The narrative surrounding autonomous vehicles has long been anchored in the promise of a safer future. For years, developers have argued that removing human error from the driver’s seat will drastically reduce traffic fatalities and accidents. While data begins to support the safety benefits of robotaxis for the general public, a different reality is emerging for those tasked with perfecting the technology. Recent data suggests that the push for full autonomy is coming with a physical toll on the human test drivers working behind the scenes.
Reports based on occupational safety data indicate that the process of refining autonomous driving systems involves significant physical risks for operators. As companies like Waymo and Zoox aggressively scale their operations and move toward international markets, the industry is grappling with the friction between rapid technological progress and the safety of its own workforce.
Occupational hazards in the driver’s seat
A review of data submitted to the Occupational Safety and Health Administration (OSHA) reveals that test drivers for major autonomous vehicle developers are sustaining injuries at a notable rate. In 2024 and 2025, test drivers for Waymo and Zoox reported more than two dozen injuries resulting from the erratic behavior of the vehicles they were operating.
The most common cause of these injuries appears to be "hard braking" or other sudden, violent movements made by the autonomous systems. When a vehicle’s software detects a potential hazard or becomes "confused" by a complex road environment, it may trigger an emergency stop that is far more abrupt than a human driver would typically execute.
For the workers inside, these sudden stops have led to cases of whiplash, sprains, and other soft-tissue injuries. In some instances, these injuries have been severe enough to sideline workers for months. This data highlights a hidden paradox in the industry: while the software is being trained to avoid external collisions, the internal experience of the vehicle remains unpredictable and, at times, hazardous for the human safety drivers required by law to monitor the systems.
It is likely that these numbers do not represent the full scope of the issue. Many autonomous vehicle developers may be exempt from certain OSHA reporting requirements based on their corporate structure or size, suggesting that the reported injuries at Waymo and Zoox could be indicative of a broader industry trend.
Gatik secures $200 million to scale autonomous freight
While passenger robotaxis face scrutiny over worker safety, the autonomous freight sector is seeing a massive influx of capital. Gatik, a startup focusing on the "middle mile" of logistics, recently announced a $200 million funding round. The investment was led by the Qatar Investment Authority and Koch Disruptive Technologies, with participation from ARK Invest and Millennium Management.
Gatik has distinguished itself by focusing on a specific, repeatable use case: moving goods between distribution centers and retail locations using autonomous box trucks. Unlike passenger robotaxis that must navigate unpredictable urban centers, Gatik’s trucks often operate on fixed routes for high-volume customers like Walmart.
This latest funding round is the largest in the company’s history, but it follows an even more significant commercial milestone. In June, Gatik signed a multiyear agreement with PepsiCo, contributing to what the company claims is $600 million in contracted revenue. The transition from a research and development phase to a commercial operator with significant revenue suggests that the "middle mile" may be the first sector where autonomous technology achieves true financial viability.
Technical milestones and hardware clarifications
As Waymo expands its footprint, the company has begun sharing technical insights gleaned from over 200 million autonomous miles driven. In a recent series of reports, Waymo outlined several "lessons" it has learned about the requirements for safe, scalable autonomy.
Chief among these findings is the necessity of multimodal sensors. Waymo’s insistence on using a combination of lidar, radar, and cameras stands in direct contrast to Tesla’s vision-only approach. Waymo argues that the redundancy provided by different sensor types is indispensable for handling the "long tail" of rare and dangerous road scenarios.
The company also clarified recent confusion regarding its custom-designed hardware. Waymo’s self-driving system utilizes a 5 nm ASIC chip designed to handle massive amounts of raw sensor data. While initial reports suggested that a single chip could deliver 1,000 TOPS (trillions of operations per second), a Waymo spokesperson clarified that this performance figure refers to the entire system rather than an individual processor.
This level of compute power is necessary to support the vision language models that are becoming a central part of the company's AI stack. These models help the vehicle understand not just that an object exists, but the context of the environment, such as interpreting hand signals from a construction worker or recognizing the intent of a pedestrian standing near a crosswalk.
Global expansion and regulatory hurdles
The momentum for autonomous vehicles is now moving beyond the United States. Waymo has announced plans to launch its robotaxi service in Munich, Germany, marking a significant step into the European market. This expansion will test the company's ability to adapt its technology to different regulatory environments and driving cultures.
However, as the technology moves into new territories, existing players are facing increased pressure at home. General Motors is currently under scrutiny from U.S. safety regulators following reports of brake problems in its electric vehicles. Federal officials are investigating hundreds of incidents, including more than 20 crashes or fires and at least six injuries linked to these braking issues.
The scrutiny of GM serves as a reminder that even established automotive giants face significant hurdles when integrating advanced software with traditional vehicle hardware. For the autonomous industry, the challenge is twofold: proving that the software can drive better than a human, and ensuring that the hardware itself remains reliable under the stress of autonomous operation.
Broad activity in the mobility sector
The broader mobility market continues to see diverse investments and leadership changes as companies prepare for a more automated future.
- Drones: Airbound, an Indian startup, raised $37 million to develop autonomous drones for delivery. The company’s design mimics a rocket-like structure, aiming to compete directly with traditional trucking for certain logistics routes.
- Maritime: Regent Craft, based in Rhode Island, secured $120 million in Series B funding and $120 million in debt capital to manufacture electric "seagliders." These vehicles operate just above the water's surface using the wing-in-ground effect. The company is increasingly targeting the defense sector for manufacturing and partnerships.
- Executive Moves: Rivian CFO Claire McDonough recently announced her resignation. Her departure comes at a pivotal time for the electric vehicle maker as it attempts to scale its R2 SUV platform and fulfill its robotaxi partnership with Uber.
The path forward
The autonomous vehicle industry is no longer in its infancy. With billions of dollars in contracted revenue and hundreds of millions of miles driven, the technology is moving toward a permanent place in the global transportation infrastructure.
However, the recent reports of injuries to test drivers serve as a sobering reminder that the "driverless" future still relies heavily on human labor. As companies move from testing to full commercialization, the focus will likely shift from simply proving the technology works to ensuring that it can be operated sustainably and safely for everyone involved, including the workers who are still behind the wheel.
The next few years will determine whether the industry can solve these internal safety issues while simultaneously satisfying regulators and expanding into complex international markets like Munich.
Filed under: AI, TechNews, Startups, ProductLaunches, Waymo, AutonomousVehicles, Transportation