Tesla has launched its newest ‘Automobile Security Report’, which seems to be into the completely different accident charges per mile of its automobiles on and off Autopilot.

The most recent information exhibits a slight enchancment.

Since 2018, Tesla has been attempting to create a benchmark for its enchancment in Autopilot security by releasing a quarterly report that compares the variety of miles per accident on Autopilot versus off Autopilot.

Right now, Tesla launched its report for Q3 2020:

Within the third quarter, we registered one accident for each 4.59 million miles pushed wherein drivers had Autopilot engaged. For these driving with out Autopilot however with our energetic security options, we registered one accident for each 2.42 million miles pushed. For these driving with out Autopilot and with out our energetic security options, we registered one accident for each 1.79 million miles pushed. By comparability, NHTSA’s most up-to-date information exhibits that in america there may be an car crash each 479,000 miles.

Listed here are how all of the completely different metrics evaluate to the info from the prior quarter and the identical interval final 12 months:

  • Autopilot Engaged: it’s at one accident for each 4.59 million miles pushed in Q3 2020. It went from one accident for each 4.53 million miles in Q2 2020 and from one accident for each 4.34 million miles in Q3 2019
  • Autopilot disengaged however with energetic security options: it’s at one accident for each 2.42 million miles pushed in Q3 2020. It went from one accident for each 2.27 million miles in Q2 2020 and from one accident for each 2.70 million miles in Q3 2019
  • Autopilot disengaged and with out energetic security options: it’s at one accident for each 1.79 million miles pushed in Q3 2020. It went from one accident for each 1.56 million miles in Q2 2020 and from one accident for each 1.82 million miles in Q3 2019

Subsequently, the info exhibits a slight discount of accidents on Autopilot from Q2 to Q3 and a good enchancment year-over-year.

As for Autopilot disengaged however with security options powered by Autopilot, it improved quarter-to-quarter, but it surely went down year-over-year.

Electrek’s Take

As ordinary, please take into account that the comparisons right here aren’t good.

Since Autopilot is at the moment primarily used on highways the place it’s simpler to build up a variety of mileage with out accidents and non-Autopilot mileage is coming from metropolis driving, the place accidents are extra probably, the 2 datasets cannot actually be in contrast.

Nonetheless, it’s nonetheless helpful to match the Autopilot mileage per accident over intervals of time, like year-over-year. The information is much from good, but it surely’s the perfect we’ve for now.

Tesla goes to wish much more information in the event that they ever wish to persuade authorities to permit them to deploy their full self-driving system.

Now that the FSD beta is out, it’s going to be attention-grabbing to see if and the way Tesla integrates the info into its security report.

FTC: We use earnings incomes auto affiliate hyperlinks. More.


Subscribe to Electrek on YouTube for exclusive videos and subscribe to the podcast.