Data likely shows Teslas on Autopilot crash more than rivals

[2101Guy]

Data likely shows Teslas on Autopilot crash more than rivals

The government plans soon to release data on collisions involving vehicles with autonomous or partially automated driving system that will likely single out Teslas for a disproportionately high number of such crashes.

apnews.com

 

[Dewg]

Data likely shows Teslas on Autopilot crash more than rivals

The government plans soon to release data on collisions involving vehicles with autonomous or partially automated driving system that will likely single out Teslas for a disproportionately high number of such crashes.

apnews.com

So what are they going to do about it?

 

[mangrove79]

From your article:

Tesla does have many more vehicles with partly automated systems operating on U.S. roads than most other automakers do — roughly 830,000, dating to the 2014 model year. And it collects real-time data online from vehicles, so it has a much faster reporting system. Other automakers, by contrast, must wait for reports to arrive from the field and sometimes don’t learn about crashes for months.

And sometimes don't even learn about crashes ever.. LOL.

 

[Dewg]

From your article:


And sometimes don't even learn about crashes ever.. LOL.

There's the solution - Tesla stops collecting data when a crash occurs. Plausible deniability.

 

[glide]

So what are they going to do about it?

Who is the “they” you are referring to?

Tesla is going to have to solve some longstanding issues they have had with their autopilot software.

NHTSA has the authority to issue a recall if Elon wants to dig in his heels. This isn’t the SEC.

 

[Dewg]

Who is the “they” you are referring to?

Tesla is going to have to solve some longstanding issues they have had with their autopilot software.

NHTSA has the authority to issue a recall if Elon wants to dig in his heels. This isn’t the SEC.

So win win for everyone. People get safer cars, the industry makes changes, and NHTSA doesn't appear to be a toothless agency.

 

[TunaBug]

Sounds to me like the metric is incidents per 1000 cars. That's probably an attempt to deal with the fact that some cars are more popular than others, but that will still work against any car that appeals to technically minded people who will use the automation system more than in other cars. I know many people who are scared to even use their car's cruise control; there are probably tons of Civics and Corollas and Escalades on the roads where those features are never used.

I would be interested to know a metric that is crashes per hours of use of the automation system. Or per miles driven instead of per time. That is probably not information that the NHTSA has available at present, but the NHTSA certainly ought to be able to require that level of reporting.

 

[sleepydoc]

According to the article, they lumped FSD, AutoPilot and TACC together. The three are dramatically different in their capabilities and intended uses. Additionally, aside from the click-bait headline, what really matters is the circumstances of the accidents. Were they purely failures of the automated systems or were they human failures and abuses of the system? If they’re because people were abusing the system then that’s a different issue.

 

[Padelford]

I want to know what model years are involved in the crashes. I suspect the data won’t be available until the report comes out. Also, Tesla management are profoundly idiotic for not having a PR department that engages with the media on Tesla’s behalf.

 

[Dewg]

It's a hard balancing act. The history of automation always results in accidents and failures, but the end result is almost always a more efficient and safer product.

Do a quick Google search on airline autopilot accidents. It sucks, but the end is still a better and safer flying experience.

Will people be hurt and killed by car autonomous features while these technologies are evolving? Yes. It's tragic, but the end result is always a better and safer experience.

I refer back to Chucks video where a trike lost a wheel and crashed into him, but his Tesla, on AP, maintained control and handled the collision safely. Had a human been driving, they may have overreacted or lost control and caused a cascading accident.

 

[AndrewZ]

Sounds to me like the metric is incidents per 1000 cars. That's probably an attempt to deal with the fact that some cars are more popular than others, but that will still work against any car that appeals to technically minded people who will use the automation system more than in other cars. I know many people who are scared to even use their car's cruise control; there are probably tons of Civics and Corollas and Escalades on the roads where those features are never used.

I would be interested to know a metric that is crashes per hours of use of the automation system. Or per miles driven instead of per time. That is probably not information that the NHTSA has available at present, but the NHTSA certainly ought to be able to require that level of reporting.

Not giving the NHTSA nearly enough credit, these are people who specifically work in Auto safety and who surely know more about the relevant metrics than us observers who don’t work full time jobs in this very specialized sector

Cumulative miles travelled with the system
engaged is one of many figures requested last year by the NHTSA, you can find an exhaustive list by reading through their data request template

https://static.nhtsa.gov/odi/inv/2021/INIM-PE21020-85284.pdf

They also requested logs per vehicle of hardware configurations, any changes, software versions and any changes, etc etc etc

 

[S4WRXTTCS]

The problem with this comparison is we don't know how many miles other vehicles traveled while on the equivalent to AP. I don't believe any of those other L2 vehicles they're comparing it with give those values.

Obviously if I drive 10K miles on AP than there is more risk of something bad happening if I only drove 1K miles. Having a 10 to 1 ratio wouldn't surprise me in the least due to the massive number of vehicles in the fleet with the technology, and that Tesla drivers supposedly drive more than average. In addition to that AP can be used over so many more roads than other similar systems.

I'd look at the crash rates per miles for all L2 vehicles regardless of whether the L2 technology was ever used.

 

[S4WRXTTCS]

"Nissan, with over 560,000 vehicles on the road using its ”ProPilot Assist,” didn’t have to report any crashes, the company said."

I find this highly suspect especially if adaptive cruise control accidents count as well.

 

[S4WRXTTCS]

Not giving the NHTSA nearly enough credit, these are people who specifically work in Auto safety and who surely know more about the relevant metrics than us observers who don’t work full time jobs in this very specialized sector

Cumulative miles travelled with the system
engaged is one of many figures requested last year by the NHTSA, you can find an exhaustive list by reading through their data request template

https://static.nhtsa.gov/odi/inv/2021/INIM-PE21020-85284.pdf

They also requested logs per vehicle of hardware configurations, any changes, software versions and any changes, etc etc etc

They requested that information, but connected vehicles are a fairly recent thing. So I don't believe a lot of manufactures will be able to provide that data as its simply doesn't exist.

It also seems weird to ask for data that pertains to L1 (adaptive cruise control) when what's being investigated is L2. My Jeep Wrangler has Adaptive Cruise Control, but doesn't even have lane keep let alone lane steering. It doesn't count as an L2 vehicle so crashes don't need to be reported despite almost identical usage as I would if it was L2 (I mostly just use adaptive cruise control).

 

[Huskyf]

"Nissan, with over 560,000 vehicles on the road using its ”ProPilot Assist,” didn’t have to report any crashes, the company said."

I find this highly suspect especially if adaptive cruise control accidents count as well.

Yes but system on public cars started on 2021 Tesla on 2014 how to compare ????

 

[Barklikeadog]

Not giving the NHTSA nearly enough credit, these are people who specifically work in Auto safety and who surely know more about the relevant metrics than us observers who don’t work full time jobs in this very specialized sector

Cumulative miles travelled with the system
engaged is one of many figures requested last year by the NHTSA, you can find an exhaustive list by reading through their data request template

https://static.nhtsa.gov/odi/inv/2021/INIM-PE21020-85284.pdf

They also requested logs per vehicle of hardware configurations, any changes, software versions and any changes, etc etc etc

It's easier to dismiss NHTSA as either a room full of fools, or room full of short sellers. NHTSA has been dealing with corner cutting, liability avoiding automakers for decades. They are not the bad guy.

 

[2101Guy]

Level 2 ADAS Crashes by Reporting Entity​

Tesla

273
Honda

90
Subaru

10
Ford

5
Toyota

4
BMW

3
General Motors

2
APTIV

1
Hyundai

1
Lucid Motors

1
Porsche

1
Volkswagen

1
Source: National Highway Traffic Safety Administration

 

[2101Guy]

From July 20th, 2021, to May 21st, 2022, there were 273 crashes involving Tesla vehicles using Autopilot, according to the report. The EV company’s crashes represent the bulk of the total 392 crashes reported during that period.

Other automakers didn’t come close to Tesla’s number of reported crashes. Honda, which sells its ADAS features under the brand “Honda Sensing,” disclosed 90 crashes. Subaru, which packages its ADAS under “EyeSight,” reported 10 crashes. Ford disclosed five crashes, Toyota reported four crashes, BMW reported three crashes, and General Motors, maker of Super Cruise, only disclosed two crashes. Aptiv, Hyundai, Lucid, Porsche, and Volkswagen each reported one crash.

 

[S4WRXTTCS]

Yes but system on public cars started on 2021 Tesla on 2014 how to compare ????

My understanding is ProPilot Assist was initially released on 2019 models.

So it should be a good 3 years. Quite a bit less time than Tesla, but not exactly nothing either.

 

[sleepydoc]

From July 20th, 2021, to May 21st, 2022, there were 273 crashes involving Tesla vehicles using Autopilot, according to the report. The EV company’s crashes represent the bulk of the total 392 crashes reported during that period.

Other automakers didn’t come close to Tesla’s number of reported crashes. Honda, which sells its ADAS features under the brand “Honda Sensing,” disclosed 90 crashes. Subaru, which packages its ADAS under “EyeSight,” reported 10 crashes. Ford disclosed five crashes, Toyota reported four crashes, BMW reported three crashes, and General Motors, maker of Super Cruise, only disclosed two crashes. Aptiv, Hyundai, Lucid, Porsche, and Volkswagen each reported one crash.

Again, I go back to my earlier statement- Adaptive Cruise/TACC, AutoPilot and FSD all fall under the 'Level 2' category but are dramatically different. They have different abilities, different use cases, different expectations and different potentials for abuse. As others have said, you also need to look at accident rates per mile driven. The above numbers make a nice clickbait headline but are really impossible to interpret or compare without further details.