Singapore’s Grab is poised to underpin Berlin-based remote-driving startup Vay to the tune of up to $410 million, underscoring a firm vote of confidence in the human-in-the-loop vectors around autonomous mobility. The deal is worth $60 million upfront with the potential for another $350 million contingent on performance, and sets Vay up to accelerate its U.S. rollout while investigating data and technology synergies alongside one of Southeast Asia’s biggest super-apps.
How the deal is structured and milestones tied
The investment is pending regulatory approval. The first $60 million offers immediate capital for building out the business, while the remaining $350 million only becomes available if Vay hits benchmarks including launching in further U.S. cities, obtaining regulatory clearances and achieving consumer-revenue targets. Grab does not do business in the U.S. and will be a strategic backer to ensure execution, rather than a direct market entrant.
The milestones-based approach is a response to the way more and more investors are funding advanced mobility: Capital is staged against tangible benchmarks, de-risking a sector that mixes software (think machine learning for mapping apps), safety (for self-driving cars) and city-by-city compliance.
For Vay, that means demonstrating repeatable unit economics and regulatory momentum beyond one flagship city—and no longer in a single market.
What Vay brings to the road with teleoperation
Vay’s service hinges on teleoperation. These tens of thousands of small, snake-like personal vehicles are operated by trained remote operators who “teledrive” them over public roads to the customer’s location and then transition control to you, the user (like a game console) as you drive normally. A driver picks up the car when you are done and takes it to another location so there is no hassle finding a parking spot. This hybrid model allows for a light hardware stack—less-than-fully-autonomous robotaxis, according to the company—that can be priced at about half the cost of ride-hailing for similar trips.
The company is already up and running in Las Vegas, Devlin says, having won regulatory clarity in Europe after developing its technology in Germany. Since the driver is also the customer, Vay is going after a different use case than on-demand ride-hailing: longer errands, suburban drives and times when temporarily having your “own” car trumps riding in the back seat of a robotaxi.
It is also an intermediate step toward full autonomy. By centralizing parts of the journey with this remote operator, Vay reduces how long a car stays on the street and balances supply between neighborhoods without having to send out an empty car that’s technically not under human supervision. It’s an execution challenge, but it scales much more like a logistics network than a pure AI bet.
Why Grab wants in on Vay’s teleoperated mobility
Grab casts it as one of the new breed of services catering to a fast-expanding segment of users who prefer access rather than ownership. As the everyday super-app that can span bike share to Uber Eats and financial services, it has a sunny interest in supporting technologies that tamp down the per-trip cost of moving people and things while extracting rich operational data.
The companies intend to look into Southeast Asia opportunities where Vay’s driving data could provide AI training for autonomous features and fleet optimization. 📈 Grab has been ramping up its play for autonomy and advanced driver-assistance, having previously invested in May Mobility and WeRide. In contrast to pure-play robotaxi projects, Vay’s teleoperation footprint could generate useful datasets today, while continuing to work on improving autonomy-readiness over time.
Market context and competition in autonomous mobility
The competition to be the first company commercializing driverless mobility is heating up. Waymo recently announced it would be expanding across the United States, including in cities where it is less established. Strategic investors are also stepping up: Nvidia last month announced a $500 million backing for Wayve, highlighting demand for end-to-end AI stacks. Into that landscape steps Vay, which provides a differentiated approach by combining human oversight with efficiency-led software, something that could resonate with investors who are seeking closer-term revenue and to eschew capital intensity.
Vay’s war chest has been growing. The startup has raised $131.8 million (per Crunchbase) from backers that include Kinnevik, Coatue, Eurazeo, Atomico, General Catalyst, Creandum and the European Investment Bank. If Vay unlocks Grab’s full pledge, it would be more than three times as much money as the company has raised in its history, offering a multiyear runway for engineering and regulatory work to enter the U.S. market.
Execution risks remain. City-by-city approval for remote driving is uneven and safety validation must satisfy both rigorous simulation and real performance on the road. Scaled teleoperation also relies on a resilient connection, highly trained operators and thoughtfully designed handoff protocols from remote drivers to end users. But if Vay can show that it can repeat the launch of a city with good unit economics, then it is less a fly-me-to-the-autonomy-moonshot play and more an actual scalable platform for mobility.
What to watch next as Vay scales teleoperated driving
Key signals will be the rate of new U.S. city deployments, regulatory approvals in more markets and evidence that Vay’s pricing edge against ride-hailing drives continued demand. And its partner activity with Grab in Southeast Asia — which includes data pipelines, mapping and fleet operations — could also hint at other products it plans to develop that live in that gap between teleoperation and high levels of autonomous activity.
For now, Grab’s up-to-$410 million bet makes a big bet not on the poles of autonomous mobility — software or human judgment — but rather on a middle road where both drivers and algorithms operate alongside each other to offer a smooth, user-friendly experience at lower cost, without having to wait possibly years for full self-driving technology to roll out everywhere all at once.