On March 18, a self-driving Uber car crashed into a 49-year old woman who was crossing a four-lane road in Tempe, Arizona. Unfortunately, the women died at the hospital. The Governor of Arizona has blocked Uber from testing its self-driving cars on public roads. On Monday, the head of Mobileye said that its computer vision system software would have detected the woman that was hit by the self-driving Uber SUV. Mobileye is a vision system company owned by Intel Corp.
Amnon Shashua, a chief executive officer of Mobileye, said that his company’s software would have efficiently detected the pedestrian who was killed by the self-driving Uber car and questioned the safety of the autonomous vehicles that will hit the roads soon.
Shashua criticized the new entrants in the autonomous vehicle field for not ensure full safety of their self-driving vehicles before hitting the public roads. It said that safety should be the main priority of these self-driving cars and the companies need to go through years of research and development to ensure that. To get more details about the safety lag of the self-driving Uber vehicle that killed the pedestrian, Mobileye took the dashboard camera video that was released by the police last week and ran its advanced driver assistance system (ADAS) through the videos.
In a blog that was published on Intel’s website, Shashua wrote that although the quality of the image of the police video was quite low, Mobileye’s ADAS technology successfully managed to detect the pedestrian approximately one second before the impact. The woman, named Elaine Herzberg, was seen pushing the bicycle across the road when the accident happened. So, Mobileye is saying that its high-tech and well-developed sensors would have detected the lady just one second before the crash. “Experience counts, particularly in safety-critical areas,” wrote Shashua.
According to Mobileye, companies like Uber are a new entrant to the autonomous vehicle field and also the self-driving technology they use is new and cannot challenge the decade-plus experience of incumbent computer vision experts. Shashua then wrote, “I firmly believe the time to have meaningful discussion on a safety validation framework for fully autonomous vehicles is now.”