Robotic systems for upper-limb rehabilitation: a systematic review of actuation technologies, kinematic architectures, and clinical translation barriers

Purpose: Robotic technologies for upper-limb rehabilitation have emerged to address the increasing demand for therapy associated with aging populations and post-stroke recovery. Despite their technological maturity, significant barriers to clinical translation continue to limit widespread adoption. This study aims to systematically analyze existing robotic architectures and identify strategic pathways toward evidence-based implementation. Methods: A systematic review was conducted following PRISMA guidelines. A structured search of the Scopus database yielded 215 records. After applying rigorous inclusion criteria focused on peer-reviewed engineering studies, 35 devices developed between 2003 and 2024 were selected for analysis. Devices were compared across eight parameters: structural typology, actuation modality, kinematic complexity, system mass, assisted movements, therapeutic function, validation status, and user interface characteristics. Results: Temporal analysis indicated a marked increase in development after 2018, accounting for 58.3% of the identified devices. DC motors were the most commonly used actuation method (50%), followed by pneumatic systems (19.4%), which offer inherent compliance, and shape memory alloys (19.4%), enabling lightweight designs (0.3–1.55 kg). Multi-degree-of-freedom systems (4–7 DOF) provide extensive workspace coverage but are associated with increased system mass (1.6–31.2 kg). Active and hybrid control modes (83% combined) are consistent with neuroplasticity-driven rehabilitation approaches. However, only 34% of devices underwent patient trials, 6% reached clinical deployment, and fewer than 10% achieved regulatory approval, highlighting a significant gap between technological development and clinical application. Conclusion: Upper-limb robotic rehabilitation systems demonstrate high technological sophistication, yet clinical validation remains the primary bottleneck to adoption. Future research should prioritize standardized reporting frameworks, patient-centered design, large-scale multi-center clinical trials, integration with telemedicine, and the development of modular and reconfigurable platforms to facilitate the transition toward accessible, evidence-based rehabilitation solutions.