The truth is on the 3rd of June. Like every day since February 2022, one of Cruise’s electric Chevrolet Bolts has been crossing the streets of the city of San Francisco in California. But while they were negotiating a left turn, the self-driving car was hit by a Toyota Prius coming from the opposite direction. Fortunately, there were no serious injuries, but the role of the autonomous vehicle was immediately questioned. Very quickly, the main fault fell to the driver of another car, a Toyota Prius driving over the speed limit (40 mph in a small zone to 25 mph, ie 64 km / h instead of 40 km / h). Especially since he was in the lane reserved for vehicles that want to enter on the right and he continued straight, thus causing the accident with an autonomous vehicle that blocked his path.
But the impact of the robotaxi operated by the company Cruise raises questions. The autonomous vehicle spotted a Toyota Prius coming up ahead. When the second one was driving in the right lane reserved for cars to turn, its software interpreted that the car would turn right behind it. As it approached to its left, the self-driving car realized that the Prius was going straight in the opposite direction as expected. He then braked hard to avoid an accident, hit from behind by a Japanese compact hybrid. And here’s the problem in this case: in their investigation into the accident, NHTSA experts estimated that the autonomous vehicle software failed to see it coming. The self-driving car’s ability to cope with sudden changes in direction has been flawed. Although this late change of direction is against the rules of the road, the accident can be avoided with a better interpretation of the data. Would a real driver have managed to avoid the accident when he saw the driver coming in his Prius? It is very possible.
Reminder to fix the program
Faced with the conclusion of the investigation, the Cruise company notes that its self-driving car was not at fault and that it was the behavior of the other driver that caused the accident. But the company will be recalling all of its robotics hubs in San Francisco to update their software to better deal with such situations in the future. Will artificial intelligence succeed in understanding and anticipating human unpredictability?