No brake llights above 136 kph (85 mph) while using regen braking

[AWDtsla]

Just tested deceleration in my ICE, got around .25G's using engine braking from 60mph in 2nd. It's a 3 axis accelerometer so technical I have to do some number crunching to remove gravity, but the the number won't change by enough to matter.

oh and that was after rev matching, without rev matching initial deceleration is much stronger.

 

[slcuervo]

My autopilot car contradicts your theory.

I posted earlier, that with reduced regen, I do not see brake lights at correspondingly lower speeds. My car is an autopilot P85D and does not show the brake lights until the speed is slow enough to produce regen force that is comparable to light braking.

Well, not really. To contradict my theory you would need full regen (60 kW) like in my videos, and high speed (above 85 mph). The way brake lights and regen work at lower speeds is perfectly fine in all Model S I have seen and tried. The problem arises above 85 mph only.

Now, I am not asking or encouraging anyone to get out there and drive at illegal speeds to test this. As I stated on the video, I am aware that the circumstances under which I claim there is a bug or some sort of error can only be tested out in German motorways or in test tracks. So I do not pretend you guys get out there, drive and shoot videos at high speed, and submit inputs. Far from this.

My initial idea was to try to get attention from some Model S owners in Germany, or at least in Europe (living near Germany), and hoping they would want to try it too. And this is how I've learned that autopilot cars do have brake lights above 85 mph under full regen! I am waiting for a video that shows this. When I get it, I´ll confirm this theory here.

- - - Updated - - -

Just tested deceleration in my ICE, got around .25G's using engine braking from 60mph in 2nd. It's a 3 axis accelerometer so technical I have to do some number crunching to remove gravity, but the the number won't change by enough to matter.

oh and that was after rev matching, without rev matching initial deceleration is much stronger.

Thanks for testing and measuring deceleration with your ICE. However, I can´t really use your inputs, since I am not after deceleration figures below 60 mph, but above 85 mph.

Like I said in a previous post, I do not try to encourage anyone in the US to drive at illegal speeds measuring deceleration. I am more after European drivers who can test this in Germany.

The point is to compare deceleration from ICE cars above 85 mph to deceleration from Model S under full regen (60 kW). This will take some time, but I hope that eventually some Model S owners from Germany and/or ICE car owners driving in Germany will share inputs...

 

[rdrcrmatt]

You need to chart rate of deceleration to find at what rate of deceleration the brake lights come on. THEN you can start investigating a bug. It might not have anything to do with over X MPH.

I've noticed the brake lights do not come on till I'm using more than 30kw of regen and I'm down to mid 30's mph. by your example the brake lights should come on at a certain regen KW, when it is actually the deceleration rate.

 

[LetsGoFast]

My brake lights come on in that scenario. I have auto-pilot sensors. It seems to me that either my car shouldn't show brake lights or the non-AP should show brake lights. One behavior or the other is clearly erroneous. Personally, given the rate of slowdown, I think displaying brake lights is the normal and expected behavior.

 

[slcuervo]

My brake lights come on in that scenario. I have auto-pilot sensors. It seems to me that either my car shouldn't show brake lights or the non-AP should show brake lights. One behavior or the other is clearly erroneous. Personally, given the rate of slowdown, I think displaying brake lights is the normal and expected behavior.

I totally agree.

I will probably get a few videos from ICE cars and from various Model S over the next few days, decelerating like I described, and we'll be able to see the differences very clearly.

 

[slcuervo]

I have built a table with deceleration times and rates from 170 down to 70 kph (105 mph to 44 mph).

Here it is:

Initial Speed (kph)	Final Speed (kph)	Elapsed time (sec)	Deceleration (m/s)
170	160	2,527	-1,1
160	150	2,551	-1,1
150	140	2,304	-1,2
140	130	2,314	-1,2
130	120	2,862	-1
120	110	2,409	-1,15
110	100	2,001	-1,39
100	90	2,229	-1,25
80	70	2,174	-1,28

I have measured time by recording the speedometer and then noting the start and end time on the video player. Not the most accurate way, I know.

The most interesting part of all these inputs is that deceleration from 150 to 140 kph is almost identical to the deceleration from 100 to 90 kph, i.e. 1,2 m/s vs 1,25 m/s.

And there are no brake lights at 150-140, but they are there at 100-90. And this, because my Model S has no autopilot. If it had autopilot, brake lights would be on while decelerating with full regen from 150 to 140 kph.

Do you guys still think this is not a bug or an error of some type?

I am still gathering inputs from other cars and will share it later when I have everything ready.

 

[woof]

I have built a table with deceleration times and rates from 170 down to 70 kph (105 mph to 44 mph).

Is this from a dual motor Model S? It possible the dual's have a different deceleration rate as they have more traction to work with.

 

[slcuervo]

Is this from a dual motor Model S? It possible the dual's have a different deceleration rate as they have more traction to work with.

No, it's my S85, classic (non AP). So that could be an explanation indeed.

I will come back later with AP deceleration figures when they send me the videos.

In any case, the table above shows almost identical deceleration rates at different speeds 150 kph and 100 kph) but different behavior from the brake lights. Makes no sense. And again: it can be a safety issue on the German autobahn.

 

[brianman]

My brake lights come on in that scenario. I have auto-pilot sensors. It seems to me that either my car shouldn't show brake lights or the non-AP should show brake lights. One behavior or the other is clearly erroneous. Personally, given the rate of slowdown, I think displaying brake lights is the normal and expected behavior.

I can think of several other reasons why they might not have matched behavior. Erroneous is an opinion, not a foregone conclusion -- IMO.

 

[slcuervo]

I can think of several other reasons why they might not have matched behavior. Erroneous is an opinion, not a foregone conclusion -- IMO.

OK, my statements are certainly not proven facts. All of this issue is driven by the fact that I perceive the absence of brake lights at high speeds as a safety issue. But you don´t all have the same perception, of course.

What are these other reasons you think could explain the different behavior? I´m curious to see what I might be missing...

 

[woof]

Lots of variables to control for.

Here are my top 3:

Tires: 21 vs 19
Motors: 1 vs 2
Sensors: AP vs Classic

Farther down the list:
Tires: Snows vs Summers vs All Seasons
Tires: Wear
Software: revision #
Chargers: 1 vs 2 (does regen even go through the chargers?)
Mass: weight of passengers and cargo

Also, you measured Delta V, which isn't acceleration, so it's hard to really know what an accelerometer would sense. (Assuming the brake lights are accelerometer based). There are some phone based apps that measure acceleration using the built in accelerometers. Not super accurate but might get a good feel using the same phone/app in different cars (make sure it's in the same position, like flat on the floor).

 

[slcuervo]

Too many variables for me. Maybe the right way for Tesla Motors to determine whether brake lights should go on or not, but that´s too much information for me.

I am the user, and as an user I am worried about safety.

All I know is the deceleration I have measured with my Model S is:

1. significantly greater than the deceleration from ICE cars (regardless of the car and even when downshifting agressively)
2. almost the same from 150 to 140 kph and from 100 to 90 kph (a difference of milliseconds only)

So I don´t understand why my car doesn´t use brake lights when decelerating at high speeds, especially if you take points 1 and 2 into consideration. And even more if you know autopilot equipped Model S do use brake lights above 85 mph.

It makes no sense, and when driving at high speeds in Germany I want to signal my strong deceleration at ANY speed. It is a safety requirement.

Details like you mention will probably only make a very small difference, measurable in milliseconds. I don´t think they change anything to my basic claim: that brake lights should also go on above 85 mph, for safety reasons.

 

[woof]

It makes no sense, and when driving at high speeds in Germany I want to signal my strong deceleration at ANY speed. It is a safety requirement.

I'm not disagreeing, I'm just trying to figure out what is the definition of "strong deceleration". I'm assuming the car uses accelerometers (or other means to calculate acceleration), and there is a fixed threshold value above which it turns on the brake signal, and below which it doesn't, regardless of initial speed. Either the car is slowing fast enough to cause the accelerometer to read above the threshold, or it isn't. Speed isn't the measurement here, rate of change of speed is. It's quite possible that above 85 MPH, your car doesn't decelerate fast enough to trip the threshold.

Details like you mention will probably only make a very small difference, measurable in milliseconds. I don´t think they change anything to my basic claim: that brake lights should also go on above 85 mph, for safety reasons.

But it's quite possible that those small differences move the acceleration of the car above or below the threshold. For example, on your car, there is only a rear motor, so the regen is only from the rear wheels. The amount of regen may be software limited to allow smaller values at higher speeds for stability reasons, as the front tires aren't braking. When measured as acceleration, it's possible this doesn't trip the brake light signal threshold. On dual motor cars, with both front and rear tires involved in regen, the software may allow more deceleration, which does trip the signal threshold.

 

[Andyw2100]

I just found this thread. I became more shocked at what I was reading as I progressed through it.

First of all, I don't understand all the posters saying the deceleration speed is no different than a manual ICE, etc., etc., or that drivers behind the slowing vehicle need to be careful/responsible, etc. All of that is irrelevant! Tesla has the autopilot cars exhibiting one behavior and non-autopilot cars exhibiting a different behavior. The OP didn't start this thread to begin a debate on which behavior is better. He was trying to help the community by letting us know about this issue, which initially he thought existed in all cars, but which he then, with the help of jpet, and other European owners, discovered was an inconsistency between AP and non-AP models. It's really as simple as that! It sounds like jpet checked with enough other owners to determine that this is a black and white split by AP / Non-AP. Wheel size, dual motor, snows vs. all-season--none of that matters at all. Just as the speed that various ICE vehicles decelerate is also irrelevant. The only thing relevant is that non-AP cars are exhibiting a behavior different from AP cars, and there does not seem to be an obvious reason for it, and it is, at least potentially, dangerous.

Perhaps Tesla does have a reason for having the AP cars exhibit one behavior and the non-AP cars exhibit another behavior. If so, I hope after the OP writes to Tesla, they will explain what that reason is. If, on the other hand, Tesla does not have a sound reason for the difference, I hope they will resolve it in a coming firmware update.

I just think the criticism of the OP, his methodology, his reasons for wanting to decelerate from speed without using the brake pedal, etc., when he came here trying to help were completely and totally misplaced.

Thanks, slcuervo, for trying to help the community!

 

[slcuervo]

Perhaps Tesla does have a reason for having the AP cars exhibit one behavior and the non-AP cars exhibit another behavior. If so, I hope after the OP writes to Tesla, they will explain what that reason is. If, on the other hand, Tesla does not have a sound reason for the difference, I hope they will resolve it in a coming firmware update.

Thanks, slcuervo, for trying to help the community!

Thanks for your comments Andyw2100!

I did contact Tesla about a week ago, but that was done online by leaving comments using the owner feedback feature of the website. That seems to do nothing, since I´ve had no reply. I know they read those comments because I have used that tool in the past for other matters, but they have not answered this one.

Next week I have an appointment at my local Service Center (which was scheduled long ago for a different matter), so I will let them know. But I doubt they can do anything about it, since top speed in Norway is 110 kph - I don´t think they´ll want to go out there with the car and check themselves.

I tried contacting owners clubs from Germany through twitter, but alas, no response.

Anyway, IMO this is a safety issue and I will keep trying to get an answer from Tesla about it.

 

[AWDtsla]

I just found this thread. I became more shocked at what I was reading as I progressed through it.

First of all, I don't understand all the posters saying the deceleration speed is no different than a manual ICE, etc., etc., or that drivers behind the slowing vehicle need to be careful/responsible, etc. All of that is irrelevant! Tesla has the autopilot cars exhibiting one behavior and non-autopilot cars exhibiting a different behavior. The OP didn't start this thread to begin a debate on which behavior is better. He was trying to help the community by letting us know about this issue, which initially he thought existed in all cars, but which he then, with the help of jpet, and other European owners, discovered was an inconsistency between AP and non-AP models. It's really as simple as that! It sounds like jpet checked with enough other owners to determine that this is a black and white split by AP / Non-AP. Wheel size, dual motor, snows vs. all-season--none of that matters at all. Just as the speed that various ICE vehicles decelerate is also irrelevant. The only thing relevant is that non-AP cars are exhibiting a behavior different from AP cars, and there does not seem to be an obvious reason for it, and it is, at least potentially, dangerous.

Perhaps Tesla does have a reason for having the AP cars exhibit one behavior and the non-AP cars exhibit another behavior. If so, I hope after the OP writes to Tesla, they will explain what that reason is. If, on the other hand, Tesla does not have a sound reason for the difference, I hope they will resolve it in a coming firmware update.

I just think the criticism of the OP, his methodology, his reasons for wanting to decelerate from speed without using the brake pedal, etc., when he came here trying to help were completely and totally misplaced.

So two different software versions behave differently, news story of the century...

The brake lights come on with annoying frequency. 1.1 m/s^2 is a gentle coasting down. Crossing my fingers all cars will get this behavior.

- - - Updated - - -

Brake light thresholds are probably my least favorite thing about the car.

 

[sillydriver]

All of that is irrelevant! Tesla has the autopilot cars exhibiting one behavior and non-autopilot cars exhibiting a different behavior. The only thing relevant is that non-AP cars are exhibiting a behavior different from AP cars, and there does not seem to be an obvious reason for it, and it is, at least potentially, dangerous.

This is a good point unless AP and non-AP cars are decelerating at different rates under software control. They are probably the same, but I note that the deceleration rate under regen on my own AP car has been reduced by one of the recent updates.

I'm actually very conflicted about this. On the one hand I want slcuervo to have brake lights on deceleration that make him (or her) feel safe. On the other, I agree with AWDtsla that the brake light threshold is very annoying: I can't shift my foot on the accelerator pedal without producing a spurious brake light pulse. If I had a choice, I would have the brake lights on my car activate only with the friction brake (although if TACC wanted to light up while on regen that would be fine). I use the rear view mirror to see what's behind me and would drive accordingly to be safe.

 

[slcuervo]

I find it really surprising that some of you can find the use of brake lights under regen braking annoying. Hey, everybody is entitled to an opinion, but to me, when I read those statements, it sounds like saying the use of blinkers when changing lanes is annoying. Or the use of lights, unless it is pitch black, is annoying. To me, it doesn´t make sense.

You see, one of the most important things to have safe driving conditions is predictability, IMO. When we have a system that is predictable, the possibility of having accidents decreases a lot. If you drive in an unpredictable way, you will cause accidents.

So, just as much as driving at night without lights (or changing lanes all of a sudden without using blinkers) may cause accidents, decelerating more than 99% of the cars out there (which is what the Model S does) without brake lights, can cause accidents.

Whether you find it annoying or not, warning others that your speed is decreasing at a rate that they don´t expect is a good thing. It can save your life.

 

[gelden]

I find it really surprising that some of you can find the use of brake lights under regen braking annoying. Hey, everybody is entitled to an opinion, but to me, when I read those statements, it sounds like saying the use of blinkers when changing lanes is annoying. Or the use of lights, unless it is pitch black, is annoying. To me, it doesn´t make sense.

You see, one of the most important things to have safe driving conditions is predictability, IMO. When we have a system that is predictable, the possibility of having accidents decreases a lot. If you drive in an unpredictable way, you will cause accidents.

So, just as much as driving at night without lights (or changing lanes all of a sudden without using blinkers) may cause accidents, decelerating more than 99% of the cars out there (which is what the Model S does) without brake lights, can cause accidents.

Whether you find it annoying or not, warning others that your speed is decreasing at a rate that they don´t expect is a good thing. It can save your life.

No one is saying that driving without brake lights is better. When the cars decelerates a lot, the brake lights should come on. The behavior should resemble ICE cars, as that is what most drivers are used to (and expect). So if the regen braking at high speeds is not more than engine braking in ICE cars, the brake lights should not be on. If they are on, they insinuate hard braking and will result in unnecessary braking of cars behind you which might result in traffic jams.
The questions is: why the difference between auto-steer cars and non-auto-steer cars. Maybe AS cars have more regen braking at highers speeds?

 

[slcuervo]

So if the regen braking at high speeds is not more than engine braking in ICE cars, the brake lights should not be on. If they are on, they insinuate hard braking and will result in unnecessary braking of cars behind you which might result in traffic jams.

The questions is: why the difference between auto-steer cars and non-auto-steer cars. Maybe AS cars have more regen braking at highers speeds?

Which is what I've proven by measuring deceleration from several ICEs. Some diesel engines with high torque come close to the Model S, but no ICE car matches it.

I'm still working on AP Model S deceleration to try to figure out why the different behavior.