For many people, there’s a disconnect between what we hear about self-driving vehicles and how we feel about them. We’re accustomed to being in control of the vehicle, and it’s often uncomfortable to think about turning over operations to machinery–particularly at high speeds and in traffic. Still, the National Highway Traffic Safety Administration (NHTSA) and other agencies say autonomous vehicles have the potential to reduce our risk on the road.
So, which is it? Do autonomous vehicles protect us, or create new risks?
One large study conducted in 2024 says the answer is both.
The study, conducted by researchers at the University of Central Florida, looked at crash data from thousands of collisions. Researchers concluded that autonomous vehicles are, in fact, less likely to crash…usually. However, those autonomous systems still need work in some areas. While the data showed self-driving cars were safer most of the time, there were two significant exceptions. First, autonomous vehicles were nearly twice as likely to crash when turning. Their performance in low-light settings was even worse. Self-driving cars were more than five times as likely to crash in dawn or dusk-like lighting.
Of course, the technology is still evolving, and those weaknesses may soon be addressed. For now, though, it seems that drivers whose vehicles have autonomous capabilities are best off keeping their hands on the wheel in certain circumstances.
Not All Self-Driving Cars are Created Equal
It’s also important to note that different autonomous systems use different technologies. That means some may be more effective than others in certain circumstances. One significant difference is the type of sensors used to determine when there are obstacles in the road, including other vehicles, pedestrians and bicyclists.
Most autonomous vehicles use some combination of ultrasonic sensors, cameras, radar and LIDAR. Each of these mechanisms works differently, and combining technologies can increase the vehicle’s ability to identify hazards and react in time. Of the five largest manufacturers of autonomous vehicles, three (Waymo, Cruise (GM) and Motional) use a combination of cameras, radar and LIDAR. A fourth, Zoox, employs those three systems along with long-wave infrared technology. Just Tesla relies on a single system.
Since 2021, the company has been phasing out ultrasonic sensors to rely entirely on its camera-based Tesla Vision system.
How Does Tesla Vision Stack Up?
There are many variables, and different autonomous technologies may be more effective under different circumstances. However, recent side-by-side testing conducted by engineer Mark Rober, showed some drawbacks to relying entirely on camera technology.
Rober’s testing involved trials in different types of lighting and visibility conditions. For each of the tests listed here, Rober compared a LIDAR-based autonomous vehicle with a camera-based Tesla on autopilot.
- The first trial mimicked a child dashing out in front of the car. Both the camera-based Tesla and the LIDAR-based car stopped in time.
- The second mimicked a child standing in the road in foggy conditions, sufficient that the human driver couldn’t see the child at a distance. Neither could Tesla’s camera-based system. The car with LIDAR technology stopped in time, but the Tesla didn’t even slow.
- In heavy rain conditions, both vehicles were slow to take note of the child. However, the LIDAR-based car stopped just in time, while the Tesla did not.
- With sun directly in the driver’s face making it difficult for a human driver to see the road ahead, both vehicles stopped just in time.
- When Rober placed a wall with a painted landscape across the road, the LIDAR-based system recognized it as an obstacle and stopped, but the Tesla crashed through it without slowing.
Autonomous Vehicle Accident Cases are Complicated
Many variables are in play in any car accident case, but that’s even more true when an autonomous vehicle or a vehicle with driver assist technology engaged is involved. In some cases, the human driver may be wholly or partly liable, despite the presence of the technology. In others, the human may be entitled to rely on what the technology should do, and the injured party may have a product liability claim. And, of course, there are all of the other possible factors that apply in any car accident, such as negligence of other drivers on the road, debris in the road, and faulty road design or maintenance.
If you’ve been in a motor vehicle accident with an autonomous or driver-assist equipped vehicle, you should speak to a Charleston car accident lawyer as soon as possible.