Return to the Table of Contents

Paratransit Guide Supplement
Scheduling and Dispatching
by Richard Schultze

NOTE:  This supplemental section provides more in-depth treatment of the material on "Scheduling, Dispatching, & Operations" found in pages 57-62 of "Paratransit for mobility-impaired persons in developing regions: Starting up and scaling up," a guide prepared by Access Exchange International.  The opinions expressed are by the author and readers will want to adapt this material to their local conditions.

Scheduling and dispatching play a vital role in paratransit.  Along with the drivers, the schedulers and dispatchers are often the only point of contact with the customers and public. 

In many cases, the scheduler and dispatcher may be the same person doing both roles.  Also, especially in small systems with a few vehicles, the drivers may be doing their own scheduling and dispatching.

Scheduling’s role in paratransit

The role of a scheduler is to develop the order of pickups and dropoffs, as effectively and efficiently as possible. 

Scheduling requires a strong understanding of the services to be provided, the system’s policies and procedures, a good sense of problem solving, and the ability to interact with customers. The have to be able to balance the customers’ need to get to their destination with the system’s need to maintain high productivity.  And, they have to interact with dispatchers and drivers after the schedule has been created.

A good scheduler has to be an expert problem solver, be detailed oriented, work well under pressure, be flexible and be persistent.

Scheduling paratransit service consists of three major steps: receiving trip requests, developing an efficient schedule, and preparing the vehicle and driver assignments.

Scheduling service is an iterative process, going back and forth between all three steps until the schedule cannot be improved.

For fixed route systems, scheduling is solved once, with fine-tuning over time, and with major service changes occurring every quarter or six months, due to demand changes.

Paratransit scheduling is much more complicated than fixed route service scheduling. The service changes on a daily basis, as a result of demand, and new scheduling solutions must be found every day. What works one day, will very likely have to be modified significantly in order to work the next day. 

Receiving the trip requests

Trip requests may be submitted: directly from the rider; from a family member, friend or caregiver acting on behalf of the rider; or, from a social service agency or government agency acting on behalf the rider.

Trip requests may come in by many methods: in-person, phone call, email, website, app, fax, letter, flag-down/hailing, or a central call taker/brokerage.

The phone (cell or land line) needs phone answering capabilities by which callers can leave messages.  The cell phone needs to receive texts.

Messages, emails, etc. should be checked first thing in the morning, and several times during the day.

The scheduler or call taker needs a routine procedure to ensure that the same, necessary information is taken for each trip request and each person. 

The following information needs to be taken for each specific trip request:

The call-taker should also ask if the rider is flexible regarding trip times or even days. Many riders can be flexible in the times of their random trips (and even subscription trips) and sometimes in the days when the trips need to occur.  Such flexibility allows the schedule more latitude in which to find a slot for the requested trip.  The scheduler should try to work with the requestor to negotiate such flexibility, if needed.  One good example is negotiating a trip requested in the peak period to instead occur in the off-peak, when buses and drivers are more available. Negotiating a trip time and date is acceptable, and doing so can improve productivity, ridership and revenue. 

The call-taker needs to stress that the trip may be shared-ride, where the vehicle is shared by other riders, often strangers to the caller.

The following information needs to be taken for each specific person requesting a trip; this data does not need to be updated with each trip request:

The paratransit system should have in place policies and procedures that state what hours and days of the week the system staff will accept trip requests, when the phones will be answered, how after-hours trip requests will be handled, if a message machine will take messages after hours, how far in advance trip requests are taken, how advance trip requests and same-day trip requests will be dealt with, and if a wait list will be used and how trips are drawn from the list.

The paratransit system should also have in place policies and procedures that state how quickly calls are answered, how long they are placed on hold, how many callers hang up due to lack of response, and how quickly the system responds to messages and requests. Relatively low-cost office technology exists to monitor the call-answering activities.

Subscription and random trips

There are two types of trips: subscription (standing order repetitive trips) and random trips (reservations for one-time trips).

Subscription (standing order) trips

Scheduled trips may be for round-trips, one-way trips, for several days a week, one day a week, etc.  The pattern will vary for each rider. 

Generally, subscription trips are for work, education, worship, or special medical treatments (like dialysis). 

Once the scheduler commits to taking the subscription trips as requested, then that commitment needs to be honored over time.  The acceptance of the subscription trip thus requires that the scheduler give significant thought to the ability to honor those trips over time.

The long term commitment to subscription trips means that the subscription trips will likely form the base for the whole schedule, with random trips placed onto and within the structure created by the subscription trips.

Subscription trip requests are submitted one time, and continue to be valid until they are discontinued, changed or the person is a chronic no-show or cancel, or until they reach a maximum time specified by the system, at which time they have to be renewed to be valid. 

Random (or reservation) trips

Random trips are much more varied and can be submitted individually or as a group. 

Random trip purposes, trip times, trip days, trip frequency, etc. will vary by the person. 

Since random trips are usually for just one round trip, there is no long term commitment required of the scheduler.

The scheduler has a great deal of flexibility in picking and choosing the random trips to develop a full and efficient schedule. 

Random trips should not be handled on a first-come first-served basis, but should be scheduled in a manner that results in the most cost-effective and fullest schedule.

Random trips can be handled many different ways:

  1. accept the trip request right then
  1. deny the trip request right then
  1. encourage the denials to call back on the requested day of service to determine if a slot has become open due to a cancellation or no-show
  2. allow anyone to call in with same-day requests, in the hopes that a slot may become open

The approach one uses in handling random trips can greatly impact the riders’ satisfaction, ridership levels, and the schedule’s cost-effectiveness.

Remember that, for paratransit services, adding riders often means adding vehicle miles, vehicle hours, vehicles and costs, unless the additional riders are in groups, or at or near the same locations as already scheduled pick-ups and drop-offs.

Developing an efficient schedule

Scheduling is not a precise science. It is more of an art that is developed with experience and strengthened with common sense, practicality and a capability for solving puzzles day after day.

In fixed route systems, scheduling is solved essentially once, with fine-tuning over time, or with major service changes resulting from major demand changes. 

Demand responsive service is much more complicated. 

The service changes on a daily basis, as a result of demand, and new scheduling solutions need to be found every day.  What works one day, will very likely have to be modified significantly in order to work the next day. 

Service design criteria dictate how the service is scheduled

The service design criteria will dictate how the service is scheduled. 

The system needs to develop a policy on the order-of-scheduling. This will specify whether trips will be scheduled on a first come/first served basis or selected to provide the most cost-efficient schedule for the day.

The system should have a policy on trip denials, which would determine which trips will be taken and which will be denied, and why.  

If the system is set up in zones, with travel to occur only within a zone, then the zone boundaries will guide the scheduler in linking pick-ups and drop-offs in a sequential order.  In a system that is set up as several feeder areas connected by express or other faster services between timed transfers in each feeder area, the sequence will be scheduled a particular way.

Circuitous routing needs to be minimized.  It is most efficient to go from one event (pick-up or drop-off) to the next in as straight and as fast a manner as possible, in order to reduce costs.  It is also most efficient for the vehicle not to zigzag all over an area, but to progress in some kind of arc from one event to another, thus minimizing miles and hours.

For demand responsive services, adding riders means adding vehicle-miles and vehicle-hours, and therefore adding costs.  Grouping nearby riders and destinations can improve efficiency.

For each pick-up and drop-off, the scheduler needs to be aware of the mobility issues of the rider(s) that will board or alight.  Does the rider require curb-to-curb service or is door-to-door service required?  Is the rider ambulatory or does the rider use a device like a wheelchair?  If ambulatory, is the rider a slow, moderate or fast walker, and is the path to the vehicle short or long, smooth or rutted?  If a wheelchair user, is it manual or electric, and will the user require assistance getting to the vehicle? 

The scheduler will have to make an informed estimate of the loading or unloading time, including the time required to access the vehicle, for each disabled and elderly rider.  Over time, this estimate will be fine-tuned based on actual experience with the rider.

Some systems may use attendants to accompany the drivers on routes that have a high number of special needs riders, or where there are known behavior concerns, particularly among cognitively and emotionally disabled riders.  The attendants help the driver concentrate on driving, and assist or take the primary role in loading and unloading passengers.

The riders should be given a time when they can expect the service to be there for them.  However, they also need to be instructed that there is a window, typically, in some countries, of +/- 15 minutes when the rider needs to be ready to be picked up. 

For example, if the scheduled pickup time is 10AM, then the rider should be ready by 9:45AM and be on the lookout for the vehicle as late as 10:15AM.  The vehicle may come before the scheduled 10AM time because one or more trips were no-shows, allowing the driver to come early. The vehicle may come after the scheduled 10AM time due to traffic delays, or delays due to riders boarding slowly.  This approach applies to pickups from the home as well as pickups at the destinations.

As much as possible, the scheduler should attempt to group riders.  Each system, however, will need to develop a policy on how long a rider will be on the system each way.  It is possible to increase ridership, up to a point, by keeping riders on for quite some time, and going to many locations to pick them up.  However, after more than 90 minutes on the system, the trips become onerous and undesirable.  

In many countries, traffic congestion will seriously slow down service.  The scheduler needs a realistic estimate of travel times throughout the service area.  The scheduler needs to rely on feedback from the dispatcher, drivers and data on late trips.  It is counterproductive to underestimate travel times, since that will cause riders to be very late and will undermine the confidence in the system and ridership loyalty.

If the system uses timed transfers between services or routes, then the scheduler needs to ensure that there is the needed coordination of arrival and departure schedule times at the timed transfer stop or terminal.

Along the routes and services, the scheduler needs to have time points.  A time point is where the scheduler assures the rider that there bus will be there, within a certain time window.  On pure demand responsive service, the time point is the scheduled time for a pick-up or drop-off. 

On a flexible route or with route deviation, there are time points along the route, and for deviations.  The bus is not supposed to pass the time point until the scheduled time; otherwise riders will miss the vehicle.  If the vehicle arrives at the time point early, it should hold at that location until the scheduled time.

Flex/route deviation services are scheduled as fixed routes, with time points and specific routings that can be clearly shown in a timetable.  The deviations are to be considered as subscription or random trips, and need to be accounted for just as in traditional demand responsive services. 

The flex/route deviation services will have extra running time (in the form of schedule recovery time) built into the schedule, such that some deviations can be inserted into a day’s schedule without too much disruption.  The service may run late for a while after a deviation, but there should be enough schedule recovery time to get the vehicle back on schedule before too long.

In some countries, there is not adequate signage for road names or addresses.  The scheduler will need to provide the driver with whatever directions or guides as needed so that the driver will not get lost.

In most DRT (demand-response transportation) systems, for pure demand responsive service, the scheduler does not provide the driver with precise paths to follow between events (pick-ups and drop-offs).  These precise directions are called “right/left directions”.  This would be much too time consuming for the scheduler. And, it is assumed that the drivers are familiar enough with their service area, and have access to maps or on-board GPS devices, such that they can determine a relatively efficient and quick path.  Sometimes, drivers do get confused or lost, so the dispatcher will have to give directions.

The system should have a ride time policy, that specifies the maximum amount of time a rider will be forced to ride, and how riders will be grouped and share rides.

The system should have a policy, that is clearly communicated to the riders, about how it will deal with return trips that are delayed (such as by a doctor’s appointment that is running late, or when a person has to stay late at work). The transit system should be notified of the late pickup.  The system will try to reschedule its trips to send a vehicle back at the updated pickup time.  This is often called a “will-call” trip.

The system should have a policy on no-shows and cancellations. These are a waste of the system’s money and disrupt the service for other riders. No-shows are when the person is not there when the vehicle arrives, or when the person waves off the van or bus and says he/she does not need the ride, or when a person cancels within a certain amount of time (often 1-2 hours) before the scheduled pick-up time.

The system needs to be strict with no-shows. After 3 no-shows in a 3-month period, some systems prohibit the rider from using the service for 3 months.  A  person that frequently cancels should be warned to stop that behavior or lose the right to use the service. 

Overbooking is not recommended as a way to work around excessive no-shows and cancellations.

Same day call-ins and insertions can improve the system’s cost-effectiveness, increase ridership and revenues, and keep the customers happy.  The same-day insertions may come from a wait list to fill gaps that occur due to cancellations and no-shows, or they may be made to try to better group passengers in a trip, for example serving nearby pickups or dropoffs. 

Care should be taken with same day insertions because a trip that is inserted without accurately taking into account the effect on other scheduled trips can seriously disrupt an otherwise good schedule.

Manual vs. automated scheduling

The scheduler needs a capacity spread sheet approach that keeps track of all scheduled trips, both subscription and random, for the demand responsive service.

There are several manual or manual/computer assisted approaches:

How do you determine which approach and technology to use? Contact a peer that you know. Do an Internet search. Contact some vendors. Use on-line product testing, if possible. Go for a site visit to see how it actually works. Ask a vendor to come demonstrate the software in person or to demonstrate on your system remotely (from their company site).

Be skeptical. No system is perfect. Some work, others don’t work so well. Cost it out. Where would the funds come from? Computerized systems can be expensive to build and may be hard to maintain and troubleshoot if you are far from the vendor.

It is beneficial to have software that covers a wide spectrum of needs, including rider information, taking trip requests, scheduling, verification of data, dispatching, reporting, invoicing, and interfacing with GPS, mobile data terminals/tablets, and security cameras.

In one often-used approach, the manual and manual/computerized spreadsheet approaches need to show each day, typically for the next 2 months.  Each day may be divided up into ½ hour slots.  Each slot shows how many trip requests can be accommodated during that slot.  This capacity is a function of how many vehicles are usually on the road during that time slot, and the scheduler’s historical understanding of how many trips can be handled per vehicle in each slot. 

Since more short trips can be handled than long trips, the long trips are weighted more than the short trips, and tend to take up more of the capacity of a given slot.  This use of the slots on the spreadsheets allows the scheduler to immediately look at random trip requests and determine if they can or cannot be handled.  The capacity of each slot is adjusted over time as more experience is gained with the approach. 

These approaches need to first show all the subscription trips for the next 2 months, since the subscription trips form the backbone of the service.  At the end of each month, the spreadsheet adds another sheet for the next month, and in this rolling manner, the subscription trips are kept in view.  Every day, as many random trips as possible are fit into the slots. 

Once the scheduler decides which trip requests can be carried for the next day, and all the slots are full or accounted for, then the scheduler goes from the capacity sheet to the actual schedule.

By the middle of the day, the schedule for the next day should be complete.  The day before the weekend, the scheduler should also prepare the weekend schedules. In addition, the skeleton of the schedules for the next 3 days should be in place.

Keeping logs

In order to improve the scheduler’s accuracy, and to deal with questions or complaints from the riders, the scheduler should keep a log.  The log should include: all trip requests, when received, the requestor, where they wanted to go and when, whether the requests were accepted or denied or put on a wait list, and any communications with the riders.

It is essential that the scheduler have a system to keep track of all the trip requests and passenger information, both those that are to be carried, and those that are denied or on a wait list. 

The time and date of all discussions with the rider should be noted in case there is some dispute or miscommunication.  This can occur frequently and the scheduler needs the backup data. 

A list of trip requests that have been denied can be an important indicator of latent (unmet) demand for the service and can be used as a tool to add or modify the services. 

Scheduling the vehicle and driver assignments

Before the schedule for a day is considered complete, the scheduler must precisely show what each driver is to do and when, from the beginning of the workday to the end.  In demand responsive service, the schedules are not usually made more than one day in advance, so it is not possible to have definite groupings of days into weekly work rosters.  However, it is possible for management to estimate its needs for vehicles and drivers for several days into the future, and to request that vehicles and drivers be available until specific schedules are made. 

Vehicle assignments and driver work assignments (called runs) for a day can be easily listed on a chalkboard, posted on a bulletin board, developed by hand in a Word document or Excel spreadsheet, or generated by a computerized scheduling system. 

A precise schedule for each vehicle and each driver needs to be developed, in order that the vehicle and driver resources can be made available each day, and so that each driver knows exactly where they and their vehicle is to be at any one time.

The vehicle and driver assignments should be assembled to minimize costs and vehicles used, provide for a reasonable work assignment, reduce overtime, have the right capacity vehicle available, and reduce deadhead.

The dispatcher needs to communicate the vehicle and driver assignments to the drivers and vehicle maintenance staff.

Fueling

Ideally the fuel depot and the vehicle storage and maintenance area (i.e., garage) are in one spot.  A vehicle would be fueled at the garage either before it pulls out for its daily run, or at the end of the day after all service is completed. 

If the fuel depot and the garage are at separate locations, then the scheduler needs to show precisely when the driver is take the vehicle to be fueled during the day. 

There should be no passengers on board during fueling and no smoking should be allowed during fueling. 

Fueling could be combined with a scheduled break. 

It is inexcusable for a vehicle to run out of fuel in the field, unless there is a mechanical failure.

Deadheading

Deadhead time (or non-revenue miles and hours) needs to be minimized. 

Deadhead occurs between leaving the garage and arriving at the first pickup, as well as between pickups when no passengers are on board. Deadhead also occurs between the last drop-off and arriving back at the garage. 

The schedules should minimize the deadhead from vehicles coming back to the garage so the drivers can take breaks. 

The schedules should also minimize the deadhead that occurs when a bus runs low on fuel and has to come back to the garage early, and then go back out into service.

Drivers

The dispatchers and the drivers have to work as a team. 

It is the driver’s very difficult role to be obsessed with quality customer service and schedule adherence, while working with the dispatcher and the riders to adjust to the ever-changing conditions. 

It takes a special person to be a driver.

Drivers need breaks. They have to go to the restroom, eat meals, stretch, and just get out from behind the wheel and rest.  These need to be scheduled and enforced by the dispatcher.  They can take breaks (or switch drivers) at the garage or in the field.

All transportation systems need to have in place a set of policies and procedures that specify driver work rules.  Sometimes this may be the result of collective bargaining.  Always, they will reflect the labor regulations (if any) in place.  And, in well managed systems, they will reflect what has been found to be practical and fair in the particular local situations.

These rules need to deal with, at a minimum:

Dispatching’s role in paratransit

The dispatcher and the scheduler may be the same or different persons.

It does take a special person to be a dispatcher. 

Dispatchers need to be able to quickly and efficiently multi-task, while keeping composed and calm, work well under pressure in fast-paced hectic environments, be excellent communicators and decision makers

Dispatchers manage and troubleshoot the delivery of trips real-time.  They control the service, except in systems where the drivers have the responsibility of dealing with the riders themselves and determining their own real-time schedules.

Dispatchers should be obsessed with keeping the operations going efficiently and as closely as possible to the original schedule. 

Dispatchers need to be well-informed of the real-time status and location of every scheduled vehicle, so drivers need to frequently report their locations, and report pickups and dropoffs when they occur. 

They need to have the authority and responsibility to make immediate changes in the service in order to try to preserve and maintain the schedule. 

Drivers rely on dispatchers for information they need during the day.  Dispatchers rely on drivers for vital information as to the status of the service.

Dispatchers need to assist drivers with directions; confirm no-shows, trip refusals, and cancellations and allow the driver to move on to the next passenger; confirm and coordinate passenger information; send out maintenance and standby vehicles; provide support to a driver during an incident or accident; and advise drivers of traffic conditions and detours and, where needed, the status of other vehicles in the system.

As soon as efficient schedules are prepared, and begin to be put into operation, they begin to unravel. 

It is often nobody’s fault. Life just happens.  It is the dispatcher’s duty to keep the service flowing as efficiently as possible, in spite of disruptions. 

The dispatcher also has a daily rhythm of duties.  These duties may vary, especially with the size of the system.

The afternoon or evening before next day’s service, the dispatcher:

The morning of service, the dispatcher:

Throughout the service day, the dispatcher:

At the end of the service day, the dispatcher:

Other technology

The most common tool that the dispatcher uses to obtain the real-time location and schedule adherence of the buses is by talking to the driver on the cell phone/radio. 

This works well if the driver is accurate about the location and the status of any problems, if the driver is honest about the location and problems, if the cell phones/radios work in an area, etc.

If there is a question about the driver, or a pattern of lateness or misinformation, then the dispatcher can have a street supervisor, if available, monitor the behavior.

Phones can also be used by the driver to take trip requests directly, and communicate with the potential rider about pick up times, status of the vehicle, fares, etc., thus bypassing the dispatcher. 

This has advantages and disadvantages, depending on the size of the fleet, the need for top-down supervision of the operations, or the desire to respond in real-time to customers and let the market-driven actions of individual operators attempt to achieve efficiencies and spontaneity that may not be possible by top-down control.

The more modern and powerful smart phones and tablets can incorporate all of these features as well as provide maps, routings, apps, data bases, client lists, etc. The driver can be in touch with the customers, the dispatchers, or both.

Vehicles can be equipped with GPS to inform the dispatcher (and the driver) in real-time about the exact location, speed and direction of each vehicle.  This is an ideal tool, but is somewhat costly to install and to maintain. Costs are coming down dramatically since many cell phones, smart phones and tablets have GPS in them.

Tablets (and the more expensive mobile data terminals) can be placed with each vehicle or each driver, and also allow the dispatcher to communicate in writing to the driver, to give new instructions, transmit maps, etc, and to allow the driver to communicate with the dispatcher.  It also allows the driver to punch in all pick-ups and drop-offs, recording the location, time, mileage from last stop, fares, etc.  This dramatically improves the amount and accuracy of service data. 

The service data is important for good record keeping, productivity review, accounting, and billing.

The transportation system needs to have in place clear and enforceable policies and procedures for how the drivers and dispatchers are to use their communication devices, in order to reduce traffic safety issues, have the desired approach to scheduling/dispatching, minimize the inappropriate use of the vehicles, and account for fares and ridership.

Participating in community emergency responses

Paratransit systems that are accessible to the disabled and the elderly are sensitive to the needs of these special populations. The drivers, schedulers and dispatchers know where the passengers with special needs live and how best to transport them. 

If there is a city wide, community or rural emergency, due to bad weather, earthquakes, flooding, riots, or other issues, the special needs populations may need to be evacuated out of the area, or moved to special facilities for their protection. 

The paratransit systems should be ready to lend their assistance to this important community response.

Date of this supplemental material: June, 2013

References 

All of these references can be downloaded from the Internet for free.

Basic references:

More complex references:

The TCRP documents can be found at: www.trb.org, then to publications, transit (TCRP).  TCRP stands for the Transit Cooperative Research Project.

 

(This Paratransit Guide Supplement added in November, 2013)

Return to the Table of Contents