Welcome back to TechCrunch Mobility, your central hub for the evolution of transportation, especially regarding the increasing impact of AI. To receive this in your inbox, subscribe here for free — simply click TechCrunch Mobility!

Supporters of autonomous vehicle technology have consistently claimed that robotaxis and other self-driving cars will lessen accident rates and enhance roadway safety. Some evidence supports this. However, it does not negate the fact that there have been consequences for those employed (or previously employed) by the firms creating this technology. 

Sean O’Kane, senior reporter for special projects, examined data reported to the Occupational Safety and Health Administration and discovered that test drivers for Waymo and Zoox experienced over two dozen injuries in 2024 and 2025 due to sudden braking or other abrupt actions taken by the autonomous vehicles. In certain instances, these employees faced months off work after sustaining injuries such as whiplash from the AVs halting unexpectedly and forcefully.

Read his complete story, which features interviews with both former and current personnel. 

There’s a crucial point to contemplate. You may be wondering, what about other AV developers? Surely, this issue isn’t exclusive to Waymo and Zoox? You’re likely correct. It is probable that test drivers for other AV developers are facing injuries too, but the companies they represent could be exempt from the reporting obligations of OSHA.

Have a suggestion for us regarding this story or others? Contact Kirsten Korosec at [email protected] or reach her via Signal at kkorosec.07, and Sean O’Kane at [email protected].

Deals!

I have been following Gatik, the autonomous vehicle company recognized for its self-driving box trucks, since 2019 when it emerged from its “stealth” phase. At that time, I was uncertain about the startup’s viability. The hype cycle had been rapidly discarding AV startups, and no one appeared near to commercializing their technology. 

Gatik not only persevered but has transitioned from research and development to commercial operations, though on a smaller scale compared to conventional human-operated delivery firms. (The company’s box trucks are unmanned and transport goods from distribution hubs to retail outlets such as Walmart.) With $200 million in new funding, it’s now aiming to expand. The funding round was led by Qatar Investment Authority and Koch Disruptive Technologies, with contributions from Millennium Management, ARK Invest, and Intact Private Capital. 

This represents Gatik’s most significant funding round thus far. However, I would argue that its multi-year commercial partnership with PepsiCo, signed in June and part of a total of $600 million in contracted revenue, is even more substantial. 

Other noteworthy deals …

Airbound, an Indian company focused on autonomous drones, secured $37 million in a Series A funding round led by Greenoaks, with participation from DoorDash, Lachy Groom, Lightspeed, and Humba Ventures.

Mubadala Capital, the alternative asset management division of Mubadala Investment Company, agreed to acquire a majority equity stake in Arrive Logistics, a truckload brokerage located in Austin.  

Regent Craft, a startup from Rhode Island creating and producing electric seagliders, raised $120 million in a Series B round co-led by Mare Liberum and AE Ventures. (The seaglider falls under a category of vehicle known as a wing-in-ground effect vehicle, or WIG.) Regent also obtained around $120 million in debt funding from Erebor Bank, established by Palmer Luckey, the founder of Anduril Industries. A point of interest about Regent, a company I have tracked for some time: The startup, which recently completed its 255,000-square-foot seaglider factory, is evidently advancing deeper into the defense sector, likely due to the available funding and partnership opportunities.  

Vista Global Holding, a private aviation entity based in the UAE, is contemplating a European IPO that might generate over $1 billion, as reported by Bloomberg.

Noteworthy reads and other updates

As per a recent YouGov survey, a larger proportion of Americans oppose police license plate cameras than support their use. What’s your take? Send me an email with your thoughts. 

Any, a startup producing electric two-wheelers from Belgium, is betting on cargo capacity. 

General Motors is under heightened examination from U.S. safety officials following numerous incidents, including over 20 crashes or fires, and at least six injuries linked to brake malfunctions in its electric vehicles.

Ford has appointed Dave Carroll as president of its energy division. Carroll, who previously worked with ENGIE North America, will take over from longtime executive Lisa Drake.

Rivian CFO Claire McDonough is stepping down and will depart by the end of October. Her tenure at Rivian coincided with a challenging yet exhilarating (ahem IPO) phase for the company. Her exit occurs at another pivotal juncture for Rivian as it embarks on some of its largest endeavors to date, including the robotaxi agreement with Uber and scaling production of its R2 SUV. 

Uber is rolling out a fresh live video streaming function allowing parents to monitor their kids during rides. 

Waymo shared ten insights gained from having their vehicles cover over 200 million autonomous miles. The foremost insight — that multimodal sensors are essential — garnered significant attention as it directly counters Tesla’s focus on camera-only systems. There were additional insights that caught my eye, including Waymo’s endorsement of vision language models (which are increasingly prominent). 

In the meantime, Waymo is advancing internationally, announcing plans to establish operations in Munich, Germany. 

One more thing …

A significant clarification regarding what I mentioned last week. You may recall that Waymo provided details about its custom silicon chip — specifically a 5 nm ASIC chip designed to manage the extensive influx of raw data prior to reaching the core “brain” of the autonomous system. Waymo stated in its blog entry:

“While these ASICs alone deliver over 1,000 TOPS of ML performance dedicated to front-end processing and ML models, we optimize across the full stack to maximize achieved performance, especially in the low-batch regimes we often operate.”

That statement led me, along with others, to conclude that a single chip could deliver 1,000 TOPS (trillions of operations per second) of computing performance. I compared it to Nvidia’s Drive AGX Thor automotive processor, noting the similar capabilities. 

However, one attentive reader reached out with inquiries, leading me to seek official clarification from Waymo. A spokesperson for Waymo indicated, “It’s for the system, not a single chip.”

This is a vital distinction.