Tesla’s infamous Autopilot mode faced its latest reckoning this past weekend after it caused a driver to careen straight into a Washington state deputy’s vehicle that was parked at the side of the road.
In a statement posted to the Facebook page for the Snohomish County Sheriff’s Office, the department said that the Tesla did pretty “significant damage” to the deputy’s vehicle, which was parked on the shoulder of the road at the time. From the photos attached to the department’s post, it looks like the Tesla rammed straight into the side of the patrol car, denting the back door and nearly ripping off a driver’s side mirror.
According to NBC News, the Tesla driver—who was slapped with a ticket over the crash—just expected that his vehicle would sense the deputy’s vehicle and maneuver around it, rather than slam into its side.
He’s hardly the first one to make this kind of mistake. Despite the fact that Tesla’s own website states that its Autopilot mode requires “active driver supervision” and “does not make [these cars] autonomous,” we’ve seen multiple fatal wrecks result from drivers seemingly ignoring those caveats. At the start of 2020, the National Highway Traffic Safety Administration (or NHTSA) opened a formal investigation into more than a dozen Tesla crashes dating back to 2016 that the agency suspects might have involved Autopilot in some way.
Meanwhile, Tesla’s headaches have continued into 2021. Last week, NHTSA announced it would be opening a probe into the case of yet another fatal crash that happened in California at the start of the month. While the question of whether or not the driver was on Autopilot is still open, the California Highway Patrol has come out saying it firmly suspects that might be the case.
We’ve reached out to Tesla for comment on the Washington crash and will update this post when we learn more.
DISCUSSION
“…it caused a driver to careen straight into a Washington state deputy’s vehicle that was parked at the side of the road.”
Everything about this statement is wrong. The Autopilot didn’t cause the accident any more than the electric motors propelling the car did. What caused this crash, which is paradoxically referenced a few paragraphs later, was the driver.
Autopilot didn’t tear the steering wheel out of the drivers hand or drive up on the sidewalk to cause the crash. It didn’t aim the vehicle at another car and emit an evil cackle while the driver responsibly battled with the rogue steering wheel as the car vengefully tried to kill him. In the absence of such stories which would actually be newsworthy we get stories like these which are the equivalent of blaming the sun for blinding light during cloudy days.
It’s the drivers responsibility in all of these incidents for reasons ranging from common sense to the pop up messages that appear on the dashboard stating that. If the car is failing to recognize another car then the driver had better have the good judgement to do something about it, an expectation that is no greater than for anyone driving a car without Autopilot.
Ironically the data suggest that Autopilot is actually quite good as a safety feature, reducing accidents many fold over when it’s not used although this must be studied more with the correct context. Somewhat credible stories on the data range from 10x as safe with it on to break even depending on what other factors are considered in parsing the data. The point being that the feature, when used correctly and with almost no additional effort from the driver, is actually probably a good safety feature (and a good subject for a story). It’s when the driver decides to take their hands off the wheel or jump in the back seat that it becomes a problem, which is true of every other car.
There’s a bias in the media that somehow makes a Tesla crashing whether autopilot was on or not versus some other vehicle into something to write about. I guess the real story here is that these clickbait stories gain any traction at all. Mission accomplished I guess.