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

GM Supercruise's 0 incident in 5.7 Million miles Vs Tesla Autopilot's 1 accident every 3.45 million miles (Safety Record Comparison)

MXLRplus

Active Member
Mar 11, 2020
1,599
2,463
Eastvale, CA
Off Topic, how does a model x replace a pickup? Did you modify the vehicle? Remove seats?
No, I just fold down the seats and I have trailers. It can carry a long object inside or a bulky object. For larger parts I bring the trailer.
There are certain things the pickup is better at, but if you want an electric work truck, the MX is the closest option so far.
 
  • Like
Reactions: outdoors

thx1139

Active Member
Aug 1, 2014
1,069
3,458
Lemont, IL
So they are comparing a system that only works on roads that they have 100% mapped and cherry picked vs any divided highway. Yeah that is a real valid comparison, not!
 
  • Like
Reactions: Krugerrand

Krugerrand

Enough of the 🐩, back to 🐈‍⬛
Jul 13, 2012
11,213
58,704
Tesla friendly place
You sound as if you're a lawyer defending Tesla's Autopilot.
If there's a car driving along besides me and it comes closer to the lane marker between my lane and theirs, I may move over slightly to the other side of my lane. But I do not slow down or step on the brake. You do? That's strange.
This is what the previous poster was talking about, that his Tesla is applying the brakes in this situation.
Again, you’re being purposefully obtuse. Depending on where my car is compared to the one in the other lane that’s encroaching (and how quickly and far it’s encroached/ing AND whether another car is directly in front, behind, or on the other side of me) determines if I lift off the accelerator, actively brake, speed up to finish getting by, do nothing, simply move over a bit or move over a lot, honk my horn and flip the bird, etc... Interestingly enough, my car on AP/NOA responds EXACTLY the same way save the last item - as required for the situation to avoid an accident. Yes, sometimes my car just moves over in the lane a bit and then re-centers when safe, sometimes it moves out of the lane, sometimes it brakes lightly and momentarily, sometimes it brakes harder and waits, etc...

Long gone are the days when AP seemed to be magnetically pulled along side tractor trailers. And I can’t remember the last time I’ve had a true phantom braking episode.

Yes, sometimes AP does not behave exactly as it should. If you pay attention, you can predict with high certainty what situations will be difficult for AP to handle. Simply override the system in those instances until such time as a new update arrives and then give it another go.

I’ve had little difficulty getting the most out of Tesla’s driver assist features, imperfections and all. AP or TACC are extremely helpful in stop and go traffic, and on long distance highway driving. AP/EAP/FSD are getting better every day at navigating more intricate and complicated scenarios.
 

hopmix

Member
Mar 29, 2021
55
44
Bay Area
Your scooter friend is what we call a squid or squirrel.
LMAO. He's a total squid! He wasn't on a scooter. Though not far from it. It was a 1969 Honda s90 which are actually quite fun to around a city at rush hour.. They way he explained it to me though was he was behind the car moving slowly holding up traffic trying switch lanes with its signal. The car moved from the center lane to the left and he then accelerated and overtook the lane (I do this 30 times a day in rush hour bay bridge traffic on a much bigger bike) in the center of the lane and something moved in front of the car and it suddenly jumped back to center but didn't see him there and it bumped him. He kind of got a little slow speed wobble and ate it. He said the car didn't just ease back over, it was a sudden jerk. Like something a human would do. Which leads me to suspect the reason why GM most likely settled is because there was a human driver in the cruise and they may or may not have taken control when the obstruction entered the lane and moved back in to the previous lane. IM not saying much of the merit of the lawsuit or the tech was just stating I knew of one accident they were involved in.
 

Goblue2000

Member
Apr 12, 2021
6
6
Ann Arbor
This guy is lying, isn't on the latest version, or is using an older hardware AP.

Tesla has essentially fixed phantom braking as of 2021.4.15, at least with AP enabled. I would think TACC wouldn't be that far off.

On the other hand, I'm still experiencing incorrect speed limits which cause the car to slow down, but this wouldn't be considered phantom braking, as there's a clear cause (incorrect map data / missed sign reading).
I'm on the latest version with Beta turned on, M3. I experience the same degree of phantom braking as the Volt I had prior. I also experience lots of quirks. Latest one, AP on working fine and a message saying its not available for the rest of the drive. Turned it off and back on with the message still displaying. Pure vision won't stop phantom braking, in fact some degree of phantom braking is appropriate. There's going to be cases where it's not 100% clear what to do - you really want the system to brake in those circumstances until it can make a decision. Especially considering they couldn't get sensor redundancy to make the decision and abandoned that. Phantom braking may go up! Fog, white truck? We've seen that movie before with tragic results and we've seen human drivers make the same mistake. A mistake that won't be tolerated by the public.
 

powertoold

Active Member
Oct 10, 2014
2,468
4,767
USA
I'm on the latest version with Beta turned on, M3. I experience the same degree of phantom braking as the Volt I had prior.

Sorry, I don't know what AP others are using. Perhaps their highways are much different than the ones I drive on. It'd be great if people had videos of these phantom braking events. I try to record every phantom braking and review them afterwards.

"Latest one, AP on working fine and a message saying its not available for the rest of the drive."

Almost sounds like you're not satisfying the nag.
 

rush6410

Member
Sep 26, 2016
637
605
Montgomery TX
I have identified one scenario that took a few slow downs for me to figure out what the car was doing. I primarily drive on a highway in North Houston. There is an HOV designated lane distinguished by double white solid lines. When I am driving in the adjacent fast lane while approaching an HOV merge area, first, the car wants me to switch lanes one over to the right. If I follow the NOA recommendations, it will maintain my speed. If I ignore the lane change recommendation to avoid the HOV merge into the fast lane, the car slows down anticipating a merge event. I assume if you do not have FSD or EAP with NOA turned on, the car would always slow down when approaching the anticipated merge.
 

dgatwood

Member
Dec 20, 2017
902
1,003
Sunnyvale, ca
You make a great point that irritates me. Most ppl can't tell you if braking was truly phantom or the car was responding to a condition that the driver is unaware of. I would suspect the later. Every "phantom braking" that I have encountered was actually the car responding to a condition. I always try to immediately identify why it applied the brakes so I can avoid the scenario which caused it. Most of the time it is a vehicle getting too close to the lane line next to you. Or, it perceives the vehicle is encroaching.

Yeah, the problem, at least with HW 2.5, is that it happens five seconds after the condition that it freaked out about has already passed and is no longer important. And because it takes it so long to start reacting, it panics and massively overreacts. HW 2.5 is just plain too slow — comically so.

HW 3 is much, much more fluid, slowing down much sooner and much less. It actually behaves like a car should behave.

Also, HW 2.5 is constantly getting confused by RADAR returns from dips in the highway, particularly if there's a shadow from a bridge or whatever that roughly coincides with it. Things will be *much* better when they *stop* using RADAR entirely (or greatly limit its use), I suspect.
 

dgodfrey

Supporting Member
Nov 27, 2016
468
4,358
Niskayuna, NY
Except that Tesla also compares AP and non AP to its own fleet.

But where I do agree that numbers can be skewed is that for now, the vast majority of AP miles are on highways which are less prone to crashes. I would like to see the miles of AP vs non AP regarding highway miles only.
It would also help to know if the Tesla accidents were on the major highways or the secondary's where it's supposedly not intended yet.
Regardless, using AP miles as some metric to overall safety when compared to all human driving is misleading as others have pointed out. Freeway miles are the safest conditions which is where the majority of AP miles are accumulated. I don't doubt AP is safer and better on average than humans, just not to the level Tesla makes it seem.
 
  • Like
Reactions: SO16

About Us

Formed in 2006, Tesla Motors Club (TMC) was the first independent online Tesla community. Today it remains the largest and most dynamic community of Tesla enthusiasts. Learn more.

Do you value your experience at TMC? Consider becoming a Supporting Member of Tesla Motors Club. As a thank you for your contribution, you'll get nearly no ads in the Community and Groups sections. Additional perks are available depending on the level of contribution. Please visit the Account Upgrades page for more details.


SUPPORT TMC
Top