Ex-Meta scientists want to bring visual AI to the factory floor

Perceptron, founded by former Meta FAIR researchers Armen Aghajanyan and Akshat Shrivastava, released Isaac 0.5, an open-weight vision model that enables robots to perceive, reason and act in complex industrial environments such as warehouses and factory floors. Trained on a million hours of general, ego and UMI video, the model combines perception and control to flexibly handle tasks like label reading, spatial analysis and sequential picking.