Uber is making a major push into autonomous vehicles with a newly announced partnership with Waabi, a Canadian self-driving trucking startup, aiming to deploy 25,000 robotaxis on public roads. The ride-hailing giant will invest $250 million in Waabi, contingent on the startup achieving undisclosed milestones.
The partnership reunites several Uber alumni in the autonomous vehicle space. Raquel Urtasun, Waabi’s founder, previously served as chief scientist at Uber’s Advanced Technologies Group (ATG), the company’s self-driving car division that was sold to Aurora Innovation in 2020. Waabi’s chief operating officer, Lior Ron, also has deep Uber roots—he founded and led Uber Freight, the company’s trucking business.
Waabi faces significant competition in the autonomous vehicle market. Alphabet’s Waymo and Amazon’s Zoox are already operating unsupervised robotaxi services, giving them a substantial head start. Even in its core autonomous trucking business, Waabi has not yet deployed fully driverless trucks without safety drivers on commercial routes. The startup has not disclosed a timeline for the robotaxi deployment or announced partnerships with automakers capable of manufacturing 25,000 vehicles.
However, Ron downplays the importance of being first to market. “I think it’s really about: Can the system scale? Can the system be mass-deployed?” he told Business Insider. “It’s not about getting the first driver out of the car or truck. It’s not about the first lane. It’s not about the first neighborhood.”
The technical challenge is substantial: autonomous trucking and ride-hailing require different AI capabilities. Trucking involves set routes and highway driving, while robotaxis must navigate dense urban neighborhoods with unpredictable pedestrians. Ron claims Waabi has been building a generalizable AI “brain” since day one that can transfer across different vehicle platforms without requiring rebuilding.
Waabi’s competitive advantage may lie in its sophisticated simulation technology. The company has developed “mixed reality testing” where an AI driver controls a real vehicle on a closed course while responding to simulated scenarios like traffic jams or lane-changing cars that aren’t physically present. This approach allows Waabi to test millions of scenarios that would be dangerous or impossible to replicate in real-world testing, potentially accelerating development while improving safety.
Key Quotes
Uber has always been great in building marketplaces, in matching supply and demand, and in pricing. That’s what created Uber, that what’s created Uber Eats, and that’s what I created with Uber Freight.
Lior Ron, Waabi’s COO and Uber Freight founder, explained why Uber is the ideal partner for deploying autonomous vehicles at scale, emphasizing the company’s marketplace expertise rather than just its technology capabilities.
I think it’s really about: Can the system scale? Can the system be mass-deployed? It’s not about getting the first driver out of the car or truck. It’s not about the first lane. It’s not about the first neighborhood.
Ron dismissed concerns about Waabi’s late entry into the robotaxi market compared to Waymo and Zoox, arguing that scalability and mass deployment matter more than being first to launch autonomous services.
Now we can test anything you can imagine — every permutation of traffic jam under the sun, every millions of different scenarios of construction zone. A motorbike cutting you off — you can never do that because you’ll be endangering the tester.
Ron described Waabi’s mixed reality simulation technology, which allows the company to test dangerous or rare scenarios virtually while the AI controls a real vehicle, potentially accelerating development while maintaining safety.
Our Take
This partnership reveals Uber’s pragmatic approach to autonomous vehicles after its expensive ATG experiment failed. Rather than competing with tech giants on AI development, Uber is leveraging its marketplace strength—exactly what Ron emphasized. The $250 million investment with milestone-based conditions shows smart risk management.
Waabi’s simulation-first strategy is intriguing but unproven at scale. While Waymo has driven millions of real-world miles, Waabi is betting that virtual miles matter more. This could either be visionary or a costly miscalculation. The claim of a “generalizable AI brain” that works for both trucking and robotaxis is ambitious—these applications have fundamentally different requirements.
The reunion of Uber alumni at Waabi suggests institutional knowledge matters in autonomous vehicles, but it also raises questions about whether this team can succeed where Uber’s ATG failed. The 25,000 robotaxi target without announced manufacturing partnerships or timelines feels more aspirational than operational, signaling this is a long-term bet rather than an imminent deployment.
Why This Matters
This partnership represents a significant strategic shift for Uber, which previously abandoned its own autonomous vehicle development by selling ATG to Aurora in 2020. By investing in Waabi rather than building in-house, Uber is betting on a partner-based approach to autonomous technology—a model that could reshape how ride-hailing companies access self-driving capabilities.
The 25,000 robotaxi target is ambitious and signals Uber’s intention to compete directly with Waymo and Zoox in the autonomous ride-hailing market. Success could fundamentally transform Uber’s business model by eliminating driver costs, its largest expense, while failure could leave the company dependent on competitors’ technology.
Waabi’s simulation-based approach to AI training represents an important trend in autonomous vehicle development. If successful, it could prove that sophisticated virtual testing can accelerate deployment timelines and reduce the need for millions of real-world miles—a potential breakthrough for the entire industry. The partnership also highlights how AI generalization—building systems that work across different applications—is becoming crucial for autonomous vehicle companies seeking to scale efficiently and compete in multiple markets simultaneously.