2022 NHTSA Finds Teslas Deactivated Autopilot Seconds Before Crashes

NHTSA Finds Teslas Deactivated Autopilot Seconds Before Crashes

A NHTSA report,  during the course of its inquiry of collisions involving Tesla vehicles equipped with the automaker’s Autopilot driver assistance system, the company has uncovered a worrisome detail: Autopilot was engaged but “aborted vehicle control less than one second before to the initial impact” in 16 of these collisions, “on average.”

NHTSA Finds Teslas Deactivated Autopilot Seconds Before Crashes
NHTSA Finds Teslas Deactivated Autopilot Seconds Before Crashes


Conspiracy theorists are already asserting that this demonstrates Tesla knowingly programs its Autopilot system to deactivate in advance of an imminent, unavoidable impact. The goal of this programming is to ensure that the data will show that the driver was in control of the vehicle at the time of the crash, and not Autopilot. The purpose of the Autopilot deactivations is unknown as of this point since the investigation being conducted by the NHTSA has not found (or made public) any evidence suggesting that they were done for malicious reasons. From where we are seated, it would be quite stupid to intentionally configure Autopilot to transfer control back to a driver immediately prior to a collision in the expectation that black-box data would exonerate Tesla’s driver aid function from responsibility for the accident. Why? Because it is impossible to have a person respond in the blink of an eye, and the statistics would indicate that the computers were helping the driver up to the moment where there was no turning back.


It wasn’t as though the drivers were given zero time to react, as the NHTSA report shows that “in the majority of incidents” among those 16 under close investigation, the Teslas activated their forward collision warnings and automated emergency braking systems. However, it is not mentioned how far in advance of impact those kicked on. In eleven of the incidents, the drivers did not avoid the accident by taking any action between two and five seconds before it occurred. This suggests that the drivers, much as Autopilot, did not notice the oncoming collisions.

Autopilot, despite its name, is a driver assistance feature. Those who believe that Tesla is trying to concoct a “blame-the-driver” defense need to remember only one thing: the ghost in the machine punting control to hapless, inattentive drivers just before impact is indicative of Tesla trying to gin up a “blame-the-driver” defense. It is designed to be used with driver supervision, which means that the driver, who may not need to intervene with the system’s operation of the accelerator, brakes, and steering, must nevertheless be keeping an eye on the proceedings. This is because it is intended to be used with driver supervision. In other words, even if Autopilot were engineered to distract drivers so that Tesla could absolve itself of responsibility in the case of a malfunction in the system, it is still the driver’s job to pay attention to what is in front of the car at all times. Even if Autopilot continued to operate all the way up until the point of accident, this would still be the case, which means that Tesla could still blame the driver in good faith.

NHTSA Finds Teslas Deactivated Autopilot Seconds Before Crashes
NHTSA Finds Teslas Deactivated Autopilot Seconds Before Crashes

Things aren’t quite as straightforward as that, obviously. The company Tesla has a history of marketing its Autopilot function in a way that gives the impression that it is capable of more than it actually is. This is generally done by ignoring the widespread misconception that the system can operate autonomously, giving it the name “autopilot.” The corporation had to be virtually coaxed into making driving reminders and safety blurbs highlighting how the driver needs to pay attention more evident within their vehicles before they agreed to do so.

The odd loss of control of the autopilot in the milliseconds or so just before to each of these disasters remains unexplained. It is quite likely that there is a straightforward procedure in place to power down the computer when it detects that a crash is about to take place. Think about seatbelts that tighten up so occupants are seated more safely, or fuel-line disconnects, or some of the fancy new suspension actions that Audi’s A8 is capable of, such as lifting one side of the car just before it’s T-boned to place more of the crash structure in the path of the impact. There are plenty of new cars that feature last-ditch shutoffs and other preemptive actions that occur just before or during impact.

In the absence of any evidence that Tesla hoped its Autopilot shutoffs, which, by the way, are not claimed to have occurred in every crash under investigation, would allow it to claim, however implausibly, that the crashes were the drivers’ faults because they were “in control” for the hummingbird-wingflap of an instance just before impact, there is no “there” there. There is no “there” there. This is just speculation, and because Tesla does not have a public relations team or any other way of answering journalists’ questions (other than tweeting at Elon Musk), we won’t know for sure until the NHTSA completes its investigation, which has since expanded to include systems on over 800,000 Model S, Model 3, Model Y, and Model X vehicles from the model years 2014 through 2022. Since the NHTSA has said that it would now submit the conclusions of its inquiry on a monthly basis going forward, we should have additional information very soon.

Leave a Reply

Your email address will not be published. Required fields are marked *