TechCrunch's latest traffic column cited data from the United States Occupational Safety and Health Administration (OSHA) declaration that the auto-pilot drivers of Waymo and Zoox had more than 24 injuries between 2024 and 2025. Most of the injuries were related to car brakes or sudden movements, and some employees stopped for months because of neck sprains, etc.
This article focuses on less discussed aspects of robotic taxi expansion. In the past, more attention has been paid to the impact of auto-driving systems on passengers and road safety, and the occupational risks associated with testing and accompanying personnel are not commonly discussed in public.
The damage is about the emergency.
Reports mention that the data submitted by Waymo and Zoox to OSHA recorded work injuries caused by sudden brakes, stopovers or unusual movements of automatic vehicles. In some cases, the test driver was clearly hit inside the vehicle, which took longer to recover.
TechCrunch also interviewed current and former employees to supplement the public statements. It was reported that the records showed that even if no open traffic accident had occurred in the automated driving system, there could be a continuing risk to the personnel in the vehicle during the testing and operation phases.
More companies or incomplete disclosure
The report also notes that this problem does not necessarily exist only in Waymo and Zoox. A similar situation may be experienced by other self-driving developers ' testers, but some companies may not be covered by the OSHA mandatory declaration, making it difficult for the outside world to obtain complete data.
This means that the current publicly visible number of injuries is more like a disclosed sample than an entire industry. As robotic taxis and undistributed vehicles expand their testing and commercial operations, related labour safety issues may receive more regulatory attention.
Waymo speaks of sensors and algorithms.
In addition to security concerns, there is also mention of Waymo, which recently summarized the experience of its auto-driving vehicles, which accumulated over 200 million miles of automatic-driving mileage. According to Waymo, multi-modular sensors are still indispensable, and this judgement contrasts with the course of Tesla's partial vision programme.
Waymo also mentioned the increasing role of visual language models in auto-driving. At the same time, the company's earlier disclosure of 5 nanoSIC has attracted outside attention. Waymo later clarified that machine learning algorithms over 1,000 TOPS refer to a whole front-end processing system, not the performance of a single chip.