Robot intelligence firm Inbolt has launched new AI vision capabilities designed to simplify programming and control of industrial robots. The system allows engineers to build robot trajectories directly from CAD models, which the vision model then matches to real-world parts on the factory floor.

The launch includes Inbolt Robot Programming and an expanded Robot Control module. The technology is designed to eliminate the need for manual “touch-ups” during commissioning, where engineers typically adjust robot paths point-by-point to account for discrepancies between digital twins and physical environments.

The system now supports Yaskawa robots, adding to existing compatibility with FANUC, KUKA, ABB, Universal Robots, and Comau. The software streams corrected commands directly into the robot’s servo loop, allowing for real-time adjustments even on moving production lines.

“Engineers build the program from the CAD, our vision model locates the real part, and the robot executes the planned path,” said Rudy Cohen, CEO and Co-founder of Inbolt. “One platform from perception to motion, on the robots manufacturers already own.”

Inbolt’s technology is currently used in plants operated by Stellantis, GM, and Toyota. The company’s vision model has also been updated to track a wider variety of parts, with a dashboard providing live performance tests for robotics engineers to evaluate station accuracy.