Does ABS work when sliding backwards?

The focus of ABS is always wheel speed differential (to some reference) and as control systems have advanced other sensed parameters such as wheel deceleration allow for more robust and effective operation. Think of wheel deceleration as being an earlier detectable predictor of an impending wheel speed differential than using wheel speeds alone. But even though that can trigger the system into action, it still goes to work on matching wheel speeds to a reference.

 

I don’t dispute that modern ABS systems take into account differences in rotational speed and rate of change of rotational speed of all 4 wheels, but I’m still skeptical that it is root cause of the system not activating when sliding backwards down a icy hill as I observed yesterday.

I think that it is a refinement of the original tech. I think it’s perfectly possible to design a ABS without a sensor on each wheel and I think the original systems were far less complicated than modern ones.

I’m still sticking with my theory that the root cause of the lack of ABS activation is the lack of change in rotational speed of the wheels prior to the event. It started at zero rpm at the maximum climb height and never changed much on the slide downwards, so no activation.

 

I think we are in complete agreement.

 

A few thoughts about this with the caveat I don't know exactly how Toyota both addressed and programmed the control systems for all these particular scenarios.

Modern ecu's are more capable that just understanding wheel speeds alone to determine corrective actions for traction control, VSC, and ABS. It is possible to lock up all four wheels, yet still have ABS activate and this is because modern ECU's can use inertial measurements along with wheel speeds to understand what is happening.

The early and basic traction and ABS systems used only wheel speeds to determine the vehicle state and action, but that's not always true today. Today's systems still target some reference, but the reference isn't always a measured wheel speed.

For example, if you floor it in 4wd and all wheels begin to spin equally, it's possible for traction control to know and cut engine output. This is because the ECU knows that if you were at a stop and all wheels accelerate at some rate with traction then there should be some matching vehicle acceleration. If there is no vehicle acceleration matching the acceleration of all four wheels then it knows the wheels are spinning without traction.

Same thing with ABS. If you brake from 30 mph and lockup all four wheels instantly, the ECU knows that the vehicle MUST be skidding because the deceleration the vehicle has experienced and the ECU has measured would not have brought it from 30 mph to 0 mph.

Again, I don't know specifically what Toyota has put in our trucks but there is ability of modern units to work beyond wheel speeds alone.

 

I’m following you, but I think you might be discounting wheel speed rate of change, the wheel speed acceleration (2nd derivative)

 

A fun little test.

Get on a muddy road, slam on the brakes, and turn off on a side road.

Repeat this with your rear locker engaged (turns off ABS) and see how far you slide past the turn-off. The front wheels lock, which means no steering.

Our 3rd Generation has Yaw and Acceloremeter data to know when the vehicle is in motion.

 

I'm not specifically talking about the issue at hand yet. I actually don't know what's going on here. In order for ABS to work in reverse, first Toyota would have had to program it in. After that, in 4wd there are interesting problems with ABS operation because all the wheels have speed relationships to one another through the drivetrain. Basically, 4wd ABS has to be inertial based because you can't lock up one wheel without locking up the other 3 as well so relative wheel speed comparisons become somewhat useless. So, in these case I feel like speeds are too slow (starting from or near a stop) for inertial measurements to be accurate enough to allow this system to know what's going on and operate even if Toyota programmed it in. Also, ABS would be very ineffective at stopping a vehicle that began sliding from a stop anyway because, well, it can't be stopped.

 

My thought process is that the trigger for ABS activation is the rate of change in rotational speed of the wheel. For example, if the wheel speed goes from 100 rpm to 10 rpm in 10ms, the computer decides this is an emergency braking situation where the tire is about to slide, and it intervenes. In the sliding backwards case the wheel speed starts at 0 rpm and stays at 0 rpm, so ABS doesn’t activate.

 

Good theory. I do know from experience that ABS and VSC can easily get mixed up and you are back to just braking and steering. Do a flat spin in the snow on a parking lot. All nannies go out the window and you are on your own lol.

 

Yes, but it's unlikely that's the only trigger for ABS to activate. It is a good one and in cases an earlier indicator, but in other cases wheel speeds alone compared to some reference (another wheel or calculated) should also trigger activation. Even when wheel deceleration triggers activation, it's still wheel speeds that become the controlled output of the control system. Therefore, wheel speeds alone should be enough to activate the system.

Thinking out loud. In this case wheel speeds are zero, but what is the reference velocity? It would have to be inertially calculated from the initial conditions of being stopped. I'm just not confident that level of inertial calculation (sliding down hill in reverse from a stop) and control programming exists on the truck to allow it to even attempt to do anything in this situation.

I think the likely case to figure if ABS works in reverse is to do this same thing in 2wd, allow the truck to roll backwards a bit, and then apply the brakes. This would get rid of the 4wd problem and allow it to operate off of proper F/R wheel speed differences indicating lock up. If it doesn't work then, then there's no ABS in reverse.

 

Yes, I think a forward slide experiment would be interesting, however it seems hard to do and I’m now slightly convinced it wouldn’t matter. Im not sure I’m a good enough driver to drive in reverse up a same or similar hill in reserve and then slide down in forward. I might be able to find an easier hill or a better driver.

I’m being a bit silly. I suppose I could find a steep icy hill that I could let myself slide down. Going forwards I always use engine braking so never really experienced the same thing. Looking back on it, tossing it in reverse might have helped, except reverse is so high (6mt).

 

While observing the behavior of a dysfunctional system can help reverse engineer a function system, I don’t think you can find “provable” inference. Your particular case is interning, albeit I think it is unrelated. I think we should find some icy hills and experiment together.

 

Good luck, I live in SE Georgia. Lol

 

I always appreciate your educated input. Do I assume correctly that you work in engineering?

 

If in 4lo (for the locker) reverse is pretty low.

I think the whole situation with sliding down the hill backwards is all just, situational. Perhaps if at the moment you slid down it a number of times, with various systems on and off, 4lo, rear locked, pumping brakes etc you would find out the answer. But again. Situational.

 

No argument, but this hill was slippery and sketchy enough that I would have chosen 1 st gear if I was going down… at least with prior intel. I was mostly engine braking with 2nd the rest of the trail

 

I do. Racecars mostly.

 

So yes, depending on brand, certain brands will disable ABS in reverse for some reason.

My 2016 F150 and 2020 GMC Sierra both had no abs in reverse.

 

Bolting them together or rewriting the code?

If coding, Hardware in a Loop, or Vehicle in a loop testing? I did HIL, but we had fun VIL bugs. A road
test (on the track) is always needed. I did a funky code inversion, and locked-up tires blew out. The test driver said it was
a heck of a wake-up call! Oops.

The paper posted earlier showed an earlier ABS with the pendulum sensor. Early Toyota ABS systems. Our third-generation Tacomas have yaw and accelerometer data. With the ABS pump on my off-road, I feel no pulsing, but backing down a slick hill and jamming on the brakes, I had steering control. Hence ABS was working in reverse. No reason for ABS not to work in reverse, but it's a software function (the ABS module has a reverse input). I'm still in awe of how well the Off-Road functions in different terrain (boy, that ABS pump is noisy!)

CAN bus makes life interesting, simpler wiring.

What are your favorite debugging tools?


Random desk photo:

 

Dynamicist and performance. Don't do much programming now and my coding is usually just for custom tools, testing, and data analysis.