Can AI agents program industrial robots without training data? Berkeley says yes
RoboticsMechanical EngineeringComputer Science
THE AI ANGLE
Writing and refining robot code through simulationUC Berkeley and NVIDIA released open-source tools that use AI agents to write and test industrial robot programs. The agents generate, simulate, and revise code through a modular framework linked to a physics engine. This approach gives engineers readable Python programs without requiring weeks of physical workcell tuning.
THE TEACHING ANGLE
Discuss whether iterative testing in a physics simulator can replace manual troubleshooting on physical factory equipment.Read the original at therobotreport.com Generate teaching or study materials
More in Robotics
- Grok Bot took unapproved actions, ignored orders to stop, and then denied itArs Technica · November 18, 2026
- Security flaw found in Java, world's third-largest programming languageBBC — Technology · November 18, 2026
- US Foods to test inventory-scanning robots at 6 more warehousesSupply Chain Dive · November 18, 2026
- AI models can be tricked into identifying civilians as military targets by telling them it’s a gameTechCrunch — AI · November 18, 2026
- OpenAI sued by California after admitting to 80 loss-of-control incidentsArs Technica · November 18, 2026