Phantom braking so bad I want to return my car

[mlaskowsky]

We live in Central Oregon and drive A LOT of rural 2 lane highways. Sometimes I feel like tesla programmers dont even know these roads exist. Phantom breaking is a huge issue on these roads. Below are the major issues. We have tried every configuration of functions being on or off and phantom breaking still persists. Worse in Auto Cruise but still pretty bad in autopilot.

Traveling on two lane rural highway. 1 lane each direction traveling about 65 MPH.

Truck in proper oncoming lane: moderate to heavy automatic breaking
Truck in proper oncoming lane at night with lights on : severe breaking
Car in proper oncoming lane: light to moderate breaking
Car in proper oncoming lane at night with headlights: moderate to severe breaking
Open road Skid marks on road: Moderate breaking
Open road with standard sage and other shrubs: occasional random moderate, moderate+ or light breaking

Is there any way to report these issues. Ver frustrating and at time scary.

These are very frustrating and really should be solved ASAP or give us complete ability to stop autobraking. The button available to turn off autobraking does not work.

Also FYI we made it from Hines, OR super Charger to Winnemucca, NV super charger on one charge. drove 55 for 180 of the 220 miles. Ended with 35 miles of range left.

 

[mlaskowsky]

I feel like I have got to the end of the line on the Phantom braking issue while using auto steer or cruise control. I bought my 22 m3lr 4 weeks ago and I have had issues with phantom braking every time I pass a truck on a 2 lane highway. I drove 300 to a service center to have Tesla work on the issue and after replacing a camera the problem continues. After Tesla looking into the issue they told me that this isn’t something that they will be able to rectify. Their suggestion for me was to step on the throttle when it happens and power through it. When I asked if I could just use normal cruise control they stated that this wouldn’t be possible. The suggest that if I continue to have these issues, that I should discontinue using any autopilot features or any auto steering. When I asked for them to refund me the $12k that I just paid for fsd they said there wasn’t anything that they can do for me. I understand that this software is in beta, but not being able to use cruse control is ridiculous. I have opened a case with the ntsb but I don’t know what else to do. This was my dream car and I want to still be excited about it but Tesla is starting to crush my faith in doing the right thing.

 

[mlaskowsky]

Trying to post video another video. I put up some videos that showed my experience. I can’t find anyone at Tesla who thinks they can fix it

 

[WhiteWi]

Trying to post video another video. I put up some videos that showed my experience. I can’t find anyone at Tesla who thinks they can fix it

Well, daaa No service center is able to do anything about it bc that is not hardware problem. It is a code problem so only whoever is responsible for AP/FSD development can do something in that regard. I would practice more and see where system works better and where it is unreliable. For example: I almost never use EAP or TACC for local roads drives, I use it 90% of the time EAP when traveling on interstates more than 100 miles. In 22k miles in our radar based M3 I had handful phantom brakes. And we had done multiple 2k miles round trips with it. In that use case I can say our EAP works great. I am not a big believer in FSD any time soon. People tend to drive aggressively so many won’t be satisfied with overcautious FSD.

 

[DayTrippin]

I bought an MY LR last year. In the last month took delivery of a Model S LR and 3 LR. The S still has the radar, the 3 is pure TelsaVision like my MY was. It was the first batch of radar-less 3/Y's and I had to sign a waiver acknowledging that the car didn't have radar before I could take delivery of the Y.

In retrospect, I absolutely should have punted and not taken delivery of the Y. It has taken almost 9 months to get the PB issues down to what is even a remotely tolerable level. I was ready to sell the Y six months in as it was so horrendous. I did recently sell but by then it had gotten better but it has been very frustrating to the new owner.

Contrast that with our pure TeslaVision 3. It has not had a single PB event in over 600 miles. In the Y, in 600 miles, I'd typically have over 10. I could never understand how other owners weren't having issues but I had so many I was willing to just use a brick on the accelerator pedal to maintain speed.

Our S with radar has been flawless. Absolutely no issues as well. It also has a benefit that the 3 doesn't. The 3's speed can vary a bit on the car it is following, sort of like it is hunting, to find the right distance/speed when there is a car in front of us. Almost like a mild pulsing on the accelerator pedal whereas the S just keeps its speed very steady. I really doubt the car in front of me is continuously pushing and and releasing the throttle all the time.

It is about a 2 mph fluctuation. It seems worse when I have a great following distance such as if it can't quite determine how far the car away really is. So it speeds up or slows down sometimes frequently and sometimes not at all.

At least seeing how the 3 can be gives me some hope but driving in a greater variety of conditions will improve my confidence (I hope). The Y had a lot of issues with semis and trees with a lot of foliage mostly when driving N or S.

 

[charlesoris]

Well, daaa No service center is able to do anything about it bc that is not hardware problem. It is a code problem so only whoever is responsible for AP/FSD development can do something in that regard. I would practice more and see where system works better and where it is unreliable. For example: I almost never use EAP or TACC for local roads drives, I use it 90% of the time EAP when traveling on interstates more than 100 miles. In 22k miles in our radar based M3 I had handful phantom brakes. And we had done multiple 2k miles round trips with it. In that use case I can say our EAP works great. I am not a big believer in FSD any time soon. People tend to drive aggressively so many won’t be satisfied with overcautious FSD.

Interesting. Is your radar on or off currently?

 

[WhiteWi]

Interesting. Is your radar on or off currently?

On of course.

 

[sleepydoc]

@sleepydoc
View attachment 783464
Autosteer is the only thing on that page labeled "beta". I believe you have shown that the manual refers to TACC as beta. It is interesting that Autosteer is called out particularly, since you would assume everything in Autopilot is beta.

Maybe there are levels of beta similar to Dante's levels of hell.

I’ve got navigate on autopilot beta (maybe that’s the 3rd level?) Nothing at al about TACC. I actually had no clue it was considered beta until someone on TMC pointed it out a few weeks ago. I’d looked at the settings screens but never read through the section in the manual and assumed it was a ‘standard’ feature like most other cars.

 

[sleepydoc]

Our 2018 Subaru Outback does basically the same thing. While driving it I have had it react to

• On coming traffic
• Construction cones
• Pedestrians
• Bushes
• Parked cars

Typically only the audible alarm goes off but it also slams the brakes.

Interesting - we haven’t had any of those issues with our 2020 Forester. Is that adaptive cruise or the emergency braking system that’s had issues? (Or can you tell?) Also, do you know if the 2018 Outback use the eyesight system or a radar system?

 

[drtimhill]

Contrast that with our pure TeslaVision 3. It has not had a single PB event in over 600 miles. In the Y, in 600 miles, I'd typically have over 10. I could never understand how other owners weren't having issues but I had so many I was willing to just use a brick on the accelerator pedal to maintain speed.

Ah now this is an interesting comparison. Have you compared the 3 to the Y on the same stretches of road, in roughly the same weather/lighting conditions? It's very hard to control all the variables (traffic, visibility etc), but you seem to indicate you are getting very different experiences from the two cars in the same circumstances. One issue with all the PB discussion here is trying to figure out if there is genuine variation in the cars themselves versus just different usage/roads etc.

 

[sleepydoc]

I’ve owned 7-8 cars with adaptive cruise from all sorts of brands. Only car that phantom braked was a 2014 CTS. Everything else was perfect. My most recent a genesis gv70 was perfect in every regard on all sorts of roads.

@WhiteWi

 

[jleeusn]

Interesting - we haven’t had any of those issues with our 2020 Forester. Is that adaptive cruise or the emergency braking system that’s had issues? (Or can you tell?) Also, do you know if the 2018 Outback use the eyesight system or a radar system?

It is the emergency braking system. I do not think that it has radar, so I assume it is the eyesight cameras.

 

[DayTrippin]

Ah now this is an interesting comparison. Have you compared the 3 to the Y on the same stretches of road, in roughly the same weather/lighting conditions? It's very hard to control all the variables (traffic, visibility etc), but you seem to indicate you are getting very different experiences from the two cars in the same circumstances. One issue with all the PB discussion here is trying to figure out if there is genuine variation in the cars themselves versus just different usage/roads etc.

I pretty much take the same route every morning for both cars. Typically the same time of day and traffic is similar. In the dense traffic TeslaVision seems to operate better but I did have the occasional issue with the Y but as mentioned nothing with the 3 or S so far (but usually in the 3 for the city more than the S).

Where I had MAJOR issues with the Y was on 2 lane roads with lots of foilage. I likely won't be on the road that was terrible for me in Alabama while driving with Y. I must have had 15-20 events in a 20 mile stretch. It was a wide 2 lane road and only using TACC, no AP. OTOH some of my worst PB events have been on 4 lane divided highway roads with the Y.

Went on a 3 hour drive today with the S and flawless again. There is definitely less variability with the S when you set a following distance than the 3/Y. It seems like I've been able to sample the extremes with pretty much the same equipment. I have seen how terrible it can be, can't imagine it being much worse and pretty pumped that so for the S and 3 have been outstanding.

 

[sleepydoc]

It is the emergency braking system. I do not think that it has radar, so I assume it is the eyesight cameras.

From what I can tell on a quick search, if it has AEB then it has eyesight (and the AEB would be vision-based.)

How often do you have false-AEB events?

 

[drtimhill]

I pretty much take the same route every morning for both cars. Typically the same time of day and traffic is similar. In the dense traffic TeslaVision seems to operate better but I did have the occasional issue with the Y but as mentioned nothing with the 3 or S so far (but usually in the 3 for the city more than the S).

Where I had MAJOR issues with the Y was on 2 lane roads with lots of foilage. I likely won't be on the road that was terrible for me in Alabama while driving with Y. I must have had 15-20 events in a 20 mile stretch. It was a wide 2 lane road and only using TACC, no AP. OTOH some of my worst PB events have been on 4 lane divided highway roads with the Y.

Went on a 3 hour drive today with the S and flawless again. There is definitely less variability with the S when you set a following distance than the 3/Y. It seems like I've been able to sample the extremes with pretty much the same equipment. I have seen how terrible it can be, can't imagine it being much worse and pretty pumped that so for the S and 3 have been outstanding.

Very interesting .. thanks for the extra detail. The variance between the 3 and the Y is strange, so say the least, particularly as they are both TeslaVision. Basically they are the same hardware and software stack, so the difference is hard to pin down. Certainly, the cameras are slightly differently placed as a result of the Y having a higher ride, but I would assume the system adjusts for that (either during calibration or with static setup that is car model aware).

Are they both the 3 and Y running the same software version?

 

[WhiteWi]

Very interesting .. thanks for the extra detail. The variance between the 3 and the Y is strange, so say the least, particularly as they are both TeslaVision. Basically they are the same hardware and software stack, so the difference is hard to pin down. Certainly, the cameras are slightly differently placed as a result of the Y having a higher ride, but I would assume the system adjusts for that (either during calibration or with static setup that is car model aware).

Are they both the 3 and Y running the same software version?

He sold his Y. So basically he doesn’t know if Y is performing same as his current M3.

 

[DayTrippin]

I can't compare now. They current owner still has issues and he is on the latest stack as am I. We both drive in Dallas. Maybe not the same exact routes but in the same area. Had rainy weather this morning and still the 3 is working well. I really hope it continues the way it has been. EVery day I trust it a bit more.

 

[jleeusn]

From what I can tell on a quick search, if it has AEB then it has eyesight (and the AEB would be vision-based.)

How often do you have false-AEB events?

I would say that we get audible alarms a few times each month and braking 3 or 4 times in 4 years.
It has occurred on two-lane roads and City streets but never on divided highways.

 

[Fourdoor]

@sleepydoc
View attachment 783464
Autosteer is the only thing on that page labeled "beta". I believe you have shown that the manual refers to TACC as beta. It is interesting that Autosteer is called out particularly, since you would assume everything in Autopilot is beta.

Maybe there are levels of beta similar to Dante's levels of hell.

I believe that this is because you don't have the option of turning TACC on or off in the menu... if you don't have the option of turning it on or off there is no reason to warn you that it is BETA... your warning that it is BETA is in the owners manual.

Keith

 

[Fourdoor]

Well, daaa No service center is able to do anything about it bc that is not hardware problem. It is a code problem so only whoever is responsible for AP/FSD development can do something in that regard. I would practice more and see where system works better and where it is unreliable. For example: I almost never use EAP or TACC for local roads drives, I use it 90% of the time EAP when traveling on interstates more than 100 miles. In 22k miles in our radar based M3 I had handful phantom brakes. And we had done multiple 2k miles round trips with it. In that use case I can say our EAP works great. I am not a big believer in FSD any time soon. People tend to drive aggressively so many won’t be satisfied with overcautious FSD.

Hold up.... you talk all the time about how you don't have PB... but you don't use EAP or TACC where PB is a common problem... have you been so conditioned by Tesla to not use the autonomous features where PB is common that you now honestly believe that PB isn't a problem?

Patient: "if I walk too much my knee gets sore"

Doctor: "use this wheelchair, and your knee will be cured!"

(several months later)

Patient: "when I walk up stairs my knee gets sore"

Doctor: "wheelchairs don't work on stairs, I am not responsible for knee pain if you don't use the wheelchair as instructed, use the elevator or don't go to upper floors"

Keith