Azernews.Az

By Alimat Aliyeva

The American company Tesla does not typically disclose detailed information about accidents involving its autonomous vehicles, but it is required to report incidents at the request of U.S. regulators, including the National Highway Traffic Safety Administration (NHTSA). According to statistics, Tesla’s robotaxis in Texas are involved in accidents nine times more frequently than conventional vehicles with human drivers across the United States, Azernews reports.

Between July and November last year, nine accidents were recorded in Austin, where autonomous Tesla Model Y vehicles have been operating since June.

Three incidents occurred in July. In two cases, Tesla’s prototype robotaxis collided with other crossovers, while in the third, a stationary obstacle was struck at a speed of 13 km/h.

In September, four more accidents were reported. These included collisions with obstacles and other cars while maneuvering in parking lots, as well as incidents involving a cyclist and even an animal, with one collision occurring at more than 43 km/h. October saw a single unexplained incident at 29 km/h, and in November, one prototype repeated the July scenario while making a right turn, according to sources.

By November, Tesla’s entire fleet of several dozen robotaxis in Austin had traveled approximately 800,000 km. This equates to one accident every 88,000 km, compared to the average for U.S. drivers, which is about one accident per 800,000 km based on police records. Some analysts suggest that including unreported accidents, the typical American driver’s accident-free mileage drops to 320,000 km — still multiple times better than Tesla’s autonomous fleet.

Until recently, an observer was seated to the right of the driver’s seat in Tesla robotaxis in Austin, with the ability to intervene in vehicle control as allowed by the ergonomics of the production Model Y and the autonomous software. Tesla has now reportedly removed observers from some prototype vehicles entirely. Reports from independent bloggers suggest that these robotaxis are now “insured” or supervised indirectly by a following vehicle.

For comparison, Waymo operates a much larger fleet of fully autonomous taxis without any in-cabin safety drivers, and its vehicles demonstrate better accident-free performance than even conventional human drivers. Unlike Tesla, Waymo provides NHTSA with detailed statistics on all incidents, including those involving injuries, highlighting a stark contrast in transparency and safety outcomes.

This raises important questions about Tesla’s rapid rollout strategy: while the company pushes forward with autonomous technology, the high frequency of accidents and limited reporting make it difficult for regulators and the public to fully assess the safety of its self-driving vehicles. Observers note that the Tesla approach relies heavily on rapid real-world testing with minimal oversight, a strategy that may be cost-effective but comes with significant safety risks.

Interestingly, some experts speculate that the removal of human observers could accelerate Tesla’s data collection, allowing the AI to “learn” from more unfiltered driving scenarios — but at what cost to safety and public trust remains an open question.