New NHTSA Details Reveal Two Low-Speed Teleoperator Crashes in Tesla’s Austin Robotaxi Fleet

Isaac Avatar

by Isaac

3 weeks ago

White Tesla Model Y in a Tesla showroom

Tesla’s Austin Robotaxi program just got a lot more transparent.

Newly visible crash-report narratives in the federal NHTSA database are giving Tesla fans and critics alike a more detailed picture of how the fleet has performed since launch. The fresh details cover 17 reported incidents between July 2025 and March 2026.

Among the most notable findings: at least two low-speed crashes involved a teleoperator remotely driving the vehicle at the time of contact.

TechCrunch broke down both teleoperator incidents in detail.

Newly visible information submitted to NHTSA showed at least two Tesla Robotaxi crashes in Austin where a teleoperator was remotely driving the vehicle. Both incidents were low-speed crashes, safety monitors were in the vehicles, and no passengers were onboard.

One July 2025 incident involved a stopped vehicle whose safety monitor requested help from Tesla’s remote assistance team. The remote operator took over, gradually increased vehicle speed, turned left, drove up a curb, and contacted a metal fence.

A January 2026 incident followed a similar pattern of remote assistance after a safety monitor requested navigation support. In that case, the remote driver took control and the vehicle contacted a temporary construction barricade at roughly 9 mph.

The full record adds useful context. Several other newly visible reports involved a Tesla Robotaxi being hit by another road user rather than causing the contact, including cases where the Tesla vehicle was stopped.

That distinction matters when readers look at a raw incident count and try to understand what actually happened in Austin.

WIRED added broader context about remote support in autonomous vehicle operations.

WIRED reported that the two teleoperator-involved crashes were both under 10 mph and that safety monitors were in the vehicles. The outlet framed the new details around a larger operational question for self-driving companies: how much a remote human can safely see, understand, and control when a driverless vehicle needs help in the real world.

The article noted that remote support teams are common across the autonomous vehicle industry, but direct remote driving can raise different questions than advisory support. Visibility, cellular coverage, camera resolution, and latency all become part of the safety picture when a person away from the car is steering it through a tight street, a construction area, or another awkward spot.

That makes the two Tesla incidents important beyond the scrape damage itself. They show where the early Robotaxi system still relied on human backup, and they give Tesla watchers a clearer look at the practical operating challenges that come with scaling an autonomous fleet.

The official NHTSA ADS crash CSV gives the story a data backbone, with some important limitations.

The local extract used for this article found 17 unique Tesla incidents across 37 Tesla report rows in the official ADS crash CSV. The visible March 2026 rows included report 13781-14630, where a Tesla ADS stopped at a stop sign and a truck behind it proceeded forward and rear-ended the vehicle. That row listed a safety monitor, one passenger, no reported injuries, zero subject-vehicle speed, and truck involvement.

Another visible March row, report 13781-14631, described a Tesla ADS stopped at an intersection in a left-turn-only lane with a flashing yellow left arrow when a passenger vehicle behind it proceeded forward and rear-ended the ADS. That row also listed a safety monitor, no passengers, no reported injury, zero subject-vehicle speed, and a passenger-car crash-with category.

Many older Tesla narratives in this downloaded CSV extract remained marked as redacted. That is why the teleoperator-specific details in this article are attributed to TechCrunch and WIRED, while the CSV is used for official count and visible-row context.

Drive Tesla Canada offered the most Tesla-friendly read of the newly visible record.

The updated NHTSA entries give readers a more detailed look at Tesla’s Austin Robotaxi record. The visible filings showed no serious at-fault Robotaxi crashes involving Tesla’s Austin fleet, while many incidents were minor contacts, rear-end impacts while the Tesla was stopped, curb or pothole events, tire damage, or another road user contacting the Tesla vehicle.

The unredacted narratives are useful for anyone trying to judge the program fairly. Previous redactions made it harder to evaluate whether an incident was caused by the Tesla ADS, by another driver, by a parking-lot object, by a road hazard, or by remote assistance.

With more narrative detail available, readers can separate a serious autonomy concern from a low-speed scrape or a rear-end contact while the Tesla was stopped. The same source cautioned against treating raw totals as a direct Waymo-vs-Tesla scorecard.

Fleet size, miles traveled, service area, reporting access, and operating model all change the meaning of a crash count.

That caution echoes the official NHTSA guidance on Standing General Order data.

NHTSA says the Standing General Order exists to obtain timely notice of crashes that may provide information about potential safety defects in ADS, Level 2 ADAS, or vehicles equipped with those technologies. The third amended order, effective June 16, 2025, changed the reporting cadence so the most severe crashes are due within five days and less severe crashes are reported monthly.

The agency also tells readers to use caution when comparing crash data between companies. Reporting entities have different access to crash information, different technologies, different operating locations, different vehicle counts, and different miles traveled.

NHTSA says those factors can affect whether raw comparisons are valid. Another key point: a company is required to submit an incident report after receiving notice of a potentially reportable crash even if it has not verified or agreed with every detail.

Initial reports may reflect incomplete or unknown information, and updated reports come later when a company receives new information.

So what do we actually learn here?

The raw numbers tell us Tesla’s Austin Robotaxi fleet has had 17 reported incidents in roughly eight months of operation. The narratives tell us the contacts have been low-speed and largely minor.

The teleoperator details also tell us remote human drivers were behind two of those events.

This data is company-reported and cannot be normalized against fleet mileage, so it does not prove Tesla’s safety rate one way or the other. It gives everyone more to work with than redacted blanks ever could.

More transparency is better for Tesla, better for the Robotaxi program, and better for anyone trying to evaluate the future of autonomous driving on real facts instead of speculation.

Tesla Cybercab Spotted on Public Roads With What Appears to Be a Starlink Mini Dish

Tesla Is Quietly Building the Cybercab Support Network Nobody Talks About

 

Join the conversation!

Please share your thoughts about this article below. We value your opinions, and would love to see you add to the discussion!

We Talk Tesla