Amazon Builds $500M AI & Robotics Hub
Amazon is building a nearly $500 million AI hub with George Washington University to accelerate robotics and logistics automation. The move comes as Amazon deploys its new "Agentic AI Squad" — teams of collaborative robots using foundation models — to autonomously handle warehouse workflows.
The deal with George Washington University is a $427 million acquisition of its Virginia Science and Technology Campus by Amazon Data Services. The campus is located in Ashburn, a major U.S. data center hub, and the deed allows for the development of a data or information technology center on the site. The "Agentic AI Squad" is powered by foundation models trained for logistics. Amazon has developed DeepFleet, a foundation model trained on millions of hours of fulfillment center data specifically to predict and optimize traffic flow for large fleets of mobile robots. This is part of a broader strategy using Amazon's own "Nova" family of multimodal foundation models. This AI coordinates a new generation of warehouse systems like Sequoia, which can identify and store inventory up to 75% faster. Sequoia uses autonomous mobile robots and gantries to deliver totes to employees, working in tandem with robotic arms like Sparrow, which handles the consolidation of millions of different items. Amazon is also testing humanoid robots, piloting Agility Robotics' bipedal robot "Digit" for repetitive tasks like moving empty totes. Amazon's Industrial Innovation Fund is an investor in Agility Robotics, signaling a strategic interest in platforms that can operate in environments designed for humans.
