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.
I don't know if Dan Noyes is researching crashes into barriers in the bay area across the board regardless of vehicle, but if they truly want to improve public safety I think they should cover the spectrum of accidents, how many barrier accidents have there been over say the last 5 years, did they result in injury or death and statistics on that. With that information they might be able to pinpoint the most dangerous areas for drivers and get changes made to those areas by highlighting them in their report and get improvements made to other locations in general. I don't use facebook or twitter so not able to contact Dan Noyes or the I-Team but maybe someone else here might want to do so.


Update: Thought since google maps was dated 10/2017 decided to check out the status of the barrier after the Hayward accident and add it to this post. Shows it was extended after the accident, albeit the front panel is in a partially damaged state. However it apparently had been hit again since the Red MS accident as the damage to the chevron plate is totally different! Wonder if the I-Team knows this. ;)

92atHesperianExitBarrierFixed - 1.jpg


From 9/2017
92atHesperianExitMS - 1.jpg
 
Last edited:
  • Like
Reactions: GatorGuy
I don't know if Dan Noyes is researching crashes into barriers in the bay area across the board regardless of vehicle, but if they truly want to improve public safety I think they should cover the spectrum of accidents, how many barrier accidents have there been over say the last 5 years, did they result in injury or death and statistics on that. With that information they might be able to pinpoint the most dangerous areas for drivers and get changes made to those areas by highlighting them in their report and get improvements made to other locations in general. I don't use facebook or twitter so not able to contact Dan Noyes or the I-Team but maybe someone else here might want to do so.

Update: Thought since google maps was dated 10/2017 decided to check out the status of the barrier after the Hayward accident and add it to this post. Shows it was extended after the accident, albeit the front panel is in a partially damaged state. However it apparently had been hit again since the Red MS accident as the damage to the chevron plate is totally different! Wonder if the I-Team knows this. ;)

Looks like that location used to have the sand barrels too, then they switched it to a Smart Cushion.
It looks like the Smart Cushion has been impacted multiple times in recent years.
aug16.jpg


I think they intentionally use that device in places where they have observed frequent impacts.

Interestingly they seem to be switching back and forth. If I flip through the years I see barrels in 2014/2015, then Smart Cushion in 2016, then back to barrels in early 2017, and now back to Smart Cushion again.

barrels2.jpg


I wonder if those Smart Cushions are turning out to be less re-usable than they first thought. If they replaced one with barrels for a while, the old one must have been damaged beyond repair. We can't really tell from photos when they either repair/reset or have to replace the whole thing.
 
Last edited:
I do think that any firmware update that modifies autopilot behavior should automatically disable autosteer and display a reminder that the feature is still in beta and updates may cause it to react differently under some circumstances.

It seems like a decent possibility from the recent videos that changes to how autopilot recognizes lanes can result in acting strange around lane splits. That reddit video makes me think the car might have been confused thinking the car ahead of it was taking an exit and trying to recenter into what it thought was the correct lane. Changes to behavior intended to keep autopilot from diving into highway exits could have strange effects on lane keeping. It doesn't help that almost all of those videos I've seen so far have very faded and potentially confusing line markings.

Repetition leads to boredom and I'd be willing to bet that most autopilot users are more vigilant the first few times they use it on a new route, but grow more and more complacent the more times they drive it. As autopilot is improved, there's a very real risk of regressions under certain circumstances and it's impossible for Tesla to anticipate them all.
 
OMG Not a fan of that hinged signage as shown on the reddit link above! It's a slim stationery fixture essentially (although it folds forward when you strike it). Maybe people would see the thin arms but how are car systems like AP suppose to see it? And what happens when you don't have a detectable car in front of you? Seems like you would definitely not want to be on any driver-assist with these. Looks like there are partial bots dots used as well (maybe some missing?). As for using GM's SuperCruise, even mapping it wouldn't work I suspect because the guards are in place only during certain hours.

Seem to think it would do a bunch of damage to your car if you hit them and the other hinged railroad-like arms (second photo) aren't any better for the same reasons. Have to think these would get broken off eventually and go flying into neighboring lanes.

redditbarrierscene - 2.jpg


redditbarrierscene - 1.jpg


The Mt. View and Hayward crashes were much less complicated roadways in comparison to the above.
 
Last edited:
All these posts about the car driving into those gore points - Is there any experienced autopilot user that would actually let their car go unguided near one of these (aside from youtube voyeurs)? I might wait to see if the car goes the way I wanted, but I can't imagine being disappointed when it picks the wrong way, it does that probably at least half of the time.
 
I think whenever there is a severe car crash due to badly marked gore points and poor lane markings, the politicians who oversee the funding and the person responsible for highway safety should see an immediate 5% pay cut for one year.

In no time, there will be no longer any car crashes due to poorly maintained/marked highway dividers.

As is, they have no accountability. Well, other than "thoughts and prayers.™"
 
I think whenever there is a severe car crash due to badly marked gore points and poor lane markings, the politicians who oversee the funding and the person responsible for highway safety should see an immediate 5% pay cut for one year.

In no time, there will be no longer any car crashes due to poorly maintained/marked highway dividers.

As is, they have no accountability. Well, other than "thoughts and prayers.™"

Out of curiosity, are you willing to let these folks increase budgets and taxes to the level where they feel comfortable they will be able to provide these services at this level of quality?
 
  • Like
Reactions: e-FTW and T34ME
@Economite


Or you go as far as it has gone here in the UK, where motorways are being converted into what are in effect 100% radar speed limited.

In part the new "smart motorways" we have here are a backdoor to automated average speed camera checks (which are already prevalent in things like construction zones or known accident hot-spots). Between on ramp and off ramp cameras read your licence plate, stop-clock your time between the two and if that equates you exceeding posted limits a fine arrives in the post automatically.

Be careful where you wish your tax dollars spent ;)
 
Out of curiosity, are you willing to let these folks increase budgets and taxes to the level where they feel comfortable they will be able to provide these services at this level of quality?

Yes of course. What can be more important than the safety of human beings. For all the fingers pointed at Tesla, I feel more fingers need to be pointed at how poorly the roads are maintained and how horrible the road markings are.

The point being, the road is so horribly marked that both AI and actual human beings constantly get confused and hit the gore point.

They should definitely fund road maintenance and safety and this is why I suggested that politicians who don't assign adequate funds for road safety should see a pay cut if someone is gravely injured or died because the roads were poorly maintained.
 
  • Like
Reactions: 22522
Yes of course. What can be more important than the safety of human beings. For all the fingers pointed at Tesla, I feel more fingers need to be pointed at how poorly the roads are maintained and how horrible the road markings are.

The point being, the road is so horribly marked that both AI and actual human beings constantly get confused and hit the gore point.

They should definitely fund road maintenance and safety and this is why I suggested that politicians who don't assign adequate funds for road safety should see a pay cut if someone is gravely injured or died because the roads were poorly maintained.
Exactly, there was no one thing that let this happen. It was a series of events that came together and this was the unfortunate result.
 
  • Like
Reactions: P85Dave
The markings of the road make a huge difference.

I watched the videos that were made, illustrating how AP will take you of course, describing they had to take over.
If you watch closely one line is very faded or non extent so the AP recognizes our sees the darker more prominent line
and takes that line or road marking as a main road and centers to it.

I have observed in my own driving that it will (ap) get lost and drift over a little then corrects it self and recenters.

If I am driving up and over a steep hill or an aggressive turn it will lose it self and drift or swerve trying to find the center again.

I realise that that is a limitation to the system and plan to take over control.
AP has limitations agreed but it has saved me multiple times from accidents that other drivers cause.

Isn't a design flaw for the AP to rely principally on the road mark.
What about if there is suddenly some snow, rain, or mud... covering the road marks.
Or if the lanes get controlled by traffic light to be shared with the opposite traffic during rush hours.
Following the car in front also is not ideal, as show with the collision with a fire truck.

I believe that AP need to be associated with some type of GPS and AI control.
This would certainly necessary for autonomous driving.
 
Isn't a design flaw for the AP to rely principally on the road mark.
Nope, that's how I drive.

What about if there is suddenly some snow, rain, or mud... covering the road marks.

Then I use the shoulders, or previous car tracks, or look for where the markings are poking through.

Or if the lanes get controlled by traffic light to be shared with the opposite traffic during rush hours.

Then I follow the lights/ markings.

Following the car in front also is not ideal, as show with the collision with a fire truck.

Following other cars is fine as long as you keep enough space for braking and reaction time.

I believe that AP need to be associated with some type of GPS and AI control.
GPS doesn't handle the firetruck or traffic signal issues you just raised. Also not as reliable in valleys or cities.

This would certainly necessary for autonomous driving.

FSD will have vision and mapping integration. However, with EAP (only form of AP on Teslas) people are still the drivers.
 
  • Like
Reactions: NeverFollow
Isn't a design flaw for the AP to rely principally on the road mark.
What about if there is suddenly some snow, rain, or mud... covering the road marks.
Or if the lanes get controlled by traffic light to be shared with the opposite traffic during rush hours.
Following the car in front also is not ideal, as show with the collision with a fire truck.

I believe that AP need to be associated with some type of GPS and AI control.
This would certainly necessary for autonomous driving.

Based on my experience, if the lines become poorly visible due to snow etc the car should provide an alert and disengage autosteer forcing the driver to take over.
 
  • Informative
Reactions: NeverFollow