Baidu launches Apollo World to train robotaxis in simulated road environments

Baidu unveiled Apollo World on November 3, a world model that simulates interactive road environments to train and test autonomous vehicles, according to TechNode (China Tech). The company says the system is already supporting Apollo Go’s global robotaxi development, and TechNode reports it can generate up to 100,000 kilometers of simulated driving per day. The launch underscores a push to scale validation of driverless systems without relying solely on costly real-world miles.

Apollo World is designed to accelerate iteration by exposing software to diverse traffic scenarios in silico, reducing time to deploy updates across fleets, TechNode notes. Baidu did not share wider availability details in the report, but the company is positioning the tool as core infrastructure for its robotaxi roadmap. Watch for technical benchmarks, new Apollo Go city rollouts, and any regulatory comment on simulation-led testing in China and abroad. [Developing. This article will be updated.]