ABS sensor issue?

Possibly. I can share the common cause of this, but you're working on a much newer model. I don't know if GM has made changes in this regard.

Back as far as the mid '90s, GM has built their front wheel speed sensors onto the backs of the front wheel bearing assemblies. They develop just barely enough signal voltage for the computer to be able to read. With a little play that develops in the bearings, which is normal, the inner tone wheel, a stamped steel ring, can move away from the sensor just enough for the signal strength to become too weak to be read by the computer. This is where you'll see the 0 mph road speed. You'll get a diagnostic fault code for a missing signal. That is very different from a fault code related to the electrical integrity of the circuit, meaning a cut wire or corroded connector terminal. A break is detected right away, then the yellow "ABS" warning light will go through its six-second self-test, turn off, then turn back on right away. With the missing signal fault code, the warning light won't turn on until the vehicle starts to move faster than 9 - 15 mph.

This problem was extremely common and could happen to a wheel bearing in as little as 15,000 miles, including to the replacement assemblies. The dealers' scrap metal bins were full of them. Funny thing is, those bearings would work perfectly fine in a car that didn't have anti-lock brakes. You could pop the speed sensor off before installing it, but you didn't have to. Even funnier, when you buy the new bearing, compare prices for one with the sensor and one without. Very often, the one with the sensor is a lot less expensive. I suspect that's because they sell so many more of them, the large quantity gets them a better price.

The next problem had to do with how the ABS Computer sets fault codes. In the earlier versions, only one code could be set at a time, at least as far as some problems. Once the missing signal developed, the fault code was set, the warning light turned on, and the computer stopped looking at the other sensors' readings. It could still detect a broken wire, (electrical problem), but it stopped looking at the other three road speeds, (mechanical problem). Since the play in the wheel bearings was so common, there was a real good chance it would happen to the other front one before the first one got taken care of. This was a much bigger problem for owners who ignored the first one for months before deciding to get it repaired. This was also frustrating for mechanics because all they had to go on was the one original fault code. They provided a repair estimate based on that, not realizing the second problem had developed. Once the new bearing was installed and the first fault code erased, the second bearing problem got detected on the next drive cycle. That could be the mechanic on a test-drive or the owner leaving the shop. The owner incorrectly assumed the car wasn't diagnosed correctly or wasn't repaired correctly. The mechanic is frustrated because he has an angry customer to work with, and he has to start all over with the diagnosis.

To my knowledge, this sensor problem didn't involve the rear wheels even though they used the same design.

When you drove around the driveway, you didn't hit 9 mph, so road speeds weren't involved yet. Below that speed, ABS isn't needed or beneficial. Also, since signal strength is reduced at lower speeds, all systems stop responding to prevent "false activation. That's another common GM problem on trucks, also due to weak signals, but the fix doesn't require replacing the sensors, just cleaning them.

Did you erase the first fault code? If you didn't, that can keep the warning light turning on. Next, see what comes up for the next fault code. Let me know what you find.

I forgot to mention that I have long suspected the high repeat failure rate might be caused by insufficient tightness of the axle nuts. 180 foot pounds was a common value for many vehicles, but some GMs called for as much as 240 foot pounds. I posted the specs below and was surprised to see this procedure. They want you to use a new nut. That means it's a "torque-to-yield" design. The threads peel a little as the nut is tightened. It won't develop the needed clamping power if used a second time.

The surprising part was the step to back the nut off, then tighten it to specs. I haven't seen that before in this application.

If you're going to replace the axle nut, another point to be aware of is there must never be any vehicle weight on the wheel bearing when that nut is not at full tightness. Doing so will instantly make them noisy. They'll make a buzzing noise like an airplane engine. A lot of competent do-it-yourselfers install the wheel / tire and set it on the ground to hold the axle from turning so they can tighten that nut. That's too late. The damage has been done. Instead, just drop a punch or screwdriver into one of the cooling slots of the rotor to hold the axle while the nut is tightened. If you're putting on a new nut after everything else has been reassembled, just jack up the one tire off the ground, then do the job. If left in "park", the other tire will hold the axle you're working on from spinning.

Please keep me updated.

That would say everything is okay electrically because no defect is detected below 9 MPH. Most likely you're looking for a missing wheel speed signal. If I remember correctly, the wheel speed sensors are not used to tell the computer when 9 MPH is reached. The ABS Computer gets road speed information from the same vehicle speed sensor used by the Engine and Transmission Computers.

Can you watch those wheel speeds on a scanner? Check for diagnostic fault codes too. If you post the code numbers, I'll look up the definitions for you. There should be diagnostic charts available too that I can post.

I apologize for not looking at this earlier. The story I gave you refers to older models with the speed sensor pressed onto the back of the bearing assembly as a round coil of wire. The photo below shows your front sensor. It's the same design GM used on their trucks since at least the late '90s. This design develops the same symptoms of loss of speed signal at low speeds, and "false activation" at low speeds above 9 mph. The solution is much easier. My red arrow is pointing to the mounting surface where rust builds up under that tab. That rust pushes the sensor up just enough for the signal to become too weak for the computer to read it by around 15 - 20 mph or less. At those speeds the computer is still watching each wheel, so a weak signal will get interpreted as a wheel has locked up and the tire is skidding.

Look for a layer of rust under that tab, and see if any scale can be chipped off the bearing. Also clean off the end of the sensor in case there's any metal chips. Those interfere with the development of a strong signal voltage. If you find scale on the bearing, be careful that as you chip it off, none of the particles fall down the hole. Even though it's rusty, it's metal that can stick to the magnet on the end of the sensor.

Did you check for scale on the hub's surface that needs to be chipped off? There's a 99 percent chance the sensor is sitting too far from the tone ring, or there's metal filings on the tip of the sensor where the magnet sticks out. Those cause a lower signal voltage to be developed.

The first time I ran into this was on a friend's '99 suburban. His was causing "false activation", meaning the ABS was kicking in at low speeds when it wasn't called for. When generating a voltage mechanically, you need three things. You need a wire, (coil of wire is much more efficient), a magnetic field around that wire, (that's why there's a magnet in the sensor), and most importantly, movement between the two. Generators work the same way. That's why we have to spin them with a belt and pulley.

In the case of speed and position sensors, the movement is a little different. The magnet develops the magnetic field, but the toothed wheel moving past the sensor isn't the movement I'm talking about. It's disturbing, or changing, the magnetic field. The result is the same. As the magnetic field changes, it "induces", or causes changing voltages to be developed in the sensor's coil of wire. In all of these cases, speed, or the rate of change of the magnetic field is the biggest variable. Generators don't develop any output voltage or current when they're standing still. Position sensors don't develop any signal voltage when nothing is moving. The voltage developed gets higher as speed increases. This is why, by up to 9 mph, the computer doesn't look at the sensor voltages. It knows the signals are going to be too low to see.

This is what's happening now, but at a higher speed. The signal voltage is too low. The other variables include the strength of the magnet, and the distance it sits from the moving component that disturbs its magnetic field. These sensors don't develop much signal voltage under the best of conditions, so it doesn't take much for a signal to become too low to read.

As in my friend's case, the false activation meant the signal was high enough to see at higher road speeds. It was dropping too low before the truck coasted down to 9 mph. The computer knew the signal was there, but thought the wheel had locked up too soon. That's why his didn't set a fault code yet.

For a learning experience, we watched the road speeds with a digital voltmeter set to "AC Volts". Be aware meters are only accurate on this scale at 60 HZ, what you'd find in house wiring. Wheel speed sensors develop that frequency at roughly a little over one wheel revolution per second, which is much less than 9 mph. At higher frequencies, the voltage displayed means nothing except for comparing to other similar sensors at the same speed. We found one front sensor to be developing 700 millivolts at 1000 engine rpm, with the other tire blocked, (0 mph). That was the only way to know we were being consistent with the two readings. With the first tire blocked and reading the other front wheel, that sensor was developing just under 1.00 volt of signal. That's normally not enough difference to worry about, but in this case, it was the clue we needed to know about. After cleaning the scale off both bearing hubs, both front sensors read around 1.20 volts at 1000 engine rpm. That was about ten years ago. We've worked on plenty of other problems since then, but none related the ABS system.

Since you have a scanner, you can read the actual wheel speed sensor voltages and road speeds with it. On other car brands it's common to find 3.0 to 4.0 volts of signal voltage. I don't think you're going to get that much, but compare the two front readings and tell me what you find.