When it comes to self-driving cars and trucks, one of the big challenges for developers is programming the vehicles to operate both predictably and safely. Right now, most autonomous vehicles (AVs) operate according to computer algorithms programmed by technical experts. But an Israeli startup says it has found a better way to teach AVs how to drive—copy the locals.
Tel Aviv-based Nexar has created a system that uses crowdsourced dashcams to study the way people drive their own cars and then “humanize” robot drivers. The company’s “Driver Behavioral Mapping” data can be fed into its artificial intelligence (AI)-based computer vision platform, and then used to help AVs stop, merge, turn, and drive the same way humans do. Although drivers in some U.S. cities have poor reputations, the law of averages “irons out” those bad habits, Nexar says.
That information is collected via cars outfitted with Nexar’s smart dashcams, which currently drive over 160 million miles per month, covering all 50 states. According to the firm, AVs will use that real-time data to “learn” how the locals drive, acclimating to the local culture.
“A self-driving car that drives only according to a raw map would be an immediate danger due to its robotic style of driving,” Eran Shir, co-founder and CEO of Nexar, said in a release. “It’s not necessarily that humans drive better than robots; it’s that AVs need a lot of human data obtained by those who have driven through a particular area. Without even being aware, we make hundreds of decisions that adapt to local conditions, culture, and comfort each time we get behind the wheel. AVs need to sync into this behavior in order to provide the most secure and comfortable ride.”
