Can a Prosthetic Hand Deliver Humanoid Dexterity?

· Source: Big Data & AI News - EE Times · Field: Technology & Digital — Robotics & Autonomous Systems, Artificial Intelligence & Machine Learning, Software Development & Engineering · Depth: Intermediate, medium

Summary

At CES 2026, Italian bioengineering firm BionIT Labs showcased Adam's Hand, an AI-based bionic prosthetic hand, integrated into Oversonic Robotics' industrial Robee R and medical Robee M humanoids. Developed since 2018, Adam's Hand is a myoelectric prosthesis that uses electromyographic (EMG) signals to control movement, featuring a single-phalanx thumb and four two-phalanx fingers, each with two degrees of freedom. The hand is designed for durability and dexterity, capable of lifting up to 100 pounds and completing over 1 million open-close cycles. It incorporates inertial measurement units and is being upgraded with pressure and thermal sensors for enhanced closed-loop control. An on-board AI algorithm interprets myoelectric signals for quick, cloud-independent responses, learning from interactions to improve consistency. BionIT Labs is also collaborating with STMicroelectronics to integrate AI processing at the edge, aiming for safer and more efficient humanoid-human interactions.

Key takeaway

For AI Engineers developing humanoid robots for industrial or human-interactive roles, Adam's Hand offers a proven solution to address critical dexterity and durability gaps. Its real-world tested design, on-board AI, and sensor suite can significantly reduce cognitive load and improve reliable object manipulation. You should consider integrating such a plug-in manipulator to accelerate deployment and enhance safety in complex environments, rather than developing proprietary hands from scratch.

Key insights

Integrating human-proven AI bionic hands into humanoids enhances factory-ready durability and dexterity.

Principles

Method

Adam's Hand uses myoelectric signals, processed by an on-board AI algorithm, to control a single-motor, multi-fingered manipulator, adapting grip without pre-selection.

In practice

Topics

Best for: AI Product Manager, Robotics Engineer, AI Engineer, Director of AI/ML

Related on AIssential

Open in AIssential →

Editorial summary, takeaway, and curation by AIssential. Original article published by Big Data & AI News - EE Times.