The ReHyb project will develop a hybrid upper-limb exoskeleton that uses its own sensing and novel actuation capabilities for generating the digital twin of the user. The advanced cognitive system will interface the digital twin to plan and control the hybrid exoskeleton so that the control system sympathetically respects user’s neuromechanical and psychological status, and intelligently assists the user’s motion. The cognitive capability accomplishes the robotic system to predict user behaviour and derive personalised interaction strategies by accessing to the digitised knowledge about the user. The hybrid of powered exoskeleton and FES on the user’s arm and hand allow the assistive interaction at the behavioural and muscle levels, which increases the interaction efficacy by improving the user model and wearability of the system, while an advanced control algorithm ensures all interaction strategies are executed safely and robustly.