The UCLA Department of Computer Science has joined the newest cohort of the Toyota Research Institute’s University Research Program (URP) 3.0, a five-year initiative supporting collaborative research in artificial intelligence, robotics and automated driving. The newly launched program includes 69 projects across 31 universities, making it the largest cohort since the program began.
Among the funded projects is End-to-End Driving Simulation with Generative Neural Rendering and Safety-critical Scenario Generation, led by Bolei Zhou and UCLA’s Vision and Autonomy Intelligence Lab (VAIL). The three-year project, running from April 2026 through March 2029, is a joint research effort between UCLA and TRI’s Automated Driving division, and is co-led with TRI researchers Rowan McAllister, Yu Zeng and Kevin McGee.
The project focuses on developing next-generation driving simulation technologies that move beyond traditional graphics-based systems. Rather than relying only on manually designed virtual environments, the research explores data-driven methods to generate realistic and diverse driving scenarios modeled after real-world conditions. The team will also investigate ways to simulate rare and safety-critical situations that are difficult to capture in practice. This method will help improve the testing and evaluation of autonomous driving systems.
The collaboration builds on both the strong foundation of open-source simulation and scenario-generation research from Zhou’s lab, which includes platforms such as MetaDrive, MetaUrban and ScenarioNet, and recent projects including Vid2Sim, SceneStreamer, Adv-BMT and AutoVLA.
The project is significant because safe deployment of autonomous vehicles depends on testing systems against realistic and uncommon edge cases that are difficult to encounter in real-world driving. By combining generative AI, neural rendering and safety-critical scenario generation into a unified closed-loop evaluation platform, the UCLA-TRI collaboration aims to advance safer and more reliable autonomous driving research while strengthening a long-running partnership between the university and industry.