Waabi's Physical AI Platform Targets Generalization, Pivoting from Highways to Urban Robotaxis
Raquel Urtasun founded Waabi to address one of the most complex challenges in modern engineering: creating reliable, all-encompassing AI that interacts safely with the unpredictable real world. Her core vision...
Raquel Urtasun founded Waabi to address one of the most complex challenges in modern engineering: creating reliable, all-encompassing AI that interacts safely with the unpredictable real world. Her core vision moves beyond building siloed vehicle functions; instead, Waabi developed a 'Physical AI' platform capable of generalized intelligence—treating self-driving vehicles as complex, rolling machines that can adapt their operational model to different environments and form factors.
From an engineering perspective, the platform's ingenuity lies in its end-to-end architecture. Instead of developing separate models for braking, steering, or object detection, Waabi built a comprehensive system capable of assessing and reasoning about the potential consequences of its actions in real time. Crucially, this framework is designed for generalization. It can be built to accept various market sensors and is not limited to the controlled environment of a highway; this structural flexibility allows the technology to smoothly scale from long-haul trucking on major arteries (like those tested in Texas) into dense urban settings, specifically exemplified by the deployment into robotaxis via Uber.
This platform approach—combining a robust, verifiable end-to-end AI brain with a sophisticated simulation environment—is key. It establishes a highly generalized capability. This contrasts sharply with older industry approaches that often treated autonomy as a set of separate, optimized modules. By betting on a generalized model, Waabi positions its technology to be hardware agnostic and deployable across multiple verticals, proving its viability not just in the controlled loop of a long-haul truck, but in the erratic chaos of city streets and robotaxi fleets.
Waabi’s platform focus on generalized Physical AI—an end-to-end system with built-in simulation and adaptable sensor integration—allows it to accelerate deployment by moving beyond the specialized constraints of long-haul trucking and into diverse urban applications like robotaxis, solidifying its position as a holistic autonomy provider.