The Impacts of Vehicle Automation on the Public Transportation Workforce (2022)

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.

45   C H A P T E R 5 This chapter presents the research team’s work to engage the transit industry during the course of the research. The purpose of industry engagement was to present transit industry stakeholders with the research team’s initial thoughts about transit automation use cases and likely affected transit jobs and obtain feedback on the identified planning and policy decisions for each use case to help the research team build assumptions into the workforce effect calcula- tor. Industry engagement comprised three main activities, and this chapter is divided into sec- tions that describe each: • In-person workshop: Held at the 2019 APTAtech: Transportation Technology Conference in September 2019. • Front-line employee survey: Administered in October through December of 2019. • Two industry-wide webinars: Held in June 2020. 5.1 In-Person Workshop In September 2019, the research team held an in-person workshop, attended by 21 par- ticipants, at the APTAtech: Transportation Technology Conference in Columbus, Ohio (see Figure 8). Attendees served in many different roles and came from several different organiza- tional types. Job titles of attendees included operational analyst, data systems manager, general manager, chief technology officer, technology policy analyst, chief strategy officer, communica- tions assistant manager, and business development director. Attendees were from a variety of organizations, including transit agencies, the USDOT, consulting firms, software companies, transit unions, and the departments of transportation (DOTs) from many different states, including Pennsylvania, Florida, Illinois, Missouri, Washington, and Tennessee. The purpose of the workshop was to obtain feedback from attendees regarding the key planning and policy decisions that will have implications on the workforce effects of transit vehicle automation. (These planning and policy decisions were listed for each use case in Chapter 3.) The workshop was organized with the following sections: • Purpose of the workshop and the study. • Brief discussion of previous examples of industry automations (e.g., manufacturing and aviation). • Summary of potential disparate impacts of automation. • Foundational assumptions related to the study (e.g., workforce effects will not include impacts of fleet electrification). • Brief presentations on each use case including: – Description and examples. – Potential operational impacts. Industry Engagement

46 The Impacts of Vehicle Automation on the Public Transportation Workforce – Supporting technologies. – Potential timelines. • Breakout sessions (discussed in more detail in the following sections). During the breakout sessions, attendees were divided into six groups—one for each use case. (At this point in the study, the research team had two separate use cases for automated MOD—one for service operated by private entities and one for service operated by transit agencies. As the study progressed, the use case for private entities operating automated MOD was dropped because it was outside the scope of the study.) Each group had a notes packet containing a series of planning and policy decisions that it was asked to discuss and then provide written notes on. (The full presentation given at the workshop is available as Attachment 2. The full set of blank notes packets used for the breakout sessions is available as Attachment 3.) Although all data collected during the workshop were qualitative, attendees provided excellent feedback, which was added to other sources of industry engagement to help the research team make its assumptions about planning and policy decisions and to fine-tune the research team’s understanding of the industry’s perspectives regarding transit vehicle automation. 5.2 Front-Line Employee Survey The front-line employee survey examined the opinions of transit employees regarding transit vehicle automation. The research team recruited employees who worked in fixed- route and demand-response bus operations, maintenance, management, and administration, including: • Operators. • Mechanics and technicians. • Fuelers and cleaners. • Street supervisors. • Trainers. • Dispatchers. • Planners/schedulers. • Managers. • Police and security personnel. Figure 8. Title slide of the APTAtech in-person workshop.

Industry Engagement 47   The survey explored if front-line employees had concerns or perceived benefits regard- ing automated transit vehicles, and if so, what those concerns were. The survey also asked respondents to identify their job titles and whether they were part of a labor union. 5.2.1 Methodology 5.2.1.1 Survey Design The survey questionnaire was carefully developed by the research team to avoid influencing respondents’ perceptions of the benefits or concerns of vehicle automation. That is, the research team attempted not to induce benefits or concerns that employees did not perceive on their own. Therefore, the survey started with introduction pages providing background information on transit vehicle automation and the five use cases. The descriptions of the use cases focused on the nature of the automated transit services and did not discuss specific potential workforce effects. Then, respondents were asked basic questions about their current job title, number of years of service, and whether they supervised the work of others (Table 9). Next, the survey asked respondents a repeated set of six questions about the concerns and benefits of each use case. The six questions were divided into three questions for perceived concerns and three questions for perceived benefits as follows: • An open-ended question that asked for the respondents’ perceived concerns/benefits of the use case. • A question that contained a list of possible concerns/benefits and asked respondents to rate how likely it was that they would experience the possible concern/benefit on a scale from one, equal to (=) not at all, to four, equal to (=) extremely. (Respondents could also select Does not apply.) Survey Question Number of Responses Percent What is your current job title? Bus mechanic/maintenance technician 12 9% Bus operations trainer 3 2% Bus operator 63 49% Bus service person/fueler/cleaner 2 2% Dispatcher/controller 6 5% Police officer or security person 1 1% Road/street supervisor or traffic controller 7 5% Short-range transit planner/schedule maker 4 3% Other 31 24% How long have you been in your current position? Less than 1 year 3 2% 1 year–4 years 25 19% 5 years–9 years 23 18% 10 years–14 years 32 25% 15 years–19 years 14 11% More than 20 years 32 25% Do you supervise the work of others? No 79 61% Yes 49 38% Not answered 1 1% Table 9. Descriptive analysis of survey participants.

48 The Impacts of Vehicle Automation on the Public Transportation Workforce • An open-ended question that asked for an explanation of ratings given in the previous question. The research team placed the open-ended question about perceived concerns/benefits of the use case before the rating questions to give respondents a chance to describe their perceived concerns/benefits in their own words before the survey introduced ideas to them in the rating question. The lists of possible concerns/benefits included in the rating question can be found in Table 10 and Table 11. Concern and Use Case Not at All Likely Somewhat Likely Very Likely Extremely Likely Does Not Apply Chance to learn new skills Use Case #1: Bus Automation for Maintenance and Yard Operations 32% 24% 15% 5% 24% Use Case #2: Low-Speed Automated Shuttles 50% 22% 9% 6% 13% Use Case #3: Automated BRT 50% 25% 10% 2% 13% Use Case #4: Automated MOD 55% 18% 9% 5% 13% Use Case #5: Automated Local Bus Service 43% 32% 9% 4% 13% Improvements in working conditions Use Case #1: Bus Automation for Maintenance and Yard Operations 45% 29% 5% 3% 17% Use Case #2: Low-Speed Automated Shuttles 61% 20% 3% 2% 14% Use Case #3: Automated BRT 63% 17% 5% 2% 13% Use Case #4: Automated MOD 55% 21% 5% 5% 13% Use Case #5: Automated Local Bus Service 61% 27% 2% 0% 11% Increases in pay Use Case #1: Bus Automation for Maintenance and Yard Operations 66% 7% 3% 1% 23% Use Case #2: Low-Speed Automated Shuttles 75% 5% 5% 0% 16% Use Case #3: Automated BRT 72% 10% 5% 2% 12% Use Case #4: Automated MOD 70% 11% 4% 2% 14% Use Case #5: Automated Local Bus Service 79% 9% 2% 0% 11% Reduced job physical demands Use Case #1: Bus Automation for Maintenance and Yard Operations 36% 27% 12% 7% 19% Use Case #2: Low-Speed Automated Shuttles 42% 22% 13% 5% 19% Use Case #3: Automated BRT 57% 17% 8% 5% 13% Use Case #4: Automated MOD 50% 21% 7% 9% 13% Use Case #5: Automated Local Bus Service 55% 29% 2% 4% 11% Reduced job stress Use Case #1: Bus Automation for Maintenance and Yard Operations 45% 26% 8% 3% 19% Use Case #2: Low-Speed Automated Shuttles 53% 25% 5% 2% 16% Use Case #3: Automated BRT 58% 18% 7% 5% 12% Use Case #4: Automated MOD 55% 21% 4% 5% 14% Use Case #5: Automated Local Bus Service 57% 27% 4% 4% 9% More desirable working assignments Use Case #1: Bus Automation for Maintenance and Yard Operations 58% 16% 1% 1% 23% Use Case #2: Low-Speed Automated Shuttles 77% 6% 3% 0% 14% Use Case #3: Automated BRT 75% 7% 7% 0% 12% Use Case #4: Automated MOD 66% 14% 5% 2% 13% Use Case #5: Automated Local Bus Service 71% 14% 0% 2% 13% Table 10. Perceived benefits by use case.

Industry Engagement 49   A PDF version of the full online survey is available in Attachment 4. The entire survey protocol was reviewed and approved by Texas A&M University’s Institutional Review Board. 5.2.1.2 Survey Distribution The online survey was developed using the Qualtrics® web-based survey platform. The research team developed a survey flyer with the survey link and a quick response (QR) code (see Attachment 5). The research team sent the flyer and link through various points of contact, including: • A representative from the Amalgamated Transit Union (ATU). • Various APTA committees. • The Conference of Minority Transportation Officials. • The research team’s contact list. Concern and Use Case Not Concerned at All Somewhat Concerned Very Concerned Extremely Concerned Does Not Apply Change in job responsibilities Use Case #1: Bus Automation for Maintenance and Yard Operations 23% 23% 17% 24% 14% Use Case #2: Low-Speed Automated Shuttles 22% 18% 23% 30% 7% Use Case #3: Automated BRT 17% 14% 20% 41% 8% Use Case #4: Automated MOD 18% 21% 16% 36% 9% Use Case #5: Automated Local Bus Service 11% 11% 22% 47% 9% Job loss Use Case #1: Bus Automation for Maintenance and Yard Operations 8% 18% 18% 38% 17% Use Case #2: Low-Speed Automated Shuttles 18% 8% 15% 53% 5% Use Case #3: Automated BRT 12% 8% 17% 53% 10% Use Case #4: Automated MOD 14% 14% 18% 46% 7% Use Case #5: Automated Local Bus Service 13% 7% 16% 56% 7% Loss of desirable working assignments Use Case #1: Bus Automation for Maintenance and Yard Operations 23% 23% 8% 27% 20% Use Case #2: Low-Speed Automated Shuttles 22% 12% 17% 45% 5% Use Case #3: Automated BRT 14% 19% 14% 46% 8% Use Case #4: Automated MOD 20% 16% 13% 45% 7% Use Case #5: Automated Local Bus Service 16% 7% 24% 45% 7% Needing additional training to succeed Use Case #1: Bus Automation for Maintenance and Yard Operations 26% 24% 15% 18% 17% Use Case #2: Low-Speed Automated Shuttles 28% 23% 17% 22% 10% Use Case #3: Automated BRT 25% 27% 8% 22% 17% Use Case #4: Automated MOD 21% 27% 14% 21% 16% Use Case #5: Automated Local Bus Service 16% 29% 9% 35% 11% Reduction in pay Use Case #1: Bus Automation for Maintenance and Yard Operations 9% 12% 24% 35% 20% Use Case #2: Low-Speed Automated Shuttles 17% 18% 15% 45% 5% Use Case #3: Automated BRT 14% 15% 14% 49% 8% Use Case #4: Automated MOD 14% 18% 14% 46% 7% Use Case #5: Automated Local Bus Service 13% 9% 16% 55% 7% Table 11. Perceived concerns by use case.

50 The Impacts of Vehicle Automation on the Public Transportation Workforce 5.2.1.3 Data Cleaning and Analysis A total of 129 valid survey responses were collected. The collected survey data were cleaned and analyzed by the research team using both qualitative and quantitative analysis techniques. The responses to the open-ended questions for each use case were coded and analyzed to identify the participants’ concerns and expectations. 5.2.2 Participants Table 9 describes the characteristics of the 129 survey participants. Among the participants who indicated their current job title, most (49 percent) were bus operators. The tenure of partici- pants in their current positions was relatively well spread across tenure categories. The purpose of the survey was to get the perceptions of the front-line workforce; of the 128 participants who answered the supervisory question (one participant did not answer), most (61 percent) did not supervise the work of others. 5.2.3 Highlighted Results The research team analyzed the open-ended questions and rating questions about the per- ceived concerns and benefits. Although there were 129 total responses, many respondents did not answer the questions about potential concerns and benefits. For the open-ended questions, between 38 percent and 55 percent of the respondents provided answers; between 42 percent and 58 percent answered the rating questions. Table 10 shows the results of the rating questions about perceived benefits of use cases, and Table  11 shows the results of the rating questions about perceived concerns. The percentages in the figures reflect the total percentage selecting the response option out of all completed responses to the particular con- cern or benefit. Overall, respondents reported that concerns were far more likely to materialize than potential benefits for all the use cases. With only a few exceptions, potential benefits were rated by 50 percent or more of respondents as extremely unlikely. The two potential benefits that had the relatively highest perceived likelihood (rated as very likely or extremely likely) were reduced physical demands and a chance to learn new skills. Generally speaking, the con- cerns that respondents rated themselves as being most concerned about (being very concerned or extremely concerned) were reduction in pay and job loss. Also, Use Case #5: Automated Local Bus Service tended to have a very high amount of reported concern (very concerned or extremely concerned) across all possible concerns. The research team identified repeated themes in respondents’ open-ended statements to the open-ended survey questions about the perceived concerns and benefits of each use case. The research team used these themes to code responses and count perceived concerns and benefits across all use cases. Table 12 includes the themes mentioned five or more times. For all the use cases, respondents were most concerned about the risk of unemploy- ment due to transit service automation, for example, “Please do not do this. Not only is it unsafe, but it takes jobs from people who need them. This will make my job hardly neces- sary or obsolete.” Also, respondents were concerned about safety issues, for example, “Unless on a dedicated arterial, the human needs to be in the seat to control for the unexpected hazards and safety.” Some respondents stated that the lack of operator-passenger interaction would reduce customer service, for example, “The connection I have with my passengers is valuable. We greet each other, they trust me, and they know me. My passengers should be seated and secured. Spoken with (not to) and they enjoy their trip in a safe and comfortable HUMAN environment. There is no replacement and certainly not value above that of human connection.”

Industry Engagement 51   On the other hand, a few respondents expected that automation would increase productivity and road safety. One respondent stated, “This may cut down on the small amount of accidents that occur on property. [This] may allow more efficient use of time doing serving or repair versus driving a vehicle.” Some respondents said that there will be “less stress from driving” and that work “might possibly be less stressful because of not having to deal with traffic.” 5.2.4 Summary of Survey Results Overall, the survey results suggest that front-line transit employees have significant concerns about transit service automation—and that those concerns are perceived as much more likely to occur than potential benefits. Both open-ended and rating questions suggest that the biggest concern front-line employees have is the potential for job loss. Although some potential benefits appeared to be recognized as possible for respondents, most respondents were skeptical that benefits for them would actually materialize. Use Case #1: Bus Automation for Maintenance and Yard Operations Theme Count of Theme Perceived Concerns Unemployment 43 No benefits 26 Unreliable technology 10 Perceived Benefits Increased productivity 16 Fewer car crashes 9 Convenient technology 6 Use Case #2: Low-Speed Automated Shuttles Theme Count of Theme Perceived Concerns Unemployment 30 No benefits 17 Safety 5 Lack of human aspect 5 Unreliable technology 5 Perceived Benefits (no themes with five or more counts) Use Case #3: Automated BRT Theme Count of Theme Perceived Concerns Unemployment 33 No benefits 9 Lower wages 5 Perceived Benefits (no themes with five or more counts) Use Case #4: Automated MOD Theme Count of Theme Perceived Concerns Unemployment 24 Lack of ability to serve people with disabilities 6 No benefits 6 Change of job duties 5 Safety 5 Perceived Benefits (no themes with five or more counts) Use Case #5: Automated Local Bus Service Theme Count of Theme Perceived Concerns Unemployment 32 No benefits 21 Safety 6 Perceived Benefits (no themes with five or more counts) Table 12. Themes extracted from the respondents’ statements.

52 The Impacts of Vehicle Automation on the Public Transportation Workforce 5.3 Industry Webinars The research team held two interactive industry webinars on June 16 and 18, 2020, to obtain feedback from the transit industry on the key planning and policy decisions that will have impli- cations on the workforce effects of transit vehicle automation (see Figure 9). (These planning and policy decisions were listed for each use case in Chapter 3.) The webinars contained live polls embedded into the slideshow using the Poll Everywhere® platform. The webinar was organized with the following sections: • Purpose of the workshop and the study. • Discussion of directly and indirectly affected transit jobs. • Foundational assumptions related to the study (e.g., workforce effects will not include impacts of fleet electrification) and a brief overview of automation-supporting technologies. • Instructions and a preview of how to participate in the live polling. • Presentations on each use case, including: – A brief overview of the use case. – Live polling of each planning and policy decision for the use case. In addition to poll questions on the use case planning and policy decisions, the webinar included some introductory questions about attendee characteristics (e.g., whether the attendee was from a rural, small urban, or large urban transit agency or some other organization) and experience with automated transit vehicles. Attendees were asked to answer the poll questions from the perspectives of transit agencies or organizations that they were working for, assum- ing that funding was not a limitation and the technology was fully mature and safe. Figure 9 presents the title slide from the June 16, 2020, webinar, and Figure 10 displays an example of a live polling slide. Attachment 6 contains the slides presented at the webinars, including all live polling questions. A total of 89 participants attended the seminars. Table 13 shows the number of attendees by agency type and the results of the introductory questions. Although the research team tried to recruit participants from a variety of geographic areas, unfortunately, there was only one attendee from a rural transit agency. Eight attendees were from a small urban transit agency, 35 were from a large urban agency, and 25 reported they were not from a transit agency. Figure 9. Title slide from the June 16, 2020, industry webinar.

Industry Engagement 53   Figure 10. Example of a live polling question and responses received during the industry webinar. There were 26 planning and policy decisions posed to webinar attendees. The planning and policy decisions could be grouped into two major categories: • Implementation questions: Questions that asked how likely and to what degree a particular use case would be implemented. For example: – How likely is it that [use case] would be used at your transit agency type? ◾ Very likely. ◾ Very unlikely. ◾ Not sure. – Transit agencies with multiple bus yards will deploy the bus automation for maintenance and yard operations to: ◾ All yards. ◾ Only select yards. • Operational and workforce impact questions: Questions that asked how use cases would impact transit service operations and ultimately the workforce. For example: – Transit agencies will mainly use [use case] to: ◾ Replace current routes or services. ◾ Create new routes or services, causing minimal impact on current transit routes. ◾ BOTH replace current services and create new services. – Will dedicated positions for human operators/attendants be needed, and if so, at what position-to-vehicle ratio? ◾ Yes—on board every vehicle. ◾ Yes—but at a reduced ratio (using teleoperators or stand-by operators). ◾ No—dedicated positions will not be needed (other staff/tech will handle tasks). The full results of all 26 planning and policy questions are available in Attachment 7. The research team used the data collected during the webinars to inform the design of the workforce effect calculator; however, since all the webinar attendees (except for one who was from a rural agency) were from either a small urban transit agency, a large urban agency, or not from a transit agency, the use of the data was limited and had to be combined with professional judgment and feedback from the research panel.

54 The Impacts of Vehicle Automation on the Public Transportation Workforce Agency Type Total Rural/Tribal Small Urban Large Urban Not an Agency No Answer Number of Attendees 1 8 35 25 20 89 Introductory Questions and Number of Responses Have you ever ridden in an autonomous shuttle? Yes 0 4 17 8 1 30 No 1 3 14 15 0 33 Not Answered 0 1 4 2 19 26 What type of agency or organization best describes where you currently work (or most recently worked)? Public transit agency or department 0 3 13 0 6 22 City or county government 0 0 1 2 0 3 Local or regional planning entity 0 1 1 0 0 2 State government or DOT 1 3 0 4 1 9 Labor union 0 0 0 1 0 1 Human services agency or similar non- profit 0 1 2 0 0 3 Private contracted transit provider 0 0 0 1 0 1 Transit software or hardware company 0 0 1 0 0 1 Other 0 0 8 13 5 26 Not answered 0 0 9 4 8 21 Table 13. Attendees by agency type and the results of the introductory questions.