Self-driving Cars Not Cure-all For Crashes

Self-driving vehicles have been heralded as eliminating driver error and providing a safe transportation future. But autonomous vehicles may prevent only a third of all car accidents if their automated systems work too much like humans, according to a recent Insurance Institute for Highway Safety study.

Motorist error is the final failure in the sequence of events leading to over nine out of 10 vehicle accidents, according to a national survey of 5,000 police reports. Only around a third of those accidents resulted from mistakes that automated vehicles would be expected to avoid because their systems have more accurate perceptions than humans and are not susceptible to incapacitation.

To avoid these mistakes, autonomous vehicles must be programmed to emphasize safety over speed and convenience. Five driver-related errors should be addressed.

First, sensing, and perceiving errors account for 24 percent of surveyed crashes while incapacitation is associated with 10 percent. Sensing and perceiving includes driver distraction, blocked visibility, and late recognition of hazards. Incapacitation is a driver impaired by alcohol, drugs, medical problems, or fatigue.

An Uber test vehicle which killed a pedestrian in Tempe, Arizona in March 2018 is an example. Its automated driving system had difficulties with identifying a 49-year-old woman on the roadside. Once it made this identification, however, the system did not anticipate that the pedestrian would cross in front of the vehicle and perform the correct evasive maneuver.

Elimination of these accidents requires sensors and systems that work flawlessly and never malfunction. Even if this were possible, the remaining two-thirds of these accidents would still occur unless all autonomous vehicles were programmed to prevent other types of predicting, decision-making and performance mistakes.

Planning and deciding errors played a role in approximately 40 percent of accidents. These include deliberate driver decisions such as speeding and illegal maneuvers which can conflict with the safety priorities programmed in autonomous vehicles. Vehicles should be designed to place safety over unsafe driver preferences.

Autonomous vehicles must also comply with traffic laws and adapt to travel conditions. Vehicle should have programmed strategies that account for behavior by other drivers involving low visibility and slowing down for high pedestrian traffic.

Until autonomous vehicles take to the road, an attorney can help seek compensation for victims of accidents by motorists who engage in these unsafe behaviors and other unsafe driving. Lawyers can also address accidents involving new technologies.