Welcome to Tesla Motors Club
Discuss Tesla's Model S, Model 3, Model X, Model Y, Cybertruck, Roadster and More.
Register

Model X Crash on US-101 (Mountain View, CA)

This site may earn commission on affiliate links.
My point exactly. Current AP tries to run but cannot walk.

This does not mean they should not do it, but current performance is aimed at the gallery instead of truly improving safety:

No stationary object detection, no radar out the back, additional cameras exposed to rain and dirt.

Is it good they are trying and pushing everyone else to do the same? Yes.
Are they aiming in the right direction and doing a good job? Sort of.
Could they do better/smarter? Yes.
 
  • Helpful
Reactions: VT_EE
To give people some perspective, Tesla already had AEB for this kind of gore point barrier back in Jan 1st of 2017.
The problem is being able to distinguish from this gore point barrier and an overpass.

Based on the feedback of myself and others, Tesla had to dial the sensitivity way down.

If you read my report, the phantom breaking was no good and would have been resulted in more rear end accidents if drivers behind a Tesla was not attentive.

8.0 (2.50.185) caution using TACC/Autosteer features

I don't understand why people refuse to understand that there would have be no accident if the driver had eyes on the road. It would have been immediately evident that AP had gone off script.

LOOK AT THE ROAD when the AP nag tells you to grab the wheel.

There was sufficient time to hard brake and survive that crash, or even avoid the crash in the first place.

The problem is that AP makes people take eyes off the road. Drivers will be attentive otherwise. This is partly a Tesla problem because the way they are marketing the capabilities. They have put a fully autonomous drive video on Tesla’s website.

Tesla haven’t figure out phantom breaking is not drivers fault. The safety board has their work cut out. They can’t let people become victim of this research project doesn’t matter how useful it can be when completed.
 
I think you should picket at cell phone companies, driving while on the phone is a much bigger problem.



He is dead because he buried his head into something else instead of looking out the window.
The cell phone companies doesn’t claim that my phone is safer because it doesn’t distract as much as others while driving. He is dead because he did not read the fine print of a product that he purchased. Pretty steep price to pay.
 
If you don't like its design then don't buy it and save your money!


Well, it seems that government may step in and decide for us.

"If you don't like its design then don't buy it and save your money" doesn't work. If government thinks it is not safe, it will get pulled. Multiple medical devices get pulled often when unintended mortality is recorded even if the device may actually save many more lives.

Now that NTSB is involved, we have no idea how this will pan out. Current design may not satisfy NTSB moving forward. If more strict regulation and guideline will placed for Tesla AP or not. Only time will tell.
 
Very helpful discussion above. Owner of a MX 100D based in the UK. First thing to say is that we have much better road design here in Europe, can’t imagine something like that existing. I use AP all the time on highways here as well as around town and in general find it adds to safety (when used with supervision...)

Also worth taking a look at some of these tests that show how good AP really can be eg:


and also


Do not mean to imply that AP is 100% perfect. Problem is even if it is 99.9999% perfect that 0.00001% error rate can result in some horrible tragedies.
 
  • Informative
Reactions: ggnykk
The problem is that AP makes people take eyes off the road. Drivers will be attentive otherwise. This is partly a Tesla problem because the way they are marketing the capabilities. They have put a fully autonomous drive video on Tesla’s website.

Tesla haven’t figure out phantom breaking is not drivers fault. The safety board has their work cut out. They can’t let people become victim of this research project doesn’t matter how useful it can be when completed.

What are you talking about.. research project? For the Darwin awards? Is Tesla trying to figure out just how careless people are? If you enable AP at 70mph and start texting on your phone, then there is a good chance you will not have a long and happy life. Same thing world happen if you didn't have autopilot. The difference is that autopilot probably saves 100s of lives a year because it cuts the odds of someone drifting into another lane while distracted. Sad but true. Distracted driving is now worse then drunk driving and Tesla didn't invent that. AP allows you to relax a little and scan farther down the road and be more aware of your surroundings. That's it.. It's not the texting and driving tool that people are clearly using it for. Don't be one of those people.
 
Current Autopilot works as designed.
1) Driver is the primary operator, not Autopilot.
2) Autopilot is not finished.

Then, for goodness sake, don't call it "AutoPilot"!
Most people would have unrealistic expectation of the term "AutoPilot".

Call it "Driver Assist" instead - as this is exactly what it is for the current implementation. If this was the term used, then you won't even see these accidents mentioned in the news.

Other manufacturers like BMW now has similar capability like auto steer, auto change lanes etc. But no one else would call their system "AutoPilot".

If you don't like its design then don't buy it and save your money!

But when Tesla uses a term like AutoPilot, I bet many buyers would be misled into buying that option without being aware of its current limitations.
 
Very helpful discussion above. Owner of a MX 100D based in the UK. First thing to say is that we have much better road design here in Europe, can’t imagine something like that existing. I use AP all the time on highways here as well as around town and in general find it adds to safety (when used with supervision...)

Also worth taking a look at some of these tests that show how good AP really can be eg:


and also


Do not mean to imply that AP is 100% perfect. Problem is even if it is 99.9999% perfect that 0.00001% error rate can result in some horrible tragedies.

Don't use AP where people and bicycles are. It's not meant for that. It's meant for highways and some local roads that are not loaded with pedestrians. Always use with caution. People need to stop treating it on pedestrians. It's stupid.
 
Very helpful discussion above. Owner of a MX 100D based in the UK. First thing to say is that we have much better road design here in Europe, can’t imagine something like that existing. I use AP all the time on highways here as well as around town and in general find it adds to safety (when used with supervision...)

Also worth taking a look at some of these tests that show how good AP really can be eg:


and also


Do not mean to imply that AP is 100% perfect. Problem is even if it is 99.9999% perfect that 0.00001% error rate can result in some horrible tragedies.

It will be good to show this test with a model X since I haven't seen too many much reference to Model X AP collision avoidance
 
If more strict regulation and guideline will placed for Tesla AP or not. Only time will tell.

^ This.

Currently the guidance ( https://www.hsdl.org/?view&did=795644 ) states the following:
To aid NHTSA in monitoring HAVs, the Agency will request that manufacturers and other entities voluntarily provide reports regarding how the Guidance has been followed. This reporting process may be refined and made mandatory through a future rulemaking. It is expected that this would require entities to submit a Safety Assessment to NHTSA’s Office of the Chief Counsel for each HAV system, outlining how they are meeting this Guidance at the time they intend their product to be ready for use (testing or deployment) on public roads.

With two high profile fatal accidents featuring autonomous vehicles in such a short period, I think there will be a lot of pressure to move from a voluntary basis to something more formal.

My take on this is some (not all) people are treating AP like a Level 3 system, when it isn't.

The guidance is clear that the manufacturer must take adequate steps to train/educate/inform users in correct operation of the system. This was highlighted in the last incident.

Sadly this doesn't seem to be effective, and without collision avoidance behavioural competency needed to make it Level 3, my view is Tesla's clearest route out of this from a regulatory perspective is to force hands on at all times. (Or in the case of the Model 3 enable the driver facing camera as with the GM system.)
 
I think you're spot on with your summary. I saw the faded line in other guy's video too. My guess is that the Autopilot latched onto the brighter white line leading left exit, thinking it was the left white line, but unfortunately it was the right white line of the exit.

This accident is an example of why I think purely autonomous vehicles are a lot farther off than some would think. In a perfect world, with standard markings that are brand new (like a test track...) the automation will work great. But get in the real world with faded lines, construction, obstructions to the markings, glare, precipitation... a human brain can figure it out, or slow/stop until they do, but automation has a hard time not following the pre-set rules.

And you are not mentioning mobile objects, such as people (kids running out between parked cars in a city), animals, debris. I say 2025 the earliest and that might be optimistic, I think we will see drone taxis before we have fully autonomous vehicles, the flying taxis could operate in authorized zones and elevations to not deal with so many variables.
 
My point exactly. Current AP tries to run but cannot walk.

This does not mean they should not do it, but current performance is aimed at the gallery instead of truly improving safety:

No stationary object detection, no radar out the back, additional cameras exposed to rain and dirt.

Is it good they are trying and pushing everyone else to do the same? Yes.
Are they aiming in the right direction and doing a good job? Sort of.
Could they do better/smarter? Yes.

The lack of the rear radar is just horrible even if there was no AP, its a 6 figure car without real blind spot detection. I can't believe they didn't add it with the second gen AP, Tesla should be ashamed.
 
Just watched the video Tesla owners almost crashes on video trying to recreate fatal Autopilot accident
ouch Tesla, seriously

Sure the driver should have been paying attention, but then again the technologies purpose is so that the driver doesn't have to pay as much attention. Can't have it both ways, 5 seconds go by really fast.

Autopilot follows the lanes that the road authority puts down and maintains. In that video, AP follows the lanes, same as with the X that hit the construction barrier (not referring to 101). The manual steering at the end is interesting as it seems to be directed to stay headed toward the barrier, may have been done to avoid getting rear ended.

Just for reference (from S 6.2 release notes):
Automatic Emergency Braking will stop applying the brakes when you press the accelerator pedal, press the brake pedal, or sharply turn the steering wheel

The lack of the rear radar is just horrible even if there was no AP, its a 6 figure car without real blind spot detection. I can't believe they didn't add it with the second gen AP, Tesla should be ashamed.
What purpose does rear radar serve besides illuminating the brake lights if they are not already on and a collision is imminent? Are you expecting the car to drive itself forward or swerve out of the way?
 
Thank you 'f-stop' for putting up the freeway crash site and the sun angles, on page 83.
Very nice work.

This must the be the L.A. Freeway that Jerry Jeff Walker and Guy Clark
sing about when
Guy and Susanna wrote, 46 odd years ago,

" ...don't you think it's time we're leavin'.
adios to all this concrete,
...if i could just get off of this LA freeway
without getting killed or caught,...down the road
... to some land that I ain't bought
."

I see, in f-stop's pictures, 20 lanes of concrete, crash barriers that can't be reset fast enough,
and parallel lines wide enough for Auto Pilot to follow, even when they represent long triangles
that should not be driven at all --- ending up at crash barriers.
All this and the sun is shining in the eyes of ten lanes of drivers going due East on the Equinox.
Whew, ... , kind of glad I can live elsewhere..
Condolences to the family of the driver who did get caught and killed,
the other's involved, the car makers, and the road crews.
Add to that the accident investigators, and all of the professionals who don't have quite enough time
until the next accident to get their work done.

Long hours on the highway just don't feed my soul,
even in my model "S"

.
 
  • Like
Reactions: NeverFollow
Just watched the video Tesla owners almost crashes on video trying to recreate fatal Autopilot accident
ouch Tesla, seriously

Sure the driver should have been paying attention, but then again the technologies purpose is so that the driver doesn't have to pay as much attention. Can't have it both ways, 5 seconds go by really fast.

I see no problem with how the system functioned. It even told him to pit his hands on the wheel with the visual warning as well. Any normal "reasonable" person would have taken over in time.

And no the system's purpose isn't so that the driver doesn't have to pay as much attention. It's to alleviate driver fatigue it's a big difference.
 
To give people some perspective, Tesla already had AEB for this kind of gore point barrier back in Jan 1st of 2017.
The problem is being able to distinguish from this gore point barrier and an overpass.

Based on the feedback of myself and others, Tesla had to dial the sensitivity way down.

If you read my report, the phantom breaking was no good and would have been resulted in more rear end accidents if drivers behind a Tesla was not attentive.

8.0 (2.50.185) caution using TACC/Autosteer features

I don't understand why people refuse to understand that there would have be no accident if the driver had eyes on the road. It would have been immediately evident that AP had gone off script.

LOOK AT THE ROAD when the AP nag tells you to grab the wheel.

There was sufficient time to hard brake and survive that crash, or even avoid the crash in the first place.

The points in bold are key and you are 1000% correct about phantom braking. It's just a question of probabilities of injury. Getting rear ended has a higher likeliness since you can't account for the driver behind you but a driver has control over going forward. I cannot stand it when people don't understand the system and the circumstances of proper usage. It's those people who then say if it doesn't function hands off then what's the point of buying it!

Use it in an NYC traffic jam a couple of times... you'll see.
 
It would be helpful if Tesla can release the results of video taping of these kinds of scenarios so amateurs don't have to risk their lives for science.

Stop it now. Nobody has to do anything. These are full grown adults with voting and drinking rights doing ridiculously dangerous and/or stupid things. Simply read the instructions and operate within them and all will be right in the world. Putting fruit on the steering wheel so you can sit in the backseat while driving on an undivided and poorly marked road it not in the instructions, FYI. Neither is repeatedly using AP on a section of confusing highway, if indeed that’s what our fellow Tesla owner was doing, while also not paying close enough attention.

Take your illogical thinking over to VW where they will happily hook you up to one of their tailpipes in the name of science.

Slightly off topic but not really, I witnessed a pretty beat up, private citizen, delivery SUV swerving in their lane through residential area only to then cut in front of another car. At the next stoplight, which I’m sure they would have run if another car hadn’t already been stopped, I rolled down my window and told the driver to put their cellphone away before they killed someone. For some reason they were angry I’d interrupted their texting session.

Pay attention when you are driving. It’s that simply. Be courteous when you are driving. It’s that simple. Follow the instructions. It’s that simple. And finally be responsible for your own actions and quit blaming others. That’s simple too.
 
What really troubled me is this: If autopilot warned the driver at t minus 5 second and 150 meters prior to the collision, why autopilot not taking any evasive actions such as slowing down and or dis-engage the autopilot? I would feel better if Tesla's autopilot at-least direct the vehicle to a safer environment.
The AP didn't take evasive action or slow down or disengage because the driver had the following distance set to "minimum."

If you are driving 65/70 mph or greater on the freeway; prudence, caution, and logic dictates you set your following distance to maximum!

On our current daily driver, we always have our following distance set to "maximum" whether driving 25 mph or 70 mph. We don't want to be a statistic.