Autonomous Cars Cannot Completely Eliminate Crashes

A new study suggests that Autonomous Vehicle technology may not have the capacity to prevent all accidents caused by human error. The research conducted by the Insurance Institute for Highway Safety (IIHS) analyzed more than 5,000 police-reported crashes in the United States. Results revealed that autonomous vehicles can only prevent about one-third of crashes. According to data, human error accounts for around 94% of crashes in the country. This includes incorrect assumptions about the actions of other drivers, inappropriate speed in certain road conditions, and wrong evasive movements. The researchers also found that human incapacitation, such as driving under the influence and falling asleep while driving, can also lead to crashes. IIHS vice president for research and study co-author Jessica Cicchino explained that while it is possible that fully autonomous vehicles are better than humans at detecting hazards, they will not suffice when it comes to preventing the majority of crashes. Despite advancements in the development of sensors, cameras, and navigation tools, autonomous vehicle technology is not able to predict or eliminate the mistakes of drivers. The researchers added that crashes can only be significantly reduced if autonomous vehicles are programmed to prevent errors related to drivers’ performance and decision-making. According to Cicchino, more crashes can also be prevented if autonomous vehicles have the capacity to follow all traffic rules and speed limits. However, she warned that if these vehicles are programmed to drive and react like humans, who tend to disobey traffic laws and speed limits, then fewer crashes will be prevented.