Locus Robotics deployed AI-powered object detection and avoidance systems on its warehouse robots, enabling them to recognize and respond to real-time obstacles — such as forklifts, people, carts, and other robots — as they navigate warehouse floors. This capability is active on deployed Locus Origin, Vector, and Array robots.

Using machine learning applied to sensor and camera data, Locus robots can classify nearby objects by type and choose context-aware responses: for example, rerouting around a forklift blocking an aisle if the pick is not the final one, or waiting safely if it is. Locus describes this as a proprietary 'vision and obstacle-detection AI agent' system, developed through its LocusLABS R&D division. The company is also building vision foundation models specifically for warehouse environments, and is developing AI-based monocular vision and a visual SLAM mapping agent for future enhancements to navigation.