Autonomous driving technology developer Waabi has released details of its closed-loop simulation system, Waabi World, which it will use to develop its Waabi Driver system.
The company says it is the most scalable, highest fidelity closed-loop simulator currently available. Described as an AI-powered environment, its capabilities include being able to design tests, assess skills, and teach the self-driving Waabi Driver to learn to drive on its own.
Waabi says the simulation is defined by four core capabilities. It builds digital twins of the world from data, automatically and at scale; it performs near-real-time high-fidelity sensor simulation enabling testing of the entire software stack in an immersive and reactive manner; the simulation creates scenarios to stress-test the Waabi Driver, automatically and at scale; and, finally, it teaches the Waabi Driver to learn from its mistakes and master the skills of driving without human intervention.
According to the company, Waabi World leverages AI to reconstruct the geometry, appearance, and material properties of real-world objects and backgrounds from sensor data such as lidar returns and camera images. This enables it to automatically recreate digital twins with the diversity, scale and realism of the real world.
Waabi asserts that for a simulator to truly replace driving in the real world, the software stack needs to behave the same in simulation as it would in the real world. This is achieved by simulating how the Waabi Driver would observe the virtual world through its sensors, just like how it would see the real world; the only way to properly test the entire stack in simulation and teach the self-driving algorithms.
Traditional sensor simulators use physics-based rendering engines that model how light interacts with the artist-designed virtual world and how the sensor receives it. However, Waabi notes that it is extremely challenging to accurately simulate all the different physical phenomena that affect each sensor (such as specular reflections in cameras, spurious lidar returns from exhaust and fog, and multi-path returns in radar to name a few). Additionally, artist-designed worlds often lack the accurate physical properties needed for fully physics-based simulation, which results in unrealistic sensor data.
Instead, Waabi World leverages AI along with simplified physics-based rendering to simulate realistic sensor data in near real time. Its AI algorithms, combined with high-quality recreated virtual worlds, learn to make the physics approximation look more realistic, while being computationally more efficient than traditional complex physics simulators.
For more detailed information on Waabi World, check out Waabi’s recent blog post.
Meanwhile, Click here to read Autonomous Vehicle International‘s exclusive interview with Waabi founder and CEO Raquel Urtasun in which she discusses the self-driving startup’s unique AI-first approach to delivering true autonomy.