Alliance Innovation Laboratory -- Silicon Valley
Senior Researcher & Manager / September 2015 - now
At AIL-SV, formerly Nissan Research Center, I joined Melissa Cefkin and Gitte Jordan, both from my IRL days to work on social aspects Autonomous Vehicles. The organization was run by Maarten Sierhuis, whom I also knew from an IRL project with NYNEX; it is a small world. For some time, I worked on the Intention Indicator (or e(xternal)HMI), a concept that arose out of the concern for the challenges Autonomous Vehicles may pose for other road users. If a pedestrian is approaching an intersection and a vehicle without a driver arrives at the same time how could the pedestrian be sure that the AV would stop and indeed remain stopped to let her cross? As part of this research I did some work with a colleague at UCSD in which she drove a regular car but was disguised as a car seat (she constructed the suit from a seat cover, and used the same to cover the passenger seat). Interestingly, it turned out that most people did not even notice that there was no driver. This experience led me to the notion of "normal traffic assumptions," the idea that when people encounter other road users, they trust that those other road users see and experience the world in the same way they do. This is a basic assumption shared among human road users, and indeed it is the very foundation of our behavior on the road, and the root of our sense of safety on the road. Moreover, and ironically, although our experiment with the seat-suit demonstrated that people will assume the same about a vehicle without a driver, Autonomous Vehicles do not share the human experience of the world, they work on entirely different principles. Encountering an AV might be usefully compared to encountering a deer on the road: you know they are quite capable, but you also know they do not understand the world in the same way we do. So when a driver encounters a deer, they slow down. But when people see an AV, which looks much like another vehicle, they won't make this assumption. Aside from the work on the intention indicator, I also conducted numerous observations of drivers and driving in various international settings, to examine if cultural differences in driving needed to be accounted for in AVs. I also conducted a number of experiments evaluating different HMIs developed for AVs.