The U.S.’ top automotive safety regulator is notching up its investigation into the performance of Tesla’s Full Self-Driving (Supervised) driving software in low-visibility conditions.
The National Highway Traffic Safety Administration’s Office of Defects Investigation (ODI) said on Thursday that it has upgraded the probe it launched in October 2024 to what’s known as an “engineering analysis,” its highest level of scrutiny. It’s a step that is often required before the agency tells a company to issue a recall.
This is one of two investigations that ODI is running on Tesla’s Full Self-Driving (Supervised) software. The regulator is also probing more than 80 instances in which Tesla’s driver-assistance software has violated basic traffic safety laws, like running red lights. The investigations come as Tesla has spent months trying to get a robotaxi service off the ground in Austin, Texas.
ODI opened this particular probe after four reported crashes in low-visibility situations, one of which involved the death of a pedestrian. The regulator has spent the last year-and-a-half exchanging information with Tesla, and appears to have identified a handful more incidents where the company’s driving software proved insufficient in low-visibility settings.
ODI also said on Thursday that it has not gotten all the information it wants from Tesla in this process. The investigative office wrote that, while Tesla began “developing an update” to fix the low-visibility problems in June 2024 — before the probe was even opened — the company has still not told ODI whether that fix was deployed, or which vehicles received it.
ODI also believes there could be an under-reporting of similar crashes due to data collection and labeling limitations that Tesla reported to the safety agency.
“In the crashes that ODI has reviewed, the system did not detect common roadway conditions that impaired camera visibility and/or provide alerts when camera performance had deteriorated until immediately before the crash occurred,” the agency wrote. “Review of Tesla’s responses revealed additional crashes that occurred in similar environments and where the system either did not detect a degraded state, and/or it did not present the driver with an alert with adequate time for the driver to react. In each of these crashes, FSD also lost track of or never detected a lead vehicle in its path.”
Techcrunch event
San Francisco, CA
|
October 13-15, 2026
This story is developing. Check back for updates.
