Rethinking ”Hands Free” Driving
“Hands Free” driving is much in the news, and not in a particularly good way. A 2019 Tesla Model S recently crashed into a tree at high speed near Houston, killing two people.The victims were reportedly found in the front passenger seat and in the rear seat at the time of the crash – i.e. no victim was found in the driver’s seat. However, Elon Musk tweeted that Tesla’s investigation had not indicated that Tesla’s Autopilot feature was engaged.
That said, a year earlier in February 2020 the National Highway Traffic Safety Administration (NHTSA) had opened 14 investigations into Tesla crashes involving its driver assistance capabilities.
Adding to the current mix is a brewing dispute between NHTSA and another federal safety agency, the National Transportation Safety Board (NTSB). NTSB has publicly questioned whether the NHTSA has been fulfilling its oversight responsibilities with regard to a number of Tesla crashes. And several members of Congress are also asking NHTSA for accelerated regulations on ADAS features.
All of this is happening against a back-drop of a rapidly growing number of automobile manufacturers offering some form of what Consumer Reports calls “active driving assistance” technology. In October 2020 Consumer Reports published an evaluation of 17 such systems, an increase from only four such systems evaluated only two years earlier. The 2020 report included vehicles in most price ranges from Cadillac, Tesla, and BMW through Honda, Toyota and Mazda.
The Consumer Reports article defined active driving assistance technology as the simultaneous engagement of adaptive cruise control (ACC) and lane keeping assistance (LKA). ACC takes control of acceleration and braking to keep a vehicle a set distance from a vehicle in front of it. LKA takes control of steering to keep a vehicle within lane lines indicated on the roadway. The article warned that active driving assistance is not the same thing as self-driving – it is only meant to support a driver by reducing stress and fatigue.
From a safety (and insured loss) perspective this is all well and good . . . IF a driver stays engaged -- eyes on the road, hands on (or periodically on) the steering wheel.
On the other hand, an April 22, 2021 Consumer Reports article documented how a test driver put a Tesla Model Y in motion on a closed test track without anyone in the driver’s seat.The test driver simply put a weighted chain on the Tesla’s steering wheel (to mimic the weight of a driver’s hand), and then slid over to the front passenger seat, and used the steering wheel dial to put the car in motion up and down a half mile stretch of test track. The Tesla’s “Autopilot” function did the rest.
If I were an auto insurer, in the emerging world of available data about ADAS features on specific vehicles -- and better yet, data about when, where and how ADAS features are engaged; here are some things I would track:
- The names which automobile manufacturers give to their “active driving assistance” technology. These names communicate concepts and, to some degree, set expectations of drivers. Consider for example:
- Tesla's Autopilot, Land Rover’s In Control, and Porsche’s Active Safe
- Compared to
- Mercedes Driver Assistance, Nissan’s ProPILOT Assist, and BMW’s Driving Assistance Professional
- Tesla's Autopilot, Land Rover’s In Control, and Porsche’s Active Safe
- The design features which monitor driver engagement when these technologies are turned on – for example, requiring a driver to periodically keep hands on the wheel. (Special mention goes to Cadillac’s Super Cruise system which includes an inward facing camera monitoring where the driver’s eyes are looking.)
- Combining traditional telematics data (speed, hard acceleration and deacceleration) with the use of “active driving assistance” features – in order to test hypotheses about risk proclivities of drivers.