Below is the uncorrected machine-read text of this chapter, intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text of each book. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.
51  Conclusions Headway-based operation is a way of operating a bus route where the main objective is to keep buses spaced at equal time intervals (e.g., every 10 minutes). In the academic literature, headway- based operation has long been recognized as a tool to address various operational problems that are common in high-frequency bus service, including schedule adherence, bunching and gapping, and buses reaching full capacity. Interest in headway-based operation has been increasing in recent years because technological improvements have made it easier to track the positions of buses in real time and to communicate with bus drivers. Knowledge of bus positions and some form of communication with the drivers are the key technological requirements for operating a bus route by headway and are typically provided by a CAD/AVL system. There are many other existing and proposed uses of ITS technologies, involving APC, EFP, PIS, SS, smartphone apps, and TSP. Operating a route by headway involves both monitoring and control. High-frequency bus routes (the main use case for headway-based operation) are known to be unstable. Many dif- ferent perturbations can put a bus ahead or behind its target headway, including traffic inter- actions, signals, and more or fewer passengers than expected at a stop. At high frequencies, interactions between buses create a positive feedback loop where the small headway deviations created by these perturbations often do not self-correct and tend to grow into larger problems over time. The purpose of monitoring is to identify problems on the route and take corrective action. Research has shown the phenomenon of bus bunching, in particular, to be non- linear, highlighting the importance of early intervention. Corrective action typically involves the application of one or more operational control strategies. Strategies currently used or proposed for headway-based operation include ⢠Boarding limits: buses limit the number of passengers that can get on. All systems practice boarding limits when buses reach capacity, but limits before reaching capacity have been proposed in the literature as a way to control dwell time variability. ⢠Holding: buses wait for a specific amount of time at a control point. The literature can be divided into optimization approaches, which emphasize passenger objectives, and control approaches, which aim to keep all buses on the route at equal headways. Holding strategies were also the most commonly used by surveyed agencies. ⢠Short turns: buses turn around before the end of the route and enter service in the opposite direction. Short turns are a common feature of schedules but can also be used in real time to fill large gaps in service. Several survey respondents reported this practice. ⢠Speed guidance: bus drivers are asked to adjust their driving based on real-time headways. Some research has suggested using speed guidance exclusively, while others found it to complement holding. Surveyed agencies used speed guidance along with other strategies. ⢠Stage vehicles: buses held in reserve for use when conditions warrant. Surveyed agencies reported using stage vehicles to fill gaps in service and to provide additional capacity at times of high ridership. C H A P T E R 5
52 Intelligent Transportation Systems in Headway-Based Bus Service ⢠Stop skipping: buses skip one or more stops to make up time. Many research papers combine holding (to slow down) and stop skipping (to speed up), a practice also found at several surveyed agencies. These operational control strategies are applicable in different situations and vary in their visibility to and impact on passengers. Table 20 summarizes the main use cases and passenger perspectives. Stop skipping and boarding limits have two symbols under passenger perspective because the reaction varies. Passengers already on board may hardly notice the control action, while those at a skipped stop or who are denied boarding must wait for the next bus and may not understand why. All of the operational control strategies listed in Table 20 depend on having accurate and up- to-date information about the status of buses on the route as well as communication links with bus drivers. Headway-based routes typically rely on one or more of the following technologies: ⢠CAD is a method for real-time monitoring and control of transit vehicles, and requires buses to be equipped with AVL. CAD allows dispatchers to monitor entire bus routes from a centralized control center and to communicate with drivers by voice or text. â An MDT is the onboard component of a CAD system, which allows bus drivers to communicate with dispatchers and may also display control information such as the forward headway or speed guidance. ⢠AVL allows agencies to track the position of buses. Most systems use GPS and update at a set frequency (e.g., every 30 seconds). ⢠An APC counts passengers boarding and alighting, information that is used in some opera- tional control strategies and can be used to track the current passenger load on a bus. ⢠EFP may be another way to track passenger boardings and alightings, depending on the agencyâs fare structure and accepted payment methods. No survey respondents were currently using EFP for this purpose. ⢠SS often include video feeds on board buses and in larger stations and transit centers that can be monitored remotely by dispatchers. Headway-based operation also requires agencies to use different performance measures. Schedule adherence is perhaps the most common performance measure in the transit industry but is not applicable to headway-based operation. The common performance measures found in the survey were, in descending order: ⢠Headway adherence, the standard deviation of headway deviations divided by the mean of scheduled headways. ⢠Bus travel times, often expressed as cycle time (round-trip travel time plus layovers). ⢠Number of trips, which can be compared to the expected number based on the target headway. ⢠Passenger satisfaction, measured through rider surveys or number of complaints. Strategy Status of Route Passenger Perspectivea Normal to Minor Disruptions Major Disruptions Visibility Impact Holding P P   Stop skipping P Â/ Â/ Speed guidance P   Boarding limits P Â/ Â/ Short turns P   aAn empty circle denotes little passenger reaction, an outlined circle denotes some passenger reaction, and a filled circle denotes substantial passenger reaction. Table 20. Summary of operational control strategies.
Conclusions 53  ⢠Headway variability, which can be measured in different ways such as service regularity or the standard deviation of headway. ⢠Bus travel time variability, which can be measured in different ways such as the 90th percentile of cycle time. ⢠Passenger wait time, which for buses is typically estimated using measures like excess waiting time. Key Findings from Case Examples Several common themes emerged from the case example interviews. Benefits ⢠Several agencies noted the flexibility of headway-based operation as an important benefit. A headway-based route can adapt to the addition or removal of buses, buses passing one another, and major disruptions that slow down all buses (e.g., inclement weather) more easily than a schedule-based route. ⢠Headway-based operation can improve the passenger experience by reducing travel time variability, out-of-vehicle wait times, crowding, and denied boardings due to buses reaching capacity. Personnel ⢠All agencies underscored the importance of bus drivers in making headway-based opera- tion work. ⢠Driver training is important. Headway-based operation is a different experience than schedule-based operation. There are different targets to watch, and guidance can change rapidly. Some drivers will operate headway routes more often than others, but training should be deployed widely to include all drivers who may be called on to operate a headway run. ⢠Dispatchers also have an important role. They are the ones who see the big picture and decide when to use many operational control strategies. ⢠Some agencies have dispatchers dedicated to HBS, while others have dispatchers monitor both headway and schedule routes. ⢠Dispatcher training is also valuable so that all personnel understand the goals of HBS, the operational control strategies available, and when to use them. Costs ⢠The agencies interviewed found it difficult to break out the cost of adopting HBS. ⢠Capital costs may be negligible if the agency has already deployed ITS like CAD/AVL for managing schedule-based service. In such cases, the initial costs are mainly software and training, which larger agencies may be able to develop partly or entirely in-house. ⢠ITS operation and maintenance costs are typically shared across the agency as well. ⢠Converting a schedule-based route to headway-based operation may reduce operating costs by improving reliability and shortening cycle times, but other agency costs such as driver and dispatcher training, as well as dedicated dispatchers (if used), could be higher. COVID-19 Response ⢠The COVID-19 pandemic hit after the survey was completed and was ongoing when the case example interviews were conducted.
54 Intelligent Transportation Systems in Headway-Based Bus Service ⢠TheBus used the flexibility of its headway-based system to adjust service levels based on observed ridership and passenger loads. ⢠Capital Metro and LA Metro had reverted to a Sunday schedule at the time of their interviews. Notable Practices ⢠TheBus does not use a fixed-target headway but dynamically adjusts the target based on the current cycle time and number of buses on the route. This is a highly successful practice because a fixed target (e.g., 10 minutes) can become unachievable if the cycle time becomes too long or if the number of buses on the route is insufficient. ⢠LA Metro runs HBS on the G Line. This route is a BRT and has features that reduce both travel time variability (dedicated ROW and TSP) and dwell times (off-board fare collection and all-door boarding). ⢠LA Metro uses long layover times on its headway-based route. Long layovers can absorb headway deviations, allowing buses to reenter service in the opposite direction at the target headway. Challenges ⢠Headway-based routes that apply holding too often experience a schedule sliding effect, where buses maintain headways but travel more slowly. TheBus experienced this problem in a pilot phase and started tracking the number of trips as a performance measure to monitor the issue. ⢠Headway-based operation complicates driver shift changes. If drivers are expected to operate a specific number of trips, then they may end up working significant amounts of overtime if the route struggles with schedule sliding. Several agencies also noted that drivers near the end of their shift sometimes ignore their headway target and dispatcher instruc- tions to reach the switch point sooner. Suggestions for Future Research The findings of this synthesis suggest that there is potential for future work on the following: ⢠Blending transit data sources. Most implementations of HBS to date rely on AVL data to track real-time bus locations and headways. APC and SS were used as real-time data sources by some respondents, while EFP data were not currently used in real time by any surveyed agency. There is some overlap between these data sources. For instance, APC, SS, and EFP could all be used to track passenger loading on a bus. As real-time use of these data sources grows, transit agencies could benefit from additional research on how to blend or validate these datasets. ⢠Passengersâ views of control actions. As the literature review shows, many different oper- ational strategies have been developed for HBS. Some of these strategies, such as short turns, have negative impacts on passengers, which have been noted in the literature. In the survey, transit agencies ranked holding and speed guidance above strategies that are more visible to passengers, like stop skipping and short turns, as shown in Table 13. Since many of the proposed strategies could substitute for one another, research on passenger perceptions of control actions (e.g., holding and stop skipping) could be informative and help transit agencies select which strategies to use. ⢠Service transitions. One challenge noted by TheBus in its case example interview was how to manage transitions in service, whether from schedule to headway-based operation or from an off-peak headway to a peak headway. Limited research has been done on headway-based operation in the context of branching routes or common corridors where bus routes interact, but no research discussed in this study appears to have studied the temporal dimension.
