Tesla’s FSD Safety Data: Progress Made, But Questions Remain
▼ Summary
– Tesla launched a new website section reporting safety statistics for its Autopilot and Full Self-Driving systems to improve transparency and reliability.
– The company claims FSD users drive 5.1 million miles before a major collision, significantly more than the average US driver’s 699,000 miles.
– Safety experts criticize the report as insufficient and untrustworthy due to Tesla’s history of deceptive safety data practices.
– The new safety hub separates highway and non-highway miles, but experts highlight flaws like excluding injury data and making unfair comparisons.
– Unlike Waymo, Tesla’s safety reports lack independent verification, making it difficult for researchers to trust or validate the data.
Tesla has launched a new online safety hub to showcase statistics for its Autopilot and Full Self-Driving (FSD) systems, marking a shift from its earlier quarterly reports. The platform aims to offer more transparent and verifiable data, a critical step for a company whose future hinges on public trust in automated driving technology. According to the site, Tesla vehicles using FSD have collectively traveled over 6.47 billion miles, with a live counter tracking this figure in real time. The company asserts that FSD Supervised enables drivers to cover approximately 5.1 million miles before a major collision and about 1.5 million miles before a minor one, figures that significantly outpace the average U.S. driver’s safety record.
Despite these impressive numbers, independent safety researchers remain skeptical. Noah Goodall, a civil engineer specializing in automated vehicle systems, noted that while the statistics appear favorable, Tesla’s history of questionable reporting makes it difficult to take them at face value. He emphasized that none of the data has been independently verified, adding that academic reviewers often hesitate to accept Tesla’s figures due to concerns about their authenticity.
One notable improvement in the new safety hub is the separation of highway and non-highway driving data. Earlier reports drew criticism for focusing heavily on Autopilot, which is primarily used on controlled highways, while ignoring the more complex and crash-prone environments where FSD operates. Philip Koopman, an autonomous vehicle safety expert from Carnegie Mellon University, described this distinction as a “good start,” but pointed out several flaws in Tesla’s methodology.
Koopman argues that Tesla’s comparison between FSD-equipped vehicles and the overall U.S. fleet is misleading. The company compares brand-new Teslas loaded with advanced safety systems to older vehicles without similar technology, a mismatch he likens to claiming a high school produces elite athletes because its students outperform the general population, including elderly or physically limited individuals. He also highlighted the absence of injury or fatality data in Tesla’s report. While Tesla cites inconsistent reporting and health privacy laws as reasons for the omission, Koopman believes the company could estimate such figures through legal claims and incident reports.
Unlike competitors such as Waymo, which submits its safety data for peer review and publishes detailed studies, Tesla’s reports lack external validation. Waymo’s approach allows independent experts to assess its safety claims, reinforcing credibility, something Tesla has yet to embrace. Goodall summarized the challenge: trusting Tesla’s numbers is difficult given the company’s track record of overstating its systems’ capabilities and safety.
While Tesla’s new safety hub represents a step toward greater transparency, experts agree that more rigorous, third-party verification is essential before the public can fully trust the company’s autonomous driving claims.
(Source: The Verge)
