Report by David Gobaud, Michael Stolte and Francesca Svarcas
CIS summer research interns David Gobaud, Michael Stolte and Francesca Svarcas joined Resident Fellow Bryant Walker Smith at the Transportation Research Board’s Future of Road Vehicle Automation Workshop held at the University of California, Irvine. After attending a variety of presentations given by industry professionals, public officials and research scholars, the conference attendees participated in breakout sessions to generate ideas and propose research topics necessary to further develop automation technology. Each CIS intern selected breakout meetings that reflect his or her research interests and later convened to participate in a legal discussion specific to automated vehicle liability and risk. The summaries that follow highlight the key points shared in the following breakout sessions: (1) Public Policy Issues, (2) Driver-vehicle Interaction, (3) Technology Needs and Constraints and (4) Legal, Liability, and Risk.
Left to right: Michael Stolte, Francesca Svarcas, David Gobaud, and Bryant Walker Smith.Public Policy IssuesThe session started with a series of short presentations followed by an interactive roundtable discussion with the participants. Some of the topics discussed were: (1) how to acquire data that will help with the development and deployment of automated vehicles, (2) the role of government, and (3) how to collaborate with industries that will be impacted by automated vehicles such as the taxi industry. The session also included a lively debate over which model of automated vehicle technology to pursue - technology based on dedicated lanes or technology that operates on existing roads with human drivers. One participant commented that we are at a “hockey stick” moment and that the model of operating on existing roads (e.g., Google’s research vehicles) is going to be commercialized sooner than most think.
Driver-vehicle InteractionParticipants in this session pondered the relationship that a driver might share with an automated vehicle. In doing so, the group agreed that the success of an automated technology as sophisticated as an automated vehicle will depend largely on human factors. Safety and ease of use were ranked as the group’s paramount concerns. While some researchers said it would be safest and easiest for the user to operate a single lever that increases or decreases vehicle automation, other participants argued that the building of an automated vehicle necessitates separate systems or features that can be turned on and off at the driver's discretion. Everyone, though, worried that too many options would confuse today’s licensed-but-untrained drivers. The group determined that future research should explore the safety risks occasioned by automated driving, considering (1) driver expectations and abilities, (2) the nature of the handoff between manual and automated control, and (3) behavioral adaptation to automation.Technology Needs and ConstraintsSafety, reliability and predictability are absolutely essential. In addition to being more or less the mantra of many of the presenters at the conference, safety was the central focus of those who attended the technology breakout session. Engineers, AI research scholars, government representatives and others interested in exactly what is holding back the commercial sale of automated vehicles assembled to propose which areas of technical research are the most important to further development. The participants considered the maintenance of automated capabilities in all weather environments to be a top priority. Additional topics promoted for further study included pervasive and reliable vehicle localization superior to GPS and verification/certification to define and test safety for various levels of automation. Legal, Liability, and RiskThe legal session began with a group analysis of a hypothetical crash scenario posed by Bryant Walker Smith. This led to effective brainstorming about further research needed to understand how liability might work in a society that has adopted advanced automation technology. Favored topics included (1) a potential large scale shift in liability from drivers to manufacturers, (2) research into existing technology for the purpose of making analogies regarding liability, (3) what “control” would really mean for a driver-passenger of a vehicle with advanced automation, and (4) the “ownership” of valuable data generated by these vehicles. Thank You, CIS!David, Michael and Francesca, having enjoyed their summer at Stanford Law School, would like to thank CIS for its generous support and encouragement of their academic growth and for the opportunity to attend the workshop as a grand finale experience of the legal and technical aspects of automated vehicle technology.