On the night of last March 18th, one of Uber’s self-driving Volvo XC90s hit and killed a pedestrian in Arizona as the “safety driver” focused on something in her lap (a phone…?) while carelessly taking only occasional quick glances up at the road. The crash prompted Uber to quit testing AVs in Arizona, and has spooked large swaths of the nascent AV ecosystem.
New information and analysis about the crash are answering the first wave of questions, and raising new ones. In past articles, we’ve analysed the available facts to grapple with the question of whether better headlamps could have prevented this crash (maybe so, but even the best headlamps won’t compensate for an inattentive “safety driver”). Now other questions have come forth. For example: could this crash have been prevented with existing driver vision technology other than headlamps?
Pedestrian deaths are up 46% in America since 2009, and 75% of them happen at night. Safety researchers want to find ways of driving that figure down, while AV researchers have been grappling with the need to teach robots how to drive in the dark and avoid people who wander into the road. How about night vision that can detect the heat of a human body? “If you have a sensor that could recognize something living, that information would be extremely useful to a computer,” says Jake Fisher, director of auto testing at Consumer Reports.