Public Transit Technologies > Personalized Public Transit       Printer-friendly version

• Introduction
• Technologies Used in PPT
• Dedicated-Ride Services
  • Taxis
  • Personal Rapid Transit
• Shared-Ride Services
  • Shuttle Buses/Paratransit
  • Traditional transit with route deviation service
• Related Service
  • Carsharing
• Assessment
  • Travel Impacts
  • Benefits
  • Costs
• Case Studies
• References

INTRODUCTION

Personalized Public Transit (PPT) or demand responsive transit serves pick-up passengers on a demand or as-needed basis. In recent years PPT has been greatly enhanced by the implementation of advanced technologies for the dispatching and routing of vehicles. Typically PPT involves the application of two technologies: advanced vehicle location (AVL) and computer-aided dispatch (CAD).

Personalized Public Transit may be deployed where fixed-route fixed-schedule transit services are not cost-efficient:

• In areas with low passenger density, such as suburban neighborhoods and rural areas.
• For people with mobility impairments: the 1990 American with Disabilities Act requires that transit agencies operating fixed-route systems should provide "paratransit and other special transportation services to individuals with disabilities at a level of service which is comparable to the level of designated public transportation services provided to individuals without disabilities .

AVL and CAD have the potential to increase the efficiency of paratransit services, and offer advantages to both operators and users. Operators may consolidate existing demand in fewer vehicles, serve a larger market, and reduce labor costs. Users could enjoy reduced advanced reservation times, reduced waiting times, and faster travel times. Fewer users would be denied rides too.

Personalized public transportation uses all kinds of transit and paratransit vehicles: buses, taxis, shared-ride vans, etc. Vehicles are dispatched on demand, and they typically provide door-to-door service, or at least a short deviation from a fixed-route. Examples of personalized public transportation are:

• Taxis
• Dial-a-Ride Services
• Buses that operate on a fixed schedule but not a fixed route.

TECHNOLOGIES USED IN PUBLIC TRANSIT

Computer-Aided Dispatch Systems (CAD)
Also known as dynamic scheduling systems, this software automates the process of assigning ride requests to vehicles. Depending on the operation, the objective may be to minimize passengers' wait time, subject to a minimum vehicle load; or conversely to maximize the vehicle load subject to a maximum wait time. Another objective may be to reduce the advance notice needed to request a ride. Some services require at least 24 hours notice, while others may be able to provide same day service. The CAD system tracks all ride requests (origin, destination, time of pick up, number of passengers), assigns riders to the available vehicles while simultaneously planning vehicles' routes, and dispatches them. Ideally, Personalized Public Transit should be available on demand in real-time, but this level of service requires Automatic Vehicle Location.

Automatic Vehicle Location (AVL)
AVL lets dispatchers know where en-route vehicles are located, so that they can pick up passengers whose ride requests were received after the vehicle left the base. This is likely to reduce waiting times and increase the number of passengers served per vehicle-mile, but it could also increase the ride time for those passengers already in the vehicle. When vehicles are equipped with AVL, the CAD system would have to be able to evaluate whether it is advantageous to assign the new ride request to a vehicle enroute, instead of to an undispatched vehicle. More on AVL.

DEDICATED RIDE VEHICLES

Taxis

Dedicated ride vehicles are the simplest case of demand responsive transit; and the most common and almost exclusive form of dedicated ride vehicle is the taxi. Because there is no ride-sharing, passengers are assigned to vehicles based exclusively on the location of the available vehicles relative to the origin of the trip.

The most widely used computer technology for taxis works as follows:

Each taxi is linked to the dispatch office by a data link that can send and receive information. The dispatch office is equipped with a geographic information system database that covers the service area. The service area is divided up into zones, typically about 4 miles square (larger for lower demand areas and smaller for higher demand areas).

Drivers notify dispatchers of their zonal location via the data link. The CAD system adds the taxi identifier to the end of the queue for that zone. When a request for service is received at the dispatch center, the first taxi in the queue for that zone is queried through the data link. If the driver rejects the offer, the trip is offered to the next driver in the queue, and the original driver remains in the queue. If the driver accepts the offer, he/she is removed from the queue and receives detailed trip information (via datalink) up to and including map coordinates.

The dispatch computer keeps a record of the entire transaction: it records the time the trip was accepted, when the taximeter was turned on, and when it was turned off. A typical system has a backup voice communication system in case the dispatch computer or datalink break down (Assessment of Computer Dispatch Technology in the Paratransit Industry, 1992).

The system has some advantages over a conventional dispatch operation:

• It provides a quick, efficient method of assigning taxis to riders
• It increases the efficiency of dispatchers and taxis (the average pickup time may be reduce up to one-half of the pre-CAD pickup time - See Table below).
• It eliminates favoritism and kickbacks between drivers and dispatchers.
• It keeps records of all transactions, which allows detection of anomalies in taxi behavior. For example it can tell when a particular taxi takes longer than average to arrive at the pickup point, or when there have been irregularities in fare collection.

However, the system can be abused. Drivers can claim to be in zones in which they actually are not. This can be detected if they show long times between accepting a rider and picking the rider up, provided that the dispatcher is alert. Another disadvantage is that the zone concept does not ensure that the closest taxi will make the pickup.

An automated vehicle location (AVL) system overcomes these disadvantages by determining the location of taxis, thus removing the zone concept altogether. Drivers can not easily falsify their position, and the closest taxi is always the first to be offered a rider. A study by the U.S. Department of Transportation (1992) estimated that AVL could reduce average passenger wait time for a taxi to 5 minutes, from the 10 to 15 minutes that it takes for a non-AVL system. There were no AVL-equipped taxi companies at the time of the study, and perhaps more importantly, there was no consensus in the industry as to whether AVL would improve productivity or reduce costs.

Industry estimates of the cost of implementing CAD in taxis range between $2,000 and $2,500 per vehicle (Stone et al, 1992), depending on fleet size, system components, and training and maintenance fees. But taxi companies report higher costs, typically in the range of $1,500 to about $5,000 per vehicle (see Table 1). CAD costs are likely to be quite lower now as the reported cost data is from 1992.

Cost savings associated with CAD systems arise from savings in labor costs, and from higher vehicle productivity. The automated systems allow to substitute experienced dispatchers with call-takers. The latter require far less training and typically command lower wages than a dispatcher. Hence, the taxi company may be able to increase the volume of operations without hiring an additional dispatcher, or keep their current volume with less dispatchers.

Link to Table 1

Personal Rapid Transit

Personal Rapid Transit (PRT) is also based on dedicated ride vehicles: single or double occupant vehicles operate on an automatically controlled, dedicated guideway. Small rail or monorail transit projects are often considered to be PRT too. This technology is currently in development, either at the planning or prototype stages. A PRT system has been implemented in a few small shopping centers in the UK (Anderson, 1998). An overview of PRT systems can be found in the Journal of Advanced Transportation, Volume 32, Number 1, (Spring 1998). For an informative on-line forum on PRT visit the PRT Debate page hosted by the University of Washigton's Innovative Transportation Technologies web site. PRT links.

SHARED RIDE SERVICES

There are two types of shared ride service: door-to-door service, which uses small buses, vans, or even shared-ride taxis; and route-deviation service, where fixed-schedule buses are allowed to go a few blocks off their route to pick up or drop off passengers.

Shuttle Buses/Paratransit

Door-to-door service is similar to standard taxi service, except that travelers may share all or part of their trip with unrelated passengers. CAD, and sometimes AVL, are used to plan vehicle routing and dispatching; the former is practically essential for companies that operate a large fleet.

An example of this type of service is provided by OUTREACH Inc. in Santa Clara County, California. OUTREACH operates as a broker of rides, that is, as an intermediary between riders and private taxi and van operators. OUTREACH books and schedules ride requests, and contracts with private providers to pick up and deliver passengers. In 1995, OUTREACH implemented a digital geographic database (DGD) and an automated trip scheduling service (ATSS). A year later 40 vans (not the entire fleet) were equipped with an AVL system.

The ATSS is supplied with the trip origin and destination of all ride requests and automatically assigns the trips to specific vehicles. It then generates vehicle trip logs, which are lists of passengers per vehicle in the order in which they will be picked-up. The system selects vehicles based on street distances calculated from the exact latitude and longitude of the origin and destination. Trip logs are communicated online to taxi and van operators. With the addition of AVL, vans enroute are also considered when selecting vehicles to serve recent ride requests.

The DGD and ATSS were found to increase the proportion of shared rides from 38% to 55%, which in turn resulted in 13% savings in transportation cost per passenger mile (Chira-Chavala and Venter, 1997). Personnel salaries per passenger mile decreased by 28%. The increase in ridesharing did not change travel times or the perceived level of in-vehicle comfort. The new systems did require new personnel skills: the staff of the paratransit agency must now include at least one computer systems analyst, with expertise on both hardware and software. The study did not include the effects of AVL technology on the systems operation.

• Madison Co., IL: ridership increased from 1.8 passengers per vehicle to 2.2 passengers per vehicle; the cut-off time for a next-day trip request was delayed from 2:30 PM to 5:00 PM.
• Golden Gate Empire Transit, Bakersfield, CA: vehicle trip length and travel time were reduced by 10%.
• Antioch, CA: ridership increased by 40 passengers per day; trip denials were reduced from an average of 2.2 per day to 1.2 per day.
• San Diego Transit, Indigo, CA: ridership increased by 2% due to the ability to schedule passengers "on the fly".

In Dakota County, (in the Minneapolis-St. Paul area), a program called Dakota Area Resources and Transportation for Seniors (DARTS) is implementing a door-to-door paratransit service which uses a Management Information System to keep track of clients, CAD, wireless communication and AVL for scheduling and dispatching, and smart card technology for fare payment.

Industry estimates of the cost of CAD vary widely depending on the manufacturer, system automation, and training and maintenance fees, as shown in Table 2. The cost per vehicle varies from about $600 for a bare-bones system to $4,000 for a fully automated scheduler that includes installation, maintenance and technical support.

Source: Kihl et al, 1996.

AVL costs also vary widely. For a 15-vehicle property, base station costs range from $20,500 to $54,000; in-vehicle costs vary from $3,250 to $5,500 per vehicle, and other costs such as training and network charges could cost as much as $10,000 per year (Kihl et al., 1996). CAD and AVL costs are likely to be quite lower now as the reported cost data is from 1996.

Share-ride operations are more expensive and more complex to run than taxi services, as indicated by the statistics in Table 3. The ratio of dispatchers and call-takers per vehicle is considerably higher, and they typically serve a lower demand. But the cost of deploying CAD in share-ride operations is comparable to the cost of deploying CAD on a taxi fleet, on a per trip basis.

¹ Before CAD deployment.
² After CAD deployment.
³ Unknown / Unavailable.

Source: Stone et al, 1992.

Traditional Transit with Route-Deviation Service

Route Deviation Systems are a hybrid between door-to-door paratransit and standard fixed-route fixed-schedule bus services. In a typical operation, the vehicle arrives at designated bus stops on a planned schedule, but in between these stops, the bus may deviate from its usual route to pick up or drop off passengers. Advanced technologies can help shorten the lead-time necessary to schedule a bus pick-up or drop-off, and also inform passengers about the arrival time of the next bus.

In theory, route-deviation operations are best suited to medium density areas, with fixed routes better suited to higher density areas and door-to-door paratransit more effective in low density areas (Daganzo, 1984). Route-deviation systems have two advantages over these other types of public transportation: they reduce walking distances between home and the bus stop, increasing its attractiveness over fixed-route services; and they satisfy ADA requirements at a lower cost per passenger mile than door-to-door paratransit.

Ruf-Bus (Friedrichschafen, Germany) is a bus transit operation that has implemented a route-deviations operation. Users could make trip request using kiosks located at suburban rail stations. The request consisted of identifying the station to which the user wished to travel. The origin (current station) and destination of the requested trip were sent to a central computer, which determined the number of the next bus and its expected arrival time. Users were prompted to accept or reject the trip. If the trip was accepted, the user received a confirmation ticket, and a bus was assigned the pickup. Buses had computer terminals on board, and after each pickup or drop-off, the terminal would display the next destination (Cervero, 1997).

The Ruf-Bus project increased ridership over the fixed-line service it replaced, but lost money due to higher operating costs. As a result, in the early 1980s, the system was turned into an operation called Flexible Operations Command and Control Service (FOCCS). FOCCS operated in three modes. During peak hours FOCCS operated a standard fixed route service, offering route-deviation service during non-peak times. On weekends and in the evenings it offered curb-to-curb service. The result of these changes was an overall drop in the level of service and operating costs. After 1987, attempts were made to use GPS in the system as well as other technological innovations, for schedule adherence. As of 1997 the system continued having higher costs than revenues (Cervero, 1997).

Omnilink, a service of the Potomac and Rappahannock Transportation Commission (PRTC), is one of the better-documented examples of the route deviation system. It serves areas in eastern Prince William County, in Northern Virginia. The service covers six routes with a total of 12 vehicles on 45-minute headways. Overtime, although ridership has increased, calls for route deviations have decreased: as people became familiar with the regular route and schedule, they found it more convenient to walk to the nearest bus stop. The demand-responsive system is estimated to yield 40% cost savings, when compared to a fixed-route fixed-schedule system complemented with door-to-door paratransit. It was also found that the modus operandi of the system was not effectively communicated to current and potential users (Farwell, 1996). In 1996, there were plans to implement AVL and CAD systems.

Another route-deviation bus scheme has been implemented by Tri-Met, in Portland, Oregon. The privately operated demand-responsive service costs approximately $7.37 per passenger trip, while it was estimated that public operation of a similar service would cost $14.61 per passenger trip due to higher labor costs. Fixed-route service for the same area would cost $17.82 per passengers trip. It was found that the demand-responsive transit serves a higher proportion of non-work trips than is typical in a fixed-route operation (Rufolo et al, 1996).

RELATED SERVICES

Carsharing

Carsharing can be thought of as organized short-term car rental.

Members of a carsharing organization access the vehicles from shared-use lots (e.g., transit station, neighborhoods, and employment centers). Fees typically cover maintenance, insurance, registration, fueling, and time use.

Urban carsharing with linkages to transit is the predominant shared-use vehicle model. However, several new models are being tested (e.g., employer-based carsharing, such as CarLink - CarLink Evaluation report; resorts; gated communities; etc.).

Existing carsharing organizations typically provide a choice of vehicle type, rate, and convenience suited to participant needs. More on carsharing

ASSESSMENT

New technologies, primarily CAD and AVL, have greatly increased the efficiency of PPT and offered advantages to both operators and users. Operators can consolidate existing demand in fewer vehicles, serve a larger market, and reduce labor costs. Users typically enjoy reduced advanced reservation times, reduced waiting times, and faster travel times. Fewer users are denied rides too. The chief implementation challenge for these technologies is cost.

Travel Impacts

To the extent that new technologies increase PPT ridership and enhance vehicle productivity, they lead to the following travel impacts:

• Reduction of total traffic
• Reduction of peak period traffic
• Shift from automobile travel to transit and paratransit

Benefits

• Increased ridership.
• Reduced waiting times.
• Reduced advanced reservation time and accomodation of last-minute requests.
• Faster travel times.
• Enhanced schedule control and quick identification of slack times.
• Labor cost savings: the CAD systems allow higher volumes with the same number of dispatchers or less.

• Equipment costs
• Labor: Setting-up of technologies, time required to learn new systems and maintenance by a system administrator.

CASE STUDIES

University of Newcastle's "Phone and Go" Service
The University of Newcastle upon Tyne's Transport Operations Research Group (TORG) and Northumberland County Council are evaluating the Phone and Go service. This service is similar to Dial a Ride schemes, except that it is available to everyone. Passengers using this service phone the Travel Dispatch Centre (TDC) in advance of their trip; a dispatcher uses computer software to book the trip and determine the most efficient route. This route information is then sent to the Phone and Go bus using GSM text messaging (similar to wireless technology) and the bus--taking into account the other booking requests--embarks to pick up the passenger. Thus far, mostly pensioners and mothers with young people are using the Phone and Go service, but the project leaders are hoping this service can help alleviate the transit problems of this rural area.

Santa Clara Valley Paratransit
OUTREACH, a private, non-profit corporation, provides scheduling and vehicles to the Santa Clara Valley Transportation Authority (VTA). OUTREACH uses an automated trip scheduling system (ATSS) and an automated vehicle location (AVL) system to enhance user services. The ATSS books client rides, assigns these rides to vehicles, and informs clients of pick-up times. The AVL system monitors the real-time status of vehicle and passenger trips. Short-term and long-term scenarios describe the benefits achieved by using this system.

SMART
The Suburban Mobility Authority for Regional Transportation (SMART) operates a paratransit service, called Community Transit, throughout Macomb County and in portions of Wayne and Oakland counties in Southeast Michigan. Community Transit offers both same day service and advance reservation service, i.e., request for trips other than for the day the call was placed. For advance reservation trips, customers request service by placing a phone call to a SMART Customer Service Operator (CSO). During 1995 and 1996, SMART implemented the Quo Vadis (QV) automated scheduling and dispatch system for the Community Transit service. Quo Vadis includes a user-friendly software interface to help CSOs accommodate customer advance reservation trip requests. An evaluation was conducted to determine the effects of Quo Vadis 1) on the productivity of CSOs (primarily in terms of the average time required to serve a customer) and 2) on the objective quality of the reservation service (in terms of shorter wait for service and faster response when being served).

Shared Taxi Services
Trudel (1999) describes how shared and subsidized taxi services provide affordable mobility in rural areas. Freund (2000) describes a demand-response service that provides mobility for elderly residents.

Andreasson I. Survey of R&D in PRT Systems. Journal of Advanced Transportation. Vol. 32, No. 1, pp. 23-34, Spring 1998.

Benson, Jeffrey L. Paratransit and the Promise of Advanced Technology: The Smart DARTS case study. Compendium of Technical Papers. (Institute of Transportation Engineers 63 Annual Meeting September 19-22, 1993).

Cervero R. Paratransit in America: Redefining mass transportation. West Port, Connecticut: Praeger Publishers, 1997.

Cervero R. Commercial Paratransit in the United States: Service Options, Markets and Performance, University of California Transportation Center (Berkeley; (http://socrates.berkeley.edu/~uctc), Working Paper #299, Jan. 1996.

Chira-Chavala, T. and C. Venter. Cost and Productivity Impacts of a "Smart" Paratransit System. Presented at the Annual Meeting of the Transportation Research Board, Washington D.C. Paper No. 97-13-66.

Daganzo C.F. Checkpoint Dial-a-Ride Systems. Transportation Research-B. Vol. 18B, No. 4/5, Aug/Oct 1984, p. 325-327.

Farwell, R.G. Evaluation of Omnilink Demand Driven Transportation Operations. Proceedings of Seminar F: Public Transportation Planning and Operations. European Transit Forum: 2-6 September, 1996.

Kihl M, M. Crum and D. Shinn. Linking Real-Time and Location in Scheduling Demand Responsive Transit. Final Report. Ames, Iowa: Midwest Transportation Center, Iowa State University, 1996.

Klein, Daniel, Adrian Moore, Binyam Reja, Free to Cruise: Creating Curb Space for Jitneys, Access, No. 8 (http://socrates.berkeley.edu/~uctc), Spring 1996, pp. 2-6.

Rufolo A.M, J.G. Stratham and Z. Peng. Cost effectiveness of Demand Responsive Versus Fixed Route Transit Service: Portland, Oregon Case Study. Presented at the Annual Meeting of the Transportation Research Board, Washington D.C. Paper No. 97-1185.

Stone J.R, G. Gilbert, and A. Nalevanko. Assessment of Computer Dispatch Technology in the Paratransit Industry. Washington D.C: Federal Transit Administration Office of Technical Assistance and Safety, U.S. Department of Transportation, March 1992.

Michel Trudel, The Taxi as a Transit Mode, Transportation Quarterly, Vol. 53, No. 4, Fall 1999, pp. 121-130.

Victoria Transport Policy Institute, Shuttle Buses, from the Travel Demand Encyclopedia, August 2000.

Wilkins V. OmniLink: A Ride for All Seasons. Bus World. Vol. 18, No. 1, Fall 1995.

Author: Glenn Blackwelder and Dimitri Loukakos. Last updated 11/07/00.