Autonomous vehicles (AVs) range from AMRs (autonomous mobile robots) such as picking machines that transport loads in a warehouse or on the dock to self-driving cars, which basically don’t exist in real life yet. The promise of AVS includes predictions of increased safety, but a new study from the Insurance Institute for Highway Safety (IIHS) and Highway Loss Data Institute (HLDI) suggests that self-driving features in cars don’t offer the improved safety levels they’ve promised.
Why would AVs be safer?
Human error is a leading cause of traffic accidents. AVs, which rely on advanced sensors and algorithms, have the potential to significantly reduce accidents caused by factors such as distraction, fatigue, and impaired driving. AVs are equipped with a range of advanced safety features, including collision avoidance systems, automatic emergency braking, lane-keeping assist, and adaptive cruise control. These technologies can help prevent accidents and mitigate their severity. AVs can process data from sensors and cameras in real time, allowing them to react to potential hazards more quickly than human drivers. They could even communicate with each other and with traffic infrastructure, potentially leading to smoother traffic flow and reduced congestion.
But are they really safer?
Most of these items don’t seem to improve safety, however. The features that IIHS considered “crash avoidance features,” including things like automatic emergency braking when a crash seems imminent, do seem to help cut down on accidents caused by human error. But adaptive cruise control, lane-keeping assist, and the like don’t fall into that category.
“Everything we’re seeing tells us that partial automation is a convenience feature like power windows or heated seats rather than a safety technology,” said IIHS President David Harkey.
This was particularly true for systems and features that were on-demand rather than automatic. Automatic crash warnings that were on all the time reduced crashes by as much as 49% in some cases. Things like adaptive cruise control, on the other hand, seemed to give drivers a false sense of security. They might even induce boredom, according to the study, and cause drivers to pay less attention than they should.
It seems that the illusion of autonomy in vehicles can actually increase danger rather than increasing safety.
Further data
Another study looked not at overall driving, but at accidents in particular. A study that analyzed crashes by self-driving cars found that they were less likely to have crashes under the most common conditions. They were twice as likely as human drivers to crash during turns, however, and five times as likely to have accidents at dawn or at dusk.
A March 2024 IIHS study found that 11 of the 14 driverless systems they tested earned a “poor” safety rating. This includes Tesla’s autopilot system. Lexus, the highest-rated system, only made an “acceptable” rating. No system was rated as good.
At this point, it doesn’t look very good for driverless cars. By contrast, AGV and AMRs are still in the “they ought to be safer” territory rather than the “we tested it and here’s the truth” stage. Information on autonomous forklifts, for example, routinely points out that the automated kind won’t be affected by fatigue or human horseplay. We don’t yet have evidence on the scale of the IIHS studies.
One cheering reflection is that warehouses and docks are more likely to provide ordinary circumstances most of the time. Roads with human drivers as well as autonomous vehicles, not to mention pedestrians, passersby, and the occasional wild animal, are probably a much tougher setting for any sort of robot to accommodate.
When your encounters with industrial automation are about Indramat drive and control systems more than autonomous vehicles, we can help. We are Indramat experts, ready to facilitate factory repair and reman of your Indramat motion control systems fast. Contact us for immediate assistance.