Researchmicroroboticscmosreinforcement learning
Cornell Builds Autonomous Microrobots With CMOS Brains
8.3
Relevance Score
In a Cornell University lab, researchers led by Paul McEuen and Itai Cohen have built autonomous 100-micron robots integrating CMOS control, photovoltaic power, and platinum-legged bubble propulsion, reported in Nature. They trained reinforcement-learning algorithms (announced early 2023) in simulation to discover efficient gaits, achieving speeds above 10 micrometers per second. The scalable wafer-based fabrication suggests potential for in-vivo diagnostics and micro-manufacturing, though power, control, and biocompatibility hurdles remain.



