Tesla's 40% Reduction Stat for its Autopilot is Faulty

Tesla’s latest Autopilot death might’ve been a case of user error, but one of the stats they tout to defend the “driver assistance” system is probably bogus.

Tesla likes to quote a National Highway Traffic Safety Administration (NHTSA) stat which claims its Autopilot tech reduces crashes by 40%. And it very well could. Humans are prone to mistakes, especially in stressful situations (aka, driving).

But the NHTSA stat wasn’t all that scientific. For one, it relied on data provided by Tesla, itself. Conflict of interest, anyone?

It also tabulates crashes by airbag deployments – particularly, 1.3 airbag deployments per million miles for Tesla model years 2014-2016. After Tesla released a software upgrade, the airbag rate dropped to 0.8 per million miles. Hence the 40% reduction.

But airbags aren’t 100% reliable, and in the original Autopilot death back in 2016, the airbags never deployed. So there’s that.

And, as Wired points out, no one knows whether Autopilot was engaged in most of these crashes. So the “driver assistance” system – software upgrade or no – is more like a correlating factor.

Read more about this controversy here: https://www.wired.com/story/tesla-autopilot-safety-statistics/?mbid=nl_050518_daily_list_p