ABB Robotics Taps Prosthetic Data to Close the Human-Robot Dexterity Gap

What happened: ABB Robotics is collaborating with California bionics firm PSYONIC to improve robotic gripping and dexterity. The approach uses real-world manipulation data from human prosthetic use. It combines PSYONIC’s Ability Hand with an ABB GoFa cobot.

Why it matters: The collaboration explores how touch and motion data from prosthetic use can train robots for delicate, variable tasks that have been hard to automate. The Ability Hand, originally built for prosthetics, pairs myoelectric control, touch sensing, and compliant mechanics. ABB states that improved handling can reduce engineering time by up to 30%.

Industry context: Grasping and dexterity are central to ABB’s Autonomous Versatile Robotics vision, covering robots that sense, reason, move, and handle objects in dynamic settings. The work supports physical AI systems that learn from real-world interaction with industrial reliability. Psyonic founded in 2015 and based in San Diego, develops bionic systems for humans and robots. 

This access gives ABB a route to richer training data. The partnership approaches dexterity as a data problem as much as a hardware one.