Experienced Injury Lawyers
During the COVID-19 Pandemic, we're offering Free Virtual Consultations

How Many Fatal Accidents Can Autonomous Vehicles Really Prevent?

Published on Jul 29, 2020 at 3:54 pm in Auto Accident.

Car on slick road on a cloudy day

Autonomous, or self-driving, vehicles identify hazards while driving on a more objective level than humans do. When a person is behind the wheel, an accident may be the result of driver error. Ideally, self-driving vehicles eliminate human error. But according to a recently published study by the Insurance Institute of Highway Safety (IIHS), although self-driving cars will be able to better identify hazards that could cause crashes, that still might only prevent around a third of all crashes.

When crashes involving autonomous vehicles happen, it can be difficult to determine liability. Much like car accidents between regular vehicles, autonomous car accidents require a full investigation in order to determine who was at fault for the collision. At Thomas Law Offices, we are dedicated to ensuring your rights are protected by figuring out who is at fault for an accident that was caused by an autonomous vehicle.

Can Autonomous Vehicles Eliminate Driver Error?

As mentioned above, the study released by the IIHS claims that autonomous vehicles would only reduce car accidents by one-third. Although these vehicles fix the number one cause of car accidents—driver error—they still might misidentify hazards or have other errors which could lead to accidents. Autonomous vehicle companies argue against that number and claim that driverless vehicles would prevent nearly two-thirds of accidents—double what IIHS predicts.

However, Partners for Automated Vehicle Education (PAVE) calls attention to how these estimations are purely speculative. The IIHS study was completed by breaking down causes for accidents with human drivers into five categories. Then, each category was evaluated to see how a self-driving vehicle could prevent accidents in that category. The categories of errors are:

  • Sensing and perceiving
  • Predicting
  • Planning and deciding
  • Execution and performance
  • Incapacitation

IIHS predicted that autonomous vehicles could reduce or eliminate crashes caused by sensing and perceiving and incapacitation errors. That’s where the one-third came from—not from direct testing, but rather from number crunching and predictions. Of course, the merit of the predictions based on data cannot be negated. They could prove to be accurate, but only time will tell. Although the overall goal is to eliminate all fatal crashes, engineers of autonomous vehicles recognize that not all accidents can be prevented.

What Could Lead to Autonomous Vehicles Crashing?

Vice President of Autonomous Vehicle Standards at Intel Corp.’s Mobileye, Jack Weast, said that “crashes will never be zero until we have no more human drivers on the road.” But we have seen times when autonomous vehicles made mistakes that have led to accidents and even death. Machine learning could help reduce accidents, but reducing accidents does not mean eliminating them completely.

In 2018, a pedestrian was killed by a self-driving Uber in Tempe, Arizona. The vehicle did not identify her as a possible pedestrian who would walk across the road. When she did cross the road, it also failed to identify her and perform an evasive maneuver. Instead, the vehicle struck and killed her. Although this is only one example, misidentifying hazards could be one of the reasons that autonomous vehicles crash and the traffic fatality rates don’t fall as much as expected.

One factor that could decrease the safety of autonomous vehicles is if programming becomes similar to how humans drive their cars. Actions like speeding, making illegal turns, and running stop signs are all decisions human drivers make when they are prioritizing their destination over safety. Autonomous vehicle manufacturers say that they intend to prioritize safety in the cars’ programming, but surveys suggest that riders want their car to perform similarly to how they would drive, which would mean prioritizing speed.

In cities where snow and ice accumulate on the roads during winter, autonomous vehicles might not react as well as humans would. People know to proceed slowly and with caution when the roads are covered with snow, but an autonomous vehicle could have trouble sensing the road if it is covered. If the sensor is bothered by falling snow as well, it could have trouble figuring out where fixed objects are.

Thomas Law Offices Can Represent You

Which side is correct? The side who thinks autonomous vehicles could reduce traffic fatalities or the side who says we can’t be sure just yet? In this case, both could be. Unfortunately, we don’t know the answer yet because the statistics we see here are predictions based on present data. We can only see over time just how many accidents autonomous vehicles cause or prevent.

In the meantime, while autonomous vehicles are still being tested and used in certain cities, you might encounter one and get into an accident. When that happens, it could be complicated to determine who was at fault for the accident since the vehicle was driving itself. At Thomas Law Offices, our Louisville autonomous car accident lawyer understands the complexities of your case and can help you through the legal process. Reach out to our office today so we can get started on your claim.