Right. As I've pointed out previously, Tesla seems to be unable to detect sizable stationary obstacles that are partly blocking a lane, especially if they don't look like the rear end of a car. In addition to emergency vehicles, Teslas on autopilot have plowed into freeway barriers and a street sweeper. That's the usual situation for first responders, who usually try to block as little of the road as possible but often don't have enough shoulder space.

It's clear what Tesla really has - a good lane follower and cruise control that slows down for cars ahead. That's a level 2 system. That's useful, but, despite all the hype about "full self driving", it seems that's all they've got.

"Full self driving" just adds some lane-changing assistance and hints from the nav system.

This is why I believe the approach of incremental improvement towards full self driving is fundamentally flawed. These advanced driver assist tools are good enough to lull users into a false sense of security. No amount of "but our terms and conditions say you need to always pay attention!" will overcome human nature building that trust and dependence.

A lot of people in here saying it is not possible to drive safely with partial self driving. I wonder, how many of those people have actually driven a car with autopilot?

I have autopilot on my car, and it definitely makes me a better and safer driver. It maintains my distance from the car in front and my speed while keeping me in my lane, so my brain no longer has to worry about those mundane things. Instead I can spend all my brainpower focused on looking for potential emergencies, instead of splitting time between lane keeping/following and looking for emergencies.

I no longer have to look at my speedometer or the lane markers, I can take a much broader view of the traffic and conditions around me.

Before you say it's impossible to be safe driving with an assistive product, I suggest trying one out.

I remember when Google was first presenting on driverless technology about 10 years ago, and they mentioned how you have to go right to full self driving, because any advanced driver assistance will clash with human risk compensation behavior.

Risk compensation is fascinating; driving with a bike helmet causes the biker and drivers around the biker to behave more dangerously.

Is society sophisticated enough to deal with advanced driver assistance? Is it possible to gather enough data to create self driving ML systems?

> "The involved subject vehicles were all confirmed to have been engaged in either Autopilot or Traffic Aware Cruise Control during the approach to the crashes,"

No mention of the deceptive marketing name "Full Self Driving" in the article.

There is about 36000 death on the road per year in the USA for 280 millions vehicles, that's 128.5 death/million vehicule/year

If we assume the number of tesla autopilot death double this year to 8 (from 4 at the time of probe launch), for about 900 thousand tesla on the road in USA, that's 8.9 autopilot death/million tesla/year.

Ratio between the numbers of 14.4.

Tesla reporting says for Q1 2021 one crash on autopilot per 4.19 millions miles vs one crash per 484 thousand miles all vehicules.

Ratio between the numbers of 8.7

All numbers are full of biases and their ratio probably aren't that meaningful but they end up in the same magnitude.

Interesting data there "Fatality Facts 2019 Urban/rural comparison":

https://www.iihs.org/topics/fatality-statistics/detail/urban...

"Although 19 percent of people in the U.S. live in rural areas and 30 percent of the vehicle miles traveled occur in rural areas, almost half of crash deaths occur there. "

I was shocked that in the USA in 2019 about 40-46% of all road death people were unbelted, while 90% of front seat people wear seat belts according to observation studies.

Incidentally tesla car will beep to no end if weight is detected on a seat and seat belt isn't clicked: I have to click the seat belt when I put my (not so heavy) bag on the passenger seat since there's no software option to disable the beeping.

Glad the regulators are looking into this. It bothers me that now Tesla seems to have no liability at all for the system not working, since it's always the driver's fault for not paying enough attention.

As Teslas get better at driving the drivers will be paying less attention inevitably, Tesla needs to start being responsible at some point

I remember that video "driver is only there because of regulatory rules". That is a flat out lie, safe to say so by now. The autopilot accidents per distance is also cherry picked, turn on autopilot everywhere , including bad weather and see how that comparison goes. And the claim that the cars have all the hardware for future fsd is quite out there too. It's a bit like saying I have the next Michael Phelps here, he just cannot swim yet.

There are a lot of good points here in the comments already about the relative safety of Tesla's system compared to other vehicles and other automated driving system -- and I think they're probably right.

The differentiating issue with Tesla's system is the way it is sold and marketed. Important operational safety information shouldn't be hidden in fine print. Subtly misleading marketing has unfortunately become acceptable in our culture, but this idea needs to stay out of safety-critical systems.

We need a mandate for clear and standardized labelling for these features, à la the Monroney sticker. All manufacturers should have to label and market their cars with something like SAE J3016. https://www.sae.org/binaries/content/gallery/cm/articles/pre...

I drive a lot across Europe: as in, really a lot, long trip across several countries, several times a year. I drive enough on the highways to know a few scary situations, like the truck driver in a big curve slightly deviating out of his lane and "pushing" me dangerously close to the median strip for example.

To me driving requires paying constant attention to the road and being always ready to act swiftly: I just don't understand how you can have a "self driving car but you must but be ready to put your hands back on the steering wheel and your foot on the pedal(s)".

I have nothing against many "recent" safety features, like the steering wheel shaking a bit if the car detects you're getting out of your lane without having activated your blinker. Or the car beginning to brake if it detects an obstacle. Or the car giving you a warning if there's a risk when you change lane, etc.

But how can you react promptly if you're not ready? I just don't get this.

Unless it's a fully self-driving car, without even a steering wheel, a car should help you focus more, not less.

Unfortunately the article doesn't mention anything about how common it is for human drivers to crash into first responder vehicles during the night. I'm not trying to downplay these cases, as hitting emergency vehicles is very bad indeed, yet ~4 such crashes per year might be in the same ballpark or even better than "unassisted" drivers that cause such crashes.

Long overdue. We're going to need a more rapid and iterative way to do this if we got to have even a chance of autopilot type technologies succeeding over the long run. Companies and regulators need to be collecting feedback and pushing for improvement on a regular basis. I still don't think it's likely to succeed, but if it did, this would be the way.

Predictable outcome of a sensing system fully based on deep learning. Rare unusual situations don't have enough training data and lead to unpredictable output.

I still think that Tesla's approach is the right one, I just think they need to gather more data before letting this product be used in the wild unsupervised.

I wonder how many accidents happened because people were fiddling with their screen to turn on the windscreen wipers and took their eyes of the road too long in poor sight. I can't believe the law allowed the controls for the wiper speed to be put into this capacitive screen.

Seeing all the near-misses and life saving interventions on YouTube involving Tesla vehicles, 1 death seems not such a bad result..

The question should be - how many lives were saved by this system vs how many would die if driven "normally"?

Anybody have more information about how many, if any, of these wrecks happened after Tesla moved to vision only Autopilot? Maybe this obvious problem has already been mostly fixed with the new system and possibly a main reason for going to full vision in the first place. It seems to me NHTSA need to move a little faster with this quickly changing technology as they are going back to examine crashes that happened starting from Jan, 2018 with this probe.

This will end up being a very good thing for Tesla. They were able to operate semi-autonomous vehicles for years while they iterated through software versions. They are faaar from done, but a likely outcome will be more stringent regulation on companies that want to do a similar approach of putting the betas in customers hands. This makes it way harder for automotive companies to get the same leeway, putting Tesla further ahead.

When I was learning to drive, my grandmother drilled into me to never swerve for an animal that jumps out in front of the car. This saved me when I was driving by the Grand Canyon and jack rabbits kept jumping, out of nowhere seemingly, in front of my car. I drilled "never swerve" into my son and it saved him on a mountain road when he hit a deer. He didn't go into the trees. When I drove in Alaska I asked why the forest was cut back from the road. They said that moose like to step out in front of cars.

I have no idea how self-driving fits into this. I don't have a feel how self-driving responds to emergencies. I'd have to experience an emergency in one. For that reason, I don't see myself ever trusting self-driving.

I don't know how Tesla ever presumes to achieve FSD when they cannot detect stopped objects in the road, especially emergency vehicles. This is incredibly disappointing.

Does anyone know if the FSD Beta has this ability?

TeslaDeaths.com estimates Tesla at ~240M miles per death as of late 2019 (1). NHTSA puts the national rat for all vehicles in 2020 at ~90M miles per death. If we exclude motorcycles, the national rate is 104M miles per death in 2020.

(1) https://www.tesladeaths.com/miles.html (2) https://www-fars.nhtsa.dot.gov/Main/index.aspx

It would be great if Tesla could actually get speed limit detection right for starters, and then slow down the car in time for a new reduced speed limit. Had my MY for just a few days and found that inability pretty annoying. It detected a 40mph road as 60, and a 60mph freeway (which is only ever 60mph) as 70. I’ve had 5 year old Garmin GPS units that do better than that offline. I mentioned this in a local owner group and a bunch others chimed in saying they see this too.

I was excited to read about Tesla's "autopilot" until I read the details. To me as a consumer, autopilot would let me get in my car after dinner in SF, set the destination as Las Vegas and wake up in the morning at Vegas. Wake me up when that exists.

Or I can tell my car, "Hey tesla, go pickup my kid from soccer practice" and it would know what to do.

The crash in the city of Woodlands, Texas, was pretty terrifying. After hitting a tree, the car caught on fire. The driver was found in the back seat, presumably because he couldn't figure out how to open the door to get out.

I believe we should have safety probes. Lots of people who have taken money from the auto industry want this specifically for Tesla. There is a strong possibility this is political punishment for wrongthink.

OK people. There have been a grand total of 11 cases in 2.5 years. NHTSA investigates a lots of things. How many regular drivers collided with emergency vehicles in the same time frame?

I wonder, how does the autopilot safety record compare to driving in similar conditions to those which AP is typically used (open highway, good weather).

I always see anecdotes about Tesla crashes, but not any statistics vs other cars. I tend to assume that this is because they aren't more dangerous.

Maybe the AI has become self-aware. It's now picking off it's enemies, slowly, strategically, one by one.

Oopsie. This strikes me as perhaps one of the growing pains of not-quite-self-driving: common sense would dictate that manual control would be taken by the driver when approaching an unusual situation like a roadside incident, but we just can't trust the common sense of a minority of people.

Tesla perhaps isn't being loud enough about how autopilot isn't self-driving, and shouldn't even be relied upon to hit the brakes when something is in front of you.

Sounds like the system should be shut off completely until they find the flaws.

> In one of the cases, a doctor was watching a movie on a phone when his vehicle rammed into a state trooper in North Carolina.

Doesn't autopilot require you to put your hands on the wheel fairly regularly? Are these incidents just a matter of people using this feature outside of its intended use case?

Ban human driving on the interstate, highways, etc. Boom, self driving now works at scale.

Craziest statistic: of the 31 serious crashes involving driver-assist systems in the U.S. since June 2016, 25 of them involved Tesla Autopilot.

Judging by the lack of a market reaction this morning, this is mostly immaterial.

NHTSA is lumping TACC with Autopilot to increase number of incidents and make the case sound more serious:

"The involved subject vehicles were all confirmed to have been engaged in either Autopilot or Traffic Aware Cruise Control during the approach to the crashes," NHTSA said in a document opening the investigation.

TACC is very different from Autopilot.

Having human drivers and assisted drivers on the same road is problematic currently.

I think the best situation would be to have 'automated' stretches of highway specially designed to 'help' self driving systems.

Only self driving vehicles would be allowed on such special highways, and everything would be built around such systems.

Did I not just say this before? [0] [1] [2], seems like the safety probe formally agrees with my (and many other's) concerns over the deceptive advertising of the FSD package and its safety risks towards other drivers.

Perhaps this is for the best.

[0] https://news.ycombinator.com/item?id=27996321

[1] https://news.ycombinator.com/item?id=27863941

[2] https://news.ycombinator.com/item?id=28053883

The Tesla FSD beta is what we all dreamed of in the 90s. It’s mind blowing. Its so crazy that it’s finally arrived, though not able to fully and reliably drive itself in every circumstance, and nobody seems to care. People only seem to be foaming at the mouth about the way way the product is labeled. If HN found the holy grail but it was labeled “un-holy grail” then HN would apparently chuck it over their shoulder.

Teslas are the safest vehicles on the road, according to the National Highway Traffic Safety Administration. (https://drivemag.com/news/how-safe-are-tesla-cars-5-facts-an...).

Teslas crash 40% less than other cars, and 1/3 the number of people are killed in Teslas versus other cars.

Indeed once a common failure mode like this is identified it needs to be investigated and fixed. Something similar happened a few years ago when someone driving a Tesla while watching a movie (not paying attention) died when they crashed into a light-colored tractor trailer directly crossing the road. So an investigation makes sense. But much of the general criticism of self-driving and autopilot here seems misplaced. Teslas and other self-driving vehicle technologies are saving lives. They will continue to save lives compared to human drivers, as long as we let them.