Managing the Transit Scheduling Workforce (2019)

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.

9 Fixed-route transit scheduling is a craft with roots back at least to the 19th century. Case in point, the Thomas Cook European Rail Timetable, a compendium of train schedules in Europe, started publication in 1873 (European Rail Timetable 2018). Since then, scheduling has changed from a predominately paper-based practice to one that is largely software-driven; however, the core requirement that a well-qualified human being perform transit scheduling has not changed. This literature review chapter discusses previous research and documents about established practices in fixed-route scheduling, emerging procedures and trends in fixed-route transit scheduling, and workforce management—particularly in the transit industry. Likely because of the relatively lower number of schedulers compared with other transit jobs (e.g., bus operators), the synthesis team found no existing research directly focused on transit scheduler workforce management. This literature review is divided into the following sections: • Introduction to transit scheduling. • Role and importance of transit schedulers. • Industry resources to train and develop transit schedulers. • Evolving practices in transit scheduling. • Workforce management. Introduction to Transit Scheduling This section discusses the basics of the scheduling process to give readers unfamiliar with scheduling an understanding of its complexity and the nature of scheduling trade-offs that require high analytical capabilities among schedulers. This section will also introduce and define the key terms in scheduling. Two prior TCRP reports are the go-to manuals on fixed-route scheduling: TCRP Report 30: Transit Scheduling: Basic and Advanced Manuals (Pine et al. 1998) and TCRP Report 135: Con- trolling System Costs: Basic and Advanced Scheduling Manuals and Contemporary Issues in Transit Scheduling (Boyle et al. 2009). These manuals are approachable for individuals new to the field of fixed-route scheduling but also provide detailed guidance on highly technical scheduling concepts. Much of the content for this chapter’s introduction to transit scheduling section is taken from these manuals. Fixed-route transit scheduling consists of a five-step process, depicted in Figure 3. (Each of these processes is further defined in the list of terms that follows.) The scheduling process is similar to building a house: phases must be performed in sequence, and each phase relies on the completion (and the quality) of prior phases. Any changes or mistakes in a previous phase will undermine the integrity of later phases. C H A P T E R 2 Literature Review

10 Managing the Transit Scheduling Workforce In addition to this five-step process, transit schedules often are published and distributed publicly in the form of general transit feed specification (GTFS) data, web pages, and paper time tables and schedule books. Key scheduling terms are defined here in alphabetical order (definitions were copied verbatim or adapted from Boyle et al. [2009]). • Block: a vehicle (or train) assignment that includes the series of trips operated by a single vehicle from the time it pulls out to the time it pulls in. A complete block includes a pull-out trip from the garage followed by one or (usually) more revenue trips and concluding with a pull-in trip back to the garage. • Blocking: the process in which trips are “hooked” together to form a block. • Cycle time: the round-trip running time plus layover time. Also known as “round-trip cycle time.” • Deadhead: the time and distance that a vehicle needs to travel in places where it will not pick up passengers. Deadheading is typically required to get buses to and from their garages (also called pulling in or out of a garage), or when bus operators need to travel from one route or point to another during their scheduled work day. • Extraboard: a group of operators who provide coverage of vacant runs and other work on a daily or weekly basis. • Frequency: the number of vehicles passing a point on a route within a given unit of time, usually expressed as X vehicles per hour. Headway is the inverse of frequency. A frequency of six buses per hour is the same as a headway of 10 minutes. • Garage: the place where vehicles are stored and maintained and where operators report for their assignments. Also known as barn, base, division, yard, and so forth. • Headway: the interval of time between two vehicles running in the same direction on the same route, usually expressed in minutes. Frequency is the inverse of headway. A headway of 10 minutes is the same as a frequency of six buses per hour. Trip Generation Headways Running times Layovers Span of service Route patterns Time points Blocking Assign vehicles to trips Garage assignments Deadheads Interlines Runcutting Assign operators to blocks Reliefs Work rules Meal breaks Rostering Package runs into weekly assignments or Cafeteria-style rostering Quality Checking Review schedule documents and data outputs for accuracy and compliance with rules Some checking may be done by third parties (e.g., an operator's union or an oversight agency) Figure 3. The five steps in the scheduling process and a sampling of activities performed within each step. Note: Although most scheduling manuals depict scheduling as a four-step process consisting of trip generation, blocking, runcutting, and rostering, the authors added quality checking to emphasize that schedules must be reviewed for accuracy and compliance with rules prior to the schedules’ use.

Literature Review 11 • Interlining: the use of the same vehicle or block operating on more than one route with the same operator, without returning to the garage during route changes. • Layover or recovery time: the time between the scheduled arrival and departure of a vehicle at a transit terminal. The terms layover and recovery are often used interchangeably. This synthesis report will use the term layover. • Relief: the replacement of one operator on a vehicle by another operator on the same vehicle. • Revenue service: when a vehicle is in operation along a route and is available to the public. • Rostering: the process of grouping daily operator runs into packages of weekly work assign- ments. The finished package is known as a roster (or a bid package). Some agencies use cafeteria-style rostering, in which operators pick their own weekly packages. • Run: a work assignment for an operator. Most often, run refers to a whole day’s work assignment. • Runcutting: the process of converting (or cutting) vehicle blocks into work assignments for operators. The finished product is referred to as a runcut. The key (and challenge) is to assign block pieces to operators in the most cost-efficient way possible. • Running time: the time it takes for a vehicle to travel the length of a route or between two specific points on a route, usually between time points. • Segment: a part of a route, usually end-capped by time points. Routes are made up of a series of segments. • Service pattern: the unique sequence of stops associated with each type of trip on a route. • Sign-up: the process in which operators select work assignments. Most transit agencies have three or four sign-ups a year. (Sign-ups may also be called bids, picks, mark-ups, and so forth). • Span of service: the length of time, from the beginning of the first trip to the end of the last trip, during which service operates on the street. • Time point: a designated location on a route used to control the spacing of vehicles along the route. Also known as nodes. • Trip: the one-way operation of a vehicle between two points on a route. Trips normally are associated with a particular directionality (e.g., northbound or southbound). Also known as a “one-way trip.” • Trip generation: the process of creating scheduled trips at the desired headways, running times, and cycle times across a route’s span of service. • Work rules: rules that define the characteristics of runs and rosters. These rules may include requirements for breaks, reliefs, time off between runs, and so forth, and are usually defined in collective bargaining agreements, written into agency policies, or based on past practice. Some rules are “soft rules,” which are informal, non-binding rules schedulers work to follow whenever possible, but there is no written requirement or consequence of failing to adhere to the rule. As described by Boyle et al. (2009), the foundations of the schedule can be categorized into two types of requirements: 1. Inputs external to developing a schedule but required for proper schedule development, for example: a. Knowledge of budgetary constraints. b. Knowledge of your transit system’s goals. c. Knowledge of the area served. d. Knowledge of your agency’s current short- and long-term objectives/goals. 2. Data elements required to actually construct the schedule, for example: a. Knowledge of relevant provisions of the collective bargaining agreement or operator work rules, if applicable. b. Route design considerations.

12 Managing the Transit Scheduling Workforce c. Service standards. d. Annual service plan. e. Service data, including running time, ridership, and operations. To date, there is no completely automated tool that consumes all components of a schedule’s foundation and outputs a complete, usable transit schedule. Human knowledge and intervention are required, which makes transit schedulers so important to transit agencies. Role and Importance of Transit Schedulers The schedule provides the blueprint for any fixed-route transit service. Transit schedules can affect the safety of transit service, operator health and well-being, and overall customer service. For example, Tse et al. (2006) reviewed 50 years of research on bus operators’ well-being and found high rates of physical and psychological health problems, at least partially related to bus operator working conditions and schedule-based time constraints. Also, operator fatigue and lack of alertness (Dorn 2003) or excessive hours worked and long split shifts (Hoang 2016) can increase the risk of crashes and collisions. Transit schedules that are accurate and consider human factors can help improve operator safety, health, and well-being, saving operators and transit agencies time and money. Transit schedules also determine the scheduled number of paid operator hours, which is the main source of transit agency operating expenses. Transit schedules determine the sched- uled distance that transit agency vehicles will operate, governing fuel and maintenance costs. Schedules also set the requirements for how many vehicles and operators are needed at any given time. Ultimately, transit schedules drive most of a transit agency’s operating expenses and also influence an agency’s capital costs. Overall, transit schedulers have an immense impact on transit agencies. Boyle et al. (2009) describes additional reasons that transit schedulers are vital to transit agencies: • Cost: Good schedulers can reduce cost by providing the most cost-efficient schedule while maintaining good service for customers. • Contracts and labor relations: Schedulers’ knowledge is important, especially when negotiating changes to work rules and pay premiums. Schedulers’ adherence to work rules (or lack thereof) can jeopardize a transit agency’s service plans and can even result in unnecessary expenses when work rules are not followed. Transit schedulers can also help estimate the potential financial impact of changing work rules. • Work environment: Good schedules can reduce the stress inherent in the job for operators, improving customer service and morale and minimizing absenteeism. • Customer service: Schedulers have a direct impact on customer service. Schedulers define arrival and departure times and set bus route capacity. If schedules are not accurate or realistic given actual operating conditions, customers will experience poor service. • Fully utilizing computerized scheduling: Good schedulers can help maximize the benefits of having computerized scheduling. Software can free the scheduler from mundane and time- consuming tasks while allowing for more what-if testing for alternatives. Scheduling software still requires good schedulers with in-depth knowledge and understanding of the scheduling process. Although fixed-route scheduling may be performed with the assistance of computers (e.g., through the use of specialized software made specifically for fixed-route scheduling or standard office software like Microsoft Excel), computer assistance does not obviate the need for highly trained and capable transit schedulers—similar to how Excel does not obviate the need for

Literature Review 13 accountants. Transit schedulers must have adequate knowledge of and an ability to utilize all relevant inputs to the scheduling process to create high-quality transit schedules. Transit sched- ulers must also understand scheduling concepts, the local context for which they are creating schedules, and the specific software application they are using to create schedules, and they should be willing to solve complex logistical problems. “Scheduling is a craft, whether executed manually or with computer assistance. New employees of transit scheduling departments need training in this craft to do their job, and experienced schedulers require retraining to fill gaps in their knowledge” (Pine et al. 1998, from the foreword). Success in transit scheduling requires continuous work and analysis, and transit agencies must “assure that their schedulers possess enough proficiency and a high degree of competence to perform their tasks” (Lehman Center for Transportation Research 2015, p. 218). Industry Resources to Train and Develop Transit Schedulers Becoming a proficient transit scheduler is not an easy task. Due to the nature of the work, transit scheduling is a niche field—mastered by a small portion of the transit workforce. In addition, because transit scheduling relies heavily on proprietary software packages, schedulers must not only learn general scheduling concepts but also learn how to apply those concepts using the specific user interfaces provided by a transit agency’s chosen scheduling software provider. The National Transit Institute’s “Introduction to Transit Service Planning” course provides a basic introduction to transit scheduling (National Transit Institute 2018). However, to gain proficiency as a scheduler, more content and hands-on training are often needed. There are a few industry-wide transit scheduler training resources, including the two previously mentioned TCRP reports, TCRP Report 30 (Pine et al. 1998) and TCRP Report 135 (Boyle et al. 2009). These reports are available at the TRB website (www.TRB.org), and they provide an in-depth review of scheduling concepts that can be put into practice using paper or Excel. The Canadian Urban Transit Association (CUTA) offers an in-person paid course called “Scheduling and Runcutting Level 1” (Canadian Urban Transit Association 2018) that teaches scheduling concepts and does not teach software-specific skills; a few private firms provide scheduling concept courses and some transit consultancies provide customized scheduling training for hire. Transit scheduling software providers often provide software training and technical assistance for their clients; however, this training may be more focused on software skills than on develop- ing knowledge of general scheduling concepts and skills that could be put into practice on any software platform. Moreover, in addition to all the technical requirements of scheduling, there is a trend in the industry to use more data and more advanced software during the schedule building process— changing what it means to be a proficient transit scheduler. Evolving Practices in Transit Scheduling Although transit scheduling has always involved designating when transit vehicles and transit operators will be at specific locations, the practice of scheduling has changed significantly through the years. Scheduling was largely a manual and paper-based process, relying heavily on limited operational data collected through small samples. However, the pace of technological change is ever-increasing and has required the transit industry to respond by updating its practices and guidance for transit scheduling. Before TCRP Report 30 was published in 1998,

14 Managing the Transit Scheduling Workforce the main reference on transit scheduling was from 1946 (according to Pine et al. 1998). TCRP Report 30 was needed, in part, to incorporate the growing amount of information technology used in transit scheduling. The next national manual on transit scheduling, TCRP Report 135, was published about 10 years later in 2009 (Boyle et al. 2009). Then, in 2015, the Lehman Center for Transportation Research published its Training Manual for Transit Service Planning and Scheduling. Even with all these manuals, the realities of transit scheduling keep changing through increases in the use of scheduling software, the proliferation of data available for scheduling, and advances in computational power and mathematical theory. Scheduling Software Scheduling software helps reduce time-consuming manual tasks, speed up the scheduling process, increase schedule accuracy by decreasing manual data entries and calculations, and allow for the testing of multiple schedule alternatives (Lehman Center for Transportation Research 2015). It is likely that the use of scheduling software will continue to increase as more firms enter the marketplace and as transit agencies need to consume and produce more data during the scheduling process. There is no list of transit agencies that use specialized software for fixed-route scheduling. According to TCRP Report 135, 22% of surveyed agencies use scheduling software (Boyle et al. 2009). A 2004 survey of Florida transit agencies found that 40% of survey respondents used scheduling software (Mistretta 2005). The use of scheduling software may be related to agency size and both the resources available to purchase the software and the expected magnitude of benefits. However, the marketplace of scheduling software is changing. Software companies are entering the market targeting small-to-medium-sized agencies and offering a lower price point. Some companies are offering scheduling software that works in standard web browsers instead of requiring software installations on local computers. As discussed previously, using scheduling software does not preclude needing capable schedulers. And, training a person on scheduling software is not the same as training someone on how to be a good scheduler—just like giving a person training on Microsoft Word does not make the person a good writer (Lehman Center for Transportation Research 2015). However, as scheduling becomes an increasingly computer-driven process, transit agencies will need to ensure their schedulers have both the scheduling and computer skills to succeed. Use of Data in the Scheduling Process Another change in fixed-route scheduling is the increased availability and use of data as inputs to the scheduling process. Although several data items are necessary to produce schedules, running times are the main data needed to maintain and improve service reliability. TCRP Report 135 found that, although 86% of transit agencies surveyed still use manual data checks as their primary source of running time data, between 40% and 50% of transit agencies surveyed use either automated vehicle location (AVL) or automated passenger counter (APC) systems to at least collect and report running time data (Boyle et al. 2009). As APC and AVL systems become more prevalent, transit schedulers will have increasingly more information available to help improve the quality of transit schedules. APC and AVL data can be used for analyzing and improving schedules and runtimes, analyzing and improving schedule adherence and headway adherence, and understanding ridership demand (Furth et al. 2006). The availability of data to transit agencies has enabled more complex analyses, including analyzing extreme values such as 95th percentile running times, identifying network slow-spots (places where congestion is high), and even discovering hidden, unknown trends.

Literature Review 15 In addition to APC and AVL data, third parties also have data that can be useful during the scheduling process. Google Maps, for example, can help estimate the travel time between any two points across different days of the week and times of the day—information that can help fine-tune deadhead times. However, having data is not the same as being able to use data. Schedulers may be overwhelmed by the sheer mass of data available, may not possess the skills to analyze the data, or may not understand data analysis concepts like distributions, variability, and measures of central tendency. Again, although software may help in data collection, reporting, and possibly even utilization, software does not currently perform all steps in the scheduling process without significant human intervention and decision making. To take full advantage of the plethora of data, transit schedulers will have to be trained in data collection, analysis, and reporting. Or, at the minimum, specialized analytical staff will have to provide useful information to transit schedulers so they can use it during the schedule building process. Continued Advancement in Optimization and Algorithms for Scheduling As the size and utility of data and the affordability and speed of computing continue to increase, researchers and software developers will have the tools available to improve methods for optimizing all steps in the scheduling process. Researchers in China developed a model to minimize personnel costs while adhering to work rules, vehicle operations requirements, and improving the continuity of operators’ work (Wei et al. 2016). Researchers in Brazil developed a model for optimizing rostering (Mayrink and Silva 2013). Researchers in Ethiopia built a linear programming model to improve bus utilization and decrease operating costs (Berhan et al. 2014). Finally, researchers developed an algorithm to build bus schedules to minimize passenger transfer time across the entire transit network (Ceder et al. 2001). These advances may not immediately affect today’s transit schedulers; however, future scheduling software will become increasingly robust as computing power and its application to transit’s scheduling challenges continue to escalate. Transit schedulers will play an important role in transit scheduling and will need to be qualified to perform increasingly analytically driven tasks for the foreseeable future. Even in the futuristic hypothetical scenario of fully automated scheduling software packages that consume historical service data and vehicle characteristics; digest work rules and agency policies; and output timetables, blocks, runs, and rosters, transit schedulers will be needed to validate key data sources, specify scheduling parameters, and review and approve schedule outputs to ensure their quality and alignment with the transit agency’s current operating context, policies, and targets. Because transit schedulers are so critical and their roles continue to evolve, it is important that transit agencies examine how they are managing the transit scheduling workforce to ensure schedulers are the best people doing their best possible work in the best way. Workforce Management In this synthesis, the term workforce management practices refers to the collection of orga- nizational practices and procedures that take place during the employee life cycle: • Recruitment. • Selection. • Training and development. • Retention. • Performance management.

16 Managing the Transit Scheduling Workforce Although not all organizations are intentional and strategic about their workforce manage- ment practices, research suggests that good workforce management practices can lead to better organizational performance (e.g., see Karami et al. 2015; Yu and Guo 2016). It is likely that transit agencies that are intentional and strategic about how they manage their transit scheduling workforce have better scheduling outcomes. TCRP Report 162: Building a Sustainable Workforce in the Public Transportation Industry (Cronin et al. 2013) describes some of the challenges facing the transit workforce. Cronin et al. states that transit industry retirements over the next 5 to 10 years are expected to exceed 50% of the workforce, and this poses a significant problem for transit agencies, not only because this is a significant loss of experience and knowledge but also because recruiting new candidates may be difficult when: • Transit wages are often less than competitive with other industries. • The industry is experiencing technological changes that require skills that may not be present in current employees or in typical applicant pools. • Those in the younger generation may have the skills needed but do not see transit as an attractive or interesting career path. These statements are all true for both the transit industry in general and the transit scheduling workforce in particular. In addition, TCRP Report 77: Managing Transit’s Workforce in the New Millennium (McGlothin Davis, Inc. 2002) looks at the challenges facing transit agencies at the turn of the century. Although now somewhat dated, the report still resonates with today’s reality that many transit jobs, including those requiring analytical skills, are difficult to recruit for and retain. Cronin et al. (2013; 2017) and McGlothin Davis, Inc. (2002) provide many different strate- gies and tools that could be employed by transit agencies to help keep their transit scheduling workforces sustainable and performing well, for example: • Establishing applicant screening processes (Cronin et al. 2013). • Targeting candidates from other industries (Cronin et al. 2013). • Implementing employee mentoring programs (Cronin et al. 2013). • Instituting internship or apprenticeship programs (Cronin et al. 2013). • Addressing training needs in conjunction with performance appraisals (Cronin et al. 2013). • Developing a knowledge management system (Cronin et al. 2013; 2017). • Assessing and improving organizational culture (Cronin et al. 2013). • Implementing a rewards program for high performers (Cronin et al. 2013). • Partnering with local schools or universities to improve the pipeline of candidates (McGlothin Davis, Inc. 2002). • Keeping the work environment interesting and up-to-date with state-of-the-art technology (McGlothin Davis, Inc. 2002). • Creating a corporate “university” to help develop the necessary skills in internal employees (McGlothin Davis, Inc. 2002). More research is needed on exactly what strategies are being used throughout the industry to manage the transit scheduling workforce and to better understand which strategies appear to help improve transit scheduling outcomes. Summary of the Literature Review Transit scheduling is pivotal to transit agencies and has been around since the advent of fixed-route transit. Even though today’s scheduling software suites can help automate some tasks, scheduling still requires qualified individuals to perform and manage the scheduling

Literature Review 17 process. Because a transit schedule impacts and influences so many aspects of a transit agency’s costs and performance, it is critical that transit agencies have the resources to adequately manage transit schedulers through the entire employee life cycle. However, not many resources are available to guide transit agencies through scheduler management, and only a few resources are available to help individuals learn and perfect the skills needed to be a proficient scheduler. In addition, as the practice of scheduling has begun to use more data and more complex and powerful software, the available resources and current scheduler management practices are growing stale and do not necessarily match the changing requirements of the scheduling job or the transit industry as a whole.