Good NCAP ratings, but a new study shows Tesla is the most dangerous brand on the road for safety outcomes. And AutoPilot is not a safety feature. Here are the receipts.
Years ago I read through all the ABS safety studies I could find. The conclusion seemed to be that ABS was a wash, the complacency factor of driving faster in slick conditions plus (my personal observation) drivers who take their foot off the brake pedal when it pulses due to ABS engaging. Those two offset the safety enhancement from shorter stopping distances as demonstrated by engineering studies. (My recollection was that I read a history of ABS, but my Zotero bibliography got corrupted by DropBox so I can't provide a reference.)
David Hemenway has an engaging 1977 microeconomics reader (many subsequent additions) in which he discusses safety, including the confounding factor that people worried about safety tend to be both better drivers and to purchase vehicles with ADAS enhancements. He employs cute terminology – "nervous nellies" vs "dangerous dans" – but I don't read in the safety literature often enough to know if that's common common parlance.
Indeed ABS was a wash, although risk homeostatis is a controversial theory. A different explanation I have heard from automotive engineers is that with ABS one retains more control authority and ends up steering into trees/poles often enough to make some crashes worse (without ABS you'd skid and would not be able to execute the steer commanded by the driver that often enough ends up hitting a tree, utility pole, or other obstacle).
One thing that is pretty clear is that AEB makes a real difference. Tesla has AEB (as do essentially all new cars), and it is important to separate safety increases attributable to AEB vs. autopilot safety claims.
Delete this. What poppycock. Look at all the articles bashing Tesla and over time you see the same sources. Almost all of your points have been debunked. You do not understand that no other automobiles have and have had the tech in the millions of Tesla on the road. Beyond required testing, Tesla collects real world data and engineers for THAT. The reasons are that EV CAN offer engineering advantages such as the massive difference in front impact. Most automakers don't have that option with ICE. And most legacy auto that make EV don't bother to engineer to that advantage, instead stuffing metal into that space because they have to. The active and passive safety are far superior for reasons that we can discuss ad nauseum and in engineering detail. But the reality is that you are least likely to be injured in a Tesla even when YOU are driving it. The next level is upon us and nobody else is even close. https://www.notateslaapp.com/news/3333/tesla-launches-live-fsd-safety-hub-reveals-daily-fsd-miles-driven
Just read how Tesla is releasing a fleet of Robotaxis. For now there is a teleoperator to help drive it. This says volumes of how poorly designed this car is. It will never be self-autonomous and at best will be trying to compete with similar costs as other car hailing services such as Lyft. I don’t believe a remote teleoperator will be as safe as an on board driver. It will not be able to compete with Waymo which unlike Tesla’s Robotaxis uses LiDar and much safer. When the first reports of injuries and deaths come up it will negatively affect Tesla share price. Musk again puts cost savings above human safety. His refusal to install LiDar will be his undoing.
I’ll start my comment with a caveat—I’m not a vehicle safety expert. That said, I believe comparing cars, safety wise, is an exceedingly difficult endeavor (the number of variables is nearly infinite especially when automation is involved). Consequently, here I just want to share my experience as a long time Tesla driver.
As an enthusiast of electric transportation, I got a Tesla over 13 years ago (the only electric vehicle available at the time). I still own that car, which doesn’t have any automating features, and I can report no incidents, except for a couple of speeding tickets. Tesla, and I believe all other electric cars, require some adjustments of driving habits. For instance, these cars offer breath taking acceleration and employ regenerative breaking. You have to be very careful when your vehicle needs just 3 sec to accelerate from 0 to 60 mph. Otherwise, you will enter the statistics in the car crash category. Regenerative breaking is a great feature which increases the driving range particularly in heavy traffic. Once more, you need to be careful. If you try to recharge your battery by excessively avoiding using the brakes you may rear end the vehicle in front of you (AEB is certainly useful in this case, but my old Tesla doesn’t have it). Another aspect is that on slippery pavement, on snow for instance, an electric car may behave differently when regenerative breaking occurs, i.e., the vehicle may be less stable (particularly in the case of rear wheel traction). Again, the drivers must adjust. Based on my driving experience, it will take some time until automation will be able to cope with such details.
Finally, I think a comparison between the safety of electric vehicles would be interesting. However, the result may be biased, because the number of Tesla cars on the US roads is much higher than any other brand.
Statistics is a well understood field. Crashes rater per car at fixed mileage allows apples to apples comparison.
Fatal Accident rate / billion miles is an apples to apples comparison . . . tabular data above shows 5.6 fatal Tesla accidents per billion miles. And note this is 2X all makes statistics of 2.8 fatal accidents per billion miles. So you are twice as likely to experience a fatal accident in a Tesla based on data.
How does this fit in with the data presented here? https://www.tesla.com/VehicleSafetyReport Based on that one could conclude that Teslas are indeed a lot safer with Autopilot technology turned on. Are they measuring different things? Or do you think these numbers are wrong/biased?
The paper in point (2) explains why that data is so misleading. For example it does not account for differences between limited access road miles and other miles which have much different crash rates. Also, any new car is MUCH better than fleet average at crashes. Teslas do not report all crashes (mostly they just report ones where air bags deploy), and so on.
The various points in the post above show that the Tesla analysis is unreasonably optimistic.
Yes, Elon puts profits before safety. All the while publicly bragging about FSD. Which is near garbage without LiDar. Explains the numerous class action lawsuits against Tesla’s FSD. How many more will die because he is too proud to mention its shortcomings?
Years ago I read through all the ABS safety studies I could find. The conclusion seemed to be that ABS was a wash, the complacency factor of driving faster in slick conditions plus (my personal observation) drivers who take their foot off the brake pedal when it pulses due to ABS engaging. Those two offset the safety enhancement from shorter stopping distances as demonstrated by engineering studies. (My recollection was that I read a history of ABS, but my Zotero bibliography got corrupted by DropBox so I can't provide a reference.)
David Hemenway has an engaging 1977 microeconomics reader (many subsequent additions) in which he discusses safety, including the confounding factor that people worried about safety tend to be both better drivers and to purchase vehicles with ADAS enhancements. He employs cute terminology – "nervous nellies" vs "dangerous dans" – but I don't read in the safety literature often enough to know if that's common common parlance.
Indeed ABS was a wash, although risk homeostatis is a controversial theory. A different explanation I have heard from automotive engineers is that with ABS one retains more control authority and ends up steering into trees/poles often enough to make some crashes worse (without ABS you'd skid and would not be able to execute the steer commanded by the driver that often enough ends up hitting a tree, utility pole, or other obstacle).
One thing that is pretty clear is that AEB makes a real difference. Tesla has AEB (as do essentially all new cars), and it is important to separate safety increases attributable to AEB vs. autopilot safety claims.
Delete this. What poppycock. Look at all the articles bashing Tesla and over time you see the same sources. Almost all of your points have been debunked. You do not understand that no other automobiles have and have had the tech in the millions of Tesla on the road. Beyond required testing, Tesla collects real world data and engineers for THAT. The reasons are that EV CAN offer engineering advantages such as the massive difference in front impact. Most automakers don't have that option with ICE. And most legacy auto that make EV don't bother to engineer to that advantage, instead stuffing metal into that space because they have to. The active and passive safety are far superior for reasons that we can discuss ad nauseum and in engineering detail. But the reality is that you are least likely to be injured in a Tesla even when YOU are driving it. The next level is upon us and nobody else is even close. https://www.notateslaapp.com/news/3333/tesla-launches-live-fsd-safety-hub-reveals-daily-fsd-miles-driven
The new Tesla FSD safety report is marketing puffery, not serious data analysis.
https://philkoopman.substack.com/p/new-tesla-fsd-safety-data
Just read how Tesla is releasing a fleet of Robotaxis. For now there is a teleoperator to help drive it. This says volumes of how poorly designed this car is. It will never be self-autonomous and at best will be trying to compete with similar costs as other car hailing services such as Lyft. I don’t believe a remote teleoperator will be as safe as an on board driver. It will not be able to compete with Waymo which unlike Tesla’s Robotaxis uses LiDar and much safer. When the first reports of injuries and deaths come up it will negatively affect Tesla share price. Musk again puts cost savings above human safety. His refusal to install LiDar will be his undoing.
I’ll start my comment with a caveat—I’m not a vehicle safety expert. That said, I believe comparing cars, safety wise, is an exceedingly difficult endeavor (the number of variables is nearly infinite especially when automation is involved). Consequently, here I just want to share my experience as a long time Tesla driver.
As an enthusiast of electric transportation, I got a Tesla over 13 years ago (the only electric vehicle available at the time). I still own that car, which doesn’t have any automating features, and I can report no incidents, except for a couple of speeding tickets. Tesla, and I believe all other electric cars, require some adjustments of driving habits. For instance, these cars offer breath taking acceleration and employ regenerative breaking. You have to be very careful when your vehicle needs just 3 sec to accelerate from 0 to 60 mph. Otherwise, you will enter the statistics in the car crash category. Regenerative breaking is a great feature which increases the driving range particularly in heavy traffic. Once more, you need to be careful. If you try to recharge your battery by excessively avoiding using the brakes you may rear end the vehicle in front of you (AEB is certainly useful in this case, but my old Tesla doesn’t have it). Another aspect is that on slippery pavement, on snow for instance, an electric car may behave differently when regenerative breaking occurs, i.e., the vehicle may be less stable (particularly in the case of rear wheel traction). Again, the drivers must adjust. Based on my driving experience, it will take some time until automation will be able to cope with such details.
Finally, I think a comparison between the safety of electric vehicles would be interesting. However, the result may be biased, because the number of Tesla cars on the US roads is much higher than any other brand.
Statistics is a well understood field. Crashes rater per car at fixed mileage allows apples to apples comparison.
Fatal Accident rate / billion miles is an apples to apples comparison . . . tabular data above shows 5.6 fatal Tesla accidents per billion miles. And note this is 2X all makes statistics of 2.8 fatal accidents per billion miles. So you are twice as likely to experience a fatal accident in a Tesla based on data.
How does this fit in with the data presented here? https://www.tesla.com/VehicleSafetyReport Based on that one could conclude that Teslas are indeed a lot safer with Autopilot technology turned on. Are they measuring different things? Or do you think these numbers are wrong/biased?
Maybe the difference is accidents vs. fatalities?
The paper in point (2) explains why that data is so misleading. For example it does not account for differences between limited access road miles and other miles which have much different crash rates. Also, any new car is MUCH better than fleet average at crashes. Teslas do not report all crashes (mostly they just report ones where air bags deploy), and so on.
The various points in the post above show that the Tesla analysis is unreasonably optimistic.
Yes, Elon puts profits before safety. All the while publicly bragging about FSD. Which is near garbage without LiDar. Explains the numerous class action lawsuits against Tesla’s FSD. How many more will die because he is too proud to mention its shortcomings?