-
Want to remove ads? Register an account and login to see fewer ads, and become a Supporting Member to remove almost all ads.
Model X Crash on US-101 (Mountain View, CA)
- Thread starter mookhead
- Start date
Barklikeadog
Active Member
I'd like to hear your thoughts on that. There is the discussion of Lidar which is certainly a hardware issue. Also for true autonomous activity, it is very likely that
the car will have to communicate with other vehicles and vice versa. You think all of this can be accomplished with some software updates?
Fellow strabismus patient here FTW! You are entirely correct that stereo vision isn’t required to successfully operate a vehicle. It isn’t even required to fly a private plane, interestingly. I have zero depth perception (was off the scale on alignment prior to surgery), and was able to pass the medical and get a private pilots license. (And yes, the medical exam included a vision check). The doc told me that strabismus patients actually can do better as they aren’t “tricked” in situations where depth perception fails.
This was unnecessary and in bad taste.
FUD is better than a false sense of security. The consumers who ask tough questions, stupid questions and even blame the product after not using as intended, they all help make the product better. But, the worst kind of consumers are the fans who will even buy poop in the box from their favourite brand.
In this forum, there were a lot of "experienced" drivers who were claiming that there is no way this could be AP when the news first broke out. And after AP's role became evident, some of them were "devastated."
I am here in the forum because I am assuming Tesla's reps follow these forums and some of the criticism and fear would register with them. And of course, I don't want to hear from the fans who cannot find a single fault on the AP tech for driving the car into the barrier and put all the blame on the driver, CalTrans and even the Sun.
My thoughts?
Tesla is going for the end goal of fully self sufficient autonomous driving. Due to that goal, they are taking way longer than other companies to develop the intermediate/ lower feature content versions of the system.
For example: a system where all cars are tagged and provide GPS position, heading, and speed is much simpler. However, it requires all cars to have that system (which is not likely any time soon) it also requires no other objects in the vehicle's path. Add a object in the road: all the car position data is no help in this case. So you have to have robust object detection. If you have robust object detection, then do you need the other car data (beyond smart traffic control and routing)?
Lidar is great in that it gives you a range to the nearest object as native data. However, it doesn't provide any context data. Traffic signals, turn signals, police lights, cross walks are not handled by it. It also can't tell you if the return is cinder block or a plastic bag. So you need to have a robust vision system to handle that type of data. If you have a robust vision system, is the Lidar still needed?
I drive without lidar or other car data so it can be done. I think one stumbling block people have is the level of precision cars need to function. From a basic navigational aspect, it only cares if there is an object that may collide, it doesn't care what the object exactly is. Similar to driving without glasses, it doesn't matter what the blob is, just don't hit it. Handling turns is more complicated of course, but lanes are just objects too.
On the side of useful things that do not incur a hardware cost are mapping strategies. Each car can report its position and speed. This provides a live map of traffic conditions. It also provides an ant-trail of normal vehicle paths. Repetitive sampling of intersections can provide a "most likely to be here" mapping of where the traffic signal are. This is great for providing hints, but in order to handle changing conditions, the vehicle must be able to navigate without those aids.
A huge advantage Tesla has is that all the cars are able to feedback route data. So baseline path and traffic control data can be processed back at the ranch if need be (current indication is that that are not doing this on the full fleet). They could set up the system such that once you have manually (EAP) driven a route 5 times, it has enough data to FSD it. Again, that is sort of cheating since the system needs to react to changes, so the baseline data should only be used as a gut check against low confidence vehicle processing.
Summary of my thoughts:
Tesla is jumping to the hardest version of the problem, which takes more work and does not yield much in the way of mostly working intermediate steps. However, once they get the vision system working, it covers (or can be updated to cover) all the edge cases. Current EAP is a limited set of lane assist and cruise control functions to provide some value to owners. Future EAP will be the FSD code set with extra driver interaction/ attention requirements.
Experienced Tesla owners with AP2, common sense, steady hand and a level head have commented. You just didn’t choose to listen to them. Instead you draw from your non experience and tell everyone else how “things work”
This was a coin flip on if AP was on or not. It makes no difference as both are fallible if you are unaware enough behind the wheel.
Family didn’t do their attorney any favors when they mentioned Walter complained 7-10 times yet couldn’t get his hands off the AP stalk.
It’s self evident to Tesla about “AP faults”’when they collect logs. They know when I and the other hundred thousand drivers “disengage AP”. That’s something Walter should have done if the system wasn’t up to the task for that particular segment.
I just find it very hypercritical for you to complain about AP, don’t intend to use it but won’t take the option off.
When you do use it (and you will with your family in the car) I don’t see any planned apologies.
No one can blame Caltrans and no one can blame the Sun. If Caltrans is not the insured party and the sun is not the insured party and if Tesla is not the insured party it is all driver. No exceptions. It doesn’t get any clearer for legal and financial liability for the initial impact.
I find your lack of ability to hold Tesla even remotely responsible in this incident astounding. Can Tesla do ANY wrong in your mind? It's a serious question.
Last edited:
bonnie
I play a nice person on twitter.
I think you're missing the other part of what the 'experienced' drivers were saying beforehand, so let me help:
Either AP wasn't on OR if it was, it was inconceivable that Walter would not have paid close attention in the very spot his family said he'd complained about. Now that we know that AP was activated, it is inconceivable that no intervention was recorded for 6 seconds prior to the crash. I can only conclude that something, such as a medical condition, made it impossible for him to react. Characterizing owners as being devastated that AP was on is incorrect. The devastation is for the family who lost a brother, husband, father.
(Spreading FUD vs. a false sense of security are not the only two options. A discussion based on facts is my preferred option. But that's difficult to do when you choose to dismiss any experience that doesn't fit your preconceived opinions.)
How about calming down and choosing somewhere in the middle? There are plenty of shorts who spread vicious non-factual and outright lying FUD about Tesla and they aren't trying to make the company better, they're trying to make the company disappear. You don't need to help them. Spreading FUD about Autopilot when you don't even have any experience with it is duplicitous at best and dishonest at worst. I don't have a problem with you questioning Autopilot, it's when that degenerates into FUD I get antsy.
I viewed the after-the-event "test" video by the volunteer owner and frankly, I can see where even a human might get temporarily fooled by the faded right line leading to the abutment. That's one scary part of the highway.
I don't have Autopilot on my car, but I've driven several loaner S for a day each and a friend's X several times, enabling AP to see what it does and how it behaves. So I have some idea, and I can see on long trips it would be a great tool to lessen driver fatigue. However I would be very careful and likely turn it off when stretches of road came up that were out of the ordinary, such as narrowing lanes coming to bridges, any construction area, coming up to major interchanges if I don't know exactly where I'm going, etc. To me that's just common sense. As someone with more wit than me once said, however, "common sense isn't so common".
The attacks on Tesla's website marketing descriptions are also disingenuous. *ALL* marketing has to be taken with a grain of salt. Of course Tesla is not going to describe Autopilot using a long list of potential and possible side effects like a drug company. Who does? (Except for drug companies because they are legally forced to). A bunch of pages back I posted a magazine ad from Mercedes that claimed "self-driving". If you believe 100% of any marketing hype for any product from any company in any industry, you are doomed to disappointment.
I do have a problem with this analysis. if you read the Tesla blog https://www.tesla.com/blog/update-last-week’s-accident
it says "In the moments before the collision, which occurred at 9:27 a.m. on Friday, March 23rd, Autopilot was engaged with the adaptive cruise control follow-distance set to minimum."
There are two ways to read this, could be more.
1) The driver engaged AP few second prior to the crash. This would mean he was attentive enough to engage AP few seconds before the crash. Its unlikely that a medical condition occured just after he engaged AP. It is also unlikely that somebody will engage AP as a result of medical condition, rather than stopping at the shoulder.
2) The driver had autopilot engaged from the most of his ride till the moments before the collision. AP got disengeged when he took over moments before the collision.
I do not buy the 6 seconds logic and I have explained in the following post.
Last edited:
bonnie
I play a nice person on twitter.
I'm curious as to the effort you think it takes to disengage AP. I strongly suggest you try AP & you might have a different opinion.
And I do 'buy' the six second assertion. Your linked 'analysis' is speculation in the absence of actual facts. The full paragraph (not just the part you pulled out) is:
In the moments before the collision, which occurred at 9:27 a.m. on Friday, March 23rd, Autopilot was engaged with the adaptive cruise control follow-distance set to minimum. The driver had received several visual and one audible hands-on warning earlier in the drive and the driver’s hands were not detected on the wheel for six seconds prior to the collision. The driver had about five seconds and 150 meters of unobstructed view of the concrete divider with the crushed crash attenuator, but the vehicle logs show that no action was taken.
A few questions for you: 1) If, as you say, he didn't engage AP until moments before the collision - why would he do that if he hadn't been actively steering his X for the time preceding that?, 2) How could he engage AP without taking some action? and, 3) if 'engaging AP' is not an action, why would he engage AP and not bother steering his X prior to that?
This assertion about disengaging autopilot is so wrong, it's scary. There are three ways that the driver can forcibly disengage autopilot on a Model X. They are: 1) depress the brake pedal, 2) push the cruise stalk forward, and 3) turn the steering wheel enough to override the auto-steer. Each of these takes but a fraction of a second, and should have been trivial for the driver to do, if he were paying attention and he was physically able (i.e. not incapacitated somehow). Drivers with actual experience operating autopilot know this.
Bruce.
bonnie
I play a nice person on twitter.
Agree & I'll go a step further:
Turning the steering wheel enough to override the autosteer is as simple as turning the wheel as you would if autosteer was not engaged.
It's not some special way to disengage that requires learning - your normal steering will override autosteer if there is a difference between the two. You'll feel a small 'bump' under your hands, but it is not some effort to wrestle it back. It is simply normal steering (which is why I've said at other times that I'm more likely to inadvertently disengage than get a nag).
First, I have driven with AP but I would not call myself experienced. Let me rephrase my initial comment. I did not mean he could not disengage AP in time. I know how easy it is. I mean could not take corrective measure in time.
I am trying to make sense of the statement from Tesla. Why would they say "In the moments before the collision, AP was engaged" vs simply saying "AP was engaged at the time of collision".
I will say this force can be variable because my experience with it is different than yours, so it is possible to have a car with a heavier "release" that causes a scary swerve. My car is in the service center now, and I have noted that as a problem. We will see if they are able to address it (I am hoping I don't just get a "that's normal" response back). I would much rather have it be almost too light like yours vs what it is now.
Since I have seen similar comments from others about the steering disengage force, it might be possible for others to have cars with the same problem as mine.
Let me rephrase my initial comment, will edit my post too. I did not mean he could not do disengage in time or had difficulty disengaging. I mean could not take corrective measure in time.
It sound like you are claiming he could not press the brake. Are you saying his reaction time was so slow that he could not take any action in the multiple seconds of clear line of sight before the impact?
Joe F
Disruption is hard.
I know this has been mentioned at least once in this marathon thread, but might have been missed by some, so I'll relate my experience here.
I too have had swerving at high speeds when disengaging AS with the wheel, and it was very bad in the early AP2 days when hit with truck lust, but found that all went away once I changed to comfort level mode. It takes very little to disengage AS on my MS, and little to no (inches perhaps) oversteer.
Comfort mode has not, at least for me so far, caused me to inadvertently disengage AS.
Kudos to those few here who are undertaking the impossible task of explaining color to those without sight from birth.
Good to know. If the SC doesn’t find anything wrong, I will try that out. I keep my steering set to Standard normally, so it hadn’t occurred to me to try that.
Similar threads
