Impacts of Laws and Regulations on CV and AV Technology Introduction in Transit Operations (2017)

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.

OPERATING POLICIES 58 O P E R A T IN G P O L IC IE S

OPERATING POLICIES 59 Planning Policy Considerations Automation of buses and vehicles providing public transit service will begin to affect many local transit operating agencies within the next 5 to 10 years. Table 1 identifies estimated time- frames for the common introduction of AV technology by transit operating authorities. Over the next 30 years, all agencies will likely begin operating with AV transit technology in some services and many transit services could be operating with AV technology in all their services by that time. In the 30-year time frame, it is likely that the dispatching and operational fleet management for some agencies will also be fully automated. Policy makers in all transit operating agencies would benefit from addressing the incorporation of AV technology into agency and regional short-range and long-range transportation plans, even as AV transit technology is still in the developmental phase today. Infrastructure decisions made today could shape what is feasible in the future and agencies will want to make capital investments that can leverage the benefits of new technology. Transit agencies must coordinate closely with partner organizations such as metropolitan planning organizations (MPOs), state and local DOTs/highway agencies, FTA, and local county/city planning departments. Two types of planning decisions could be shaped by AV transit technology: 1. Long-Range Transit Planning, focused on planning decisions within the transit agency, and 2. Regional Planning and Coordination, focused on the coordinated planning needed between transit agencies and other regional planning organizations40. Long-Range Transit Planning AV transit technology should be included as a matter of policy dialogue within the top levels of any transit agency’s management during the near term to give direction to agency plans for 10, 20 and 30 years in the future. AV transit technology will likely provide the opportunities for fundamental changes in the way the public transportation is delivered, reducing costs and improving efficiencies. As such, transit agencies will want to begin thinking about how to prepare for these changes by: 1. developing or revising strategic visions to consider potential changes to service delivery (e.g., demand-response networks versus line-haul configurations), 2. identifying opportunities and threats posed by the new technology, and 3. identifying potential strategies for managing the changes. The transit operating agency’s long-range service expansion plans would benefit from considering the potential impacts of AV technologies on passenger service levels, potential capital investments and infrastructure needed to deliver AV transit service, and costs and benefits of these changes. AV technology may be applied to transit in a variety of service roles, which need to be better understood through technical research projects and associated policy studies. Key decisions about the required infrastructure in each of the settings are needed well in advance of the adoption of the vehicle technology, and long-range planning by transit 40 Note that NCHRP 20-102 (09) Providing Support to the Introduction of CV/AV impacts into Regional Transportation Planning and Modeling is currently on-going and anticipated for completion in 2018.

OPERATING POLICIES 60 operating agencies will need guidelines for these new AV transit applications as soon as possible. Types of AV Transit Services – Current thinking within the transit industry has discussion centered on the following types of automated transit services. High-frequency BRT or express bus routes, operating in semi- or fully- protected environments (e.g., dedicated transitways or managed lanes) – AV transit using L3 automation in transitway applications such as BRT lines is anticipated to be widely available in the 2020 – 2030 time frame. Securing the necessary right-of-way for transit lanes, whether within an existing roadway or for a new facility, is challenging and costly and will need considerable advanced planning. BRT stations must be planned to consider surrounding land uses and key connections to other transit lines and multimodal connections. For BRT or express bus routes that will operate on dedicated or managed lanes, special attention will need to be paid to the access points or for situations where transit vehicles will merge into and out of mixed traffic. Corridor planning studies and alternatives analyses should integrate AV technology into the modal and alignment screening process that leads to the selection of a Locally Preferred Alternative. First-Mile/ Last-Mile Applications in protected environments (e.g., campuses), in dense urban districts, or suburban/rural settings with low-speed operations (e.g., less than 35 mph) – AV transit is being discussed actively around the world as ideal technology to provide First-Mile/Last-Mile (FM/LM) connections to high-capacity transit lines and commuter rail service. L4 automation in FM/LM circulation service is possible today in campus environments. FM/LM L4 service in mixed traffic at low speeds on city streets within dense urban districts is anticipated to be widely occurring in the medium term (2025 -2035) time frame. In addition, Level 4 AV transit systems could enable expansion of frequency and coverage for what are typically less productive feeder bus routes, and in low density rural areas in the medium to long term. Conventional fixed-route transit operating in mixed traffic in unprotected environments like arterial streets – These environments could present some of the greatest challenges to full deployment of driverless transit vehicles. The typical city bus encounters numerous conflicts and obstacles including pedestrians and cyclists, illegally parked vehicles, and turning vehicles. These routes also carry heavier passenger loads and typically have frequent stops. In the near term, L2 automated driver assistance system (ADAS) features of AV technology could benefit bus operations by supporting schedule adherence and more effectively utilizing transit signal priority (TSP) to improve travel times and reliability. Long-range plans will need to consider the operating cost savings as well as the vehicle and capital costs of AV technology when setting policy goals for service expansion of AV technology into general bus lines. Other Long-Range Planning Considerations – As long-range plans begin to define the transit services of the future; additional details will need to be defined to better understand the capital improvement needs and long-term financial impacts. Service changes (i.e. demand-response) enabled by AV transit technology could affect a whole range of capital assets owned and maintained by agencies, including busways, stations and stops, vehicles, communications systems, and maintenance facilities and equipment. An integral part of service planning is addressing how public transit agencies would pay for AV technology deployments. Considerations must be given to the potential for AV technology

OPERATING POLICIES 61 having long-term operating cost savings but significantly higher up-front capital infrastructure costs. Examples of capital and operations and maintenance cost implications are for AV transit system dedicated lanes and stations with additional right-of-way, or a diversified vehicle fleet with more small vehicles to maintain along with larger traditional coaches. Some current research indicates that the on-vehicle equipment could perhaps pay for itself by reducing out-of- pocket costs for liability claims due to crashes involving public transit vehicles41. Finally, long-range planning of major transportation systems and infrastructure always greatly benefits from public outreach. This key element of the planning process should fully solicit input from customers, stakeholders and the public. These stakeholders must be informed of the benefits and potential impacts of AV technology. The transit agency’s planning process should therefore begin to include surveys and focus groups to determine what the transit users think about AV technology and the deployment of self-driving vehicles in public transit service. Regional Planning and Coordination AV technology deployment will require significant planning and coordination between transit agencies and other regional transportation agencies. This coordinated planning will most likely need to be done by the MPO. The development of the region’s Long-Range Regional Transportation Plan (RTP) will bring together the transit operator(s) with the highway operators and local governments responsible for arterial and urban streets and communities. Interagency discussions at the regional level should identify roles and responsibilities, opportunities for integrated technologies and shared facilities, needed infrastructure improvements, and potential safety impacts. The financial plan element of the region’s Constrained Long-Range Plan (CLRP) will require a realistic assessment of the capital and operating costs (and savings) associated with AV technology, expected funding sources, and trade-offs between investments in AV transit systems versus traditional rail and bus services. Some of the key policy areas where AV technology for transit would benefit from being addressed as part of the regional transportation planning process include: • Commitment to deployment of intelligent transportation systems infrastructure to enable and V2I communications • Preservation of right-of-way for dedicated lanes, and/or conversion of existing freeway and/or managed lanes • Congestion management • Planning for safe pedestrian and bicycle facilities in proximity to AV transit routes • Enhanced mobility of seniors and individuals with disabilities through coordinated health and human services plans • Air quality impacts of AV transit systems • Planning and design of intermodal hubs • Potential for land use changes 41 http://orfe.princeton.edu/mwg-internal/de5fs23hu73ds/progress?id=IIwQlPRptnVw3pMVwC5ypYo9uC- ocLTKfudSflQAdF4,

OPERATING POLICIES 62 AV transit lines will impact urban areas as provisions are made to accommodate the new technology. For example, potential changes to land use and development might occur when AV FM/LM transit service within dense urban districts is implemented. Curb lanes, for example, may need to be strategically protected for use by FM/LM circulators to berth in stations (suitably designed for ADA compliance). This will impact future curb parking, pedestrian facilities and building access in the immediate vicinity. AV transit may have significant impacts in the suburbs. By 2040 (and probably before, given the pace of development investment), it is likely that AV transit technology will be able to provide on-demand public transit service in real time using L4 unmanned vehicles to more cost- effectively serve very low density areas. This could significantly increase the ease and convenience of access to high-capacity commuter rail and BRT lines while reducing the need for large parking facilities at rail stations. It is currently quite a challenge to assess how AV technology will impact travel behavior, roadway capacity and traffic congestion in regional planning models since no reasonable scale real-world systems are deployed. Some predictions indicate that in the long term, AV adoption in privately-owned vehicles (or major adoption of shared-ride aTaxi services) may double the current carrying capacity of freeway facilities. Even without substantial gains in maximum capacity, the reduction in crashes will increase throughput and reliability of traffic facilities. Some researchers believe that such a future may (further) reduce the attractiveness of public transportation modes, particularly for suburban commuters. Others have expressed that expected increases in “no occupant vehicles” (NOV) moving through the system will prove counter-productive to decreasing congestion as vehicle trips escalate faster than person trips42. At the current time, the amount of uncertainty regarding the future transportation system “look and feel” is, at least in the opinion of the authors, at an “all-time high”; at least in the modern era since the completion of the interstate highway system. As prediction models are developed and refined based on case studies of real-world deployments, planners will better understand how particular AV transit supply services will affect mode choice and support the long-term mobility goals of a region. It seems reasonable that the availability of automated FM/LM transit services would help to make public transit a more attractive choice, but without success stories that document the actual mode-shift, the perceived benefits are still in the realm of conjecture (i.e. some handful of agencies are going to have to “go first”). Policy decisions of the local transit operating agency concerning AV technology deployment must be considered in the context of this uncertainty, and points toward the increased importance of coordinated planning across a region. All local and regional decision-making parties, including elected officials, will need to support AV technology introduction to put appropriate projects in the transportation improvement program (TIP), and then achieve successful deployments. This support must also exist within the board of directors for the transit agency itself for agency policy to be created. Many authorities have boards appointed by elected officials who have keen interest in union support, which may generate a reluctance to fund automation systems too quickly. 42 http://www.theatlantic.com/politics/archive/2016/01/will-driverless-cars-become-a-dystopian- nightmare/459222/

OPERATING POLICIES 63 Operating Policy Considerations Transit agency operations policies include the following: • Where and when services will be offered • What safety practices will be followed by employees and patrons • What activities of patrons are allowed or disallowed • How ADA and Title VI (civil rights) requirements are addressed • How union arrangements are handled • Employee rules, regulations, and duties • Emergency protocols • Technical operational agreements with local agencies such as signal priority Automation technology will affect all areas of operational policy. Operating agency management and boards will need to continuously assess how these areas are affected as AV technology development progresses. As safety and reliability issues are resolved by AV transit technology developers over time, it will become clearer how specific policies will need revision. Some example areas of operations policy are discussed in the following sections. Level of Automation Limits and Boundaries of Deployment NHTSA/SAE is anticipating that automated roadway vehicles will initially be offered for specific levels of automation by the manufacturers, but only within certain geographic areas (i.e. geo- fenced) and on certain classifications of roadways. The same will apply to transit services. If these AV buses from that manufacturer are to travel on other roadways not located inside the defined geo-fence network, then the vehicles will only be able to operate with a human driver onboard in L2 or L3 levels of automation. Policy decisions concerning staffing levels may be required to allow even these limited deployments in the example cited above, since human operators may still need to be present to assume the driving tasks over portions of the vehicle’s travel. This transition from automated driving to human driving requires provisions such as expanded roadway right-of-way, removal of curb parking and/or installation of additional lighting for safety and security of passengers. Policy decisions of the public transit operating agency would need to consider these aspects, along with coordination of the changes affecting urban districts/developments, local governments and regional transportation agencies. It is likely that AV transit policies will evolve in a regional manner, similar to how legislative rules and regulations on AV driving are evolving on a State-by-State basis in the U.S. Right-of-way, infrastructure and workforce deployment aspects will require policy decisions on capital investments, staging of operational personnel and interagency coordination that is significantly different than necessary with manual public transit vehicle operations.

OPERATING POLICIES 64 Policy Implications of Multi-Sourced System Components Automated vehicle technologies and connected vehicle technologies are being developed in parallel paths. The result of these parallel development paths will be procurements that could purchase different subsystems from different source-suppliers, and thereby impose some board policy aspects of decision concerning the risks of subsystem integration – particularly in the near to medium term for agencies that decide to become “early adopters” of the advanced technology systems. Management or board decisions at the local transit operating agency level could be involved in determining what kind of subsystem technologies should (or can) be purchased, since complete AV roadway transit systems may not be offered from a single source in either the near term (or possibly even long term). As market forces apply over the next 10 years, suppliers will be purchased by other suppliers, go out of business, change service and equipment models, upgrade to new technologies, terminate support for old technologies, and so on. Not unlike purchase of technology for any other purpose, policies will need to be established to deal with the realities of an emerging niche market. Operating-Fleet Management Plan The benefits of AV technology deployment provide good prospects for meeting customer service objectives while also meeting sustainability goals. With the flexibility that AV technology will provide over the long term, it will ultimately be possible to operate a diverse fleet with different vehicle platforms (size, weight, propulsion systems, and capacity) sized to meet the demand patterns of the service area. Over the long term, this will allow a high level of optimization in fleet operations when inefficiencies can be dynamically addressed throughout the day by dispatching smaller vehicles into service to replace larger vehicles. Operating-fleet management plans are typically based on ridership forecasts into the foreseeable future, but the paradigm shift that AV transit technology will bring also raises the importance of other policy- related factors such as: • Environmental and sustainability benefits can potentially provide a key policy rationale for the deployment of progressive levels of automation in roadway vehicle transit operations under a local transit operating agency’s policy goals and objectives. • The complexity of vehicle automation technology and diverse fleet mix scenarios that allow the type of optimization of operations currently envisioned will change capital costs as well as operating and maintenance costs in ways not be fully understood when early adopters deploy AV technology in the near to medium term. • Trends toward combining transitions to AV roadway vehicle technology with transitions to electrical propulsion systems will require new types of infrastructure for electrical power distribution. This will be combined with the need for new vehicle storage facilities and battery recharging requirements within the fleet operating plan. These and many other similar issues will become common considerations in fleet management planning. Policy makers will need to begin understanding and addressing the related new technical challenges as AV implementation progresses in the years to come.

OPERATING POLICIES 65 Safety Policy Considerations Implementing a safety program with a level of complexity which could equal those of the most complicated of rail systems is a major issue of AV public transit deployment. The complexity of AV technology will likely require the establishment of a rigorous Safety Management System43 under the FTA requirements. Of course, as with most FTA regulations, compliance will be a major factor in receiving federal funds for transit vehicles and infrastructure funding. Transit properties that only operate buses typically do not maintain a rigorous safety assurance program on the level of those properties that currently operate rail systems. In the future, all transit operating agencies operating AV technology will likely be required by FTA to put into place an SMS that is appropriate for operating an automated transit system. Local Agency Safety Programs Start with Policy For a comprehensive safety assurance program to be introduced, there needs to be a policy- level decision to commit the resources and to empower the employees who are assigned safety responsibilities to act when warranted. Under U.S. federal law, the local agency is responsible to execute the safety program, and to work under the guidance and regulatory jurisdiction of their state’s designated safety oversight agency. Safety Analysis and Risk Assessment – In part, the local agency’s safety program (in particular, the risk assessments of potential unsafe hazards), will likely be compared to that of the safety of vehicles under the control of a human operator. Policy makers should become actively involved in the discussion of safety and risk assessment if their plan is to prepare their agency for early adoption. Although there has not been any official assessment by FTA (or any other federal agency for that matter) of safety comparisons between human-driven roadway vehicles and machine-driven roadway vehicles, there have been some speculative assessments that machine-driven vehicles will likely be held to a much higher bar (e.g., “five times safer”). At this point in time, there has been no attempt to define precisely how safety will be measured and tested, and what minimum levels of safety will be acceptable for public transit applications. A policy-making entity like a board of directors will certainly have to deal with the public’s perception of their personal safety when riding in an AV transit vehicle. When AV transit is first deployed, the transit patrons will most likely only view this risk from their perception of safety and not from the statistical probability analysis. Policy makers should weigh both aspects considering such concerns to reassure the public that the system is safe. There would also have to be policy considerations of demonstration projects, public information programs, and focus groups to ensure that the transit users are prepared to utilize the new technology when it is deployed. Crash Liability Determinations in AV Operations – The risks and liability concerns for existing manually operated buses are well understood and incorporated into existing policy and contract terms and conditions (including collective bargaining contracts for vehicle operators). AV technology is heavily dependent on software program logic, sensors, and computers. Some of the related questions raised in the industry stakeholder meetings on this topic were: 43 Refer to Working Paper #2 Safety Assurance Consideration, Chapter 2 USDOT Safety Initiatives

OPERATING POLICIES 66 1. How will crashes that are caused by something within the operating environment but not caused strictly by an AV equipment or system failure be judged regarding operational liability of the transit agency? Was the AV technology not “good enough” to sense and safely respond to the hazard, and therefore the agency bears some responsibility in a way that is different from a human-operated vehicle? 2. If an accident is judged to be a software control logic failure, will that failure be judged differently by the courts if the vehicle is owned and “operated” by a public agency as compared to a vehicle privately-owned and “operated” by an individual? Will the operating agency be liable up to its legal limits, or will the manufacturer that employed the software programmer be fully liable? Can such protections be reasonably put into place to protect the operating agency? 3. How will the individual liability of the vehicle operator be judged when accidents occur under semi-automated levels of driving – either under automated driver assistance mode (L2) or L3 automation in which the AV technology may have given an alert that the onboard operator must retake control, but the operator did not respond? 4. Will definitions of liability for the vehicle manufacturer, the individual operator/attendant, and the public transit operating agency be determined by: a. The US legal system (courts)? b. Congress through legislative action? c. A combination of these two along with investigations of state or federal agencies to assess what caused the accident? In summary, policy aspects of safety assurance will likely have significant new guidance that comes from the current initiatives of the USDOT concerning “highly automated vehicles”. Contracted AV Taxis and TNC Services Considerations There has been considerable discussion within the transit industry about the contracting of ancillary services to TNCs such as Uber and Lyft. These contracts are being explored for FM/LM access services, and some aspects of demand-responsive transit services. There are a growing number of federal grants to transit agencies to test these contracted services under “mobility-on-demand” initiatives of FTA. Policy makers are being challenged to assess how TNCs and automated taxi service (a-taxis) can be integrated into the local public transportation system. The major issues affecting AV transit deployments that were discussed in recent workshops centered on the impacts of TNCs on public preferences for travel modes and the resulting implications for public transit operations. Other topics discussed included the trends in urban density/land use which may fuel TNC mode choice and the implications for long-range transportation planning and management, especially when the operational impacts of AV technology are considered as commuter person-trip patterns result in empty vehicle deadheading throughout the roadway system. The following is a sampling of the discussion topics now being considered within the industry: 1. Will TNC firms like Uber that deploy AV technology bring the demise of transit buses? Many feel that TNC firms deploying AV technology are more advanced than transit

OPERATING POLICIES 67 agencies in this field, but the need for transit buses and line-haul transit is not likely to go away as there will still be enough patrons with less money and enough time. 2. The use of TNC on-demand services could be a good solution for transit agencies to providing cost-effective access to high-capacity transit corridors with BRT or rail services, especially in low density areas where transit infrastructure is limited. But when will TNCs begin to view transit providers as competition more than as clients for their contracted services? 3. Will regional transportation plans view as undesirable the “rivers” of conventional-sized AV automobiles (i.e., solo or shared-ride services operated by TNCs) flowing into the city’s urban core during peak commute hours? Will the higher occupancy AV transit vehicles operating in dedicated transitways be more manageable for traffic congestion purposes? If so, then does the first-mile/last-mile TNC application make the most sense for transit? 4. Empty vehicle movement must be understood for TNC services when used for typical suburban commute trips. For example, typical park-and-ride transit operations have up to 50% deadheading of empty vehicles moving in the opposite direction from the peak commute direction to supply commuter service from the suburban perimeter into the urban core in the morning peak period, and in the opposite direction for the evening peak period. Empty vehicle trips for a-taxis and AV TNC autos will significantly increase vehicle volumes on the regional roadway systems with related congestion as empty vehicles are redeployed to serve the next commuter’s peak period directional trips. In contrast, the use of higher capacity AV transit vehicles would serve to reduce rather than increase congestion. Potential Regulatory Issues – If local transit operators can satisfactorily address these issues and can integrate TNCs into their overall transit services, there could be critically important policy issues concerning to how federal regulations and laws will be incorporated into the related contractual agreements. There are currently some contractual issues with TNCs for the provision of contracted services to public transit agencies that have been identified in industry discussion groups such as related sessions at the recent APTA conference. For example, it was reported that TNCs are reluctant to comply with FTA requirements such as driver drug testing. There may be little incentive over the long term for TNCs to partner with public transit agencies, particularly when these private companies could be subject to FTA regulations like ADA and Buy America requirements. In addition, the Title VI implications and the typical fares for TNC services could be difficult to satisfactorily align with typical transit fares. With the multiple demonstration projects now underway by transit agencies contracting to TNCs for local area services, there will be near-term understanding gained of the policy benefits and impacts of transit agency and TNC partnerships. The FTA regulatory aspects will need to be explored carefully as these demonstration projects proceed. Findings on Operating Agency Policy Service Planning – Policy decisions of the local transit agency concerning service planning will chart the course for when and how they will venture into AV transit technology deployments. Some may choose to be early adopters of advanced automation, and other agencies may decide to wait until full L4 automation is proven and readily available from multiple vehicle

OPERATING POLICIES 68 manufacturers. These decisions drive the agency’s input into the regional transportation plans and TIP funding commitments. Key to the policy decisions will be how AV technology is planned for deployment in ways that effectively increase transit mode choice, such as improving accessibility to high-capacity transit corridors through first-mile/last-mile transit services. Such deployments of AV technology by some agencies could be a major departure from their typical transit applications and legacy route configurations. These types of changes will require a policy-level decision to launch the planning process. AV Technology Applications – NHTSA is expecting AV technology to be offered by the manufacturer as safety certified for designated levels of operation (such as L4 fully automated driving) on designated classifications of roads within defined area boundaries (i.e. inside the geo-fence network). Policy decisions to implement AV transit will need to also address the reality that, when driving of the vehicles outside of the defined geo-fenced area, portions of the operations will necessarily require a human driver. Policy-level decisions will be required to accept the added complexity and associated risks with AV technology deployments in consideration of the many benefits that will also accrue. The integration of new technology into the transit operations, and the coordination with other governmental entities for placement of system equipment within the built environment, is a complexity many bus operators have never undertaken. Other issues like cybersecurity, large data management and creation of a sophisticated operations center need policy-level decisions to proceed with implementation. Safety Program – All public transit operating agencies down to the smallest bus operator who deploy AV technology will be required to establish a rigorous safety assurance program. Policy makers will need the necessary resources and give authority to personnel assigned safety responsibilities to execute a SMS acceptable to FTA and to the appropriate state safety oversight agency. AV Taxis and TNCs – The primary policy issues that may be faced at the local level concern the full compliance with federal law and regulatory requirements for public transit services when providers are using federal funds. The acceptance by the TNCs of the associated terms and conditions in the local transit agency’s contract concerning regulatory compliance could be problematic. Other aspects affecting the policy decision to utilize TNC contracted services include the potential competitive relationship between transit and TNC type of ride-sharing car services, if concerns of some in the transit industry prove to be correct.

OPERATING POLICIES 69 Research Projects and Policy Studies on AV Transit Operating Agency Policy – Policy decisions made at the local transit operating agency level will become increasingly more complicated due to the uncertainties of when, how and in what way AV transit deployments will affect passengers, employees and overall operations. The following key policy studies would be beneficial for undertaking based on the considerations and findings of this working paper: 1. Long-Range Planning Benefit/Cost Analysis Guidelines – Preparation of guidelines for near-, medium- and long-range service plans could be developed by this policy study’s resulting framework for decision making in long-range planning studies. The typical decision framework would typically consider the service benefits and operational/cost impacts of AV transit technologies, including the evaluation of operating fleet and infrastructure capital investments at a planning level of detail. The methodology would account for a variety of variables, including vehicle sizes and fleet requirements, types of service (first-mile/last-mile, high-frequency BRT corridor service, arterial street fixed-route service, etc.), and the associated levels of manpower needed to support the operations – refer to supporting work from research project recommendations noted in Working Paper #3. The policy study would address possible scenarios to be addressed, the sensitivity of the variables, and the associated benefit/cost implications through suggested analytical methodologies and assumptions. 2. AV Transit Service Types and Operational Planning Parameters – A policy oriented research study would be beneficial to develop definitions of service types that will become possible with progressive implementation of AV Transit technologies, including operational concepts, passenger service characteristics, vehicle fleet alternatives, dispatch complexity and other operational aspects. The preliminary service types described in the Working Papers could serve as the point of research initiation, including first-mile/last-mile (FM/LM) campus circulators; FM/LM urban district circulators connecting to high-capacity transit corridors; FM/LM systems in suburban and rural areas connecting to high-capacity transit corridors; high-frequency bus rapid transit lines; conventional arterial street corridor transit services; and regional express commuter bus lines. The complete spectrum of operational strategies to be addressed could include fixed-route, flex-route and demand-response (with real-time dispatching). 3. Benefit/Cost Analysis of Conversion from L3 to L4 Transit Operations – This research project could develop a benefit/cost analysis of L3 operations with an operator onboard in comparison to L4 operations with no onboard operator required. The comparisons should consider the reduction in operating costs of removing the employee from the vehicle, the reductions in insurance premiums, and associated reduction in legal costs and repair costs as crashes are reduced. Transitions and related safety and insurance/legal cost impacts under L3 operations could be researched to properly reflect the man-machine control transition issues as transitions to and from automated driving modes are continually occurring. This research might also leverage findings for moving from L0 to L2 (basic ADAS services) in drawing comparisons of incremental benefits from one tier of operation to another. 4. Cybersecurity Issues Affecting Transit Operating Agencies – Research would be beneficial to assess the necessary risks and technically feasible mitigations of securing the vehicle communications systems, the onboard processors supporting the communications with infrastructure, and the supporting software for AV control and localization which must be periodically updated through program downloads. In addition,

OPERATING POLICIES 70 research could address the risks and mitigations that are practical cybersecurity protection for the supervisory control systems and the operations control center operational support software. 5. Big Data Management– Research which analyzes the “Big Data” implications of video/audio, command and control signaling, vehicle system diagnostic data and “black box” onboard operational data recording for each vehicle in the AV operating fleet is a very important area requiring further study, as well as the assessment of rescaling of an agency’s data processing capabilities for small, medium, and large transit operating agencies. 6. Operations Control Center Features and Scale – A conceptual design study would be beneficial to define an operations control center and supporting local operations centers that could support an AV transit system for a hypothetical small, medium and large operating agency. Quantification of OCC personnel, roles and responsibilities, work stations, remote operator positions, communications and passenger support personnel, as well as management personnel could be accomplished for the conceptual OCC facilities. A description of the facility size and functional areas required to maintain AV transit operations for the small, medium, and large agencies would also be beneficial if prepared.