Design Methods and Optimization Tools for the Development of Upper Limbs Exoskeletons

This thesis aims to provide an in-depth analysis of innovative design strategies and advanced modelling techniques in the field of upper limbs exoskeletons. It focuses on two key areas: first, the development of optimization tools for designing passive balancing systems that enable the creation of cost-effective and lightweight exoskeletons to prevent injuries among industrial workers while performing repetitive overhead tasks; and second, the application of conventional robotic systems techniques to produce a complete device for robot assisted therapy, specifically targeting rehabilitation for post-stroke and orthopaedic patients. The research not only addresses the complexities of replicating human limb anatomy, but also offers practical insights for developing user-friendly exoskeletons, either for injury prevention or to foster recovery of arm function after impairment. The proposed strategy strengthens the theoretical foundations of exoskeleton design and offers significant potential for real-world advancements, aiming to greatly improve the quality of life for both workers and injured patients thanks to the usage of innovative upper limbs exoskeleton solutions. The essence of this thesis is embodied in Chapter 3 and Chapter 4, where the proposed systems are presented. Before developing novel devices, the research started by tackling a detailed examination of upper limbs exoskeletons available in the scientific literature including the analysis of the human upper limb anatomy to understand the natural movement of the human art. After identifying key design principles for the development of exoskeleton systems, the current technologies in the field have been reviewed and a classification of upper limbs exoskeletons has been provided by leveraging the state of the art. In detail, this study explores devices that find application in both industrial and healthcare fields, with a focus on shoulder-elbow exoskeletons for helping industrial workers in executing overhead tasks and medical devices for the rehabilitation of the wrist. Chapter 3 deploys the Shoulder-Elbow Exoskeleton (SEES), focusing on its analysis and preliminary design. The SEES is a passive upper limb exoskeleton intended to support workers in industrial environments by assisting in repetitive tasks and reducing the risk of injuries. Its main purpose is to compensate for gravity loads on the human arm. The system is implemented via a 6-Degrees Of Freedom (DOFs) kinematic model (5-DOFs for the shoulder and 1-DOF for the elbow) and employs passive elastic elements to achieve gravity compensation, minimizing both weight and costs of the overall structure. This chapter introduces a detailed analytical tool to aid in the exoskeleton design, examining its kinetic-static behaviour and optimizing the design of the elastic springs to balance gravity across several arm movements. Different kind of balancer models, i.e., with 1-DOF or 3-DOFs, and various arrangements of the springs, either linear or torsional springs, have been evaluated. One optimal configuration is proposed as a case study, and results are validated using a multi-body simulation tool for specific tasks. Chapter 4 focuses on the Wrist EXOSkeleton (W-EXOS), detailing its mechanical design and performance evaluation as a 3-DOFs device for rehabilitating orthopaedic and poststroke patients. The W-EXOS covers the 93.3% of the human Range Of Motion (ROM) and can simulate specific wrist movements (i.e., pronation-supination, radial-ulnar deviation, and flexion-extension). The device has been designed with a handle as end-effector and is powered by electric motors via an efficient cable transmission system performing high torque-to-weight and torque-to-volume ratios. Its kinematic structure includes three rotational joints with non-perpendicular axes, allowing for a compact design and effective mass distribution. Theoretical modelling facilitated the evaluation of the exoskeleton and the human joints matching, assessing the device ROM and torque for each joint. The performance assessment of the system included a position control test and a Virtual Reality (VR) serious game trial involving voluntary healthy subjects. The VR test has been performed in two conditions, namely enabling and disabling the exoskeleton assistance while the subject was asked to complete specific orientation tasks of the wrist while wearing the device. Obtained results showed that the system significantly improved performance and reduced muscle stress by approximately the 30% when the exoskeleton assistance was provided. Additionally, the W-EXOS was proved to be adaptable in the integration with different exoskeleton systems. Indeed, the handle can be replaced with a hand exoskeleton, allowing the combined motion of the human wrist and hand; then, the overall system can be attached to a rehabilitation station or a shoulder-elbow exoskeleton, enabling the motion of the upper limb within its natural workspace and the simulation of bimanual tasks in both configurations. Alongside its theoretical and practical insights, this thesis promotes accessibility and encourages broader adoption of the proposed innovative techniques within the scientific community by sharing the codes in the Appendix. The code for the parametric design tool of the SEES is attached to be run in the software Matlab and customize the system as needed. Since the model is fully parametric, features of the user (e.g., the arm weight and length), of the exoskeleton (e.g., the dimension and material of links, the type of balancer and its configuration), and the simulated movement can be customized. The Matlab code for the kinematic model of the W-EXOS, of particular interest due to the non-perpendicular axes scheme, is provided to evaluate the matching with the axes of the human wrist joint (perpendicular) while performing a specific movement. Further works include the active prototyping and experimental testing of the SEES, along with the clinical trials of the W-EXOS, the latter being already tested in the laboratory to prove the system functionalities.