Intersubjective AI-driven multimodal interaction for advanced user-centric human robot collaborative applications

Manufacturers increasingly rely on flexible production systems that draw on the strengths of both machines and human workers. While collaborative robots have begun to fill this space, uptake remains limited. For example, current systems deliver uneven performance and often force operators to adapt to the robot rather than the other way around. Programming also remains too complex, typically requiring specialised engineers and leaving little room for the tacit knowledge of experienced staff. The EU-funded JARVIS project aims to bridge these gaps with a set of reusable, AI-driven tools that improve interaction. It will introduce multimodal interfaces for control and programming, embed social skills into different robot types, and demonstrate scalable applications across industry.