absence of this type of conditional eligibility, NYCT should ensure that decisions made for applicants with variable/episodic disabilities take into consideration the applicant's travel abilities on days when the impacts of the disability/health condition are most severe.

3. It is recommended that contractor staff that interview and/or assess the functional abilities of applicants be asked to determine the need for a PCA based not only on assistance needed to travel on the fixed route system but also on the need for assistance with other tasks when at a destination. If contractor staff continue to only consider the need for a PCA in transit, NYCT should be sure that Certifiers are aware that recommendations from assessors on PCAs only consider this very narrow need and that Certifiers consider other types of PCA assistance that the traveler may need at their decision.

4. As it is unclear to AAR applicants that there might be additional requirements made of them after submitting the written application, it is recommended that NYCT include the time required to arrange and conduct in-person interviews/assessments as part of the total processing time. Use of the service should be extended to applicants if it is not possible to review the paper application, arrange an in-person assessment and make a final determination in 21 days. It would be reasonable to "suspend" the processing time, though, if caused by the applicant not meeting a scheduled in-person assessment or delaying the scheduling of an assessment. Since in-person assessments are requested on an "as needed" basis, are not required of all applicants, are an additional step required by NYCT for only certain applicants, and it is not clear that an application may not be complete upon the submittal of a written application, the "application" should be considered complete when all information requested in the paper application required of all applicants is provided by the applicant. In the alternative, NYCT could clarify that an application is not considered complete upon the submittal of the written application, putting applicants on notice as to when the 21-day clock begins.

5. It is recommended that NYCT revise its policy for service suspensions to take into consideration the frequency of use of the service and the percentage of trips no-showed or late cancelled. A frequency of no-shows and late cancellations that shows a clear "pattern or practice" should then be the basis for a proposed suspension of service. It is also recommended that public input be obtained for this policy revision.

6. It is recommended that NYCT revise its policy regarding no-shows and late cancellations to only count cancellations made within several hours before the scheduled pick-up time toward a possible suspension of service. It is also recommended that public input be obtained for this policy revision.

The review team collected information about telephone access to NYCT's AAR service for this part of the review. Telephone access for placing or changing trip reservations or checking on the status of a ride is an important part of ADA Complementary Paratransit operations. The inability to get through on the phone to place trip requests without significant delays could discourage people from using the service and could therefore be considered a form of capacity constraint.

The review team conducted the following activities:

• Gathered consumer input,
• Reviewed performance standards for wait and hold times,
• Reviewed the design of the phone system,
• Observed reservations and command center personnel handling of calls,
• Reviewed phone system monitoring (Automatic Call Distribution) reports, and
• Made calls to the AAR reservations line to hear how the system directs callers and to test access and hold times.

At the August 1, 2003 meeting, AAR riders and representatives raised the following concerns related to telephone access:

• One consumer stated that between 4:00 and 5:00 PM on weekends the call center will often stop taking calls for reservations,
• One customer said that when they got through at the end of a weekend day that they were told to call back later,
• Consumers indicated that callers using TTD were disconnected,
• Consumers noted that calls can take a long time while the call takers are scheduling trips, and
• One consumer representative claimed that AAR did not respond to telephone calls that were still in queue at 5:00 PM.

Neither of the two complaints filed with FTA in 2000 included issues relating to telephone access.

In NYCT's internal complaint tracking system, AAR had a total of 68 complaints classified as "telephone problems" for a recent 12-month period (August 2002 to July 2003; only 14 since January 2003). These complaints could include those relating to telephone access.

NYCT has established the following service performance standards for AAR service:

The reservationists must respond to an average of 95% of all incoming calls within a maximum of 2 minutes after the call has been answered by the automatic call distribution system (ACD).

NYCT's AAR has one telephone number for all voice communications with the general public (toll free 1-877-337-2017 or 1-646-252-5252). AAR has 11 "T-1" lines, each with a capacity of 24 telephone lines. At the time of the on-site review, AAR's phone system allowed a total of 212 calls to be active in queue (beyond that, the caller would get a busy signal). These lines were available to all AAR and on-site contractor staff, with no allocation among the organizational units. The number of workstations in the various units: 72 for reservations, 50 for transit control, and 12 for eligibility determination, establish the limits of hardware to accept calls. In addition to these workstations, the rest of the AAR staff shares these telephone lines. AAR uses a call management system from Avaya (Version 6, purchased from Lucent Technologies in 1998).

When a caller dials the public number, there is a menu with the following choices:

• "1" - get information on eligibility, appeals, and applications
• "2" - make a trip reservation
• "3" - change a trip reservation
• "4" - cancel a trip
• "5" - find out about today's trip ("Where's my ride?")
• "0" - talk to an operator or repeat the menu choices

AAR accepts reservations between 7:00 AM and 5:00 PM every day. The eligibility determination unit accepts calls on weekdays between 9:00 AM and 5:00 PM. If the caller presses "1," "2," or "3," when the eligibility or reservations units are not open, the call is forwarded to the transit control unit. Customer information on telephone access and service is provided in the "Access-A-Ride Paratransit Service User's Guide" (Attachment G) and the "MTA New York City Transit Telephone Directory for Customers with Disabilities" (Attachment H).

First Transit staff performs the trip reservations and initial scheduling process at a central office located in NYCT offices at 2 Broadway. A NYCT supervisor oversees the management of this function by the contractor.

The work schedule for the week of September 7-13, 2003 showed that First Transit employed a total of 164 reservationists. Most reservationists (132) were scheduled to work the full period when trip requests were accepted (from 7:00 AM to 5:00 PM) four days a week. Seven reservationists worked five-day weeks from 8:00 AM until 5:00 PM and provide additional call-taking capacity at the end of the day. The remaining 16 reservationists worked six-hour part-time shifts, four-days a week: either from 9:00 AM to 3:00 PM, 10:00 AM to 4:00 PM, or 11:00 AM to 5:00 PM. Table VI.1 below shows the total number of reservationists scheduled by day of the week for the week of September 7-13, 2003. It also shows the number of reservationists scheduled at 7:00 AM, when the reservation lines were opened each morning, and at 5:00 PM, when the reservation line is closed. As shown, 68 reservationists were scheduled for weekend days (with 62-64 available at the beginning and end of the day). On weekdays, between 80 and 116 reservationists are available (with between 63 and 100 on-duty at the beginning and end of the day).

A thorough training program appears to have been established to train new reservationists in the use of the automated reservations and scheduling system and in AAR policies. A training facility with its own dedicated version of the reservations/scheduling system has been set-up to allow new reservationists to practice using the system in a simulated but "real-life" environment. NYCT also has a "Reservations Manual" that addresses all aspects of the reservations and scheduling processes. The manual includes a detailed "script" for reservationists to follow when answering customer calls.

The review team collected and reviewed data from AAR's call management system for seven months: February 2003 to August 2003. The review team also made some sample calls to AAR to check on the response time.

Data collected by the call management system includes:

• Number of incoming calls,
• Average time to answer (hold time),
• Maximum time to answer,
• Average call length, and
• Number of abandoned calls.

Data is available by the hour and by the day. The system generates data separately for reservation lines and dispatch lines. Table VI.2 presents summary data for AAR's reservations telephone lines. Table VI.3 presents summary data for AAR's Control Center telephone lines.

The review team looked at call-data for individual days and 30-minute periods for these seven months to see if there were time periods that had consistently poorer performance, as well as to determine the reason for the maximum values for answer time. The maximum values for the reservation unit's answer time generally appeared to be exceptions, e.g., AAR reservations unit was short staffed on that particular shift. Overall, the average hold times presented in the daily reports show that the average time to answer - measured in 30-minute periods - is well below AAR's standard of two minutes. This indicates that telephone access is not an impediment to using the AAR service. However, AAR does not collect data to explicitly show AAR's performance relative to its standard, i.e., the 95th percentile for hold times.

The review team made a total of 34 test calls between July 31 and September 7, 2003. These calls included a random sampling by day of week and time of day, between 9 AM and 5 PM, including three calls between 4:30 and 5:00 PM on a weekend day. The time to answer (reach the initial menu) ranged from three to five seconds. Once the call made the menu selection, only two of the 34 calls had hold times greater than 15 seconds (24 seconds at 4:20 PM and 64 seconds at 4:40 PM).

1. AAR's phone system capacity appears to be adequate to handle calls in a timely manner. Average hold times and call times for calls to reservations are under twenty seconds. An exception to prompt response times is occasional 30-minute periods with long average hold times. These infrequent periods of poor performance seems to result from, unforeseen shortage of reservationists staffing. Overall the system performance appears to be excellent and not an impediment to using AAR service.

2. AAR's call management system does not provide data to explicitly present AAR's performance relative to its standard. The system generates data by 30-minute period and by day for average and maximum telephone hold times and call times, but not by AAR's telephone answering performance standard of an average of 95 percent of calls answered within two minutes after the call is transferred to the reservationists.

• AAR should collect data from its call management system so that it can explicitly measure performance against its standard for answering reservations calls.

In this portion of the compliance review, the team examined how trip requests from riders were handled. Particular attention was given to whether Access-A-Ride (AAR) uses any form of trip caps or waiting lists and whether there was a pattern or practice of denying a significant number of trip requests. The following information was gathered and analyzed:

• Input from customers and advocates was obtained through telephone interviews, a review of recent court documents, and a review of comments and complaints on file at NYCT;

• Reservations and scheduling policies, practices, and performance standards were reviewed;

• Service reports prepared by NYCT showing the number of trips requested, scheduled and denied were examined; and

• First-hand observations of the handling of trips were made, and staff members were interviewed about the ability to accommodate trip requests.

Consumer input, as detailed in Section III, was gathered through interviews with representatives of the disabled community, review of a report provided by Disabled in Action of Metropolitan New York and AAR complaint records for the first 7 months of 2003.

Reservations issues identified by representatives of the disabled community and the Disabled in Action Report included:

• Calls can take a long time when call takers are scheduling trips. Representatives also indicated that it takes a long time to confirm ride status when a vehicle is late.

• Representatives noted that the number of trip denials has declined in recent months. One customer indicated that times offered are usually good, but on occasion are too early or too late. Representatives questioned the turnover rate and training of call takers as possible contributing factors. Seventy-six complaints (1.1%) filed with AAR during the first 7 months of 2003 related to call takers being unhelpful.

• Consumers described missed trips from vehicles going to the wrong place as a problem. Four percent of the complaints filed with AAR were for wrong addresses and an additional 2.9% of complaints related to incorrect reservation information. AAR has been addressing the problem of wrong pick-up locations at major centers by establishing standard pick-up locations.

Disabled in Action of Metropolitan New York also cited use of standby lists for reservations and trip reservations lost in the system (AAR sometimes has no record of rides scheduled with customers) as issues.

NYCT standards that guide the reservations and scheduling process include a goal of 0% denials; an "on-time" pick-up window of between five minutes before the pick-up time scheduled with the customer and 25 minutes after the scheduled pick up time (-5/+25). The NYCT goal is that vehicles arrive on-time 95% of the time. NYCT's goal is to schedule 100% of the trips using the following travel time standards:

The Access-A-Ride User's Guide advises customers that they can apply for subscription service if they use the service for three or more trips per week. Subscription service is subject to availability limits.

Trip requests are accepted from 7:00 AM until 5:00 PM, seven days a week, 365 days a year. Trip requests can be placed from four days in advance until 5:00 PM the afternoon before the day of service. However, AAR defines "next day" to begin at 4:00 AM rather than 12:01 AM. Therefore, a caller who equests service for the following day can get a trip only if it is after 4:00 AM. For a trip between midnight and 4:00 AM, a caller must call two days in advance (e.g., Monday for a trip on Wednesday at 1:00 AM).

The "Reservations Manual" (page 1) notes that, for trips within the same borough, riders cannot book return trips that are less than 90 minutes after the going trip pick-up or appointment time. For inter-borough trips, returns cannot be requested less than 2.5 hours after the going trip pick-up or appointment. The practice of limiting the times riders can book return trips after the going trip, does not appear to provide service that is comparable to fixed route service as required by 49 CFR 37.121. A person using fixed route service can begin their return trip upon arrival at a destination. Whereas it appears reasonable to allow for the pick-up window (25 minutes on the going end and 5 minutes on the return) plus travel time to the destination, in scheduling such trips, and it may be reasonable to allow a small amount of time for the customer to conduct business at the destination, minimum requirements of 90 minutes and 2.5 hours do not appear to provide ADA eligible customers with comparable service.

As calls are received, reservationists are trained to request and record information and schedule trips as follows:

• Confirm that the person is calling to place a reservation by asking, "Are you booking a reservation?" This allows the reservationist to quickly determine if the person is calling for another reason and to transfer them to the appropriate area if they are calling for eligibility information, to check on a ride for that day, or for other reasons.

• Request the caller's AAR ID number and enter it into the trip-booking screen.

• Verify the caller's name (as it appears when the ID# is entered) to ensure that the ID number given is correct.

• Ask if the trip is beginning at the caller's home address, and if so verify the home address, cross-streets, and home phone number that appears in the system (reservationists are instructed to start the trip booking process by loading in the "primary" (home) address). If the trip is originating at a different address, that address is entered into the system in place of the home address and the cross-streets and phone number at that address are verified. In some cases, different originating addresses are already in a recent trip history file and reservationists are able to simply select and automatically load in these addresses. In other cases, reservationists type in and then go through a process of geo-coding the address.

• Request the destination address. Again, if the destination address is one that is in the rider's recent trip history file, the address is selected and automatically loaded into the system. If not, the destination address is entered and geo-coded. The cross-streets at the given destination as well as a phone number are also verified (if the address is selected from an existing list) or requested (if this is a new destination).

• The date/day of the trip is requested and entered. If callers state a date, the day of the week is verified. If callers state a day (e.g., "next Monday"), reservationists confirm the date with the caller.

• Ask, "Do you have an appointment?" If the caller indicates that there is no appointment, the reservationist asks, "What time would you like to be picked up?"

• Determine if the person will be using any mobility equipment by asking "Will you be using any type of equipment?" The system automatically loads the primary mobility aid(s) used by the caller and this is modified as needed.

• Determine if the caller will be traveling alone or with an attendant or companion(s) by asking "Will you be traveling alone?" Again, the system automatically indicates if the person is authorized to travel with an attendant and this information is modified as needed.

Consistent with the Reservations Manual, the Access-A-Ride Paratransit Service User's Guide advises rider's requesting trips to have the following information:

• ID Number;
• Trip Date;
• Desired arrival time;
• Trip origin and destination addresses;
• Whether accompanied by a Personal Care Attendant (PCA) or guest;
• Telephone number at destination; and
• Any special instructions.

The reservationist has three options for scheduling the trip using the "Trip Time & Type" portion of the trip-booking screen. The choices are:

• The appointment/drop-off time;
• The pick-up time; or
• The pick-up time with earliest departure.

Using the information provided by the caller, the reservationist selects the appropriate scheduling approach.

If the trip is scheduled based on the stated appointment time, the system will look for scheduled pick-up times that will ensure that the rider gets to their destination from 30 minutes before to no later than the appointment time. If the trip is scheduled based on a pick-up time, the system will look for available times that are anywhere from an hour before to an hour after the requested pick-up time. And, if the "pick-up time with earliest departure" scheduling option is selected, the system will look for pick-up times that are an hour after but not before the requested time.

Once the above information has been requested, verified and entered, the reservationist clicks on a "Schedule Trip" button and the system attempts to schedule the trip. If a scheduling option exists, the system will display the one best-scheduled pick-up time. This time is then offered to the caller. If the rider accepts the time, the reservationist clicks on the "Accept" button and the trip is scheduled. If the time is not acceptable to the rider, the reservationist clicks on the "Client Refusal" button and the trip request is stored in the system as a refusal.

If no scheduling options can be identified by the system, a "Trip Failed to be Scheduled" window appears. This window gives the reservationist the choice of either placing the trip in a "standby" category, recording it as a denial ("failed due to capacity" button), or offering the caller supplemental taxi service.

The "Reservations Manual" (page 15) indicates that trip requests can be placed in "standby" as long as the request is received before 2:30 PM one day in advance. It was explained that this cut-off time was established to give schedulers time to review and clear the standby list and to call all riders on the list back with final scheduled times. The "Reservations Manual" also indicates that taxi service is to be offered after 2:30 PM the day before service if no trip options are identified by the system.

If trips are placed on "standby," reservationists let callers know that AAR schedulers will continue to work on the trip request, that it will be scheduled, and that a scheduler will call them back with a final scheduled time.

If no scheduling option for a trip request is identified by the system, and if the trip is not placed in "standby" mode or the rider is not offered a taxi, the reservationists will click on the "fail due to capacity" button and the trip request will be recorded as a denial.

Return trips, or additional legs of trips, are booked in a similar way. Information about origin and destination addresses, phone numbers at each location, cross-street information, and travel times are requested, verified and entered. An appropriate scheduling method (i.e., drop-off, pick-up, or pick-up with earliest departure) is then selected. A scheduling option is then generated and the time offered to the caller.

Once all legs of the trip being requested are scheduled, reservationists call-up a "trip verification" screen that displays summary information about the trips. This information is then read-back to the caller to ensure that the addresses and times in the system are correct. The "Reservations Manual" also instructs reservationists on several other "Things to Remember," including:

• Confirming whether trip times are AM or PM;
• Getting information about the section of Queens for trips requested to or from that borough;
• Asking for the type of destination facility;
• Asking for information about the exact entrances to places like malls and hospitals; and
• Getting information about exact terminals for trips to or from airports.

This additional information is then entered into comment fields attached to the origin and destination of trips.

Finally, after completing the trip booking process and verifying trip information with callers, reservationists are instructed to inform callers "to be outside 5 minutes prior to their pick-up time and allow 25 minutes after the pick-up time for the driver to arrive."

Three assessment team members sat with several different reservationists and observed and recorded the handling of calls on Wednesday afternoon, September 10, and at various times throughout the day on Thursday, September 11. Telephone "splitters" were used to allow assessment team members to listen to the calls. In addition to observing the handling of trip requests, team members discussed reservations policies and procedures with each reservationist. A First Transit reservations Supervisor and the NYCT Control Center Manager and Reservations Supervisor were also interviewed. Tape recordings of randomly selected calls made on Monday, September 8, 2003, were also reviewed. These interviews and observations provided the following additional information about reservations procedures or practices that were not detailed in the "Reservations Manual:"

• Reservationists indicated that callers are limited to two round-trip requests (or multi-legged-trip requests) per call. Riders must call back if they need to book more than two trips.

• In most cases, only one pick-up time is generated and offered to callers for each trip requested. If this time is not acceptable to the caller, reservationists indicated that they would usually tell the caller that they must call back to see if a different scheduling option can be identified. For example, if a rider asks for a 9:00 AM pick-up, is offered an 8:00 AM pick-up, but doesn't want to be picked-up that early, they cannot ask the reservationists handling their initial request if a later pick-up time is available or ask the reservationist to try again using a later requested pick-up time (say 9:30 AM) to see if this will give them an option more to their liking. Instead, they must call back and make a new request.

• Some reservationists indicated that in some cases they would try a new time if the original trip offer does not appear to be responsive to the caller's needs. For example, one reservationist noted that if it were close to the end of the day she would look for a better time rather than asking the rider to call back and make a new request.

• Once the rider's ID# and trip date are entered into the system, reservationists will call-up a listing of all trips by that rider for that day. This procedure is used to avoid duplicate bookings. It also appears to be used to determine if riders are accepting trips and then calling back for the same trip to see if pick-up times more to their liking can be obtained.

• The practice of using the "standby" and "taxi" options appeared to be slightly different from the procedure described in the "Reservations Manual" and seemed to vary somewhat by reservationist. Because the taxi option did not meet the needs of riders who required accessible vehicles, some reservationists indicated that they would offer taxi service to ambulatory riders if no trip option could be identified after 2:30 PM the day before service. These reservationists indicated that "standby" service was offered up to 4:00 PM to riders who used wheelchairs since the taxi option would not meet their needs. The First Transit Reservations Supervisor also confirmed use of this practice of using the standby option up to 4:00 PM for riders who used wheelchairs. There appeared to be differences in the way that trip requests from riders who used wheelchairs, received after 4:00 PM, that could not be scheduled, were handled. Some reservationists were observed placing these requests on standby. Others indicated, that they would offer taxi service to these riders and did not put any trips on standby after 4:00 PM.

• There also appeared to be some difference in the way that reservationists used the "pick-up time with earliest departure" scheduling option. Some reservationists appeared to use this often. Others indicated that they only used this scheduling option if callers specifically said they could not leave earlier than a certain time. Others indicated that they used it primarily for return trips from dialysis centers or return work trips.

All reservationists did appear to closely follow the basic script for booking trips.

The process of verifying information was also observed to be very thorough. In some cases, though, riders were asked to verify so much information in a very short period of time that they did not seem to be fully focusing on the information being given by the reservationist.

Callers did not always appear to be clear on trip time limitations. For example, calls were observed where the rider stated a requested pick-up time but did not indicate that this was the earliest time that they could leave until after a scheduled pick-up time was offered. At that point, because the policy is to only offer one time, the person would be told that they would need to call back.

One call was observed where the rider indicated that the home address in the system was incorrect. The rider indicated that he had moved "a year ago." The reservationist then entered the new home address into the trip-booking screen. The reservationist did not, however, complete a form to ensure that the address was corrected in the rider file (something done only by the Eligibility Determination Unit (EDU). The reservationist also did not tell the caller that they needed to call a different number (the EDU number) to make the address change. This rider's home address was then looked-up with the help of a supervisor the following day and it appeared to still be the old incorrect address.

NYCT tracks each trip request through to its ultimate disposition. Table VII.1 Presents the number of trips scheduled and denied each month from January 2001 through June 2003.

As can be seen in the table, NYCT reports a large reduction in the number of trip denials in April of 2003, with zero denials during the month of June. During the previous periods denials ranged from 445, or 0.2% of scheduled trips, in January 2001 to 3,523, or 1.7% of scheduled trips in October of 2003.

For the month of May NYC reported 272,480 trip requests. Of this number they reported 2 denials, 6,515 (2.4%) customer refusals, and 17,370 (6.4%) early cancellations, leaving 248,593 (91%) of requested trips scheduled.

For a more detailed review of the results of the reservations process the review team focused on reported data for one day, May 15, 2003. The disposition of trip requests for that day is presented in Table VII.2

Although no trip denials were recorded for the sample day, of the 277 customer refusals, 24 (8.7%) were for trip offers that were between 61 and 65 minutes after the requested trip times. These offers of pick-up times more than one hour from the requested time are denials of service. Early cancellations are customer cancellations made before 5:00 PM on the day before service. Early taxi vouchers are authorizations given the day before service for customers to use taxis and be reimbursed by NYCT. On May 15 of the 11,576 trip requests, 10,037 were scheduled for the operators and 28 were authorized taxi vouchers.

The review team also looked at all trip offers for requested pick-up times on May 15, 2003. Of 5,278 requests for pick-up times, 108 (2%) were offered trips that were between 61 and 65 minutes after the requested trip times. This number includes the trip refusals discussed above.

As part of the observation of the reservations process, assessment team members recorded the handling of a total of 151 trip requests. Table VII.3 below summarizes these observations. As shown, no trip requests were denied due to an inability to identify pick-up times within an hour of the requested times. It is also important to note that most trip requests (44%) were placed only one day in advance and only 10% were placed a full four days in advance. This suggests that riders do not feel that they need to call as far ahead as possible to schedule trips.

A total of 19 trip requests (13%) were not accepted by riders and were recorded as refusals. This observation was somewhat different than the call handling statistics reported by NYCT. NYCT "Paratransit Operating Statistics" reports for the period from February through July 2003 indicated that riders refuse only 2.5% of all trip requests. This 2.5% is consistent with NYCT's report for May and the sample data day of May 15. Some of the difference is due to the fact that trip requests observed were only non-subscription requests. If subscription trips, for which there would be no refusals, are factored in, the difference would be narrowed somewhat. Still, the number and percentage of trip requests that are refused appeared to be higher than the service reports would suggest. NYCT indicated customers would often make multiple calls for one trip request and refuse the trip offer in hopes of getting a more desirable time on a subsequent call. In performance reports NYCT counts multiple refusals from a customer for the same trip as only one refusal. This would account for the higher number of refusals during the reviewers' observations.

1. AAR does not accept next day trip requests for the period 12:00 Midnight to 4:00 AM. AAR defines "next day" to begin at 4:00 AM rather than 12:00 Midnight. For a trip between Midnight and 4:00 AM, a caller must call two days in advance (e.g., Monday for a trip on Wednesday at 1:00 AM).

2. NYCT offers trips between 61 and 65 minutes after the pick-up time requested by the customer. One occurrence of this practice was identified while observing reservations. A scheduled time was offered that was 61 minutes after the requested time. The review team also looked at all trip offers for requested pick-up times on May 15. Of 5,278 requests for pick-up times, 108 (2%) were offered trips that were between 61 and 65 minutes after the requested trip times. Of those 108 trip offers, 24 were classified as customer refusals. This was 8.7% of 277 customer refusals on May 15. This practice does not comply with 49 CFR §37.131(b) and these offers of pick-up times more than one hour from the requested time should be classified as denials of service. This practice apparently results from using the pick-up window in the scheduling software in a way that considers the beginning of the pick-up window (5 minutes before the scheduled time) rather than the scheduled time, when offering trips to customers. NYCT indicated that they are in the process of reprogramming the software to discontinue this practice.

3. The practice of limiting the times riders can book return trips to more than 90 minutes after the going trip pick-up or appointment time for intra-borough trips and 2.5 hours after the going trip pick-up or appointment for inter-borough trips, does not appear to provide service that is comparable to fixed route service as required by 49 CFR §37.121.

4. Other than as noted above, no trip denials were noted during observations of reservations. A total of 151 trip requests were observed and scheduling options existed and trips were offered in all cases. In reviewing NYCT data for May 15, no denials, other than as addressed in Finding D.1, were identified. NYCT reported 2 denials during the month of May and none between June 1 and August 31.

5. Waiting lists and trip caps also do not appear to be used. Some trips are placed on standby, but in these cases riders are told that the trip will be provided and that a scheduler will get back to them with an exact pick-up time.

6. Reviewers observed a thirteen percent (13%) trip refusal rate by riders. This observation appears to be significantly higher than the percentage of refused trips reported by NYCT . It is recommended that NYCT examine the way that trip refusals are captured to be sure that all refusals are appropriately reported.

7. A detailed training program and call handling script have been prepared and reservationists appear to be well trained and were observed to follow the established script in most cases.

8. Reservationists were observed to diligently repeat back and confirm key trip reservation information with callers. Reservationists also consistently conducted a final verification of the trips scheduled at the end of each call.

9. In some cases, reservationists repeated back and verified trip requests information very rapidly and it was not clear in some cases that callers were focusing on all of the information being repeated.

10. The current practice of offering only one possible pick-up time in response to a trip request and not searching for alternate times within the one-hour before or after the requested time may not be consistent with the concept of "negotiate[ing]" times as detailed in 49 CFR §37.131(b)(2) of the USDOT's ADA regulations.

11. Some reservationists made limited use of the "pick-up time with earliest departure" in responding to customer needs in scheduling trips. While some reservationists often used this option, others indicated that they only used this scheduling option if callers specifically said they could not leave earlier than a certain time. Also, procedures for using the "pick-up with earliest departure" scheduling option did not appear to be clearly addressed in the Reservations Manual. Use of this option for all trips to which it applies could result in trip offers that are more responsive to customer needs and avoid seeking additional trip times when the customer can't use the offer.

12. Some reservationists appeared to continue to place trip requests from riders who could not be served by the supplemental taxi program (i.e., required an accessible vehicle) and which were received after 2:30 PM the day before service, in "standby" mode until 4:00 PM. Others appeared to offer "standby" status up to 5:00 PM. Procedures for handling these situations do not appear to be addressed in the "Reservations Manual."

13. Procedures for when to search for options to the one time given by the system did not appear to be clearly addressed in the Reservations Manual.

14. There appear to be enough reservationists on-duty throughout the week to handle calls from riders in a timely way.

1. AAR should revise its policy to accept all requests for pick-ups after midnight, and before 4:00 AM, on the following day.

2. NYCT should continue to reprogram the scheduling software to discontinue the practice of offering pick-up times trips between 61 and 65 minutes after the pick-up time requested by the customer.

3. NYCT should revise its policy limiting the time that AAR customers can book return trips to make service for such trips comparable to fixed route trips. NYCT might consider offering trips with an earliest pick-up time some fixed period, such as 30 minutes, after the drop off time estimated by the scheduling software.

4. It is recommended that NYCT review the way that trip refusals are recorded to be sure that all refusals are appropriately reported.

5. Given that the average call servicing time is about 2 minutes 30 seconds and that the established goal is 4 minutes, and that hold times are very low, NYCT should consider having reservationists repeat back and verify key trip information more slowly and deliberately in order to avoid potential trip reservation errors.

6. NYCT should consider offering customers more than one option in negotiating pick-up times with customers during the reservations process.

7. It is recommended that NYCT develop reservations and trip booking procedures that better identify trip needs of callers so that trip offers are appropriate to these needs. Procedures should also address appropriate negotiation of trip times that respond to stated rider needs. For example, if a caller initially just states a return pick-up time without constraints but subsequently indicates that the time requested was an earliest departure time, it might be appropriate for the reservationist to search for new options using the "pick-up with earliest departure" function. The "Reservations Manual" should be modified to address these procedures.

8. NYCT should clarify in its "Reservations Manual" how trip requests from riders who could not be served by the supplemental taxi program (i.e., required an accessible vehicle) and which were received after 2:30 PM the day before service should be handled and should then communicate a consistent policy to all reservationists.

9. NYCT should review and revise the Reservations Manual to clarify procedures on when to search for options to the one time given by the system, did not appear to be clearly addressed in the "Reservations Manual."

In this portion of the compliance review, the team examined how trip requests from riders were handled. Particular attention was given to whether Access-A-Ride (AAR) uses any form of trip caps or waiting lists and whether there was a pattern or practice of denying a significant number of trip requests. The following information was gathered and analyzed:

• Input from customers and advocates was obtained through telephone interviews, a review of recent court documents, and a review of comments and complaints on file at NYCT;

• Reservations and scheduling policies, practices, and performance standards were reviewed;

• Service reports prepared by NYCT showing the number of trips requested, scheduled and denied were examined; and First hand observations of the handling of trips were made, and staff was interviewed about the ability to accommodate trip requests.

Consumer input, as detailed in Section III, was gathered through interviews with representatives of the disabled community, review of a report provided by Disabled in Action of Metropolitan New York and AAR complaint records for the first 7 months of 2003.

Reservations issues identified by representatives of the disabled community and the Disabled in Action Report and a review of complaints on file with AAR included:

• Schedules were described as illogical. Routing was described as indirect with vans transporting customers in a zigzag pattern. 3% of customer complaints were for scheduling issues.

• Customers said that AAR changes subscription trips without notifying the customer. Pick-up times are changed from that agreed upon at the time of the trip request resulting in drivers and customers having two different times for the same trip. It appears that subscription routes are well designed but when changed result in poor performance which is corrected after a period of time. 1% of AAR complaints related to subscription problems.

• Customers said that some rides scheduled with AAR appear to be lost in the system. On the day of the trip AAR would have no record of the requested trip.

• Trips have been as long as 3 hours exceeding NYCT's standard of 155 minutes (2 hours and 35 minutes) for trip length. Travel times are long in the evenings causing some customers to avoid use of the service. 8.5% of AAR complaints were for long trips.

• One customer felt that schedules are not adjusted to traffic conditions in autumn with people returning to school and work from summer vacations, resulting in longer trips than in summer.

The Disabled in Action Report also raised issue with the use of standby lists for reservations; and limits on schedule for return trips (customers must allow 1.5 hour from a pick-up for a going trip to a pick-up for a return trip within borough, and 2.5 hours between boroughs).

NYCT standards that guide the scheduling process include a goal of 0% denials; an "on-time" pick-up window of between five minutes before the pick-up time scheduled with the customer and 25 minutes after the scheduled pick up time (-5/+25). The NYCT goal is that vehicles arrive on-time 95% of the time. Also NYCT's goal is to schedule 100% of the trips using the following travel time standards:

It is currently NYCT 's policy to limit scheduling trips as standing order or subscription trips only for customers who make the same trip (between the same origin and destination at the same time and day(s)) at least three times a week. NYCT plans to revise this policy to include trips that are made once a week or more.

The day-to-day responsibility for scheduling AAR trips is handled by a private contractor, First Transit. The First Transit schedulers are co-located with NYCT staff at 2 Broadway in Manhattan. NYCT employs three scheduling managers and a scheduling director to manage the Scheduling and Subscription Units of First Transit (see Figure VIII.1). As will be described in the following section, the NYCT scheduling director and managers also work closely with the eight contract carriers to assist with any scheduling-related issues. The NYCT scheduling director reports to the command center officer, who also oversees the reservations and Control Unit (dispatch) functions for NYCT.

The First Transit Scheduling Unit is responsible for overseeing the development of schedules using ADEPT software. The Scheduling Unit includes:

• 1 scheduling manager,
• 1 scheduling supervisor, and
• 10 schedulers.

All 10 schedulers work Monday through Thursday, 6 work Friday, 5 work Saturday, and 4 work Sunday. Both the scheduling supervisor and scheduling manager work Tuesday through Thursday and one or the other is on duty Friday through Monday. As shown in Table VIII.2, coverage is generally provided from 8:30 or 9:00 AM until 6:30 or 7:00 PM, depending on the day.

The Subscription Unit is responsible for working with individual and agency-related subscription (standing order) trips, which are provided on an ongoing basis. The Subscription Unit includes:

• 1 subscription supervisor and
• 2 subscription assistants.

Table VIII.2 shows the scheduler positions and shift coverage of the First Transit Scheduling and Subscription Units.

First Transit Scheduling Unit

As described in the Reservations Section of this report, First Transit reservations staff members, using the StrataGen ADEPT paratransit scheduling and dispatching software, schedule most trip requests. First Transit schedulers, on an on-going basis, review trip reservations and the Scheduling Unit schedules trips that were put on the standby list.

Each day, half the schedulers work on finalizing schedules for the following day, while the other half work on scheduling trips that will be made from two to four days into the future. Each day schedulers are assigned to work on schedules for a particular carrier or carriers, but will be assigned different carriers on different days.

During a typical day, a variety of reports are used to manage the scheduling process. For example, several times a day a "Point Coding List Report" is run to indicate addresses that were not in the ADEPT database that were manually added when the reservation was made. One of the schedulers is assigned to check the geo-coding to be sure it was done correctly. (All schedulers are trained in this capacity.) An example of a Point Coding List Report from September 7, 2003 is shown in Attachment I. (Note: Names and client ID numbers have been removed from sample reports.)

Another report is the "Multi-Boro Report," which highlights routes that appear to zigzag between boroughs. This report is run several times a day. If a person-trip falls into this category, it will be removed from that route and reassigned to a more direct route, if possible. About 150 trip segments are reviewed daily. A sample page from a Multi-Boro Report from September 12, 2003 is included in Attachment J.

Throughout the day, schedulers work on trips that have been left on the stand-by list (the trips that will be provided, but were not scheduled during the call to the reservations staff). When standby trips have to be forced into the schedule manually, the schedulers allow for 60 minutes for intra-borough trips and 90 minutes for inter-borough trips.

Once a trip is scheduled, a scheduling assistant calls the customer to confirm the trip. On September 16, 2003, during the site visit, there were 58 trips that had to be manually scheduled with callbacks for the next day (more than 10,000 trips were scheduled for that day). When a customer is contacted either via telephone or voice mail to confirm the trip, it is recorded in the trip history file. If a customer is not reached and there is no voice mail, additional attempts will be made throughout the day. Passengers are instructed to call if they have not heard back to confirm the trip. Schedulers indicated that they are unsuccessful in contacting only a few passengers. If a passenger cannot be reached, the trip is still scheduled.

During the month of May 2003, 248,593 trips were scheduled. The number scheduled per day ranged from 2,499 on May 26, to 10,535 on a weekday. For a sample day, May 15, 10,065 trips were scheduled and an additional 28 (0.3%) customers were authorized to use taxi vouchers.

Each day at noon the trips are "re-optimized," which means the computer re-batches trips to optimize route assignments in order to create more efficient and effective schedules. Each day's schedule is re-optimized a total of four times (one to four days in advance). In most cases both random and standing order/subscription trips are re-optimized (a few program-related subscription trips have been permanently assigned to a route or black car voucher service). A sample "Re-optimization Report" is included in Attachment K. Table VIII.3 shows an excerpt from that report.

First Transit Subscription Unit

The Subscription Unit handles all requests for subscription trips. A letter describing the subscription process is provided to passengers who would like to set up a subscription trip. A subscription request form is completed that details the travel needs of the customer. At the time of the review, subscription trips accounted for about 16% of all trips and were only allowed for passengers traveling to/from the same origin/destination pair at the same times at least three times a week.

Subscription trips are not usually anchored to a particular route unless there have been complaints about the way a trip has been performed or the customers have special needs that are better served by anchoring the subscription trips to the same route. Anchored trips may shift to different routes during optimization. The promised times are retained. There are only a few group trips where several people travel together from the same origin, or to the same destination. This fall, NYCT plans to relax the requirements to allow subscription trips to be established for repetitive trips that occur at least once a week (e.g., every Monday at 1:00 PM with a return trip at 5:00 PM).

In December 2002, NYCT set up a system for providing black car vouchers for some passengers who have standing order reservations for appointments at the Veterans Administration hospitals located in the Bronx and Brooklyn. This practice has been done to provide service for passengers returning home from dialysis. Uncertainty as to when these passengers will be able to begin their trip can cause scheduling problems requiring dispatch of a vehicle for a passenger who was not ready to be transported at their scheduled return time.

NYCT Scheduling Management Unit

NYCT provides management, quality control, and oversight to the First Transit reservations, scheduling and command center functions. The NYCT scheduling director reports to the command center officer. He, along with three scheduling managers, are responsible for overseeing the performance of First Transit and for managing issues related to scheduling at each of the eight carriers. The director's specific duties include overall management of the Scheduling Unit, as well as monitoring the master route history. He also oversees schedule development and execution for TFM and MV Transit. He has been working closely with MV, a relatively new carrier, to improve that carrier's performance.

The three managers report to the director of scheduling and have specific duties. One is assigned to oversee the development of holiday schedules, billing resolution, daily scorecard reporting of carriers, managing the Subscription Unit, and monitoring schedules for RJR and PTC. The other two managers are responsible for managing the First Transit Scheduling Unit. One also monitors schedules for Atlantic Paratransit and Maggies; the other monitors schedules for American Transit and Star Cruiser.

The NYCT Scheduling Management Unit troubleshoots when problems arise with either the scheduling function or the carriers. For example, if a carrier is having difficulties with particular schedules, the NYCT manager will work with the carrier and First Transit to identify the source of the problem and resolve it. They are also responsible for identifying various ways to improve service.

Two members of the review team met with the command center officer and the NYCT scheduling director to discuss the functioning of the scheduling personnel employed by NYCT and First Transit. One member of the review team observed the scheduling staff while they worked on September 16, 2003 and met with the First Transit scheduling manager. It appeared that the First Transit staff worked together as a team in their assigned roles. Reports were routed, schedules reviewed and modified, and calls were being made to update customers with respect to their trip status. That day a record for the number of trips scheduled was set - 10,814.

In February 2002, NYCT began phasing in an updated computer-assisted paratransit scheduling and dispatching system, moving from the PASS system to StrataGen's ADEPT software program. The first implementation phase (from 2/1/02 to 5/1/02) centered on transitioning from PASS to ADEPT. The implementation of the second phase, on 2/24/03 resulted in additional computer program modifications designed to make the AAR system route trips more efficiently and effectively. The time line for these changes is described below:

• April 2002 - Changed outside-in to inside-out engine logic,
• July 2002 - First trip on each route at start of the first trip's window,
• September 2002 - Zones increased from 5 to 9,
• November 2002 - Re-optimization in place,
• February 2003 - Implemented affinity settings to routes, and
• November 2003 - Planned implementation of multi-batch logic.

Several parameters have been adjusted to attempt to improve how trips are assigned by the automated system. One approach has been the use of affinity settings described above), which anchor carrier routes to a particular home zone or zones. A matrix imbedded in ADEPT assigns routes to the carriers' home zones first, but may assign trips to a route outside the home zone if needed (and as permitted by the affinity settings). The home zone preference is accomplished by assigning a weight to routes depending upon the trip origin. The routes operated in the zone of the trip origin receive a favorable weight whereas those operated in a distant zone receive an unfavorable weight. According to NYCT and its operators use of the affinity factor has resulted in much more efficient assignment of trips to routes. The nine zones used by the affinity factor are shown on a map included in Attachment N. The zones are:

• Zone 1 - The Bronx,
• Zone 2 - Manhattan North (above 67th Street),
• Zone 3 - Manhattan South,
• Zone 4 - Brooklyn West (west of Flatbush Avenue),
• Zone 5 - Brooklyn East,
• Zone 6 - Queens North (North of Grand Central Parkway),
• Zone 7 - Queens South,
• Zone 8 - Rockaway, and
• Zone 9 - Staten Island.

In order to more accurately calculate route times, on-street ride checks are performed to determine actual travel times. Adjustments are made to system speeds used by the scheduling software on an on-going basis based upon the ride checks. In addition, NYCT has established travel times for approximately 10 problematic areas, typically river crossings. Different times are used for direction of travel and for time of day using six time periods. These speed zones, which have been input into the scheduling parameters, were developed to address widely divergent travel times by time of day and direction of travel in some areas. Subscription trips have been run through the ride check program as well and some adjustments were made to reflect the updated information.

Another approach to improving scheduling has been the development of "Peter Pan" trips. Trips are selected for analysis based on customer complaints (e.g., the ride is too long, the routing is too circuitous, etc.). The NYCT scheduling staff will review these trips and determine how these trips may best be scheduled to avoid future problems. The analysis can include ghost runs or on-the-road travel time tests and so on.

NYCT also sets parameters in the schedule that limit the trip duration based upon distance traveled as described in NYCT policies. Another feature built into the software is use of a -5/+20 minute window for scheduling trips rather than the -5/+25 NYCT standard. The additional five-minute cushion built into the schedule should assist in achieving better on-time performance.

NYCT managers indicated that the number of routes scheduled each day is based upon historical experience for the particular day of the week. New routes are added when the NYCT Scheduling Management Department identifies the need for more service in a particular area. Some service increases have been accomplished with existing vehicles with added hours; others require additional vehicles and personnel. In March 2003, as part of the effort to eliminate trip denials, NYCT reports that it added a large number of new routes. The NYCT scheduling staff works with carriers to make these changes as needed.

1. The scheduling process appears to be well structured. NYCT and First Transit staff appeared highly knowledgeable of the scheduling process and software. Staff appears to be pro-active in reviewing schedules, addressing scheduling problems, and seeking out improvements to the scheduling process and resultant schedules.

2. Computer software modifications made over the past year seem to have significantly reduced earlier problems encountered during the transition from the PASS to the ADEPT software system.

3. Use of the beginning of the pick-up window (5 minutes before the scheduled time) in the scheduling parameters rather than the scheduled time appears to have resulted in offers of schedule times to customers that are more than 60 minutes from the requested time. NYCT staff indicated that they are modifying the software parameters to prevent trip offers of more than 60 minutes from the requested time.

4. Subscription trips are handled by a separate unit and appear to be adequately staffed for the current volume. Changes to the subscription program (allowing subscriptions for one trip per week instead of limiting subscriptions to three per week) could increase the burden on this department as they process and monitor subscription requests.

5. Schedulers' use of The Point Coding List Report, Multi-Boro Report, standby list, and other reports appears to be effective in refining schedules.

6. By working with the operators, NYCT appears to effectively address scheduling issues that affect service performance.

7. Active efforts to improve schedules through the application of the affinity factor, on-street travel time checks, and the use of "Peter Pan" trips appear to have had a positive effect on the quality of the schedules.

8. Scheduler's do not appear to review actual service performance to identify significantly late or long trips but instead rely upon customer complaints and the Multi-Boro Report and operator feedback to identify such trips and "ghost runs." Review of actual performance of significantly late, long and missed trips should be done to assist schedulers in identifying the cause of such trips, such as travel time assumptions, geo-coding errors, etc., and further reduce the number of significantly long, late and missed trips.

9. Assigning trips with uncertain pick-up times (such as return trips from dialysis) to the black car voucher program appears to provide better service for the passengers and removes potential problem trips from scheduled routes.

10. There appear to be adequate First Transit schedulers on duty throughout the week to handle trip-scheduling responsibilities, including customer callbacks, as needed.

1. NYCT should continue to reprogram the scheduling software to discontinue the practice of offering pick-up times trips between 61 and 65 minutes after the pick-up time requested by the customer.

2. NYCT should consider regular review of actual service performance to identify significantly long travel times, significantly late trips, and missed trips. Scheduling procedures could then be modified based with a goal of eliminating significantly long and late trips.

3. NYCT should consider preparing an additional report based on scheduled ride time to identify excessively long trips that might not be captured in the "Multi-Boro Report."

The DOT ADA regulations for ADA Complementary Paratransit service indicate that capacity constraints can result from poor service quality. Specifically, they note that missed trips or the provision of untimely trips or significantly long rides can constitute capacity constraints. On-time performance and on-board ride times were therefore examined as part of the review. In addition, this section of the report addresses delivery of transportation service and dispatch operations to identify procedures and practices that could affect service delivery. These aspects of service provision were assessed as follows:

• Consumer input was obtained on each issue through telephone interviews and through a review of complaints filed with FTA and with NYCT;

• NYCT's service policies, procedures and standards in these areas were reviewed;

• The review team visited each of the eight principal operating contractors and interviewed managers;

• The dispatch function was observed and dispatchers were interviewed;

• Drivers were interviewed;

• Information on operations resources was collected;

• NYCT's on-time performance and travel time reports were reviewed;

• Actual pick-up and drop-off times reported on a randomly selected day were used to tabulate on-time performance to compare to reported performance; and

• The review team analyzed trip length, including a comparison of travel times between ADA Complementary Paratransit trips and comparable fixed route trips.

Consumer input, as detailed in Section III, was gathered through interviews with representatives of the disabled community, review of a report provided by Disabled in Action of Metropolitan New York and AAR complaint records for the first 7 months of 2003.

Transportation issues identified by representatives of the disabled community and the Disabled in Action Report and a review of complaints on file with AAR included:

• Vehicles sometimes arrive early and leave before the pick-up time. One customer cited a vehicle arriving at 10:30 and departing at 10:37 for an 11:30 pick-up. 1.3% of complaints on file with AAR were for early drivers.

• Visually impaired customers are often stranded because the driver doesn't approach the customer and the customer is unaware that the vehicle has arrived.

• People have been dropped off by as much as 1.5 hours early resulting in customers discontinuing use of the service.

• Consumer representatives indicated that there is no effort to contact customer no-shows with the exception of one operator, RJR, which does it sometimes.

• Customers are categorized as no-shows when the vehicle arrives late and the customer is no longer waiting for the ride. A consumer provided an audiotape, which purports to demonstrate dispatchers directing drivers to classify missed trips as customer no-shows. Verification of what is said on the tape has not been possible because of the poor quality of the audio. 4.4% of AAR complaints were for disputed customer no shows.

• Missed trips from vehicles going to the wrong place are a problem. AAR has been addressing this problem by establishing standard pick-up locations. 26.3% of customer complaints on file with AAR were for missed trips, 4% were for incorrect addresses and 3.2% were for stranded customers.

• Pick-ups are late in the evening. When a second vehicle is dispatched for a late pick-up it often arrives 1.5 hours late. 15.8% of customer complaints were for late pick-ups.

• AAR provides vouchers for Yellow Cab service for ambulatory customers who are stranded (missed trip). The vouchers are usually authorized after the trip is significantly late (1 hour). The number of vouchers available is limited by budget. Since Yellow cabs are not accessible to wheelchairs, the vouchers are not available to wheelchair users. 2.3% of AAR complaints related to taxi vouchers.

• Trips have been as long as 3 hours, exceeding NYCT's standard of 155 minutes (2 hours and 35 minutes) for trip length. Travel times are long in the evenings, causing some customers to avoid use of the service. 8.5% of customer complaints were for long ride times.

• One consumer representative indicated that NYCT oversight and coordination with operators appeared limited.

The Disabled in Action Report also raised the following additional transportation related issues:

• Pressure to accept early pick-ups.

• Late pick-ups - particularly during the afternoon; drivers appear to be reporting passenger pick-ups before they occur.

• Unreliable service.

• Stranded riders.

• Inadequate back-up service - long wait for approval, pre-payment of taxi fare, delayed reimbursement, limited to ambulatory customers.

• Inexperienced drivers - low driver pay rates and limited training, poor knowledge of routes. 6.8% of customer complaints were for rude drivers, 3.6% were for failure of drivers to assist customers, and 1.9% of AAR complaints were for reckless drivers.

• Inaccessible sedans.

• Poor van maintenance - high entry step, lift failures.

On-Time Performance Policies and Standards

NYCT has a performance goal of 95% for on-time AAR pickups. This goal is included in the operator's contracts as a performance standard (article 104 C.2.). "On time" is defined as any time within the window of -5/+25 minutes from the pickup time agreed upon with the customer. For example, for a negotiated pickup time of 8:00 AM, the vehicle is considered on time if it arrives between 7:55 and 8:25 AM.

The Access-A-Ride Paratransit User's Guide advises customers to be ready to travel from five minutes before until 25 minutes after the scheduled pick-up time. The Guide also indicates that the driver will leave after five minutes if the customer is not at the pick-up location. The User's Guide advises customers to cancel trips before 5:00 PM on the day before service and advises customers that a cancellation after that time will be considered a late cancellation.

The User's Guide further instructs customers to call AAR if the vehicle does not arrive within the 30-minute window. The AAR transit control officer will obtain the estimated time of arrival from the driver to which the trip is assigned, or will dispatch another vehicle and advise the customer of the newly estimated arrival time. If a vehicle did not arrive within the 30-minute pick-up window and is not readily available at the time of the customer's call, the customer can request authorization to use a taxicab or car service with the expense to be reimbursed by NYCT. Authorization is confirmed by AAR transit control officer issuance of an authorization number.

AAR's contracts with the operators provide definitions for the following performance measures:

• Contractor Early - The vehicle operator arrives more than five minutes before the scheduled pick-up.

• Late Pick-Up - AAR pickup that occurs past the scheduled pick-up window.

• Late Trip Cancellation - Customer cancellation after 5:00 PM the night before a scheduled trip.

• No-Fault No-Show - The contractor fails to provide the scheduled trip to a customer for documented reasons beyond the control of the contractor. NYCT indicates that vehicle arrivals after the pick-up window, 25 minutes past the scheduled pick-up time, and before 45 minutes after the scheduled pick-up time are categorized as no-fault no shows when the driver waits five minutes for the customer, and the operator is unable to locate the customer.

• Customer No-Show - The customer has not presented him/herself, declines the trip or is unable for any reason to take the trip when the vehicle operator has arrived and waited for the customer in accordance with contract standards and procedures. Section 3.22 of Attachment 1 of the contracts further indicates that the operator shall complete pick-ups within a 30-minute window running from 5 minutes before to 25 minutes after the pick-up time scheduled with the customer.

• Missed Trip (Contractor No-Show) - Any trip not performed other than a Late Trip Cancellation, Customer No-Show, or No-Fault No-Show. NYCT further describes this to include a no-show reported by the operator more than 5 minutes before or 45 minutes after the scheduled pick-up time, or the vehicle never arrives at the correct location for the scheduled trip, or the vehicle arrives at the correct location within 5 minutes before and 45 minutes after the scheduled pickup time, but does not wait 5 minutes.

Regarding trip duration, NYCT's goal is to schedule 100% of the trips using the following travel time standards:

Each of the eight operators provides service throughout the service area, New York City. Through the scheduling process, passenger trips are assigned to routes based upon nine service zones (see Attachment N). Similarly, routes are assigned to carriers to concentrate carrier operations within geographic areas to promote efficient delivery of service.

The operators entered into five-year contracts during August 2001. Several of the operators were providing AAR service through earlier contracts. NYCT is in the process of purchasing fuel in bulk and making it available to the operators at lower cost than direct purchases. Operators are reimbursed for their direct cost of fuel by NYCT.

The review team interviewed the Project Managers for each of the eight operators. The results of the interviews are reported as follows:

Access-A-Ride Command Center

Customer calls for same day travel information are directed to the AAR Command Center at 2 Broadway in Manhattan. The Command Center is staffed and managed by First Transit under a contract to NYCT. NYCT staff oversees the Command Center Operation. Command Center staff is responsible for:

• Accepting service day (Late) trip cancellations from customers;
• Addressing customer inquiries on status of rides;
• Reassignment of passenger trips from one carrier to another;
• Carrier authorization to extend route times or open new routes;
• Assignment of passenger trips to supplemental carriers;
• Authorization of vouchers for customer use of taxis and black cars; and
• Other same day service issues as they arise.

Reviewers observed three call takers and general operations in the Control Unit during the morning and afternoon on September 11. Reviewers observed transit control officers handle 55 calls. Calls observed are summarized in Table IX.1

Late Cancellations - The control officers were observed coding the trip as a late cancel, requesting the customer's reason for canceling the trip and entering the reason into the trip record. In one instance the customer called while the driver was at the pick-up location. The control officer contacted the operator's dispatcher and the dispatcher indicated that he would code the trip as a no-fault no-show. According to one of the control officers the procedure is to telephone the operators dispatcher if the no-show has been received close to the scheduled pick-up time so that the driver can be redeployed. Other than this one instance none of the cancellations observed at the Control Center appeared to be within an hour and a half of the scheduled pick-up time. One customer, who had both a same day and future trip cancellation, was transferred to reservations to make his future day cancellation, after completing his same day cancellation. [back to Table IX.1 - Command Center Observations]

Ride Checks - When customers called to check on the status of their ride, the control officer would review the status of the ride in the scheduling dispatch system. If the system indicated that the ride was on time and the call time was before the pick-up window had elapsed generally the control officer would indicate that the ride was on time and, when appropriate, remind the customer of the pick-up window. 19 of the 25 ride-check observations fell into this category. If the route appeared to be running late, the control officer would place the customer on hold and contact the operator's dispatcher to obtain the current status of the route and estimated time of arrival of the vehicle. Six of the twenty-five ride check observations were for routes that were running late. The six calls to dispatchers to get current estimated arrival times took from one to four minutes to complete. [back to Table IX.1 - Command Center Observations]

Route Extensions and New Route Additions - The practice observed for opening route extensions or new routes was observed as follows. If a dispatcher for the contract carrier observes one or more routes running late and has been unable to use lag time on the route or reassignment of trips to other routes to restore on-time performance within a short period (one hour +/-), the dispatcher may request a route extension or new route to rescue customers and the service. (A route extension is an extension of the time an active route is authorized by NYCT to continue in service. A new route is initiation of service on a new route beyond those previously authorized for the service day.) The dispatcher makes the request to a transit control officer in the call center. The transit control officer completes a form for the route authorization and reassigns the passenger trips to be moved to an open or dummy route. The request form is then referred to a supervisor for review and authorization. The supervisor reviews the status of the late routes on the scheduling and dispatch software to confirm that they are running late and avoid authorization of service that is not needed. If the route is not running late in the system the supervisor will not authorize moving passenger trips and associated new routes or route extensions. The review team observed a Command Center supervisor while she reviewed and authorized one of the four requests from operators to increase service. Initially she approved a new route for only two of the three trips requested because the route for the third trip did not appear to be running late. The reason for reassigning the third trip was to permit a driver to end his route early thereby resulting in no net additional vehicle hours to the operator. Because the reason for the request was not clear the supervisor initially declined authorization but changed her decision upon clearer presentation of the reason. Based on observations of operator dispatchers they will often request authorization to reassign trips before the route appears to be running late in the system. They will do so based upon tight travel time allowances for the route, traffic conditions, and driver's experience and ability to recover from delays. Because of the time required to authorize increased service and the urgency of dispatching vehicles to avoid further service delay, several operators indicated that they will dispatch drivers before authorized, even at risk that they will not receive authorization and will not be reimbursed for the additional service.

NYCT indicated that it plans to improve coordination between the Control Center and Operator dispatchers by assignment of a dispatcher from each operator to the Control Center. The proposed change in dispatch operations was scheduled for late September and early October. [back to Table IX.1 - Command Center Observations]

Missed Trip - One customer called to report a missed trip. She made the trip on her own. The trip record contained an incorrect address and no phone number for the pick-up location. [back to Table IX.1 - Command Center Observations]

Wrong Department - Three calls were identified as being for the wrong department. One, for trip reservations was transferred. Two were from the same caller who had a same-day trip issue. The cause of the problem appeared to be incorrect or incomplete information provided by the Eligibility Department. The caller was referred back and forth between the two departments until the control officer referred the issue to a supervisor. The customer, who was applying for eligibility, lives out side of the AAR service area. He was scheduled by the Eligibility Department for an in-person assessment and, apparently through a form letter, directed to reserve a trip with a black car service, which only operates within the AAR service area. NYCT personnel rescheduled his appointment and instructed him on how to obtain a ride on accessible service to a transfer point with the AAR service provider. An internal e-mail indicated that this is an on-going problem. It wasn't clear that the systemic problem was corrected by providing different form letters for service applicants who live outside of the service area or by some other means. [back to Table IX.1 - Command Center Observations]

Wrong Customer Identification Number - In one instance a customer called and indicated a driver was there to pick-up a customer with a different name who was going to the same destination at the same time as she was. No trip record could be found for the caller and the driver left the location. Apparently the trip had been booked under the wrong ID number and as a result the customer did not receive her trip and the trip was recorded as a no-show for the customer whose ID number was used. [back to Table IX.1 - Command Center Observations]

Schedule Change Requests - Five calls were observed in which customers were seeking same day changes in their scheduled trips. In one instance, a customer who had a 2:41 PM scheduled pick-up requested a pick-up as soon as possible at 10:00 AM because she was not feeling well. In an effort to accommodate this same day request for a change in her trip the control officer telephoned several carriers, including back-up carriers before one would accept the trip. The carrier who accepted the trip scheduled the pick-up for 10:30 AM.

Another caller indicated that they would be unable to meet their pick-up time because the doctor was delayed. The control officer indicated that the procedure was to not accept request for later pick-up times more than one-hour before the scheduled pick-up time. Because the request was made more than one hour in advance, the caller was requested to call back later. The trip was not changed in the trip record. As a result the operator was not informed that the trip was no longer needed and therefore was unable to use that driver to complete another trip. Additionally, the customer was required to make at least one more call to reschedule their trip. [back to Table IX.1 - Command Center Observations]

Operators

Atlantic Paratrans - Between nine and ten dispatchers were on duty on the afternoon of Monday, September 15. Each dispatcher was managing about 20-22 runs. Some time was spent observing the procedures used by each to manage runs. Most dispatchers were observed polling drivers frequently to check on the status of pick-ups and drop-offs. Actual times were then entered into the automated system to allow estimated pick-up and drop-off times to be updated. These dispatchers were also identifying trips that were potentially late and were moving trips to other available runs. The Dispatch Supervisor on duty was also observed facilitating reassignment of passenger trips between dispatchers. He would periodically ask if anyone had trips that they couldn't handle and would then find another dispatcher with routes with slack time.

Two dispatchers, however, appeared to be somewhat behind in managing assigned routes. One dispatcher was observed polling for and entering actual times for 12:00-1:00 PM pick-ups at 2:50 PM. The other dispatcher had several runs behind schedule (some by up to 65 minutes), but appeared to be slow to reassign trips or seek the assistance of the Dispatch Supervisor. On one run that was 65 minutes behind schedule, she stated that, "the driver should have let me know sooner."

Dispatchers and drivers appeared to handle customer "no-shows" properly with drivers radioing dispatch while at the pick-up location, dispatchers attempting to contact riders, landmark information being requested and entered into the no-show screen, and drivers waiting at the pick up location for at least five minutes within the pick-up window before classifying customers as no-shows.

Several early arrivals and early "call-outs" were noted (more than were observed at other providers). As was the case at other providers, dispatchers would typically just state that, "this is AAR calling; your vehicle is outside waiting," rather than indicating that the driver was early and if the rider could go early it would help. Dispatchers were not observed to be pressuring riders to leave early and drivers waited if riders indicated they were not ready to go, but it was easy to see how some riders might interpret the calls as pressure to leave before their scheduled pick-up times.

Dispatchers noted that between 3:00 and 5:00 PM the automated system slowed down. When dispatchers would attempt to add actual pick-up times, the system would take about 20 seconds to process the information and update subsequent times. During this time, dispatchers were not able to move to other screens or scan other runs. Several dispatchers indicated that the system does slow regularly and noted that when things are busy and every second counts, this delay in updating runs can have an impact on efficiency of dispatching and, in turn, on-time performance. One dispatcher noted that when slowdowns occur, he tends to put-off updating runs so that his ability to scan runs and perform other tasks is not affected.

In general, the dispatch operation appeared to be well organized and managed given its size. There was some variation in apparent dispatcher skill (which is not unusual for a large operation).

Based on a work schedule for the week of September 14-20, 2003, Atlantic Paratrans employed 19 dispatchers, three Assistant Operations Managers and eight Dispatch Supervisors who worked directly on the dispatch and driver pull-out/pull-in functions. Dispatcher schedules are shown in Table IX.3 below.

Two Assistant Operations Managers work morning shifts from between 4:00 AM to between 1:00-2:00 PM. The third Assistant Operations Manager works from 11:00 AM to 8:00 PM. Two managers work Sunday shifts, but none are scheduled on Saturday.

Five of the Dispatch Supervisors work morning and early afternoon shifts beginning between 4:00 AM and 6:00 AM and running through 3:00 PM. Three morning Dispatch Supervisors work Saturday mornings and one works Sunday mornings. One Dispatch Supervisor is then scheduled from 2:00 PM to the close of the operating day all seven days a week.

Table IX.3 - Atlantic Paratrans Dispatch Staffing, Based on September 14-20, 2003 Work Schedule

D
[Link opens in a new browser window]

Weekday morning dispatchers begin reporting at 4:30 AM. By 6:00 AM, eight dispatchers are on duty and by 7:15 AM, 10 dispatchers are on duty. Six afternoon/evening dispatchers report between 12:00 noon and 2:00 PM and work until between 9:00 and 11:00 PM. During the afternoon peak period (3:00-5:00 PM), six to eight dispatchers are on duty. On Saturdays and Sundays, there are five dispatchers during the morning, six to seven during the mid-day, five through the afternoon peak and two to four dispatchers during the later evening hours.

Given that there are about 185 peak-hour runs on weekday mornings, each dispatcher typically manages about 19-20 runs during the peak.

Maggie's Paratransit (Maggie's) - The Maggie's operations center has eight phone lines that come in to a call distribution system. Callers are asked to indicate which extension (or functional area) they are trying to reach and calls are then directed to those lines. Calls to Dispatch are directed to the next available dispatcher. Dispatcher workstations are arranged in a single room, which allows calls to easily be directed to the dispatcher handling the route (or rider) in question. Given the number of Dispatch workstations, the number of incoming lines appeared adequate to handle expected call volumes.

The morning Dispatch Supervisor indicated that he (or another lead dispatcher) reviews all assigned runs first thing in the morning. Adjustments are made as needed and final manifests are then printed out for drivers. He estimated that some adjustment is made to "about 3-4 of every 10 runs." He also noted that he and other dispatchers meet with someone from the scheduling group at least once a month. He noted that they also bring issues to the scheduling group's attention as needed. He indicated that a good working relationship exists between the Maggie's dispatchers and the schedulers at the control center. The Dispatch Supervisor said that problems are addressed and corrected as they bring them to the schedulers' attention.

The Dispatch Supervisor indicated that the formal procedure for requesting and getting approval for "rescue" routes or extended route hours can make performing trips on-time a challenge. He noted that the information has to show the run late (past the 25-minute pick-up window) before a request for a "rescue" route can be made. Then, approval can take an additional 10 minutes. To make the process more "workable," the Maggie's Dispatch Supervisor indicated that they would make the changes internally to extend a run or add a rescue run when they know it is needed. Sometimes, the approval process will be done after the fact. He indicated that they are willing to take the chance that the extension/rescue would be approved. He noted that while the process can be cumbersome, there is a good working relationship with the control center and they rarely will have requests denied.

On the morning that the operation was observed, all six morning dispatchers were on duty. Each dispatcher was handling between 12-14 runs. It did not appear that drivers were reporting each pick-up and drop-off as they occurred. Instead, dispatchers were polling drivers periodically to get information on pick-ups and drop-offs. Several dispatchers indicated that they tried to poll each run at least once every 45-60 minutes. Additional attention would then be given to runs that were identified as being behind or drivers who needed assistance.

Maggie's dispatchers appeared to have very good control of all runs. The dispatch room appeared calm and "under control" throughout the morning. All dispatchers appeared to know the status of each run he or she was managing. Several passenger trips that were running slightly behind schedule were reassigned in a timely way. A few runs were observed to be running 10-20 minutes behind the scheduled time, but all appeared to be running at least within the pick-up window.

Reviewers observed that the StrataGen system shows trips to be running "red" (i.e., a problem) even if the estimated drop-off time is before the appointment time. Dispatchers noted that they system builds in some time (believed to be about 5 minutes) to allow riders to exit vehicles and get to their final destination. The drop-off will be flagged in the system if the estimated time is less than five minutes before the appointment time.

The Maggie's dispatchers indicated that in Queens there are many roads that have similar names. The Forest Hills section of Queens was called-up on a map on one of the workstations and dispatchers showed that there were three street segments marked with "68." They said that one might be a "street," another a "road" and another an "avenue," but that the screen just showed the number. They said that this sometimes causes a problem when trying to give drivers directions to an address. This issue was pursued back at the control center the following day. It appears that the system has the full street segment names stored but that there was a display problem at Maggie's. There was speculation that perhaps the Maggie's dispatchers were not displaying the information in the correct way. The control center manager indicated that he would contact Maggie's and address the problem.

Dispatchers noted that Nextel coverage on Staten Island could also be a problem sometimes. They also noted that some drivers are not familiar with all parts of Staten Island and that if radio coverage becomes a problem they may not be able to reach the dispatch center for assistance.

Reviewers observed the handling of three no-shows. In all cases, dispatchers followed a standard routine of asking the drivers for the time they arrived at the pick-up location. They then checked that time against the promised pick-up time and the 25-minute pick-up window. Dispatchers attempted to reach the customers by phone in each case. In one case, the promised pick-up time was 8:22 AM and the driver radioed-in at 8:24 AM. The driver indicated that they had arrived at 8:02 and had waited for 22 minutes (to be seven minutes inside the pick-up window). The dispatcher called the customer's home number and the person who answered indicated that the rider was not going that day. The dispatcher then authorized the driver to no-show the customer and to proceed. In the second case, the promised time was 9:10 AM and the driver radioed-in at 9:20 AM. The driver reported that he had arrived at 9:12 AM and had waited for eight minutes within the window. The dispatcher attempted to call the customer but found that the number provided had been disconnected. The dispatcher then authorized the driver to no-show the customer and to proceed. In the final case, the driver arrived at 9:24 AM for a 9:05 AM promised pick-up. The driver went to the door and was informed that the rider had left "just one minute before." The dispatcher authorized the no-show since the driver had arrived within the pick-up window.

One instance of an early arrival for a pick-up was also observed. In this case, the driver arrived at 7:14 AM for a 7:28 promised pick-up time. The driver requested that the dispatcher call the customer to see if they were ready, which the dispatcher did. The rider indicated that they were not ready and would be out "at the scheduled time." The dispatcher informed the driver and the driver waited for the rider. While there was no overt pressure on the rider (and the driver waited without expressing any issues), it is possible to see how some riders might feel pressured by these calls. When the rider answered the phone, the dispatcher simply said, "This is Access-A-Ride calling; your vehicle is outside waiting." The dispatcher did not say the vehicle was early and ask if it was possible for the rider to board early.

As with other providers, there appeared to be some misinterpretation of the schedules. Two times appear on the manifests, the time promised to, or negotiated with, the customer, and the scheduled time, which is the time that the driver is actually expected to arrive at the pick-up location as estimated by the scheduling software. As dispatchers were discussing assigned runs for the day and talking about runs that appeared to be too tight, they mentioned that runs had promised times that were only a few minutes apart. They would then indicate that there was no way that the driver could get to both locations in the time allowed. When it was pointed out that the scheduled times appeared to allow more appropriate times between pick-ups, they would acknowledge this, but would indicate that they manage runs based on the promised times. A few of the dispatchers indicated that they do not like to "work late in the window." They said that if they allow drivers to make pick-ups "late in the window" it is too easy to start running behind if there are any problems. They appeared to try to perform trips based on the promised times and to reserve the 25 minute time period following the promised time as a cushion, or contingency, to address service delays. While it is important for dispatchers and drivers to be aware of the customers expectations, as reflected by the promised time, and arriving at pick-up locations early in the pick-up window rather than late makes it easier to stay on schedule, drivers and dispatchers should also be aware that the schedule times provide important guidance to drivers in setting the sequence and time of their stops in a manner that maximizes service efficiency.

Based on the work schedule that was effective for the week of August 18-24, 2003, Maggie's employs 10 full-time dispatchers and one part-time weekend dispatcher. Dispatcher work shifts are shown in Table IX.4 below. There are six weekday morning dispatchers that have staggered work shifts starting between 4:30 AM and 9:30 AM. Before 6:00 AM on weekdays, there are two morning dispatchers. A third dispatcher reports at 6:00 AM and a fourth reports at 6:30 AM. By 9:00 AM there are between four and six dispatchers scheduled each weekday. Three of the morning dispatchers work four-day weeks and have shifts that extend until 5:00 to 7:00 PM.

Three afternoon dispatchers, with staggered shifts that begin between 12:00 Noon and 1:30 PM, replace the three morning dispatchers. This enables six to eight dispatchers to be available weekdays at 1:00 PM and five to six dispatchers to be available up to 5:00 PM. The number of dispatchers is reduced by one each hour between 5:00 and 7:00 PM. Two to three dispatchers are available at 8:00 PM, one to two at 9:00 PM and one dispatcher is then available at 10:00 PM. No Dispatch coverage is provided after 10:00 PM on weekdays.

On weekends, there are only two dispatchers. One works a 6:00 AM to 4:30 PM shift and the second reports at Noon and works until 10:00 PM. As a result there is only one dispatcher on duty for most hours of the day on weekends with the exception of the mid-day (Noon to 4:30 PM) when the shift overlap allows both to be on duty.

Table IX.4 - Maggie's Dispatch Staffing Based on August 18-24, 2003 Work Schedule

D
[Link opens in a new browser window]

MV Transportation (MV) - The manager noted that there are nine total phone lines and that five lines are dedicated to the dispatch function. If all five lines are in use, callers get a busy signal. In the past, if dispatch lines were busy they were directed to voice mail. The manager indicated that the use of voice mail was a problem and that the voice mail was eliminated.

MV managers indicated that the night dispatchers are assigned the job of reviewing route schedules received from First Transit. This staff either moves trips between runs or highlights trips and runs that appear tight and will need to be watched. The manager noted that having the night dispatchers do this did not always work well, though, as they were sometimes unable to focus enough attention on the task. He indicated that MV was reconsidering this approach to reviewing and revising routes.

MV drivers did not appear to radio-in each pick-up and drop-off. Instead, dispatchers were observed periodically polling drivers for actual pick-up and drop-off times. Observations of MV dispatch on Friday afternoon, September 12, 2003, between 3:00 and 5:00 PM indicated that dispatchers were somewhat behind on polling drivers. They appeared to be focused primarily on getting pull-out mileage and times from drivers. Even though several runs on the dispatch screen appeared to be running behind (highlighted in red), little attention appeared to be given to determining the exact status of those runs, moving trips between runs, requesting rescue routes or route extensions, or otherwise managing runs. At 4:45 PM, during the afternoon peak period, dispatchers appeared to be 60-90 minutes behind on polling drivers for actual times.

On a second visit (Wednesday afternoon, September 17, 2003, between 3:00 and 5:30 PM), dispatchers were paying more attention to managing runs, but reviewers observed several service problems that were causing a number of routes to run late. One breakdown (Bus 922/Route 272) and one accident (Bus 83/Route 133) were reported by radio during the two and a half hours on-site. Dispatchers also noted that there had been a breakdown in the morning as well and that some morning runs scheduled to be back at 3:18 PM were running late and causing afternoon pull-outs to be delayed. Trips on at least three runs on one dispatcher's screen (of the 20 runs being managed) were running 50-60 minutes late.

Reviewers also observed the handling of three no-shows. In two instances, the AAR procedures were followed, but in one case, the no-show was reported and entered after the vehicle had left the pick-up location and it appeared that the driver had only waited three minutes for the customer, rather than the five minutes required by the procedures. In one case, the driver reported arriving at 4:10 for a 4:00 PM promised pick-up. The driver radioed-in at 4:25 PM and the dispatcher attempted to call the rider (with no success). The trip was authorized as a no-show at 4:27 PM and the dispatcher requested a landmark from the driver. In a second case, the driver reported arriving at 4:25 PM for a 4:20 PM promised pick-up. The driver radioed-in just before 4:45 PM and the dispatcher authorized a no-show at 4:45 PM after not being able to contact the rider by phone and after recording a landmark. In the third instance, however, the driver reported a no-show while being polled by the dispatcher about the status of the run. The driver simply reported the time he had arrived and the time he had left and the dispatcher entered this into the system. The trip had a promised time of 4:05 PM, and the driver reported arriving at 4:20 PM and departing at 4:23 PM.

According to the work schedule in effect at the time of the visit, MV employs 13 dispatchers and four Dispatch Assistants. The Dispatch Assistants check drivers in and out and are the first point of contact for all calls to the dispatch center. As calls are received they will either look up the status of the trip using a computer workstation located in the driver check-in area or will transfer the call to the dispatcher handling the run to which that trip has been assigned. Dispatchers manage runs and assist drivers as needed. Table IX.5 below shows the work schedules for dispatchers and Dispatch Assistants.

Four or five morning dispatchers report between 5:45 and 6:45 AM on weekdays. One additional dispatcher reports at 12:30 PM to assist with covering lunch breaks. Two to three afternoon dispatchers report between 2:00 and 2:30 PM to replace the morning dispatchers. Two or three evening dispatchers report between 9:00 and 11:00 PM and work through the night.

During weekday morning peak hours, MV has four to five dispatchers on duty. Only three to four dispatchers are on duty, however, during the afternoon peak (3:00 to 5:00 PM). With about 85 peak hour runs, this then suggests that each dispatcher is managing between 17 and 21 runs in the morning and 21 to 28 runs in the afternoon.

On Saturdays, only one dispatcher is scheduled between 7:30 and 9:00 AM. This increases to two dispatchers at 9:00 AM, to three dispatchers at 1:00 PM and to four dispatchers at 3:00 PM. Four dispatchers remain on duty until 5:30 PM, three remain on duty until 10:30 PM and two are then scheduled to work through the night.

On Sundays, two to three dispatchers are scheduled between 6:00 and 7:30 AM, three are on duty from 9:00 AM to 1:00 PM, four dispatchers are scheduled from 1-2:30 PM and then three dispatchers remain on duty throughout the evening. Overnight, two dispatchers are scheduled.

Table IX.5 - MV Transportation Dispatch Staffing Based on September 11, 2003

D
[Link opens in a new browser window]

One assistant dispatcher is scheduled in the early morning on weekdays to cover morning pull-out. An afternoon assistant dispatcher starts work at 12:00 noon and the overnight assistant dispatcher reports at 8:30 PM. On Wednesdays, Thursdays and Fridays, a second afternoon assistant dispatcher reports at 1:00 PM to help with afternoon pull-ins and pull-outs.

On weekends, only one assistant dispatcher is scheduled (from 6:00 PM to 2:30 AM). For the rest of the days, other staff must cover the front desk, check drivers in and out and handle ride status and other calls. Given that regular dispatch staff levels are already low on Saturday mornings, it appears that MV may also be low on the total number of dispatchers/assistant dispatch staff at some times on the weekends.

PTM Management (PT) - PT dispatchers are assigned approximately 17 to 20 routes. A review team member observed the PT dispatchers from 6:50 to 8:30 AM on Monday, September 15, 2003. Dispatchers monitor service using the scheduling software. Dispatchers check the driver's location relative to their route schedule every 30 minutes and enter the driver's last actual pick-up or drop-off time into the computer. PT indicated that drivers often have trouble reporting locations when dispatchers are busy. As a result dispatchers often contact the drivers. PT also maintains a "dispatch turnover log" on which they record vehicle breakdowns, service no-shows, routes that are more than 45 minutes late and driver checks. The form provides for hourly driver checks (see Attachment O). The computer automatically updates the estimated times for the rest of the route based upon the performed time entered by the dispatcher. Routes that are running late will appear in red on the display. This includes routes that are running later than five minutes before scheduled drop-off or appointment times. Dispatchers can manually override the estimated times generated by the computer, but not the times promised to customers.

PT dispatchers indicated that the Command Center calls with cancellations less than one hour before the scheduled time. The dispatcher can then notify the driver of the cancellation. If the cancellation is requested more than one hour in advance, the cancellation is communicated to PT through the scheduling and tracking software.

Dispatchers performed time checks and monitored route schedules for both lateness and slack time. At PT the reviewer observed 29 actions. These included 22 time checks, one early arrival at a pick-up location, two cancellations, one trip moved and one customer no-show. The moved trip resulted from the wrong customer boarding the vehicle and taking the trip, resulting in the original customer being late. Additionally, there was one instance in which the driver was directed to the wrong location. When checked using the Geographic Information System (GIS), the wrong location continued to appear. Eventually, through direct discussion between the dispatcher and the customer the driver was redirected to the correct location. The customer said that she has had the same problem before and, more than once, requested that the erroneous directions be corrected. It appears that the incorrect information is the result of a problem with the GIS.

Dispatchers indicated that they moved trips based upon how late the route is running, how tight the route is and how good the driver is. PT dispatchers focus on service up to two hours in advance. By monitoring service and taking corrective action, no more than one late or missed trip occurs on a route as a result of a service problem. A PT dispatcher indicated that some routes are very tight, such that a five-minute delay in boarding a passenger can result in the remainder of the stops scheduled for the route being late. The software system offers suggestions on routes to move trips to. The PT dispatcher understood the -5/+25 pick-up window. PT dispatchers require drivers to wait five minutes for the customer after the beginning of the pick-up window. The PT dispatcher indicated that when the vehicle arrives early, they will call the customer before the pick-up window but if the customer fails to show for his/her ride the dispatcher will not make a second call to the customer within the pick-up window.

TFM (Transit Facility Management) - TFM has a T1 line with 24 paths with one direct dial number. The line is shared with its bus company. Seven lines are assigned to AAR operations, four for dispatchers, two for supervisors and one general line. All TFM drivers are equipped with Nextels.

TFM operates approximately 50 peak period routes with three dispatchers. Dispatchers are assigned approximately 15 routes apiece. A review team member observed the TFM dispatchers from 7:00 to 9:00 AM on Friday, September 12, 2003. Dispatchers monitor service using the scheduling software. Dispatchers check the driver's location relative to their route schedule every 30 minutes and enter the driver's last actual pick-up or drop-off time into the computer. The computer automatically updates the estimated times for the rest of the route based upon the performed time entered by the dispatcher. Routes that are running late will appear in red on the display. This includes routes that are running later than five minutes before scheduled drop-off or appointment times. Dispatchers can manually override the estimated times generated by the computer, but not the times promised to customers.

Dispatchers were observed performing time checks and monitoring both schedule and monitor routes for lateness and slack time. At TFM the reviewer observed 31 actions including 22 time checks, one trip cancellation, three customer no-shows, one no-fault no-show and four transfers of trips from one route to another. The 31 actions over the two-hour period for approximately 15 routes indicate that there were one or fewer contacts with each driver per hour. (One driver was contacted four times during this period.) For one customer no-show the vehicle arrived 10 minutes before the pick-up time scheduled with the customer. The dispatcher called the customer 10 minutes before the scheduled time and got no answer. This call was made outside of the -5/+25 minute pick-up window that the customer is supposed to be available for a ride. The dispatcher instructed the driver to categorize the customer as a no-show if they did not appear before the scheduled time.

Dispatchers indicated that they moved trips based upon how late the route is running, how tight the route is and how good the driver is. By monitoring service and taking corrective action, no more than one late or missed trip occurs on a route as a result of a service problem. The software system offers suggestions on routes to move trips to. The TFM dispatchers seemed competent and appeared to coordinate well with one another in reassigning passenger trips from one route to another. In one instance a trip was moved from one route to another and a second trip was moved from the receiving route to yet another route to address a late route. Also, two trips were moved to a rescue route.

Star Cruiser - The telephone system at Star Cruiser has six lines. All six lines are available on telephones throughout the operation. Managers reported that the number of lines was adequate for the number of dispatch stations and for administrative needs. Staff reported no issues regarding phone capacity.

The morning Dispatch Supervisor reviews runs transmitted by the control center first thing each morning. Some "fine-tuning" of the schedules is done each day. This includes moving some trips between runs or making minor adjustments to scheduled times.

The dispatchers on duty were observed regularly scanning the status of runs. Both dispatchers indicated that they review runs when they first come on duty to identify those that appear tight and those that have more slack. They then will spend more time tracking runs they expect will need assistance. During observations drivers did not report each pick-up and drop-off. They did, however, contact dispatch if they needed assistance. Otherwise, the dispatchers monitored the status of each run with periodic polling.

Reviewers observed dispatchers handling two customer no-shows. In both cases, dispatchers asked the drivers for the time they arrived at the pick-up location. They then checked that time against the pick-up time promised to the customer and the 25-minute pick-up window allowed for arrival of the vehicle. Dispatchers attempted to reach the riders by phone in both cases. They were not successful in either attempt. Dispatchers then asked the drivers to describe a landmark and landmarks and the arrival times were entered into the system. In one instance, the driver arrived for a 10:00 am pick-up at 9:49 AM and the dispatcher approved the no-show at 10:06 AM. In the second case, the driver arrived at 9:18 AM for a 9:12 AM pick-up and waited until 9:26 AM. Just as the dispatcher was authorizing the driver to proceed, the rider appeared and boarded.

Both dispatchers indicated that some runs could be tight. They noted that if there were any unexpected delays in riders boarding vehicles, traffic, or other delays, drivers would begin to fall behind. Both dispatchers felt that having a floater vehicle was very helpful since it was not always possible to identify another vehicle with slack time in the area to accept reassignment of the customer. Both felt a second floater would be useful.

Observations during the morning peak period indicated that several runs were tight. Dispatchers appeared to identify pick-ups that were identified by the system as potentially late, however, and proactively assigned these passenger trips to other drivers or to the floater vehicle.

Reviewers also observed both dispatchers keeping the window that shows cancellations open (in the lower corner of the dispatch screen) and to constantly scan the window for new cancellations. These were then communicated to drivers in a timely manner. Dispatchers indicated that they usually communicate cancellations at the same time they poll drivers on run status - to minimize the number of driver interruptions. Dispatchers also indicated that they often count on trip cancellations to open-up slack time in runs that can be used to keep the system on schedule.

Both dispatchers indicated that getting authorization for "rescue routes" was sometimes difficult and that the authorization might not be timely. They indicated that requests for rescue routes are only entertained by the control center when a pick-up is 30 minutes behind. One dispatcher said "That then only gives us 15 minutes to do the trip" - meaning that this was how long they had before being penalized for the late pick-up. Both noted that a more timely process for getting rescue routes approved would help.

Dispatchers received several calls from the control center asking about the status of rides. Reviewers observed some cases, in which riders were calling the control center well within the 25-minute period after the promised time. The dispatchers indicated that they get a lot of calls like that and that some riders either do not understand the "pick-up window" or simply do not like to wait for the full 25 minutes after the promised time.

Reviewers also observed dispatchers using the time promised to the customer when communicating with drivers. For example, they might say, "Have you made your 9:00 pick-up?" if the trip in question had a 9:00 promised time but a 9:15 scheduled time. This seemed to reinforce to drivers that the time they needed to be watching (and performing) was the promised time. And this then seemed to be causing some confusion in terms of whether the schedules were reasonable. When pointing out runs that were tight, the two dispatchers typically would note multiple trips with similar or very close promised pick-up times. They did not seem to pay attention to the scheduled times which were more spread-out and more reasonable.

The dispatcher on duty indicated that a Scheduling Supervisor from the NYCT meets with dispatch staff once a week to review scheduling and dispatch issues. This was felt to be very positive and constructive.

Based on a work schedule provided dated September 1, 2003, Star Cruiser employs two full-time Dispatch Supervisors and five full-time and one part-time dispatcher. One Dispatch Supervisor works a daytime shift (5:00 AM to 5:00 PM weekdays and 7:00 AM to 12:00 PM on weekends - 70 hours total). The second Dispatch Supervisor works a night shift (6:00 PM to 6:00 AM Wednesday through Sunday - 60 hours total). These and other dispatcher shifts are presented in Table IX.6 below. In addition to the Dispatch Supervisors, there are two morning dispatchers (one that reports at 6:00 AM and a second that reports at 8:00 AM). There is also a mid-day dispatcher, two afternoon/evening dispatchers, and one late night/early morning dispatcher. On weekdays during peak operating hours, there are a total of two to four staff scheduled - one to three dispatchers plus a Supervisor. During the evening and early morning hours on weekdays, there are two to three people working in dispatch. On weekends, two people cover dispatch at all hours except 6:00 AM to 7:00 AM when only one person is on duty.

During weekday peak periods, each dispatcher manages 17-28 runs. On the day the operation was observed, two dispatchers were handling the primary dispatch functions and were managing 55 runs between them. One was handling the odd-numbered runs and the other the even-numbered runs. The Supervisor on duty managed other aspects of the operation.

Table IX.6 - Star Cruiser Dispatch Staffing Based on September 1, 2003 Work Schedule

D
[Link opens in a new browser window]

Operators

Atlantic Paratrans - Two assessment team members visited the Atlantic Paratrans of NYC, Inc. operations center on Monday, September 15, 2003. Reviewers interviewed staff, including senior managers as well as operations staff and drivers. Driver training, turnover, availability and run coverage was examined, and vehicle availability and condition was analyzed.

The Atlantic Paratrans Manager noted that schedules are typically received between 6:00 and 7:00 PM on the evening before the service day. The evening dispatchers then review the schedules after the evening peak (usually between 9:00 and 11:00 PM). Some adjustments are needed, but the Manager noted that the schedules they receive are "pretty good." It was noted that if schedule adjustments are needed, trips are moved within or between runs but no trips are left unscheduled on open runs.

The Manager also noted that the central schedulers seem to be continually revising and improving the schedules and they have been getting better over time. He also indicated that there is a good working relationship between the schedulers at First Transit and NYCT and the Atlantic dispatchers and Supervisors.

It was the Manager's opinion that there were operating issues in 1999 and 2000 when the system grew rapidly but that in the past few years, service quality has improved significantly. He noted that in the 1999-2000 period, Atlantic Paratrans' operation grew by about 100 vehicles in a nine-month period. Driver recruitment and other operating problems emerged during this period of rapid growth. He also noted that a tight labor market at that time contributed to difficulty in recruiting and retaining drivers.

Vehicle and run records for August 18-24 were reviewed. At that time, Atlantic Paratrans operated a fleet of 210 AAR vehicles. On August 25, 2003, ten additional vehicles were assigned to Atlantic Paratrans increasing their AAR fleet to 220 vehicles. Between 196 and 221 weekday runs were assigned to Atlantic Paratrans at that time. Peak service requires that 184-187 vehicles be available. About 113-117 weekend routes were typically assigned.

The Atlantic Paratrans Manager noted that they operate one "999" vehicle (a floater vehicle). He noted that this vehicle is "under the control of the central control dispatchers."

Vehicle availability reports for August 18-25, 2003, were reviewed to determine if enough vehicles were available to cover peak service requirements. Summary information from these reports is provided in Table IX.7 below.

There were enough vehicles available each day to cover peak period requirements. Prior to the receipt of 10 new vehicles, there were 3-6 vehicles available as spares beyond those needed for scheduled runs. On August 25, with the 10 new vehicles, there were 17 vehicles available as spares.

As of September 8, 2003, Atlantic Paratrans employed 373 drivers with average tenure of 3.74 years. 333 (89.3%) of Atlantic Paratrans drivers had worked for Atlantic Paratrans for more than one year. A review of the daily operations sheets for the week of August 18-24, 2003, which show driver assignments by run and the use of extra board drivers, indicated that no runs were cancelled due to a lack of drivers. There appeared to be an adequate number of available extra board drivers to cover scheduled days off as well as service-day absences. The Atlantic Paratrans Manager indicated hat he couldn't remember the last time they had cancelled a run due to the lack of availability of drivers. Atlantic Paratrans does not appear to have a problem with driver recruitment or retention.

New drivers receive nine days of classroom training. Three of these days are spent on disability awareness and passenger assistance techniques training. Trainees are then "apprenticed" to experienced drivers for on-the-road training. On-the-road training lasts between one and two weeks depending on the skill of the trainee. It was noted that only about two to three of ten trainees in each class successfully complete the training course and that as a result, the company is constantly recruiting drivers.

Drivers who complete the training course are started on 6.5-hour runs but are paid for a full eight hours. This gives them time to review schedules and plan their day. They are gradually moved up to runs of between 8 and 10 hour runs in duration.

Starting pay for drivers is $10.25 per hour. During the first year of employment the pay rate gradually increases to $12.38 per hour. Based on the current contract, maximum pay is $15.00 per hour. Drivers also receive health benefits ($15 individual or $59 family weekly contribution). There are six paid holidays, one personal day and a paid birthday leave each year. Drivers also get one week of paid vacation after one year and two weeks of paid vacation after two years.

The manager noted that drivers who are hired for eight-hour runs are paid the full eight hours even if some of the hours on that run are trimmed, or removed during scheduling by First Transit. Drivers also receive overtime if runs are extended beyond eight hours. The Manager noted that there are about 1,800 hours of overtime paid per week.

The Manager indicated that the workforce is relatively stable and that most turnover is due to transfer by drivers to other parts of the company. Atlantic Paratrans is a major school bus operator and also a fixed route operator and paratransit drivers often transfer to better paid positions in these other divisions. Even with this internal migration, however, the Manager did not feel that driver turnover was a significant problem at the current time.

Maggie's Paratransit - Two assessment team members visited the Maggie's Paratransit Corporation operations center on Monday, September 15, 2003. Reviewers interviewed staff, including senior managers as well as operations staff and drivers. Driver training, turnover, availability and run coverage was examined, and vehicle availability and condition was analyzed.

Maggie's Paratransit operates a fleet of 115 AAR vehicles. About 84-85 vehicles are needed to cover peak-hour runs on weekdays. Maggie's is not authorized to operate a "floater" vehicle. About 30-31 spares are therefore available.

Vehicle availability reports for the week of August 18-22, 2003, were reviewed to identify the number of vehicles available for service each day compared to the number of peak-hour vehicles needed to cover all assigned routes. The information collected is provided in Table IX.8 below.

As shown, 23 to 24 spare vehicles were available during peak periods on these days. Maintenance staff and operations managers indicated that they always have enough vehicles to cover assigned runs.

As of September 8, 2003, Maggie's Paratransit employed 169 drivers. Average period of employment for these drivers was slightly less than one year with 99 (59%) of its drivers having less than one year's experience.

Managers indicated that trainees are paid a $6.00 per hour. New drivers then start at $11.00 per hour. Maximum hourly wage for drivers is $15.00. After 90 days, drivers also are eligible for medical coverage ($17 per week individual, $57 per week family). There are also five paid holidays and drivers also get their birthdays off with pay.

New drivers receive 10 days of training. Drivers spend the first day, as well as days three thru seven, in a classroom setting. Sensitivity training and passenger assistance is taught on days three and four. The second day and days nine and ten are spent on the road. On day two, trainees ride with experienced drivers and observe what they do. On days nine and ten, trainees drive and perform all other duties and are observed by driver-trainers. The second day "ride along" is used to allow trainees to get a first-hand knowledge of the job early in the training to be sure that the job is what they thought it would be.

MV Transportation - Two assessment team members visited the MV Transportation, Inc. operations center on Friday, September 12, 2003, and again on Thursday, September 18, 2003. Reviewers interviewed staff, including senior managers as well as operations staff and drivers. Driver training, turnover, availability and run coverage was examined, and vehicle availability and condition was analyzed.

MV Transportation operates a fleet of 110 AAR vehicles. It is typically assigned about 134-148 runs per weekday, about 41 runs on Saturdays and about 51 runs on Sundays. On weekdays, about 85 to 87 vehicles are needed to cover peak operations.

Four vehicles that were returning from morning runs were inspected. All appeared to be generally clean and in fair to good condition given age and mileage. A few minor problems were noted, however. The tailpipe on one vehicle was noted to be loose (the rear hanging bracket was missing). The Quality Control officer saw the problem, but the vehicle was not taken in for maintenance before the next run. The door to the battery tray (mounted on the passenger-side skirt) of one vehicle also appeared to not be secured (broken latch). Again, the vehicle remained in the yard for the next run.

With the help and guidance of NYCT contract management staff, MV Transportation recently implemented a more efficient vehicle turn-around procedure. Doorways at the facility were modified to allow vehicles to pull-through and be checked-in and checked-out in a more efficient manner. MV Transportation staff indicated that there had been some difficulty in the past with turning vehicles around in the time allowed by the schedules. This reportedly has improved.

The maintenance supervisor indicated that because some vehicles run "24 hours," getting them cleaned every day is sometimes a problem. Drivers also indicated that minor maintenance issues are sometimes not addressed because the vehicles were needed for scheduled runs.

Vehicle availability reports for weekdays between August 19 and August 25, 2003, were reviewed to determine the number of vehicles available for service each day compared to the number of peak period vehicles needed to cover all assigned runs. The information collected is provided in Table IX.9 below.

Vehicle availability records for weekends are not prepared since far fewer vehicles are needed on weekends and availability is not an issue. In August 2003, MV Transportation operated a fleet of only 100 vehicles. In the last month, 10 additional vehicles have been added to the fleet and as noted at the beginning of this section, their fleet is now 110 vehicles. In August 2003, however, it can be seen that the maintenance shop was reporting far fewer vehicles available for service on some days than were needed to cover peak period routes. On Tuesday, August 19, 2003, only 77 vehicles were reported ready for service while the peak requirement was 85. And on Monday, August 25, 2003, only 71 vehicles were available while 85 were needed. The MV Transportation Manager indicated that maintenance staff often reports vehicles as unavailable for service but that many of these vehicles have only minor problems that do not prevent them from being used if necessary. When needed, these vehicles can be placed in service. He indicated that MV Transportation "has enough vehicles, it is a matter of going to the shop to get the buses released." He indicated that he has been working on developing better communication between his Quality Control Officer and his Maintenance Manager.

Based on an interview with the MV Transportation Manager, it was clear that this provider has gone through significant staffing changes in the past year. The current Manager has been in that position for about a year. He was a former NYCT employee who indicated he had come out of retirement to help turn the operation around. He indicated that in the past year, since he started, he had replaced the Maintenance Manager, and three of the Dispatch Managers and about half of the dispatchers were new employees.

A review of the MV Transportation driver-training curriculum indicated that new drivers receive at least 10 days (80 hours) of training. This includes nine days of classroom training and at least one day of "behind-the-wheel" training. Additional behind-the-wheel training is scheduled as needed. Three of the nine classroom days are spent on disability awareness and passenger assistance techniques training.

As of September 8, 2003, MV Transportation had 250 drivers with an average of one year of employment. Approximately half the drivers had more than one year of employment. The Assistant Manager indicated problems having enough drivers to cover runs on Sundays, indicating that MV Transportation may have difficulties with driver recruitment and retention.

PTM Management (PT) - Operators can review schedules through their computer system as First Transit develops them during the day before service delivery but may not alter the schedule. Operators normally receive access to their schedules for service between 5:30 and 7:00 PM on the evening before service. On evenings before holidays schedules are sometimes delivered before 5:00 PM and when there are computer system problems the schedules can be delivered as late as 3:00 AM on the service day. After the schedules are handed off to the operator, the operator is free to alter the schedules. Operators address modifications to the schedules in different ways. PT's night dispatcher reviews the next day's schedules and flags potential problems. Operators have the latitude to move passenger trips from one route to another among the routes assigned to them or to change estimated times in the schedules but not the times promised to customers. Extension or addition of routes requires approval by the AAR Command Center. Common schedule problems encountered are: geo-coding errors, address errors, problems with some of the manual insertion of passenger trips that override the software parameters onto routes. Operators print the drivers' manifests on the night before service.

PT drivers arrive 15 minutes before their first pull out to review schedules and prepare for their runs. They are not paid for this time.

The new computer scheduling was activated in May of 2002. There were problems with the schedules during an implementation period of approximately one year. The overall effect of the new schedules has been an increase in productivity. PT's Project Manager also indicated that schedules have improved with trip times that better reflect traffic conditions and denser trip patterns on routes. The denser trip patterns resulted from assigning trips to routes based upon an "affinity" factor in the scheduling process that groups trips that are near to each other on one route.

NYCT also faxes a "trim slack" report to each operator. The report identifies the time that each route will begin or end. The operators then notify their drivers when their route begins if there's a change in the start time.

The PT morning dispatchers review their routes and enter any cancellations onto the manifests of drivers who have not yet reported to work. Some operators are authorized by NYCT to operate "floater" or unassigned drivers to be available to respond as the need arises. PT has one floater for eight hours on Fridays. When problems arise during the service day dispatchers can add service by extending routes (having drivers work beyond their assigned work hours by working late or beginning early) or adding routes. PT indicated that many routes are about five hours long providing opportunity to extend routes to as much as eight-hour work shifts. Prior approval of the AAR Command Center is needed for route extensions. PT also indicated that routes must appear late in the computer system before the Command Center will authorize a route extension and that it can take two to three hours after a problem has been identified by the dispatcher to get authorization. It is the practice of PT to extend or add routes prior to getting Command Center approval in order to minimize delay to customers. By doing so the operator is making best efforts to serve the customer when there is a service problem but is taking a risk that the "rescue" will not be approved and the operator will not be paid. PT indicated that on occasion they have a stranded passenger and return the trip to the command center for reassignment to another carrier.

Star Cruiser Transportation (Star Cruiser) - Two assessment team members visited the Star Cruiser operations center on Thursday, September 11, 2003. Reviewers interviewed staff, including senior managers as well as operations staff and drivers. Driver training, turnover, availability and run coverage was examined and vehicle availability and condition was analyzed.

Star Cruiser operates a fleet of 60 AAR vehicles. It is typically assigned about 70 runs per weekday and 55 vehicles are needed to cover peak-hour runs. Star Cruiser also operates one "floater" vehicle to assist with same day service issues and has four spare vehicles.

Star Cruiser has implemented a very efficient vehicle turn-around procedure. As vehicles return from runs a maintenance supervisor inspects them. Vehicles are then cleaned, fueled and made ready for the next run. Maintenance issues reported by drivers are also handled as needed.

Maintenance staff and operations managers indicated that they always have enough vehicles to cover assigned runs and runs rarely, if ever, go uncovered due to a lack of vehicles.

Seven randomly selected vehicles were inspected at pull-out on September 11, 2003, for general condition and cleanliness. All appeared to be in very good condition. It was noted that Star Cruiser has a 509/19A Compliance Officer who supervises pull-out to make sure all vehicles that leave the yard meet state and federal requirements and are in good condition.

As of September 8, 2003, Star Cruiser employed 112 drivers with an average term of employment of 1.4 years. 72 drivers (64%) had been employed for more than one year. Star Cruiser managers indicated that the workforce seemed to be relatively stable and that turnover was not a major issue.

It was noted by managers that drivers start at $10.50 per hour and receive $12.00 per hour after one year. After 90 days, drivers are eligible to receive medical benefits (paid 100% by the company) and contributions to a retirement account.

Star Cruiser management reported that new drivers receive 85 hours of classroom training and at least 34 hours of on-the-road training. Classroom training includes "Passenger Assistance Techniques" instruction. It was noted that about half of all new hires successfully complete the full training. To discourage applicants from going through the training and then going to another paratransit or transportation company, Star Cruiser holds the pay due for training time in an account and gives this to the employee only after they have worked for several months.

Transit Facility Management (TFM) - Operators can review schedules through their computer system as First Transit develops them during the day before service delivery but may not alter the schedule. Operators normally receive access to their schedules for service between 5:30 and 7:00 PM on the evening before service. On evenings before holidays schedules are sometimes delivered before 5:00 PM and when there are computer system problems the schedules can be delivered as late as 3:00 AM on the service day. After the schedules are handed off to the operator, the operator is free to alter the schedules. Operators address modifications to the schedules in different ways. The afternoon dispatcher for TFM reviews the schedules, as First Transit develops them, and flags potential problems. TFM is adding a scheduler to review schedules between 6:00 and 9:00 PM and make adjustments and corrections as needed. Operators have the latitude to move passenger trips from one route to another among assigned routes or to change estimated times in the schedules but not the times promised to customers. Extension or addition of routes requires approval by the AAR Command Center. Common schedule problems encountered are: geo-coding errors, address errors and problems with some of the manual insertion of passenger trips that override the software parameters onto routes. Operators print the drivers' manifests on the night before service.

The TFM dispatcher supervisor reviews routes when he reports to work at 9:00 AM and fixes problem routes by transferring trips to other routes. Some dispatchers review the routes they're dispatching before they begin work. In the event of absences, other dispatchers work over time and/or supervisors perform dispatcher functions.

All of the operators have union drivers but with different unions and contracts. Currently there is no provision for driver review of schedules before pull out at TFM. TFM is renegotiating the contract with its drivers union. The negotiations include provision for a paid 15-minute pre-trip inspection, which would include review of schedule manifests by drivers.

TFM indicated that they had some problems with drivers checking in at the dispatch window on time but pulling out of the garage late for their route. The dispatch window and the yard are currently located away from each other. TFM is reorganizing its space to locate the dispatch window nearer to the yard to improve supervision and correct this problem.

The new computer scheduling was activated in May of 2002. There were problems with the schedules during an implementation period of approximately one year. The overall effect of the new schedules has been an increase in productivity. According to TFM's Project Manager passenger trips increased by 25-33 % from 12,000 +/- in December 2002 to 15-16,000 in March of 2003. During the same period the number of routes increased from 68 to 71 (4.5%). The Project Manager attributed the increased efficiency to better fit of routes to peaks in demand, improved schedules and denser trip patterns on routes. The denser trip patterns resulted from assigning trips to routes based upon an "affinity" factor in the scheduling process that groups trips that are near to each other on one route. At TFM the productivity increased to 1.7 person trips per hour.

NYCT also faxes a "trim slack" report to each operator. The report identifies the time that each route will begin or end. The operators then notify their drivers when their route begins if there's a change in the start time.

According to the TFM Project Manager the time required from identification of need by the dispatcher and approval by the Command Center of these "rescues" can significantly delay service to affected customers. It is the practice of TFM to extend or add routes prior to getting Command Center approval in order to minimize delay to customers. By doing so the operator is making best efforts to serve the customer when there is a service problem but is taking a risk that the "rescue" will not be approved and the operator will not be paid. TFM reported that on occasion "rescues" that have been dispatched have not been approved.

Driver Interviews

While conducting site visits to the contract carriers, the review team interviewed a total of 53 drivers. Drivers were randomly selected as they finished or before they began their runs. Team members interviewed a mix of new and veteran drivers, with from two weeks to 13 years of experience as AAR drivers. Team members asked each driver a series of questions that covered the following topics:

• Training received,
• Condition of the paratransit vehicles,
• Quality of daily manifests,
• Level of dispatch support provided, and
• Their understanding of operating procedures (particularly the on-time performance window and no-show procedures).

Table IX.10 presents the number of driver interviews by carrier and by the range of time drivers have worked for AAR.

Following are some of the key observations gathered from the driver interviews:

• Training was helpful and good. Each carrier trains its own drivers.

• Most drivers said that their carriers maintained the vehicles well. The general exception was that the opinions of MV drivers' on maintenance varied greatly.

• Drivers said most vehicles were fine. Older vehicles (more than two years) had relatively more problems, though still were in good condition.

• Many drivers thought that their daily manifests were too tight. Some also thought that the routing was often not logical.

• There was an inconsistent understanding among drivers of the difference between the estimated time and the negotiated pickup times listed on the manifest.

• Not all drivers knew the correct pickup window of -5 minutes/+25 minutes.

• Most drivers understood the correct procedures for passenger no-shows. Most drivers also understood the correct procedures if they arrive early at a pickup address.

• Most drivers knew that they were supposed to get dispatcher permission to re-arrange the order of pickups and drop-offs on the manifest. But other drivers made changes and told the dispatcher after the fact or not at all.

NYCT Reports

The assessment team reviewed the disposition, or coding, of trips from what was scheduled by NYCT on the evening before service to the ultimate performance of the trip. The team reviewed AAR's performance reports, AAR's data file for a randomly selected weekday, Thursday, May 15, 2003, and a sample of manifests as completed by drivers for each of the operating contractors for the same service day. The purpose of this review was to affirm that NYCT's performance reports are representative of the information being reported by drivers and recorded in AAR's database and properly reflect service performance.

Each of the operators prepares computer reports on the disposition service performance and submits the reports to NYCT. The actual passenger pick-up and drop-off times reported are taken from the information recorded by drivers on their manifests. According to NYCT managers, analysts, assigned to each of the four Compliance Managers, review and reconcile performance reports. Analysts review a sample of trips and customer no-shows. If incorrect reporting of information by the operator is identified, the operator is advised and further actions are taken to correct the problem if warranted. NYCT managers also observe whether morning and afternoon driver pullouts onto their routes are on time.

Sample manifests from 24 randomly selected routes (three from each of the eight major operators) were examined to verify accuracy of pickup and drop-off data that had been entered into NYCT's database. The random sample consisted of 203 completed trips taken by clients on May 15, 2003. Of the 203 trips, pickup times on the manifests varied from that in NYCT's database in 18 instances, or approximately 9% of the time. Drop off times varied in 10 cases, or about 5% of the time.

Of the 18 differences:

• Seven listed ranges of times on the manifest, but a single pickup time had been entered on the master spreadsheet,

• One had a pickup time that was unreadable on the manifest, but had an entry in the spreadsheet.

• Two had pickup times on the manifest that had been transposed when transferring data to the master spreadsheet,

• Three indicated earlier pickup times on the manifest than reported on the master spreadsheet, and

• Five indicated later pickup times on the manifest than reported on the master spreadsheet.

In the eight cases that had ranges or an illegible number it is NYCT's default procedure to have the operator use a time entered into the scheduling system by the dispatcher when performing time checks on drivers. This procedure would account for these eight differences and should result in representative times in the NYCT database. Likewise, the transcription errors and early pickup times should not cause the data reported by NYCT to be unrepresentative of service.

Of the five cases above (2.5% of the 203 trips sampled) with later pickup times on the manifest than that reported on the master spreadsheet, time differences ranged from 3 to 10 minutes later than reported, with an average difference of 6 minutes. Only one case was later than 25 minutes from the promised pickup time. This potentially represents an underreporting of late pick-ups by 0.5%.

Of the 10 cases with manifest drop-off times that differed from the master spreadsheet:

• One had no drop-off time on the manifest, but had an entry in NYCT database,

• One had a drop-off time that was unreadable on the manifest, but had an entry in the NYCT database,

• Six (three pairs) had drop-off times on the manifest that had been transposed when transferring data to the NYCT database, and

• Two indicated earlier drop-off times on the manifest than that reported on the NYCT database.

As with pick-up times, none of the differences observed appear to skew the NYCT reported data. Accordingly, the NYCT database appears representative of the information provide by drivers on their manifests and has been used as a basis for further analysis of service performance.

AAR categorizes the disposition of scheduled trips as follows:

• Late Trip Cancellation (CX_L) - Customer cancellation after 5:00 PM the night before a scheduled trip.

• Customer No-show (NSCNS) - The customer has not presented him/herself, declines the trip or is unable for any reason to take the trip when the vehicle operator has arrived and waited for the customer in accordance with contract standards and procedures.

• No-Fault No-Show (NSNNS) - The contractor fails to provide the scheduled trip to a customer for documented reasons beyond the control of the contractor. The Contract Standards of Performance (Article 104) further reserves the right for NYCT to declare either a contractor or no-fault no-show, as applicable, when the contractor is unable to appear at the designated location within 45 minutes after the scheduled pick-up time. NYCT has indicated that operator vehicle arrivals after the pick-up window but within 45 minutes of the scheduled pick-up time are categorized as no-fault no shows.

• Contractor No-Show (NSSNS) - Any trip not performed other than a Late Trip Cancellation, Customer No-Show, or No-Fault No-Show. NYCT further describes this to include a no-show reported by the operator more than 5 minutes before or 45 minutes after the scheduled pick-up time, or the vehicle never arrives at the correct location for the scheduled trip, or the vehicle arrives at the correct location within 5 minutes before and 45 minutes after the scheduled pickup time, but does not wait 5 minutes.

• Served (S) - The trip was completed.

• Taxi - AAR authorized the customer to use a taxi or black car to complete the trip.

Trip Disposition

49 CFR §37.131(f)(3)(B) prohibits transit entities from limiting service availability through patterns or practices that result in substantial number of missed trips. For purposes of measuring performance against the regulatory criteria, missed trips are defined as trips that were not served when the customer was available for the trip throughout the 30-minute (-5/+25) pick-up window. It would appear that all of the trips categorized as contractor no-shows and no-fault no-shows are missed trips. Some taxi trips may be missed trips if the customer chose not to use the authorization and make the trip. Because of lag time in customer and black car payment requests and possible failure to request reimbursement for completed trips, AAR could not confirm the number of authorized taxi trips that were completed. Also some customer no-shows may be missed trips if the pick-up location was incorrectly recorded on the manifest and/or the vehicle driver unknowingly went to the wrong location, in spite of AAR's efforts to avoid or minimize such problems.

NYCT provided the reviewers a table, Paratransit Summary Statistics, identifying the disposition of trips for each day during the month of May (Attachment P). For the month, of 248,593 trips scheduled 15,578 (6.3%) were late cancellations, 7,165 (2.9%) were customer no shows, 1,769 (0.7%) were no-fault no-shows, 390 (0.2%) were carrier no-shows, and 223,692 (90.0%) were completed by contract operators. During the month 747 vouchers were redeemed for trips completed by taxis or black cars (Attachment P). Because there can be a substantial lag from the date of travel until the voucher is redeemed, these vouchers are not reconciled by travel date. As a result the number of trips made by voucher cannot be precisely identified. However, the redeemed vouchers during May equals 0.3% of scheduled trips during May. NYCT officials indicated that they subtract multiple requests for the same trip and trips for which taxi vouchers were issued from their performance reports. As a result the reported figures in Attachment P will be greater than the numbers presented for a sample day as discussed below.

The review team selected May 15 as a sample day to perform a more detailed review of trip disposition. Table IX.11 presents the disposition of trips for that day. As noted above these numbers do not include trips for which taxi vouchers were authorized.

Reviewers looked at the number of taxi vouchers authorized to determine how many of the trips that were not completed by operators might have been completed through the use of a voucher. Vouchers issued on May 15 for trips to be made on May 15 are presented in Table IX.12. It appears that the taxi vouchers issued for the Taxi category were coded incorrectly. As one might expect, most vouchers were authorized for the no-show categories. Same day vouchers were authorized for 1.7% of the scheduled trips. Applying the 0.3% of vouchers redeemed during the month of May results in estimated use of the vouchers in approximately 20% of the time that they are authorized.

Assuming 20 % of the vouchers are actually used results in estimated disposition of trips as presented in Table IX.13. For example if 20 vouchers were authorized for carrier no-shows it is estimated that four would actually be used to complete the trip, reducing the number of trips not completed because of carrier no shows from 37, as presented in NYCT's data, to 33. 38 trips categorized as served were incorrectly categorized and, based upon review of a log of messages for taxi voucher authorizations, were typically some form of no-show or late cancellations. Vouchers were not used for 30 of these 38 trips categorized as completed. This constitutes 0.3% of scheduled trips. As a result these other categories are slightly understated in Table IX.13.

The 314 customer no-shows were reviewed to determine whether the vehicle reportedly arrived at the pick-up location between the scheduled pick-up time (5 minutes into the pick-up window) and the end of the pick-up window (25 minutes after the scheduled pick-up time). In all 314 instances the reported vehicle arrival time was within this twenty-five minute period. According to the trip logs customers disputed some of the no-shows as potential problems with things such as wrong addresses and passengers who were delayed in appearing when the vehicle arrived.

A no-fault no-show occurs when the operator arrives after the end of the pick-up window (25 minutes after the scheduled time), but less than 45 minutes after the scheduled time. From a customer service perspective these trips are missed trips, the vehicle did not arrive within the agreed upon time and the trip was not completed. These trips are captured in a separate category for contractual purposes. Operators are not penalized for late trips until they are more than 45 minutes late. Of the 118 trips categorized as no-fault no-shows in 12 (10%) instances the vehicle arrived more than 45 minutes after the scheduled time. This could be due to circumstances beyond the operators control and appropriately categorized as a no-fault no-show. For 59 trips including the 12 above, (50%) the vehicles arrived between 25 and 45 minutes after the scheduled time. These are no-shows for which the contractor is not penalized. For 17 trips (14%) the vehicle arrived more than 5 minutes early or before the pickup window communicated to the customer. These early arrivals could be carrier no-shows if the vehicle departed early or disputed no-shows if the vehicle awaited the customer. Finally, on 42 trips (36%) the vehicle arrived within the pick-up window. These could be trips where the driver arrived at an incorrect address or otherwise missed connections with the customer. From the perspective of customer service at least 50% of the no-fault no-shows are missed trips by the contractor and the remaining 50% are potential missed trips. Accordingly, for the sample day there were an estimated 82 (0.8%) missed trips. Whereas 49 CFR 37.131(f)(i)(B) states that there be no substantial number of missed trips, it is important to take particular care in defining and categorizing no-shows so as to accurately identify "missed trips" so that appropriate corrective actions can be taken to eliminate patterns of missed trips.

On-Time Performance

NYCT provided a report on Paratransit On Time Performance for the period February through July 2003 (Attachment Q). The report identified trips completed and late trips by month for each of the eight operators. Not included were trips completed by taxi or black car or the three back-up carriers.

NYCT measures only late pick-ups. It does not measure late drop-offs or early pick-ups. Early and on-time trips for the period were 92.3% of completed trips ranging from 91.5% in June to 93.8% in July. The number of late trips dropped from a high of 19,083 in May to 13,606 in July. Since an increase in completed trips from 178,554 in February to 225,750 in March, the number of completed trips appears to be stable. It appears that there has been a steady improvement in on-time performance. This could be attributable to seasonal traffic conditions as well as service improvements reported by NYCT.

Among the operators RJR has the best on-time performance, ranging from 96.6% in May to 98.2% in March. Contributing to this performance level is an experienced operator and favorable operating environment. RJR's primary service area is Staten Island. TFM has the lowest on-time performance ranging from 84.6% in June to 89.2% in March. TFM is a newer contractor.

In order to develop an independent estimate of on-time performance, the review team assessed AAR trips for a sample day of Thursday, May 15, 2003. Using an AAR data file the reviewer identified on-time, early, and late pick-ups and drop-offs for 8,968 trips completed by the eight operators. On-time pick-ups are those with a reported pick-up time within the pick-up window provided to customers from 5 minutes before until 25 minutes after the scheduled time. Early pick-ups are those performed more than 5 minutes before the scheduled time. If the pick-up occurs more than 25 minutes after the scheduled time it is considered a late pick-up. Drop-offs are considered late if they occur after the appointment time or the drop-off time given to the customer. Early drop-offs have not been considered although dropping customers excessively early for appointments would not likely meet the test of comparability to fixed route service. Information is reported on drop-offs only for those trips that the customer requested an appointment time.

For the sample day, 91.8% of pick-ups were on-time or early. This percentage is comparable with the 92.3% reported by NYCT for the six months from February through July, indicating that the sample day is representative of overall service performance. 6,634 of the pick-ups were within the pick-up window and 1,603 (17.9%) of the pick-ups were early. A concern with the large number of early pick-ups is the potential for customers to be pressured to accept a trip before they are ready. If this were to occur it could discourage customers from using the service and could be considered a constraining practice, which is not allowed by the USDOT ADA regulations. It may be appropriate for NYCT to monitor early pick-ups to assure that customers are not pressured to accept trips before they are ready.

On-time drop-offs of 89.4% is comparable to the rate of on-time pick-ups. Of particular concern with drop-offs is the need to meet appointment times such as medical appointments or the beginning of work or school. For an employee this means that they might expect to be late once in every ten trips or approximately once every two weeks. Chronic lateness could jeopardize riders' employment or require rescheduling of appointments with the associated inconvenience. Given the importance to customers of meeting appointment times, NYCT should consider monitoring performance for drop-offs as well as pick-ups.

The reviewers looked at the extent to which pick-ups or drop-offs are late. Of the completed drop-offs, 6% were more than 10 minutes late and none were more than 35 minutes late. This results in 94% of drop-offs made less than 10 minutes late. This means that every 4 weeks they will be late by less than 10 minutes once, and between 10 and 35 minutes once.

3.9% of completed pick-ups were less than 10 minutes late. As indicated in Table IX.15, 4.3% of all pick-ups were more than 10 minutes late with 0.6% more than 35 minutes late and 0.1% more than 1 hour late.

It should be noted that these late trips do not include trips completed through the use of vouchers. As discussed in the Trip Disposition section of this chapter, of the 167 vouchers authorized for May 15 it is estimated that approximately 20% (33) were used. Since vouchers are typically offered after trips are late, as indicated by NYCT practices and the logs for taxi authorizations, these estimated 33 trips are most likely late and possibly very late. If these trips were added to the late trips in the NYCT reports it would increase the late trips by about 0.4%.

At MV Transportation, a chart showing on-time performance for the months of August and September 2003 was on display in the administrative offices. Figure IX.1 below shows the on-time performance that was shown on that chart. Apparently data was not available for September 3. The information for this one operator provides an indication of the extent to which on-time performance varies by day. The chart showed that weekday on-time performance in August typically was between 88-94%. In September, weekday performance was somewhat lower, running typically between 84-90%. On weekends, though, much lower on-time performance was noted.

Figure IX.1 - On-Time Performance Noted by MV Transportation, August 1-September 16, 2003

D
[Link opens in a new browser window]

Table IX.16 shows the weekend on-time performance charted by MV Transportation during this period. As shown, Saturday on-time performance often ran only 79-85% and Sunday on-time performance was only 84% on August 24 and September 14, and only 83% on September 7, 2003.

MV Transportation management noted that on weekends MV Transportation is required to cover all of the service area because fewer providers operate during the late evening/early morning hours. Drivers are less familiar with the full area. And, because there are fewer total vehicles on the road at this time, the opportunity to reassign trips between vehicles if there are in service problems is reduced. Managers also noted that they had some issues covering runs with available drivers on Sundays. All of these factors have a negative affect on weekend on-time performance.

AAR On-Board Ride Times

As part of this review, the team analyzed AAR's on-board ride time (travel time) performance including a review of policies and procedures, a sample day of travel time data and a comparison of the travel time of actual ADA Complementary Paratransit trips with comparable fixed route trips.

Policies and Procedures

NYCT's goal is to schedule 100% of trips using the travel time standards in Table IX.17.

NYCT has no stated goal for actual performance in meeting the scheduling travel time standards and does not measure trip duration.

The scheduling parameters limit the duration for passenger trip schedules. Also each operator is subject to liquidated damages for poor on-time performance. The scheduling limitations and late trip disincentives are intended to prevent substantially long trips.

NYCT's categories overlap. For example, trips of three miles are included in both the first and second categories. The first category includes trips greater than or equal to zero miles and trips less than or equal to three miles and the second category includes trips greater than or equal to three miles and less than or equal to six miles. For analytical purposes, the trip categories were taken to mean: < 3 miles, >3 to 6 miles, >6 to 9 miles, and so on. The mileage refers to map mileage calculated by ADEPT (and for the reviewer's analysis based on MapQuest routing).

Review of On-Board Travel Time

A sample day - Thursday, May 15, 2003 - was randomly selected for this analysis. First, the average trip length was calculated based on completed trips operated during that sample day. Actual pick-up and drop-off times were used, not scheduled travel times. Reported odometer readings were used for mileage, recognizing that trips may not have been direct if they involved shared rides. In addition to overall averages, the data was aggregated into categories coinciding with the travel time standards set by NYCT. For example, trips < 50 minutes, trips from 51 to 65 minutes, and so on. Within each category, the data were scanned to look for lengthy trips that exceeded the ride time standards (based on actual on-board travel time and actual odometer readings).

Table IX.18 summarizes the results of this analysis. Using reconciled data for May 15, 2003, there were 8,968 completed trips. Of those, 8,774 trips (97%) were included in this analysis. The 194 trips that were excluded were trips that did not have recorded odometer readings and/or actual pick-up/drop-off times. This figure includes trips that were provided by taxi, black car and contractors that did not reconcile data.

The average travel time for all trips was 42 minutes; the overall average trip distance (based on odometer readings) was 9.3 miles. 68% of all trips were 50 minutes or less. Of those, the average travel time was 36 minutes and the average trip distance was 5.8 miles. 81% of all trips were 65 minutes or less and 95% of trips were 95 minutes or less. 23 trips (less than 1%) were more than 155 minutes (2 hours and 35 minutes), which is the maximum travel time standard for any trip set by NYCT.

The second part of the analysis included a comparison of actual on-board travel times for a sample of long ADA Complementary Paratransit trips with fixed route travel times as a means of assessing whether these paratransit trip lengths are significantly longer than comparable fixed route trips between the same origin and destination at the same time of day. For this analysis, a sample of 70 long AAR trips was selected from trips provided on May 15, 2003. The average travel time for the sample was 105 minutes and the average actual trip (odometer) mileage was 23 miles. The average trip mileage calculated by MapQuest for the sample trips was 13 miles.

NYCT's customer service staff assisted the review team in developing fixed route trip itineraries to compare with the actual on-board travel times for the sample of AAR trips. The fixed route itineraries were based on travel from the same origin to the same destination at the same time of day and day of week as the comparable AAR trip. The shortest fixed route trip time was calculated for each trip as well as a trip time using only busses.

A software program ("OTIS") was used to generate fixed route itineraries. Fixed route customer service personnel use this software to plan trip itineraries for customers. The following information was entered for each AAR trip:

• Origin address and borough,
• Destination address and borough,
• Desired departure or arrival time, and
• Date.

Other assumptions included:

• Minimize transfers,
• Use all modes (bus, subway, rail), and
• No restrictions based on disability.

The computer typically generated multiple itineraries. For analysis purposes, the first itineraries were selected for the "fastest" trip and "bus-only" options. The itineraries included the following information for each segment of the trip:

• Departure location (i.e., bus or subway stop),
• Departure time,
• Subway line or bus route number and head sign,
• Fare,
• Arrival location (i.e., bus or subway stop), and
• Arrival time.

The OTIS program calculated:

• Walk time and distance from origin to stop,
• Walk time and distance from stop to destination,
• Total fare, and
• Total trip time (including walking to and from stops).

The reviewers added 5 minutes to the trip time calculated by OTIS to each fixed route trip to account for wait time at the initial bus stop. The wait time assumes buses arrive every 10 minutes and passengers arrive randomly throughout the 10-minute period, resulting in an average waiting time of 5 minutes.

Using trip information from the sample drawn for May 15, 2003, Table IX.19 shows the results for the "fastest trip" and "bus-only" options generated by OTIS compared to the AAR travel time. The table shows the time of day that the trips were made, the origin and destination for each trip, the estimated Map Quest and actual AAR mileages, the travel times on AAR, and the OTIS estimated door-to-door travel times for the "fastest trip" and "bus-only" options. The table also indicates the number of transfers required to complete the trip on fixed route.

The set of columns on the right side of the table compares the AAR travel time with the "fastest route" and "bus-only" options. A minus sign (-) means that the AAR travel time was less than the comparable fixed route travel time. The table also includes ratios of paratransit to the fixed route times. The AAR and fixed route trips have the same travel time when the ratio is 1.0. Ratios less that 1.0 mean that AAR is quicker; ratios greater than 1.0 mean that the AAR trip is longer than fixed route.

The table shows that 43 (61%) of the AAR trips took less time or the same time as the bus-only trips. 10 (14%) of the trips could not be made using buses only (designated by N/A). The table also shows that 21 (30%) of the AAR trips were faster than or equal to the "fastest route" fixed route trips. The analysis indicates that when the AAR trip is very long, the comparable trip on fixed route is also long. For example, trip #46 was 124 minutes on AAR. However, the comparable trip itinerary for the "bus-only" option showed a total travel time - including walk time to/from the stops at each end plus 5 minutes wait time at the origin - of 211 minutes, which is 87 minutes longer. Similarly, for that same trip, the "fastest route" travel time was 174 minutes, which is 50 minutes longer than AAR. The Map Quest mileage indicates the driving distance for that trip was about 21 miles. The AAR trip was 22 miles long based on reported odometer readings.

The bold entries in the table indicate trips that have ratios greater than 1.5 (in other words, the AAR trips were more than 50% longer than comparable fixed route trips. Trip #1 indicates ratios of 1.8 for "bus-only" and 2.7 for "fastest route". Upon closer analysis, however, it should be noted that this trip is still relatively short (a paratransit ride time of 32 minutes and a bus-only travel time of 18 minutes). It is unlikely that a paratransit trip of 32 minutes would be considered significantly long even if longer than a comparable fixed route trip, given the characteristics of a shared route system and added time for some riders to board and alight vehicles. In contrast, trip #36 has a "fastest route" ratio of 2.1. The AAR travel time for that trip was 103 minutes; the "fastest route" travel time was 48 minutes (for a difference of 55 minutes). The "bus-only" travel time was 113 minutes, which is 10 minutes longer than the trip on AAR.

The six entries that are both bold and shaded indicate AAR trips that appear to be significantly long compared to both "bus-only" and "fastest route" trips. These trips also appear to have mileages that are considerably longer than the direct driving distance calculated by Map Quest. For example, trip #68 had an AAR travel time of 207 minutes - in excess of NYCT's standards. The estimated travel time for the "bus-only" option was 138 minutes; the estimated travel time for the "fastest route" trip was 76 minutes. The odometer mileage was 40 and the Map Quest mileage was estimated to be about 18. These results indicate that this trip was significantly long compared to both fixed route itineraries.

Table IX.19 - Travel Time Comparison: Access-A-Ride vs. Bus-Only and Fastest Travel Time

For complete information of this table, go to D
[Link opens in a new browser window]

1. Dispatch

   1. Dispatch procedures and operations generally appear to be effective with good coordination between the First Transit Control Center and the Operator's dispatchers.

   2. Routing of customers to different departments for each service issue can consume more staff time and discourage customers from resolving their concerns and using AAR service than addressing as many issues as practicable with one staff person. Customers who call the control center for reasons other than a same day service issue are transferred to another AAR department. In one instance a caller was rerouted to reservations to make a next day trip change after making a same day cancellation. In another instance a customer who was provided incorrect or incomplete information was routed back and forth between the control center and eligibility before resolving the customer's problem.

   3. Observed calls from customers inquiring as to the status of their ride appeared to be addressed effectively, with responses in less than four minutes. However, operator dispatchers indicated that the control center would often call to check on the status of rides that were operating within the "on-time" window. This practice can unnecessarily consume staff time and distract dispatchers from more important activities.

   4. The wording of dispatcher calls to customers announcing driver arrivals before the pick-up window could be construed to encourage, or require, customers to board the vehicle before the beginning of the pick-up window. In some instances when drivers arrive at pick-up locations before the pick-up window, dispatchers were observed telling customers that their "vehicle is waiting for them."

   5. In general, operators were observed applying procedures for customer no-shows effectively. In one instance a driver classified a customer as a no-show without dispatcher approval. In several other observations the dispatcher called the customer before the beginning of the pick-up window but not during the pick-up window. Since the customer may not be expecting the vehicle until the pick-up window begins, the early call could result in a missed connection with the customer.

   6. Same day changes in customer schedules were generally handled very well. Reviewers observed control center dispatchers making a positive effort to accommodate real time requests by customers for schedule changes.

   7. The practice of leaving a scheduled trip unchanged, when customers indicate that they will be late, until one hour before the scheduled pick-up time, limits utilization of the slack time created by the rescheduled trip.

   8. Some dispatchers appeared unclear that the scheduled time used in the software is intended to guide the driver in performing their trip. Conversely, dispatchers all seemed sensitive to the importance of the promised time in the schedules when monitoring service.

   9. Most dispatchers appeared to very effectively track performance of their routes and move trips as needed. Dispatchers poll drivers for location rather than drivers calling in their locations as indicated in AAR procedures. Polling drivers permits dispatchers to communicate with drivers at a time of their choosing, thereby avoiding interruptions from drivers when dispatchers are busy addressing late runs or no-shows.

   10. Reviewers observed some dispatchers at MV as much as 60-90 minutes late in polling drivers and devoting little time to monitoring late runs and taking corrective action.

   11. It appeared that dispatchers did not consistently report incorrect customer or trip information to the responsible department sufficiently for appropriate corrective action or that the responsible department was not always following through to correct incorrect information. Reviewers observed a number of instances of incorrect information causing service problems. This included incorrect customer IDs, addresses, phone numbers and GIS locations. In one instance an applicant for eligibility could not schedule service because of incorrect information on a form letter. Some of these problems with misinformation were recurring.

   12. The time required for operators to obtain authorization from the control center for recovery routes (route extensions or additions to address service problems) may contribute to late and missed trips. The use of route extensions and additions appears to be an effective way of addressing real time service problems. However, the time from identification of a late route by an operator's dispatcher to the route appearing late in the scheduling/dispatch system to request and authorization for recovery could be extensive. Most operators indicated that the time is excessive and they dispatch vehicles in advance of control center authorization to minimize delays to customers.

   13. The number and assignment of dispatchers appeared to be adequate for each of the operators. Most operators maintain a ratio of no more than about 20 runs per dispatcher most of the time. MV Transportation appears to have dispatchers handle up to 28 runs each on Monday and Friday afternoons. Managing 28 runs at one time could reduce the dispatcher's effectiveness in maintaining schedule control of the runs.

   14. The skill of the dispatchers appeared to be good but varied among operators and among dispatchers, which is to be expected in a large operation with personnel having differing amounts of experience. While some operators, particularly those with more experienced dispatchers, appeared to be very effective at monitoring their runs and quickly taking corrective actions when delays occurred and working cooperatively, others, particularly at MV, appeared to have difficulty tracking their runs and identifying and correcting problems.

   15. Slowdown in computer operation was reported to occur frequently by Atlantic Paratrans dispatchers and was observed on Monday afternoon, September 15, 2003. During this time, dispatchers were not able to use the system for other functions.

   16. Communications equipment appeared to be adequate to support dispatch operations.

   go to top of page

2. Transportation

   1. Operator's practice of reviewing and revising schedules before and during the service day appears to positively contribute to service delivery.

   2. Not all drivers understood the difference between the estimated time and the negotiated pick-up times listed on the manifest. This could cause confusion in running their route.

   3. Not all drivers knew the correct pick-up window of -5 minutes/+25 minutes. This could contribute to drivers encouraging passengers to leave early, late pick-ups and miscommunication with customers.

   4. Some drivers made changes to their routes and told the dispatcher after the fact or not at all. This failure to obtain prior dispatcher approval could cause routes to run late, by either inefficient changes or dispatcher having incorrect information on the status of the run.

   5. Operators appear to have sufficient equipment to provide service.

   6. MV Transportation appears to have a high percentage of its passenger equipment assigned to maintenance and repair. In August of 2003, 14 to 29% of MV Transportation's fleet was assigned to maintenance and repair and insufficient vehicles were assigned to meet peak service requirements. Since that time the MV fleet has increased from 100 to 110 vehicles.

   7. Operators appear to have an adequate force of drivers to perform assigned routes.

   go to top of page

3. Performance

   1. For a sample day of May 15, 2003, 94.7% of scheduled trips were completed, 3.3% were Customer No-Shows, 1.2% were categorized as No-Fault No-Shows, 0.4% were Carrier No-Shows and 0.4% were other.

   2. Taxi and black car vouchers were issued for 167 (1.7%) of the scheduled trips for May 15, 2003. Voucher redemptions accounted for 0.3% of scheduled trips for the month of April 2003. Based upon this redemption rate it appears that approximately 20% of the vouchers were used. Because of lag time in redeeming vouchers it is not possible to accurately estimate the utilization rate. Based upon these numbers, 1.4% of trips otherwise classified could be missed trips. These are trips that were not completed by the customer and the vehicle did not arrive within the pick-up window agreed to with the customer.

   3. It appears that more than 50% of the trips categorized as No-Fault No-Shows or 0.2% of scheduled trips may be missed trips. In total missed trips for May 15, 2003, are estimated at 0.9% of scheduled trips including Carrier No-Shows (0.4%), No-Fault No-Shows (0.2%), and unused vouchers 0.3%

   4. Taxi vouchers appear to effectively provide a backup service to prevent customers from being stranded. However, the apparent low utilization of the vouchers by customers (estimated at 20%) and authorization of some vouchers for trips that are already late may indicate that the vouchers have limited effectiveness in reducing the number of missed and late trips. The late authorization for customer use of vouchers, the initial customer outlay of money for the taxi fare and time and effort required for reimbursement may limit the attractiveness of vouchers to customers.

   5. For the sample day of May 15, 2003, 91.8% of pick-ups were on-time or early. This is consistent with on-time performance of 91.5 to 93.8% as reported by NYCT. For May 15 0.6% of pick-ups were more than 35 minutes after the pick-up window or 1 hour after the pick-up time negotiated with the customer. 0.1% of pick-ups were more than 1 hour after the pick-up window or 1 hour and 25 minutes after the negotiated time.

   6. On-time performance internally tracked by MV Transportation indicated some performance issues on weekends. Managers at this provider indicated that because there were fewer vehicles on the road on weekend evenings, the opportunity to reassign trips between runs if in-service problems developed was more limited than on weekdays.

   7. NYCT does not have a performance standard nor does it track performance for on-time drop-offs. On-time drop-off performance is often more critical a concern to consumers than on-time pick-up performance, particularly for trips to appointments for work, school, medical services, etc.

   8. NYCT's goal is to schedule 100% of trips using distance-based travel time standards. The distance parameters overlap potentially causing confusion. NYCT does not have a travel time policy for actual performance of trips and does not measure performance based upon AAR trip travel times.

   9. Based on reconciled trip information for May 15, 2003, the average travel time for all trips was 42 minutes; the overall average trip distance (based on odometer readings) was 9.3 miles. 68% of all trips were 50 minutes or less. Of those, the average travel time was 36 minutes and the average trip distance was 5.8 miles. 81% of all trips were 65 minutes or less and 95% of trips were 95 minutes or less. 23 trips (less than 1%) were more than the maximum travel time standard for any trip set by NYCT (155 minutes or 2 hours and 35 minutes).

   10. The average travel time for sample of 70 long trips from May 15, 2003, was 105 minutes and the average travel distance was 23 miles. The average trip mileage calculated by MapQuest for the sample trips was 13 miles.

   11. The analysis identified approximately six trips from the sample of 70 long trips that appeared to be significantly long considering the trip distance, AAR travel time and comparable fixed route travel times. An additional 16 long AAR trips from the sample that took more than 30 minutes longer than fixed trips by the "fastest route" could also be considered significantly long.

1. Dispatch

   1. NYCT should consider having call takers respond to routine requests without transferring customers to other AAR departments. NYCT should consider having call takers, or unit managers, act as customer agents in resolving issues rather than referring customers to other departments. This practice can reduce the number of incidents of customers making multiple calls to resolve issues.

   2. NYCT should consider instructing control unit dispatchers to avoid the practice of contacting operator dispatchers to check rides that are being performed within the pick-up window.

   3. When vehicles arrive early, operator dispatchers should be instructed to inform the customer that their vehicle has arrived early and is ready for them if they chose to leave early. NYCT might consider providing such guidance in the form of a script for the dispatchers.

   4. NYCT should remind drivers that they are not authorized to classify a customer as a no-show without dispatcher approval.

   5. Operator dispatchers should be instructed to call all no-shows during the pick-up window, including those who may have already received an early call, before classifying the customer as a no-show.

   6. NYCT should consider changing the procedure for early customer requests (made more than an hour before the trip time) for unavoidable changes to same day trips. It is recommended that the trip be cancelled, so the slack time is made available for use by the dispatcher for another trip, and that a will call trip request be created.

   7. NYCT should consider reinstructing dispatchers on the significance of the different trip times used in the scheduling software.

   8. NYCT should encourage MV to consider expanding its staff of dispatchers and providing additional training for its dispatchers. Additional training might also be appropriate for dispatchers at other operators as well.

   9. NYCT should emphasize the importance of clearly communicating incorrect information, such as bad street addresses, incorrect GIS locations, etc., to the responsible department. The staff responsible for correcting errors should follow up with the department or customer who has identified the problem to assure that it has been corrected.

   10. NYCT should take steps to authorize rescue runs more expeditiously to avoid late and missed trips. Centralized dispatch coordination, as proposed by NYCT, should facilitate speedier approval of rescue runs.

   11. It is recommended that NYCT analyze the computer system to identify and eliminate slowdowns or to allow other functions to be performed on the system while run times are being updated.

   go to top of page

2. Transportation

   1. NYCT should consider refresher training for drivers to include: reading and understanding the estimated and negotiated times on the manifests; understanding the pick-up window; and procedures for obtaining authorization for route changes.

   2. NYCT should encourage MV Transportation to increase focus on vehicle maintenance to assure adequate fleet availability to meet peak service needs.

   go to top of page

3. Performance

   1. NYCT should consider increasing its focus on early authorization of "rescue runs," floater vehicles and other flexible capacity to reduce the number of missed trips and late pick-ups.

   2. It is recommended that NYCT review the apparent low on-time performance at certain times on weekends. Floater vehicles, more slack time spread throughout runs, or other solutions to late night weekend performance should be considered.

   3. NYCT should consider establishment of a performance standard for on-time drop-offs and track performance against the standard.

   4. NYCT should revise its travel time standard to eliminate overlapping time categories.

   5. NYCT should consider applying its ride time standard for scheduling, or other revised standard, to performance of service.

   6. NYCT should measure travel time performance and review trips with significantly long AAR travel times when compared to fixed route trips with the same trip characteristics. The analysis should be performed to determine the cause of the long ride times and to identify and implement remedial action to eliminate significantly long trips.

Information about the adequacy of resources available to provide the ADA Complementary Paratransit service as required by the USDOT ADA Regulations was collected and examined as part of the review. This included:

• Input from consumers;
• Information about the vehicle fleet and the availability of vehicles to cover scheduled routes;
• The number of drivers and the availability of drivers to cover scheduled routes; and
• The operating budget for the service and the process used to estimate funding needs.

Following is a summary of observations in each of these areas.

Consumer input, as detailed in Section III, was gathered through interviews with representatives of the disabled community, review of a report provided by Disabled in Action of Metropolitan New York and AAR complaint records for the first 7 months of 2003.

Resource issues identified by representatives of the disabled community and the Disabled in Action Report and a review of complaints on file with AAR were concern about inconsistent vehicle maintenance and a high driver turnover rate that they attributed to low pay rates. One indicated that drivers were insensitive. Otherwise there appeared to be no concerns directly related to resources.

NYCT Managers indicated that the vans in the passenger fleet are configured with five passenger seats and two spaces for wheel chairs. The passenger fleet is owned by NYCT and leased to each of the operators for a fee of $1 per vehicle per month. The equipment is insured by MTA as part of a bulk insurance policy.

NYCT managers indicated that they assign vehicles to operators sufficient to serve peak route assignments plus spare vehicles equal to or exceeding 10% of the number needed to meet peak service requirements. Assignment of vehicles to each of the operators is presented in Table X.1.

A randomly selected weekday, Wednesday, August 20t, 2003, was selected to review equipment availability against equipment needs. Since the size of the AAR fleet has been expended, the fleet size and assignment at the time of the field assessment is also presented in the table. The "% Spares" column represents the percentage of vehicles in the current fleet not needed for service on the sample day. Assuming peak service requirements have not changed significantly from August 20, 2003, to September 12, 2003, the overall percentage of spare vehicles is 23% and ranges from 13% for RJR to 35% for Maggies. In all instances the spares exceed 10% and in all but one instance they exceed 15%. Accordingly the fleet capacity appears adequate to serve current needs and future service growth.

On the sample day MV, PTM and Transit Facilities Management had only one or two vehicles available beyond what was needed for service. PTM had 92% of its fleet available for service on August 20, 2003, indicating a potential shortage in equipment. Subsequent expansion of its fleet by ten vehicles should have adequately addressed this potential shortage. MV and Transit Facilities Management had 86% and 79% of their fleets available for the service day, respectively. The number of vehicles unavailable for service indicates potential equipment or maintenance problems. MV managers indicated that some vehicles reported as unavailable could be placed in service if needed. MV drivers also noted problems with equipment maintenance.

NYCT transportation and maintenance supervisors inspect 20% of the fleet each month during unannounced field surveys. NYCT also reviews pre-trip inspections twice a month and reviews preventative maintenance logs of each of the operators. Ten dollars in liquidated damages is assessed to the operator for each violation of contract requirements for the vehicles.

As discussed in other parts of this report staffing of the call center, scheduling and dispatch functions appears to be adequate and not limit service. Capabilities of dispatchers appeared to vary, which is to be expected in a large organization in which the experience of the dispatchers also varies significantly. In addition, as observed during this review, staff involved in management of AAR service at all levels appeared to be professional, highly experienced and dedicated to provision of public service.

Reviewers analyzed driver tenure based upon drivers who were employed on September 8, 2003. The results are presented in Table X.2.

The lowest tenure rates belong to some of the newer operators. The current contracts have been in effect since August of 2001, approximately two years before September 2003. Maggies, one of the newer operators, had average driver employment of less than one year. Atlantic, a long time contractor, has the highest tenure rates with an average period of driver employment of 3.7 years. Only 10.7 % of its drivers have worked for less than 1 year, while 78.3 % have worked for more than three years.

Each of the operators employs union drivers. TFM, Maggies, Atlantic and MV all have labor contracts with ATU. ATU driver hourly compensation is $11.00 initially, $12.38 after one year and $12.73 after two years. Drivers also receive four paid holidays and one paid sick day and are eligible to participate medical insurance, sharing the premium costs with the operator. PT also reported starting salaries of $11.00 per hour with annual increases and co-paid medical insurance, but no retirement program.

School bus operators (which pay approximately $15 an hour), fixed route operators and other AAR contract operators compete for drivers. TFM operates fixed route services, which draw from their AAR driver pool. TFM reported driver shortages and problems in retaining drivers with an attrition rate of 80% from application through testing and training to two weeks of operation. With a recent training class they reported losing 16 of 20 applicants during drug testing, an additional applicant during training and 2 during the first two weeks of operation. PT indicated that turnover is normal and that recruitment efforts have improved with increases in unemployment. Short periods of tenure for drivers at Maggies and MV Transportation may also be an indication of problems with recruiting and retaining drivers. Given the apparent stability of the labor force of the three long-term operators: American, Atlantic, and RJR, the stability of the newer contractors labor force may be expected to improve over time. This improvement may be accelerated with higher unemployment rates increasing the available work force.

TFM and PT recruit drivers through newspaper ads, TFM also uses bumper stickers and social service agencies. PT also uses dispatchers as back-up drivers when needed.

Table X.3 shows NYCT Paratransit budget and expenditures between 2000 and 2003.

1. Financing of AAR services appears to have strong support from MTA and public officials as evidenced by an increase in the operating budget from $88,860 million for the year 2000 to $153,889 million for 2003. Funding does not appear to limit delivery of ADA Complementary Paratransit services.

2. The NYCT budget process appears to effectively address the anticipated market demand for service through the use of a demographic based model and trends in use of AAR service.

3. The present fleet appears adequate in size to serve current demand with a cushion for increased demand.

4. Staffing of the call center and scheduling appears to be adequate and not limit service. Staff involved in management of AAR service at all levels appeared to be professional, highly experienced and dedicated to provision of public service.

5. MV dispatch staff appeared to be taxed managing assignments of 28 runs per dispatcher. All other dispatch functions appear to be adequately staffed.

6. The availability of drivers for some of the newer carriers may limit service. The stable labor force of long-term carriers indicates that current limitations of labor may be corrected by the passage of time.