-
Want to remove ads? Register an account and login to see fewer ads, and become a Supporting Member to remove almost all ads.
Reverse made my car go forward...
- Thread starter superh0
- Start date
Kitfox
Member
Very condescending, but given the fact that there are people who are just trolls and jerks, I can understand your wariness. I also recognize that everything so far is wholly anecdotal, and that's not something I can solve yet, but I'll try going forward.
I can't speak for others, but in my situation, the vehicle was definitely in reverse.
Order of operations:
Apply brake.
Because I'm still exploring the car, push stalk up to half-up (before the secondary switch engages). No change from park, and I was very closely paying attention to this because I wanted to understand the controls better.
Push stalk fully up. Observed vehicle change to reverse. Backup camera is on.
Note that Hold is on. (This and...
Release brake pedal. (...this may be in reverse order.
Reapply brake pedal briefly to release hold. Note that hold turned off.
Release brake pedal.
Car started rolling forward.
Immediately reapply brake pedal.
Reobserve that car is in reverse.
Gingerly release brake pedal and car creeps backward.
Finish reverse from parking space and go on my way.
Two factors that could confound things would be the mild downhill the vehicle was on, and it also occurs to me that the half-up followed by full up may play a factor in this.
I personally like to have good faith in the behavior of my vehicle, and also I like to understand how systems work. I'm a tech person by trade, so I need to understand system models very well. I'm also very happy to dig with iterative testing and to potentially even make it a habit of recording in-cab on my phone every time I try to back out.
To answer a prior question that was brought up in this thread, I found this post when I was searching Google for "tesla model 3 hold in reverse", since I was trying to understand what caused it (still not sure) and what the proper response to it was. That was after the garage jerky backup from hold. That easily could've been too much pedal pressure or extra pressure caused by the backward acceleration. Since I'd like to avoid suddenly jerky jumps, I worked to look it up and found the "Press the brake again when on hold" directive.
Overall, I see enough evidence that this might be an actual issue somehow that I am happy to dig and gather data and hopefully get enough to be able to submit a bug report. If it's something as silly an edge case as the half-up, then that's notable, as experienced Tesla drivers probably won't do that, and if they ever did it likely wouldn't occur to them. Until I went back to recall specifics, I'd forgotten about that, and that's a lead at least.
If it's something that is an actual issue, it's better to get it fixed and safe before there's an actual accident instead of just people complaining about it on forums. A happyTesla driver refers more. An unhappy one turns people off. So even a silly ding that never makes the news could still end up being a detractor in the making. I'd rather have happy Tesla drivers about.
I can't speak for others, but in my situation, the vehicle was definitely in reverse.
Order of operations:
Apply brake.
Because I'm still exploring the car, push stalk up to half-up (before the secondary switch engages). No change from park, and I was very closely paying attention to this because I wanted to understand the controls better.
Push stalk fully up. Observed vehicle change to reverse. Backup camera is on.
Note that Hold is on. (This and...
Release brake pedal. (...this may be in reverse order.
Reapply brake pedal briefly to release hold. Note that hold turned off.
Release brake pedal.
Car started rolling forward.
Immediately reapply brake pedal.
Reobserve that car is in reverse.
Gingerly release brake pedal and car creeps backward.
Finish reverse from parking space and go on my way.
Two factors that could confound things would be the mild downhill the vehicle was on, and it also occurs to me that the half-up followed by full up may play a factor in this.
I personally like to have good faith in the behavior of my vehicle, and also I like to understand how systems work. I'm a tech person by trade, so I need to understand system models very well. I'm also very happy to dig with iterative testing and to potentially even make it a habit of recording in-cab on my phone every time I try to back out.
To answer a prior question that was brought up in this thread, I found this post when I was searching Google for "tesla model 3 hold in reverse", since I was trying to understand what caused it (still not sure) and what the proper response to it was. That was after the garage jerky backup from hold. That easily could've been too much pedal pressure or extra pressure caused by the backward acceleration. Since I'd like to avoid suddenly jerky jumps, I worked to look it up and found the "Press the brake again when on hold" directive.
Overall, I see enough evidence that this might be an actual issue somehow that I am happy to dig and gather data and hopefully get enough to be able to submit a bug report. If it's something as silly an edge case as the half-up, then that's notable, as experienced Tesla drivers probably won't do that, and if they ever did it likely wouldn't occur to them. Until I went back to recall specifics, I'd forgotten about that, and that's a lead at least.
If it's something that is an actual issue, it's better to get it fixed and safe before there's an actual accident instead of just people complaining about it on forums. A happyTesla driver refers more. An unhappy one turns people off. So even a silly ding that never makes the news could still end up being a detractor in the making. I'd rather have happy Tesla drivers about.
While I can see the unwanted behaviour that has been noted, I don't think this is any different or worse than any other car. My wife's Highlander acted different than my old Taurus, My Taurus wouldn't roll backward from a stop when releasing the brake on the same incline that my wife's Highlander would roll backwards and it took a little more effort to manage the brake/accelerator pedals in the Highlander in that situation.
The Tesla does what it does and while yes you could call it an "issue" it is the same "issue" that a lot of other cars have in varying scenarios. So there isn't anything to fix(because of a safety standpoint)...unless you want to fix it for every car that exists. I do agree with learning how the car works for you though, but again same goes for every car.
And before anyone goes and jumps on me, this all tracks with the OPS original post.
To the OP's specific question...just learn the car and the required pedal pressures for your normal driving scenarios...Big thing is to make sure the car is on hold after you put it in reverse and then you just have to apply the correct pressure on the accelerator to overcome any incline/decline.
The Tesla does what it does and while yes you could call it an "issue" it is the same "issue" that a lot of other cars have in varying scenarios. So there isn't anything to fix(because of a safety standpoint)...unless you want to fix it for every car that exists. I do agree with learning how the car works for you though, but again same goes for every car.
And before anyone goes and jumps on me, this all tracks with the OPS original post.
To the OP's specific question...just learn the car and the required pedal pressures for your normal driving scenarios...Big thing is to make sure the car is on hold after you put it in reverse and then you just have to apply the correct pressure on the accelerator to overcome any incline/decline.
Kitfox
Member
The thing that took into consideration was that I did what was apparently the same thing twice (taking my foot off the brake) and a different result occurred each time.
But the main thing that I consider different about a Tesla than any other car is quite simple:
If there is a drive by wire weirdness on another car, it won't get any attention until it's caused problems (accidents, damage, angry people), generated a recall, and bad brand perception. Then the recall involves the car in the shop and replacing control modules and all kinds of annoyances.
If there is a drive by wire weirdness on a Tesla, and they are able to find out about it while it's still just a "Huh, weird" situation, they can correct it OTA. No shop visit, no recall, no bad stuff happening, and look good as a result.
I figure, complain about things you can't fix. If you can fix it, fix it. If I can get any information on this, then that's something I can help fix. That's a benefit to me as a driver, and to everybody else as a driver, and to Tesla as a brand, and to safety overall. I've personally only seen unexpected behavior once, so I can't and won't say I've personally seen a pattern. But with other reports, I'm happy to look into what I can find out. If I do find something, my attitude is "Great! Now it can be fixed!" and that means improving things. I'm not a bitter "ha ha, look at why this sucks" person. Nothing is perfect. Everything can be better. So I do what I can to contribute to that betterness, rather than contributing to bitterness.
I think it's a glitch in creep functionality. I've noticed a few times that sometimes creep doesn't activate right away after you switch from Park to Drive or Reverse. If you are on an incline, it may even start rolling in that direction just like if creep is not on. After you apply and release the brake again (or switch gears) it would start creeping normally. It doesn't happen all the time, though, just occasionally.
darth_vad3r
Well-Known Sith
Has anyone actually had “reverse” make their car go forward and not “gravity made my car go forward while I was in reverse”?
Kitfox
Member
If it's a timing thing (not activating right away), that gives me more data to do testing with. One of the downsides to playing close attention to things for testing is that it changes user behavior and makes it more difficult to replicate things.
I can't say it was not gravity, as there was a downhill involved (and trust me, I know it only takes a tiny incline to roll something like a car), though the fact that creep overcame the power of gravity on the second brake release still makes it an inconsistent behavior. If gravity pulls it forward downhill on the first brake release, gravity should do the same on the second one as well, and the third, and any others, sans accelerator application.
I can't say it was not gravity, as there was a downhill involved (and trust me, I know it only takes a tiny incline to roll something like a car), though the fact that creep overcame the power of gravity on the second brake release still makes it an inconsistent behavior. If gravity pulls it forward downhill on the first brake release, gravity should do the same on the second one as well, and the third, and any others, sans accelerator application.
darth_vad3r
Well-Known Sith
Just turn off creep — problem solved. Your foot will be on the accelerator to break hold and you can finesse it enough to overcome gravity. (Yes I know, not solved, “avoided”, but creep is an abomination that is the second thing a time traveler should kill if they go back in time).
Brake pedals should be for stopping only, not going. Creep is “degenerative braking”.
Accelerator pedal for one pedal driving is fine. Regenerative braking is where it’s at.
One-pedal driving with regen on accelerator pedal fails safe (e.g. it stops / slows your car if you have a heart attack or pass out).
One-pedal driving with degenerative braking (creep) fails UNSAFE, pushing you into the intersection if you have a heart attack or pass out.
Last edited:
Kitfox
Member
One of the things we need to take into consideration though is whether Hold -> Accelerator will possibly end up going the wrong way like creep "did". Bit of a difference between having your foot over the brake and having it on the accelerator for response time.
darth_vad3r
Well-Known Sith
If it goes the “wrong way” because of gravity, you just press the accelerator a tiny bit more to go “the right way”.
If the brake isn't on (Pedal, Hold or Parking) AND you aren't pressing the accelerator, it will roll whichever way gravity tells it.
In Gear, once you press the accelerator, it will cancel Hold, and then as long as you are accelerating faster than gravity, it will go in the direction of the gear (forward or reverse). (Tip, cancel Hold by pressing accelerator, not the brake)
If you aren't accelerating faster than gravity (Right root, or creep) then Gravity Wins and the car will go whichever way gravity tells it.
When you want to stay stopped, put foot on brake, press until Hold engages. The car will now not go anywhere. Now put it in Park if you want.
No gears to worry about, No clutch. It's very. very simple.
In Gear, once you press the accelerator, it will cancel Hold, and then as long as you are accelerating faster than gravity, it will go in the direction of the gear (forward or reverse). (Tip, cancel Hold by pressing accelerator, not the brake)
If you aren't accelerating faster than gravity (Right root, or creep) then Gravity Wins and the car will go whichever way gravity tells it.
When you want to stay stopped, put foot on brake, press until Hold engages. The car will now not go anywhere. Now put it in Park if you want.
No gears to worry about, No clutch. It's very. very simple.
Another possibility as to why the car rolled forward the first time, but not the second time is that you abruptly let off the brake pedal on the first time, as opposed to gingerly releasing the brake pedal the second time.
The car in creep only provides so much torque. When you let go of the brake abruptly, gravity overwhelms the amount of torque provided from creep. Whereas when you ease off the brake, the torque provided by creep is able to catch the car from moving rolling forward. It's similar to a manual car: abruptly let go of the clutch, and the engine stalls. Gently let go of the clutch, and the vehicle starts to move.
Why are you pressing the brake after putting it in Hold? Just push the accelerator to release the hold. The car will go immediately from a hold position to an accelerated action in whichever direction you want to go. There is no need for hitting brake again to release a hold.
And by the way, that you experienced this “roll forward in reverse” immediately after releasing the hold without engaging the accelerator is hard evidence that your car merely rolled due to gravity, not because of an anomalous bug in the software.
And by the way, that you experienced this “roll forward in reverse” immediately after releasing the hold without engaging the accelerator is hard evidence that your car merely rolled due to gravity, not because of an anomalous bug in the software.
Kitfox
Member
Revisiting this now with more details...
Perhaps driver's training has changed, but back in my day it was "Keep your foot over the brake while backing out and let the creep do the work, because moving your foot from the accelerator to the brake takes too long if something like an idiot in a car or on foot or a child or pet gets in your way."
The first time releasing hold, it went forward. The second time releasing hold, it crept backward.
Given that circumstances were relatively the same each time, it should have reacted identically. Either roll forward from gravity both times, or creep backwards from creep both times.
Moving onward...
I did too much testing for my wife's happiness with how much time I spent doing it. Same parking spot as before, same settings (Creep On). I did variations of "Half-up/Full up" on the gear stem and did multiple sets of each test, as well as working with timing of holding the gear stem. Also tested on less decline.
Reproduction steps - Inconsistent reproduction:
On a decline (Facing downhill)
Press brake enough to engage hold.
Move gearstem partly up and hold for just over one second. Do not release gearstem down.
Move gearstem fully up. - R is displayed on screen.
Release brake - Hold is on.
Press and release brake to release hold
Expected results:
Car creeps backward.
Actual results:
Car moves forward even though gear indicator states "R"
Re-engage Hold and re-release Hold.
Car creeps backward as expected.
Reproduced seven out of fifty attempts each test set (3.7 degree decline and 1.1 degree decline). The speed of releasing the brake may theoretically have something to do with it. It's very difficult to get everything precisely identical
Was not able to reproduce through 50 attempts today on 2019.32 software.
Perhaps driver's training has changed, but back in my day it was "Keep your foot over the brake while backing out and let the creep do the work, because moving your foot from the accelerator to the brake takes too long if something like an idiot in a car or on foot or a child or pet gets in your way."
The first time releasing hold, it went forward. The second time releasing hold, it crept backward.
Given that circumstances were relatively the same each time, it should have reacted identically. Either roll forward from gravity both times, or creep backwards from creep both times.
Moving onward...
I did too much testing for my wife's happiness with how much time I spent doing it. Same parking spot as before, same settings (Creep On). I did variations of "Half-up/Full up" on the gear stem and did multiple sets of each test, as well as working with timing of holding the gear stem. Also tested on less decline.
Reproduction steps - Inconsistent reproduction:
On a decline (Facing downhill)
Press brake enough to engage hold.
Move gearstem partly up and hold for just over one second. Do not release gearstem down.
Move gearstem fully up. - R is displayed on screen.
Release brake - Hold is on.
Press and release brake to release hold
Expected results:
Car creeps backward.
Actual results:
Car moves forward even though gear indicator states "R"
Re-engage Hold and re-release Hold.
Car creeps backward as expected.
Reproduced seven out of fifty attempts each test set (3.7 degree decline and 1.1 degree decline). The speed of releasing the brake may theoretically have something to do with it. It's very difficult to get everything precisely identical
Was not able to reproduce through 50 attempts today on 2019.32 software.
There is nothing clever/sinister/bug like going on here. IT'S JUST GRAVITY.
If acceleration due to gravity, is greater than acceleration due to engine, then gravity wins. If acceleration due to engine is greater than gravity, then engine wins. If they exactly balance, the car doesn't move.
If you are facing downhill, then if you want to go backwards, you have to accelerate faster than gravity.
If 'Creep' is turned on, then creep applies 'a little bit' of acceleration. That little bit, may not be more than gravity, so you will go in the direction of gravity. Works the same as that in ANY car I have driven.
Try a hill-start in a manual/stick car. Throttle, Clutch, Foot-brake, Hand-brake all needs to be coordinated to go in the right direction.
Well, firstly, your handy dandy driver training from the colonial era was probably not referencing an electric car with a gas pedal that also acts as a brake.
Second, I’m not sure how geriatric you are, but I’m 37, and I can stop the car on a dime when moving foot from gas to brake. Don’t know which paraplegic wrote the rule book you read, or during which world war it was written, but honestly, unless you’re either A) 90 years old, B) have a broken leg, or C) are backing up at 60 MPH, then it should not be a problem to move your foot from the gas to the brake when needed.
Also, you only need to tap the accelerator to release the hold, and then you can either use momentum to carry you, or you can continue pressing accelerator slightly.
Anyway, I mean w/e. You do you. I’m just not quite sure what the problem is here.
DopeGhoti
Active Member
The problem seems to be, as someone else was posting, this behavior which was at least somewhat repeatable:
Car configuration: Creep set to on
- Apply foot pressure to the Brake pedal
- Set car into Reverse. Observe the camera come on, and the large friendly 'R' on the MFD.
- Release pressure from Brake pedal.
- Observe car rolling forward.
- Reapply pressure to the Brake pedal
- Release pressure from Brake pedal.
- Observe car creeping backward
darth_vad3r
Well-Known Sith
To avoid this problem, time travel and prevent creep from being invented.
Failing that, turn creep off.
Kitfox
Member
This thread seems to have some people up in arms. Like I said before, if there is a problem, it's better to find and fix than have it become a bigger one. Not everybody is a troll or a troublemaker when they find a problem. Take problems found as a way to improve and make things better.
Got in touch with a friend who has a model 3 and caught him before accepting the .32 upgrade.
Continued testing and changed it up a bit:
Creep is on.
In park, apply brake. Car is on light INCLINE (gravity pulls backward)
Put car into drive.
Push gearstem up half for close to long enough to go into neutral but not quite. Finish pressing fully up and car goes into reverse.
Release brake.
Car moves forward against gravity until brake is reapplied and re-released, then car moves backwards with creep and with gravity.
Reproduced 6 times out of 80 attempts, so there has to be a factor that is hit and miss.
Turn creep off.
Repeat steps above but use accelerator to remove hold.
Again car moves forward instead of backward. Release accelerator and car drifts backward. Reapply accelerator and car moves backward under power.
Reproduced three times out of 33 tries (The friend got sick of doing it after the third time it happened.
Upgraded car to .32.
No longer occurs in fifty more tests. I'd say that some of the "various bugs" fixed in .32 might include this, which honestly makes it a moot point, which is good. I'm happy to call it fixed unless somebody can provide a new set of directions to reproduce it at least somewhat reliably.
IMO:
"Don't use a feature of your car" doesn't feel like an appropriate answer. Either the feature should work consistently as expected, or it should not exist. If it does not work consistently as expected, that should be corrected or the feature removed.
We have a tremendous number of smart people in this forum, and awesome cars that just keep getting better. Better comes from improvements, not ostriching. Like somebody said, "Before you tell me to not worry about it and just enjoy the car...", though part of my enjoyment comes from helping find issues that I know can be fixed so that they do get fixed, and so that people are aware of them until they are.
Got in touch with a friend who has a model 3 and caught him before accepting the .32 upgrade.
Continued testing and changed it up a bit:
Creep is on.
In park, apply brake. Car is on light INCLINE (gravity pulls backward)
Put car into drive.
Push gearstem up half for close to long enough to go into neutral but not quite. Finish pressing fully up and car goes into reverse.
Release brake.
Car moves forward against gravity until brake is reapplied and re-released, then car moves backwards with creep and with gravity.
Reproduced 6 times out of 80 attempts, so there has to be a factor that is hit and miss.
Turn creep off.
Repeat steps above but use accelerator to remove hold.
Again car moves forward instead of backward. Release accelerator and car drifts backward. Reapply accelerator and car moves backward under power.
Reproduced three times out of 33 tries (The friend got sick of doing it after the third time it happened.
Upgraded car to .32.
No longer occurs in fifty more tests. I'd say that some of the "various bugs" fixed in .32 might include this, which honestly makes it a moot point, which is good. I'm happy to call it fixed unless somebody can provide a new set of directions to reproduce it at least somewhat reliably.
IMO:
"Don't use a feature of your car" doesn't feel like an appropriate answer. Either the feature should work consistently as expected, or it should not exist. If it does not work consistently as expected, that should be corrected or the feature removed.
We have a tremendous number of smart people in this forum, and awesome cars that just keep getting better. Better comes from improvements, not ostriching. Like somebody said, "Before you tell me to not worry about it and just enjoy the car...", though part of my enjoyment comes from helping find issues that I know can be fixed so that they do get fixed, and so that people are aware of them until they are.
Similar threads
