Dexterity :: Robots to Perceive, Reason & Act in Unstructured Environments

Dexterity develops Physical AI systems that give industrial robots human-like dexterity for complex logistics tasks. Founded in 2017 and headquartered in Redwood City, California, the company is led by CEO Samir Menon and focuses on deploying real-world autonomous solutions rather than laboratory demonstrations. Its technology enables robots to perceive, reason, and act in unstructured environments, handling edge cases that typically defeat conventional automation.

The company’s core offering is a fleet of AI agents that work together to control robotic arms and mobile platforms. These agents run the Foresight world model, a large-scale Physical AI system trained on more than 100 million real-world autonomous actions. Foresight allows robots to understand physical dynamics, plan sequences, and make decisions in under 400 milliseconds while maintaining a perfect safety record.

Dexterity’s primary robot-based product is its autonomous truck-loading solution, which uses high-payload robotic arms to load and unload trailers in live distribution centers. The system handles mixed pallets, irregular packages, and dynamic conditions without human intervention. Additional deployments include fully autonomous piece picking and depalletizing for parcel and 3PL operations. All solutions are designed for continuous production use and integrate with existing warehouse infrastructure.

Major customers include FedEx, Sagawa Express, Sumitomo Corporation, and Mech. These deployments have driven over 100 million autonomous actions across multiple countries, with zero safety incidents. Dexterity positions its technology as ready for immediate enterprise adoption, not future pilots.