Microrobot navigation training drops from days to under 10 minutes
A new learning framework lets researchers train autonomous microrobot navigation policies in under 10 minutes, shortening the loop for testing robot designs and navigation settings. It runs thousands of simulated vascular environments in parallel, producing roughly 190,000 training transitions per second, then transfers the learned policy to different microrobot types and scenarios without additional training. Across evaluated scenarios, the method cut action variation by at least 33.7% and increased obstacle clearance by at least 2.1%.