A new analysis of a high-profile government report published Friday pours cold water on a claim that Tesla has long touted: use of the company’s semi-autonomous “Autopilot” technology reduces crashes by an estimated 40%.
The figure, published in a January, 2017 report by the National Highway Traffic Safety Administration, was based on data supplied by Tesla in the wake of a fatal collision. But an analysis of the underlying information used to arrive at that conclusion appears to be flawed, according to the data analysis firm Quality Control Systems Corp. The new research, based on previously undisclosed data that the firm obtained after filing a Freedom of Information lawsuit against the federal government, found that key information was missing.
This includes the actual mileage at the time key Autopilot features were installed, which was missing for the majority of the Tesla vehicles included in the original study. The timing of when Autopilot features were installed was also unclear for many of the vehicles, making drawing conclusions from the data difficult, the new report said. The lack of meaningful safety data from Tesla is made more glaring by the fact that the government has largely failed to collect this information on its own its own, as The Information reported last year. Tesla customers, meanwhile, are often confused about the limits of the technology, which has played a role in a number of high-profile collisions.