General Robotics Launches GRID to Run 50-100 Robots From the Cloud
Updated
Updated · Computerworld · Sep 14
General Robotics Launches GRID to Run 50-100 Robots From the Cloud
1 articles · Updated · Computerworld · Sep 14
Summary
General Robotics unveiled GRID, a cloud-based intelligence layer that develops, tests and deploys skills remotely instead of embedding most intelligence inside each robot.
CEO Ashish Kapoor said one centralized layer can manage 50 to 100 robots, with GPU-backed models, simulations, data and agents used to orchestrate industrial machines across sites.
In one biolab example, AI agents built a hybrid simulator for fluid handling and transparent materials, cutting work that Kapoor said would have taken at least a year to minutes.
Latency and data-security demands remain key hurdles: robots still need a millisecond-level edge control loop, and regulated customers such as automakers may resist putting proprietary designs in the cloud.
General Robotics is betting those limits can be handled through commercial cloud, on-premise and edge deployments as industrial sectors such as ports and energy adopt more robots over the next 18 to 36 months.