At the recent technology demonstration, Allan LAGASCA, Director, Smart Industrials Segment & Robotics Segment Strategic Program at STMicroelectronics, presented a commercialized robotics demonstrator featuring the 600F Robotic Hand powered by N6 AI Gesture Recognition Control.
The demonstration showcased how embedded microprocessors and inertial measurement unit (IMU) modules enable physical AI execution, bridging the gap between trained digital AI models and real-world task execution.
Demonstration Overview
Key Takeaways from the Presentation
1. Commercialized and Industrialized Platform
The demonstrator is a fully commercialized, compact platform designed for easy deployment across various customer applications. Unlike traditional PLC setups reliant on external camera vision, this unit houses an integrated IMU module and microprocessor directly inside the setup, enabling precise gesture-based manipulation using onboard control algorithms.
“This is a fully commercialized and fully industrialized platform… It’s smaller than the camera vision across the PLC, and you can put everything in here.” — Allan LAGASCA
2. Physical AI: The Bridge Between Data and Execution
While digital AI algorithms can process data, physical execution requires a hardware bridge. Humanoid robots operate by mimicking human actions—such as folding, cooking, or playing piano—which are recorded as training data. Once uploaded into the onboard system, the robotic hand executes the pre-programmed commands right away.
“The reason why they call this physical AI is because regardless of how much AI data you have, you need someone or something to execute. This is the physical form of AI to execute all the data.” — Allan LAGASCA
