Modeling and Experimental Validation of FeatherExo: A Soft Exoskeleton for Gait Assistance

This master thesis has been conducted at the Rehab Technology IIT-INAIL Lab of the Fondazione Istituto Italiano di Tecnologia in Genoa, in the context of the FeatherExo project. Its goal is to define a soft-exosuit with possible applications both in rehabilitation fields, in the case of individuals recovering from motor impairments, and as an assistive tool for older or physically disabled users. The object of interest of this study is the Feather actuator, a novel device whose performance and limitations have been investigated in order to enhance its integration in wearable robotic systems. The aim of this thesis is to provide experimental and analytical insight into the actuator behavior, enhancing the technology foundation of the project and helping in its development. The thesis addresses the modeling of the system from a kinematic point of view. A kinematic model has been formulated to describe the motion behavior of the system and understand its mechanical properties and interaction with the elastic components. Experimental investigation is coupled with analytical modeling to establish a framework that brings together practical testing with comprehension. A significant contribution of this work is the design and implementation of a testbench, which serves as the main platform for all the experiments. It has been developed through a rapid prototyping approach that combines Simulink with a Speedgoat real-time target machine, enabling the automatic generation and execution of real-time applications. The testbench facilitates precise and reproducible test conditions, with uniform data acquisition and the possibility to carry out many characterization tests. Initially, the testbench has been employed for the characterization of the elastic band that will be integrated in the actuation system. The same control software has been used for validating the force sensing strategy used in the actuator, based on the measurements of the elastic component's displacement. The validation procedure tested the accuracy of the sensing elements under a range of loading and operating conditions. In wearable robots, accurate force estimation is required for safe human–machine interaction. Another important outcome of this work has been the development of an efficiency map of the actuator, obtained through the Simulink-Speedgoat setup. The map was developed by testing the actuator's performance under a wide variety of operating conditions, e.g. load and speed sweeps. The resulting dataset enabled complete analysis of the actuator efficiency and energy conversion behavior, offering valuable information on power consumption, which is a key design parameter for long-term operation. The combination of these contributions — kinematic modeling, testbench design, elastic band characterization, sensor validation and efficiency mapping — provide the foundation for this thesis. They provide both functional support to further development and empirical knowledge of the behavior and limitations of the Feather actuator. The results provide support to the technological basis of the FeatherExo project by measuring strengths, weaknesses, and potential areas for optimization of the actuator design. In conclusion, the thesis demonstrates that combining experiments with modeling is necessary to advance innovative actuators for exosuit use. It improves the understanding of the Feather actuator, supports its integration, and contributes to future progress in assistive robotics.