Embodiment through AI-supported proprioception in prostheses (EmKIPro2)

Focus:	Artificial Intelligence Transfer
Type of funding:	Project funding programmes
Programme:	CZS Transfer
Funded institution:	University of Applied Sciences Kaiserslautern

The team from Kaiserslautern and Trier universities is developing a user-friendly, intelligent leg prosthesis that uses artificial intelligence to provide the user with precise feedback on movements and ground contact.

Goals

The project focuses on the development of a leg prosthesis that uses artificial intelligence (AI) to detect the condition of a leg/foot prosthesis, e.g. movement and ground contact, and transmits this sensation to the user.

Unlike conventional systems, the new prosthesis is designed to use the deformation of a flexible carbon fiber foot to capture information about ground contact. This makes it lighter, more robust and more cost-effective. The complex deformations are interpreted with the help of AI and transferred to the user's skin using a novel approach with high-resolution electrical impulses. These electro-tactile signals are also to be individually adapted with AI support, giving the user an optimal feel for the prosthesis.

Surrogate models - AI-based substitute models that approximate complex physical processes efficiently and precisely - are being developed to quickly generate realistic training data.

The goal is a user-friendly, intelligent leg prosthesis that is perceived as part of the body.