The ReHyb platform comprised four primary technological subsystems, designed to operate synergistically:

1. Exoskeleton Modules:
   The system featured two distinct exoskeleton variants. The ReHyb-SL, a lightweight, spring-loaded orthosis, provided gravity compensation and minimal-burden support for patients with mild to moderate impairments. The ReHyb-HP, a powered, multi-degree-of-freedom exoskeleton, delivered active torque assistance tailored to individuals with more severe deficits. Both devices were developed for high wearability, rapid donning and doffing, and compatibility with diverse anthropometric profiles, facilitating use in clinical and extended training sessions.
2. Functional Electrical Stimulation (FES):
   A configurable FES module delivered neuromuscular activation via a 32-electrode array targeting the forearm and hand musculature and four independent channels for biceps and triceps stimulation. The system supported spatially distributed stimulation patterns and adaptive intensity control, coordinated with voluntary motor activity to reduce muscle fatigue while promoting neuroplasticity.
3. Multimodal Sensing and Digital Twin:
   Surface electromyography (sEMG), inertial measurement units (IMUs), and additional physiological signals such as heart rate and electrodermal activity were integrated to monitor motor intent, motion kinematics, and user state. These data informed the construction of a digital twin, a computational model capturing biomechanical and neuromuscular dynamics, which enabled individualised calibration of assistance levels and dynamic adaptation of therapy.
4. Interactive Training Environments:
   Virtual and augmented reality modules, incorporating serious gaming scenarios and task-oriented exercises, provided engaging, goal-directed training. These environments facilitated repetitive, functional practice while maintaining patient motivation, aligning therapy with daily living skills and rehabilitation objectives.

The system was developed through a structured, multi-phase approach encompassing requirements elicitation, prototyping, iterative refinement, and clinical validation. The initial phase involved participatory design with patients, clinicians, and caregivers to identify target use cases, define three representative patient profiles (mild, moderate, severe impairment), and establish clinical and technical specifications for all subsystems.

Subsequent phases produced progressive alpha and beta prototypes of the exoskeleton modules, FES apparatus, sensing framework, and cognitive architecture, including the digital twin. Iterative laboratory testing and stakeholder feedback guided improvements in ergonomics, wearability, stimulation comfort, and the synchronisation of voluntary and assisted movement. Parallel research addressed neuromechanical modelling and spasticity assessment, which informed the design of adaptive, safety-aware control strategies accounting for the non-linear, time-varying nature of human neuromechanics.

In the final phase, the fully integrated platform was deployed in multi-centre clinical evaluations at Schön Klinik Bad Aibling (Germany) and Ospedale Valduce (Italy), where its feasibility, usability, and therapeutic efficacy were systematically evaluated with both healthy participants and post-stroke cohorts.