Neuro-Mechanical Modeling and AI-Driven Exoskeleton Control for Rehabilitation Robotics: A Comprehensive Review

Abstract

In recent years, rehabilitation robotics has become a promising area of research to enhance mobility and quality of life for people with neurological and musculoskeletal impairment. Lower-limb exoskeletons are one of the most promising advances in this field, combining advanced control strategies with principles from biomechanics to improve gait rehabilitation and functional recovery. This review discusses the state of the art in neuro-mechanical modeling approaches and Artificial Intelligence (AI) control techniques for rehabilitation exoskeletons. There are several important topics, such as musculoskeletal modelling, Central Pattern Generators (CPGs), neural control mechanisms, machine learning algorithms, reinforcement learning, sensor fusion and adaptive control strategies. Additionally, recent developments in the design of exoskeletons, in the field of actuators and in the human-robot interaction are provided. The review also covers the latest difficulties with models and their accuracy, real-time applicability, safety, personalization and clinical application. Last but not least, future research directions are proposed such as the incorporation of digital twins, sophisticated AI architectures and multimodal sensing devices, which can enhance rehabilitation results. The purpose of this review is to give the researcher and practitioner a structured overview of the state-of-the-art and future developments in Artificial Intelligence (AI) assisted rehabilitation exoskeletons.

Keywords: Neuro-Mechanical Modeling; Bipedal Locomotion; Exoskeletons; Rehabilitation Robotics; Central Pattern Generators; AI Control; Gait Analysis

Citation

Mansour MM. Neuro-Mechanical Modeling and AI-Driven Exoskeleton Control for Rehabilitation Robotics: A Comprehensive Review. WebLog J Robot Appl. wjra.2026.h1004. https://doi.org/10.5281/zenodo.21862008