Sviluppo di un esoscheletro intelligente per la mobilità in ambienti esterni

In recent decades, active lower limb exoskeletons have been developed for rehabilitation, but their application in real-life conditions remains limited due to high costs, lack of environmental adaptability, non-intuitive controls, and insufficient open-source resources. This Ph.D. thesis aims to address these challenges by proposing an intelligent exoskeleton, OpenExo, for assisting individuals with lower limb impairments, particularly in agricultural settings. The thesis focuses on: 1. **Mechanical Design**: Developing a robust, safe, low-cost, and open-source structure capable of supporting users and adapting to outdoor environments. It incorporates standard components and integrates computer vision and EMG-based interfaces. 2. **Adaptive Gait Planning**: Using an RGB-D camera and advanced algorithms for obstacle avoidance and safe trajectory generation to ensure adaptive and secure movement in dynamic environments. 3. **EMG-Based Motion Interface**: Leveraging motor primitive analysis for accurate motion intention recognition using minimal EMG channels, enabling intuitive control and adaptability to various locomotion tasks with 70% classification accuracy mid-gait. The resulting OpenExo prototype, built on the ROS platform, provides an open-source framework for broader accessibility and innovation, facilitating real-world use beyond clinical settings.