Tesla’s Accident Rate on Autopilot Nearly 9 Times Better than National Average

As Tesla says it’s getting closer to releasing its Full Self Driving product, it continues to report the safety benefits of Autopilot.

In the company’s Q3 update, Tesla provided safety numbers related to accidents occurring when Autopilot was in use. 

According to Tesla’s Q3 2019 shareholder letter:

“During Q3, we registered one accident for every 4.34 million miles driven in which drivers had Autopilot engaged. This compares to the national average of one accident for every 0.5 million miles based on NHTSA’s most recent US data.”

Based on those numbers, cars operating on Autopilot were nearly nine times less likely to experience an accident than the national average.

Looking at Tesla’s previous safety reports, the miles per accident has ranged from 2.87 million miles in Q1 2019 to 3.34 million miles in Q3 2018. This year’s third quarter saw a 23% decrease in accidents per miles driven with Autopilot.

 

[mikebixby]

Could this be skewed a little since Autopilot is primarily used on freeways, and the accident rate is being compared with all driving scenarios? Not trying to put a damper on the news - just trying to compare similar items.

 

[Glamisduner]

Could this be skewed a little since Autopilot is primarily used on freeways, and the accident rate is being compared with all driving scenarios? Not trying to put a damper on the news - just trying to compare similar items.

That and I tend to use AP on less busy freeways, or else in stop and go traffic.

 

[mswlogo]

Such misleading data.

What do you want to bet the statistics of cars driving on Adaptive Cruise Control is 9x better than average too (per mile).

ACC or AutoPilot is typically used in good conditions on longer routes.

Most accidents happen 1 mile from home, where you’re way less likely to be using ACC or AutoPilot a mile from. In fact it’s recommended you not use AutoPilot with cross traffic. How convenient.

What is the accident rate on highways normalize per mile for all cars? That’s the number we want. Since many Tesla’s have AutoPilot and most likely enabled on the highway, Tesla should you look statistically better.

I’m sure insurance companies know the real numbers.

 

[SOULPEDL]

Agree, but the quarterly improvements make a strong argument for rapid improvement.
And don't discount all city driving, I'm using it everywhere I can including parking lots now.

 

[electracity]

Misconception 7: To convince us that they are safe, self-driving cars must drive hundreds of millions of miles | Driverless car market watch

7.3 Choosing the right independent variable

The number of vehicle miles traveled may not be the best unit against which to measure accidents. The quickest way to accumulate test miles is to drive on highways. But highway driving is safer than driving on many other streets. Of the 3.148 billion vehicle miles traveled in the US in 2015 about a quarter occurred on rural or urban interstate highways; however only about 12% of fatal accidents occurred on these roads. Because the average speed varies between different road types, driving time may be a better basis for measuring failure rate than driving distance. It is well known that urban driving is the most difficult scenario for self-driving cars (and probably also for human drivers, although this may be reflected more in the crash statistics than in the fatality statistics because average speeds are lower). Because average speeds differ greatly, using time instead of distance will have a significant effect on the reliability rates and might significantly reduce the number of test miles (while also reducing the incentive for self-driving car companies to accumulate test miles on what many regards as the “easiest” roads – interstate highways!).

Conceptually, driving is most difficult, when the environment changes, when actions need to be taken. Just driving straight is not much of a challenge. We should look for other measures that come closer to the complexity of driving and the risks associated with driving. A simple proxy that would be easy to calculate for self-driving cars and not impossible to estimate for human driving statistics could be the number of intersections which are passed. This measure contains both an element of distance and risk because a large percentage of accidents occur at intersections. Therefore the number of intersections passed might also be a good independent variable against which to measure reliability rates.

Just using overall miles driven and not differentiating between the different types of environments which the cars are driven in and not differentiating between the risk levels of the routes is intellectually lazy. Over the years we have collected enormous amounts of information about traffic accidents and the situations in which they occur; we need to carefully go through this data and develop much more refined measures for understanding and assessing the reliability of human as well as autonomous driving.

 

[nlmunro]

Could this be skewed a little since Autopilot is primarily used on freeways, and the accident rate is being compared with all driving scenarios? Not trying to put a damper on the news - just trying to compare similar items.

Completely agree - I only use on clear motorways and slow start/stop traffic. I think it works well in those scenarios but this sort of statistic undermines credibility.

 

[SageBrush]

Such misleading data.

What do you want to bet the statistics of cars driving on Adaptive Cruise Control is 9x better than average too (per mile).

ACC or AutoPilot is typically used in good conditions on longer routes.

Most accidents happen 1 mile from home, where you’re way less likely to be using ACC or AutoPilot a mile from. In fact it’s recommended you not use AutoPilot with cross traffic. How convenient.

What is the accident rate on highways normalize per mile for all cars? That’s the number we want. Since many Tesla’s have AutoPilot and most likely enabled on the highway, Tesla should you look statistically better.

I’m sure insurance companies know the real numbers.

This is deaths rather than accidents, found from a quick search:

Ironically, the part of driving that people fear the most turns out to be the safest part. Federal transportation data have consistently shown that highways are considerably safer than other roads. (You can see the detailed numbers here.) For instance, in 2007 0.54 people were killed for every 100 million vehicle miles driven on urban interstates, compared with 0.92 for every 100 million vehicle miles driven on other urban highways and arterials, and 1.32 killed on local urban streets.