Tesla releases new Autopilot accident data report showing slight improvement

Tesla has released its latest ‘Vehicle Safety Report’, which looks into the different accident rates per mile of its vehicles on and off Autopilot.

The latest data shows a slight improvement.

Since 2018, Tesla has been trying to create a benchmark for its improvement in Autopilot safety by releasing a quarterly report that compares the number of miles per accident on Autopilot versus off Autopilot.

Today, Tesla released its report for Q3 2020:

In the 3rd quarter, we registered one accident for every 4.59 million miles driven in which drivers had Autopilot engaged. For those driving without Autopilot but with our active safety features, we registered one accident for every 2.42 million miles driven. For those driving without Autopilot and without our active safety features, we registered one accident for every 1.79 million miles driven. By comparison, NHTSA’s most recent data shows that in the United States there is an automobile crash every 479,000 miles.

Here are how all the different metrics compare to the data from the prior quarter and the same period last year:

• Autopilot Engaged: it is at one accident for every 4.59 million miles driven in Q3 2020. It went from one accident for every 4.53 million miles in Q2 2020 and from one accident for every 4.34 million miles in Q3 2019
• Autopilot disengaged but with active safety features: it is at one accident for every 2.42 million miles driven in Q3 2020. It went from one accident for every 2.27 million miles in Q2 2020 and from one accident for every 2.70 million miles in Q3 2019
• Autopilot disengaged and without active safety features: it is at one accident for every 1.79 million miles driven in Q3 2020. It went from one accident for every 1.56 million miles in Q2 2020 and from one accident for every 1.82 million miles in Q3 2019

Therefore, the data shows a slight reduction of accidents on Autopilot from Q2 to Q3 and a decent improvement year-over-year.

As for Autopilot disengaged but with safety features powered by Autopilot, it improved quarter-to-quarter, but it went down year-over-year.

Electrek’s Take

As usual, please keep in mind that the comparisons here are not perfect.

Since Autopilot is currently primarily used on highways where it’s easier to accumulate a lot of mileage without accidents and non-Autopilot mileage is coming from city driving, where accidents are more likely, the two datasets can not really be compared.

Nonetheless, it’s still useful to compare the Autopilot mileage per accident over periods of time, like year-over-year. The data is far from perfect, but it’s the best we have for now.

Tesla is going to need a lot more data if they ever want to convince authorities to allow them to deploy their full self-driving system.

Now that the FSD beta is out, it is going to be interesting to see if and how Tesla integrates the data into its safety report.