AI-HAND is a cross-border research and innovation project in the Interreg Germany–Netherlands programme area aimed at improving home-based hand rehabilitation after traumatic injuries as well as in chronic functional impairments, based on robust evidence. The goal is to develop and validate a safe, clinically deployable system that enables patients to perform high-frequency, structured training at home while making the quality of exercise execution objectively verifiable.
At its core is an integrated “closed-loop” approach: a conversational AI supports the delivery of protocol-based exercise programmes, explains therapy goals, reminds patients of training schedules, and responds to warning signs using rule-based logic. In parallel, laptop-based computer vision captures hand and finger movements in real time, derives clinically interpretable kinematic parameters, and identifies erroneous and compensatory movement patterns to provide immediate, quality-assured feedback. Where medically appropriate, an available assistive glove is conservatively integrated to support assist-as-needed active movements; an optional VR module can enhance task-oriented practice scenarios and motivation.
The project will deliver an implementable prototype with practice-oriented maturity in the range of TRL 6–7, building on an initial development stage of TRL 3, harmonised evaluation and outcome standards for cross-border application, and automated multilingual training and progress reports to support therapists, physicians, and case management. AI-HAND thus lays the foundation for scaling, transfer into routine care, and improved return-to-work pathways.
