Distracted Drivers Learn to Skirt Limits Set by Partial Automation Systems, IIHS Says
We know that drivers are more likely to multitask when using semi-autonomous driving systems, but new research from the U.S. Insurance Institute for Highway Safety (IIHS) shows that some manage to do so even while playing by the rules of the systems’ attention requirements.
Together with the Massachusetts Institute of Technology’s AgeLab, the IIHS conducted tests over a four-week period using a 2017 Volvo S90 sedan equipped with Pilot Assist. It found that drivers were much more likely to check their phones, eat a sandwich or do other visual-manual activities while using the system than when driving unassisted.
- Also: Distractions at the Wheel: Three Questions Answered
- Also: Two Fatal Crashes Put Ford’s BlueCruise System Under Investigation
The tendency to multitask also increased over time for some drivers as they grew more comfortable with the technology, while others were more distracted while using the system from the start.
A separate study involved a 2020 Tesla Model 3 with Autopilot. The IIHS and AgeLab noticed that many drivers—all of them with no prior experience in a Tesla—quickly mastered the timing interval of its attention reminder feature so that they could prevent warnings from escalating to more serious interventions. Some people used this skill to continue engaging in distracting behaviours, punctuated by quick moves to stop the alerts.
At the time, Autopilot relied exclusively on a torque sensor in the steering wheel to monitor whether the driver was paying attention. The driver could indicate they were still engaged by nudging the steering wheel, toggling a dial or tapping the turn-signal stalk.
In 72 instances, the driver didn’t respond fast enough to prevent the alerts from escalating. Sixteen of these escalations persisted through the entire sequence to result in the driver being locked out of the system.
“In both these studies, drivers adapted their behaviour to engage in distracting activities,” IIHS President David Harkey said. “This demonstrates why partial automation systems need more robust safeguards to prevent misuse.”