Mobility on Demand and Automated Driving Systems: A Framework for Public-Sector Assessment (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.

107 REPORT APPENDIX A: STAKEHOLDER ENGAGEMENT On April 17, 2019, at the Transportation Research Board’s 2019 International Conference on Demand-Responsive and Innovative Transportation Services, the National Cooperative Highway Research Program (NCHRP) hosted a stakeholder engagement session. The session was hosted by Gustave Cordahi, who also presented, and Diego Lodato, who co-presented for segments of the meeting. The purpose of this listening session was to 1) share findings from the recently published state-of-the-practice assessment, Mobility on Demand and Automated Driving Systems: A Framework for Public-Sector Assessment, 2) discuss the impact assessment framework—a structure for assessing the impacts of MOD and ADS deployments, 3) share MOD-ADS use case scenarios outlining the circumstances of public-agency engagement with specific MOD providers, and 4) to collect useful and actionable feedback and contact information from a diverse set of invested stakeholders in attendance. The session was a success: many helpful and productive comments were collected from both the scheduled audience participation activities as well as discussions generated during the presentation. Below is a summary of stakeholder feedback from the session. PRESENTATION RESPONSES AND DISCUSSION The session began with some preliminary questions from the audience, a few of whom were new to some ADS-related terminology, likely because attendees from multiple countries were present. Cordahi explained the differences between ADS and advanced driver-assistance systems, establishing a common baseline for the stakeholders in attendance. Slides of the presentation listed examples of SAV business models. These examples included the SAV business model titles, the proposed vehicle ownership and network operations, descriptions of the business models, and current, non-AV examples. In response to these slides, stakeholders began a discussion expanding upon the subject. One stakeholder mentioned that the examples shown seemed “fine for some current business models,” but encouraged the group to collectively think about what the potential role of government and public-sector entities might be in the business model examples shown. The same individual pointed out that it would be difficult to identify a business model without thinking through the public-sector aspect of the model, which was not reflected in the examples on screen; for example, asking whether the city would be the fleet manager for the SAV service. Another participant indicated that the business model of a local government does not operate on the mission of getting people “from point A to point C” but of connecting people with transportation networks. They suggested that “newer experience will be more productive” —that is, a government might use a business model in which AVs are used for the peripheral routes and supporting transportation networks to decrease traffic congestion, which could also be reflected in the examples on screen. Additionally, a participant mentioned the unique struggles that the SAV business models would present in rural areas: jurisdictional lines may obstruct the practical concerns of installing an SAV system. Several attendees voiced concerns that the public-sector and nonprofit entities should be more prominently included in the SAV business model example considerations. This, in turn,

108 REPORT prompted a productive conversation regarding accessibility issues, raising questions including: “Who will be able to access these vehicles? How will we ensure they are accessible to users of different wheelchairs, for example?” Other questions that were important to the stakeholders in attendance included: relevant aspects of equitable service, the environment, and the impact of necessary but potentially limiting policy considerations of SAV business models. Specifically, stakeholders agreed that the purpose of policy drafting and revisions needs to be clear and agreed upon before acting, and that those involved in policy decisions may need to determine whether it might be more helpful to remain within current policy guidelines and change certain, smaller aspects of those policies to accommodate new technology. Audience members also emphasized the importance of flexibility in policy and noted that space should be set aside for the private sector to innovate and bring developments to the table: “[If we] enable the private sector to succeed at what it does and provide services, [then that will] make it more equitable for everyone.” Connected to this idea is the expanded concept of flexibility in reference to the business model examples as a whole; that is, establishing frameworks instead of rigid policies to allow for fluctuating requirements between different technologies. Further questions from the session’s audience surrounding the SAV business model examples included: 1) asking about the potential role of the transit authority or operator in regulation, 2) increasing the clarification of vehicle ownership between government and public authorities, 3) the variety of service levels, and 4) the possibility of opening a free market. Finally, there was a request for a sector-wide specification of terminology, making vocabulary common across entities. For example, attendees recommended clarifying the differences between “business” and “operating” models. Later in the presentation, session participants expressed interest in learning more about the consequences and implications of AVs being integrated into the transportation system. Many participants also requested further elaboration on current MOD and AV deployments after Cordahi described examples, such as carsharing programs in Berkeley, California , and Vancouver, British Columbia. Specifically, these MOD and AV partnership programs involving developers and property managers were practical examples of the challenges of implementing and regulating SAVs. The concept of ridesharing was also important to participants, who discussed the “equation” of balancing impact on the personal vehicle owner and the road network, and this ignited a conversation about travel time as a crucial aspect of exploring new technological possibilities: travel behavior and utility of travel will inevitably be different when SAVs become commonplace. For instance, an individual participating in ridesharing may experience longer travel time, but the quality of that travel time may be improved. However, this may also drastically change the way travel is discussed and estimated: travel time, as opposed to trip distance, is commonly used as an estimate for trip length, but this metric may be affected by future technological implementations. As the presentation discussion continued, it became more focused on accessibility, an important concern that was touched on earlier in the presentation but not explored in depth by session attendees. One individual pointed out that both visible and invisible disabilities exist, as well as the potential future for AV services that could accommodate the needs of people with disabilities. The speaker pointed out that people with disabilities that are not immediately apparent are, for the most part, excited about the potential of MOD and AVs and what technology can accomplish, since many of those people are already using similar capabilities to assist in their daily lives. However, individuals with more visible disabilities, such as those using

109 REPORT wheelchairs and other mobility devices, were concerned about physical implications and were cautious about assistance for getting in and out of a vehicle, for example. Another participant, though, noted that labeling groups of individuals with disabilities into those with disabilities visible and invisible could easily lead towards a “separate but equal” system. Yet another audience member, representing a public-private partnership between a local government and a private company, stated that his organization’s services offered separate but equal accommodations for those with disabilities (visible or invisible) and those without. In summary, one participant commented that “perfect mobility service is not a perfect driver but a perfect service provider. Our vision is to create an environment where the mobility system serves paratransit,” potentially leading to the elimination of paratransit because the entire transportation system is accessible. USE CASE AND ROUNDTABLE FEEDBACK After the presentation, session attendees were encouraged to participate in a written activity at their tables, which sparked further discussion and provided feedback for the team. Each table was given a specific MOD and AV use case to explore, and each table group recorded their feedback on provided charts for a variety of different impact categories. The use cases included: On-Demand Automated Shared Rides, Flexed-Route SAV, and Automated First- and Last-Mile Service. The group for the latter use case offered detailed feedback, described below. At the Automated First- and Last-Mile Service table, the group offered extensive explanations for each of their suggestions and comments on the provided activity chart. A representative of the group shared that a heated discussion had taken place regarding whether automated buses were a realistic possibility, as bus drivers do far more than just drive. Currently, drivers perform both driver and service tasks, and AVs have only been able to perform driving-related tasks so far. The service aspect remains, so some group members suggested that a human attendant should be present who can perform service tasks—such as assisting elderly or blind passengers. The group also discussed utilizing AVs in bus rapid transit. In this use case, there would not be human drivers or pedestrians, so AVs could drive considerably faster than on a traditional roadway. AVs may also be able to travel close to one another due to lack of human error and may not require expensive new infrastructure, creating a mode of transportation akin to high- speed rail but with bus-like AVs. One use case that this group did not elaborate upon in their chart included sidewalk delivery robots. The machines were small, safe, and nonthreatening, and the group noted that a large-scale deployment of these robots could remove larger vans that congest city streets. In terms of less publicly accepted AVs, the group described trials in which qualitative data was collected on AV opinions. Before individuals had exposure to AVs, they typically harbored negative assumptions about them. However, once they were able to board and ride one, those negative opinions disappeared. This, the group pointed out, indicates that trying to convince the public of the safety benefits of AV may be less helpful than simply introducing as many people as possible to operational AVs in person. In terms of the economy and labor, the group cited job creation and upskilling maintenance workers to operate EVs as potential impacts of AV implementation. AVs require maintenance,

110 REPORT such as battery charging, battery replacement, tire rotation, tire pressure, and other standard vehicle maintenance. The AVs may need inductive charging or to be swapped out with another for maintenance to be performed; the group noted that AVs could not run constantly. Additionally, in some extreme climates, air conditioning and heating could use as much power as the motor of the AV itself, considerably reducing the operating ranges of AVs. Further comments for this group are included in the representation of their completed activity charts in Table 18, Table 19, and Table 20. STAKEHOLDER NOTES ON MOD-ADS USE CASES Table 18, Table 19, and Table 20 summarize the activity charts completed by attendants of the stakeholder engagement. These notes summarize feedback on various MOD and AV use cases.

111 REPORT Table 18. MOD and AV Use Cases: Various from Stakeholders Additional Use Cases Description Bus Rapid Transit If AVs are separated from human drivers and pedestrians, the vehicles can travel much faster than human drivers, and more safely. Potential huge efficiency and travel time improvement. Universally accessible vehicles People with disabilities that cannot access a sedan vehicle could use accessible AVs. It is important that AVs are “born accessible” with low floors, ramps, and places to secure chairs. Emergency vehicles SAVs as ambulances, can they ever be human-less? (I.e., people assisting the person in need are no longer needed) Currently, the human vehicle operator has two tasks: 1) to drive and 2) to provide other services. If we automate task one, we still need to provide task two sometimes. Sidewalk AVs for delivery Small delivery robots may remove the need for “career vans.” Due to low speed and mass, they are safe to deploy and do not scare people. “ALL” Universal design—requiring or otherwise encouraging that all vehicles with ADS be accessible for people with disabilities, people with luggage, people with strollers, and people with grocery carts or other wheeled containers. Standards for accessibility for human- machine interfaces Encouraging regional approaches so that travel is not hindered by limiting services to county, city, or other jurisdictional lines which restrict regional mobility. If vehicles are not all accessible: we perpetuate a “separate but equal” transportation that remains inherently unequal.

112 REPORT Table 19. MOD and AV Use Cases: On-Demand Automated Shared Rides Impact Category Relevant Impact Assessment Questions Behavioral Shifts What incentives need to be in place to influence or encourage travelers to use an on-demand automated shared ride? Economics and Labor N/A Environment and Energy N/A Equity and Inclusion and Accessibility Should not be assumed Infrastructure and Land Use N/A Operational Efficiency What will on-demand automated shared rides do when there are no passengers? Public Perception What level of education is required to influence travelers to use automated shared rides? Public Transit Ridership On-demand Technical Accomplishments N/A Traffic Congestion N/A Traveler Safety N/A Table 20. MOD and AV Use Cases: Automated First- and Last-Mile Service Impact Category Relevant Impact Assessment Questions Behavioral Shifts Little needed; just education, exposure to AVs, demystification Economics and Labor New jobs in maintaining AVs Environment and Energy May need induction charging or to swap vehicles (especially in heat and cold  climate control uses as much energy as a motor) Equity and Inclusion Accessible; may need voice activation/Braille for doors Infrastructure and Land Use Need to separate humans from AVs so that AVs can travel faster than eight mph Public Perception Need to accept that AVs will sometimes injure or kill due to unpredictable human behavior Public Transit Ridership Will this improve economies of transit hubs? Technical Accomplishments AVs currently very expensive due to LIDAR; solid-state LIDAR is coming Traffic Congestion Should improve if replacing SOVs at park and ride Traveler Safety Need remote surveillance, help/panic buttons

113 REPORT STAKEHOLDER ATTENDANCE Twenty-seven individuals are recorded as participating in the stakeholder engagement. The following list comprises the affiliations of stakeholder attendants: • CCTA • Goin • Liftango • MCA • MemEx • Miami-Dade Transportation Planning Organization (TPO) • Michigan DOT • Mobility Carsharing • Metropolitan Washington Council of Governments (MWCOG) • Ph.D. Student • Pivot Strategies, LLC • PTV Group • Rutgers CAIT • Schweiger Consulting • San Francisco Municipal Transportation Agency (SFMTA) • Share Now • Shared-Use Mobility Center • Texas A&M Transportation Institute.