HARMONY · Human–robot Adaptive Role MOdulation with Neuro-inspired trajectorY

AI and robotics can make industrial workplaces more adaptive, but robot acceptance depends on comfort and naturalness of interaction. Traditional kinematic approaches fail to reproduce human-like motor control, which humans associate with trust and comfort.Drawing on the kinematic theory of rapid human movements, I will model motion as sequences of strokes, i.e., unimodal motor primitives arising from neuromuscular dynamics. By treating strokes as the building blocks of action, complex motor behaviours can be synthesized through their composition. The novelty lies in embedding these anthropomorphic traits within the trajectory-planning mechanisms of the robot, while Human-Robot physical Interaction (pHRI) takes place.In HARMONY (Human–robot Adaptive Role MOdulation with Neuro-inspired trajectory) I will design neural networks such that stroke parameters emerge from data, increasing flexibility compared to fixed-parameter models. Since collaboration requires adaptation to a partner’s dynamics, I will integrate bio-inspired adaptive mechanisms to modulate the robot’s movement and, therefore, its leader/follower role in response to human actions. This enables the robot to reduce both physical and cognitive workload for the human partner, fostering more intuitive physical collaboration. The project advances robotic control, aligning with MSCA goals on human-centred robotics. Impacts include safer industrial and agricultural workplaces, reduced ergonomic risks, and new healthcare applications for rehabilitation and exoskeletons. Hosted at IDSIA (USI–SUPSI, Prof. Roveda) with an eight-month secondment at UNISA (Prof. Chiacchio), the HARMONY project combines robotics, neuroscience, and AI for scientific, societal, and technological impact.