2017 Tesla Model 3: Stopping Distance Retest After an ABS Firmware Update
by Dan Edmunds, Director of Vehicle Testing
We test 60-0 mph panic stops because it's an important safety criteria. Cars that can stop in a shorter distance are less likely to rear-end the car ahead. In cases where a crash is inevitable, they can bleed off more speed before contact is made. Besides the brakes themselves, three criteria affect braking performance: the grip available from the tires, the surface friction and quality of the road, and the proficiency of the antilock brake system (ABS).
Our initial 2017 Tesla Model 3 full-ABS panic-stop measurement of 133 feet was indeed below average. But it didn't seem outrageous, especially given the characteristics of the standard 18-inch tires. They're hard (with a treadwear rating of 500), they run at an abnormally high tire pressure (45 psi), and their 235 mm tread width isn't generous considering the car's 3,884 pounds of weight. It all seemed to fit, but that was before Consumer Reports reported a 60-0 mph stopping distance of 152 feet, nearly 20 feet longer than ours.
Location is the main difference between our tests. Our Southern California track is smooth and well cared for. Consumer Reports' test track is in Connecticut, where there's real winter. It isn't open to the public, but I'm imagining it has a suitably grippy surface that is looked after. But with a certain amount of unevenness and maybe even a few cracks brought about by the rigors of winter, CR's has the inevitable surface irregularities we know our track doesn't have.
I didn't suspect ABS calibration after our test because, frankly, that's hard to screw up. And the obvious safety implication involved means that ABS calibrations are typically optimized to the nth degree before a car goes on sale. But ABS tuning nevertheless seemed more likely after Consumer Reports announced its alarming figures, especially considering my suspicions about its track. Tesla removed any remaining doubt by announcing the imminent release of a new over-the-air firmware update that it said would address the issue.
We recently confirmed that our car was updated with version 2018.18.13, the firmware version that according to Tesla "improved the calibration of the antilock braking program to be more effective across variations in usage and environmental conditions." We took this as our signal to take our Model 3 back to our test track.
With this new firmware, CR saw a near 20-foot improvement to 133 feet — the same distance we measured in our initial pre-fix testing. Tesla's revised ABS calibration did indeed lead to shorter stops on its test track, and the significant magnitude of the improvement suggested that Tesla's initial Model 3 ABS tune wasn't very good.
Thing is, we know our test track is smooth and utterly free of cracks. And since our initial results were identical to CR's improved ones, we had no idea if we would see any improvement on our home turf.
But we did.
Our first stop with the revised firmware measured 123 feet, exactly 10 feet shorter than our previous best. The second stop came in slightly better at 122 feet. In terms of feel, the initial "bite" of the brakes was more positive, and the ABS system pulsing was smoother throughout the stop. Our Tesla Model 3's stopping distance is now right where we'd expect it to be. Not better or worse than average, but solidly midpack.
As welcome as this was, it was still hard to take. Our result tells us that Tesla missed the ABS calibration mark in nearly ideal conditions, not just less-than-perfect ones.
In our track test video, I struggled with this and tried to come up with an alternate explanation. I theorized that the handoff from regenerative to friction braking might be in play — a unique issue that electric vehicles must deal with. This possibility seemed especially plausible in light of the Model 3's rear-axle-only regenerative braking and rearward weight bias.
EVs attempt to recover as much energy as they can using regenerative braking, a scheme wherein the electric drive motor is reconfigured to be a generator when the driver wants to slow down. Regenerative energy recovery is something you want to do as much of as possible to maximize range. During a panic stop, however, it's best to switch it off and let the brake pads and rotors work alone with the ABS system to execute the shortest stop possible.
This alternative explanation was apparently bunk. Tesla has since reached out to say my handover theory was not the issue. The carmaker put it all down to ABS calibration. I'd say it was a particularly bad ABS brake calibration if we could measure such a sizable difference on such a smooth surface.
The second biggest positive to come out of this was the ease with which our car was updated. Over-the-air firmware updates just happen, and Tesla is the only car company that swears by them. Because of this, no trip to the service center was involved. We got notified that an update was about to occur overnight via an on-screen message, and it was done when we got in the car the next morning. With any other company, this would have been a formal recall that could have dragged on through a lengthy review process before a dealer-installed remedy was made available.
The only trouble with what is being called "the brake fix" is that it wasn't described in the on-screen "Release Notes" that typically accompany major updates. At first we didn't know if our car had been updated or not. It turns out that 2018.18.13 was indeed the one, but that download came with no explanation or description whatsoever. I received confirmation from Tesla public relations after the fact.
If you own a Model 3, your car has almost certainly been updated. Check your car's software version number to be sure. The "18" in the middle of the number sequence refers to the 18th week of the year, and the 13 refers to the 13th software iteration of that week. Any sequence number that reads 18.13 or higher means you are up-to-date.
This is a critical update, but don't expect to notice any difference in daily driving. The ABS calibration only comes into play during very hard emergency stops or when braking aggressively on unusually uneven, dirty or wet surfaces. The main exception, of course, is wintertime driving on snow and ice, but we're months away from knowing how that might have changed.
It's great that such a significant problem resolution was so easily made with no effort on our part. But this should not have happened in the first place. It's kind of like watching a driver get all sideways in a race and then miraculously keep it out of the wall. "Great save! What skill," they'll say. I'm the one who will reply, "Skill? Aren't you forgetting that he botched the corner in the first place?"
Dan Edmunds, director of vehicle testing @ 6,437 miles