Shared Control Techniques for Assistive Robotics Applications and Human-Robot Interaction Scenarios

This study presents a comprehensive study of shared control techniques in assistive robotics and human-robot interaction scenarios, with a particular emphasis on the integration of human-machine interfaces based on bioelectric signals. The aim is to address the growing need for assistive systems to improve the quality of life for individuals with motor disabilities, while simultaneously advancing collaborative robotic applications in several fields, including healthcare, rehabilitation, and industrial manufacturing. This thesis at first presents a review of the fundamental principles of assistive robotics, focusing on the importance of autonomy in daily activities for individuals with disabilities. The field of assistive robotics, including prosthetics, exoskeletons, and robotic wheelchairs, is receiving increasing attention also due to the possibility of integrating human-machine interfaces based on physiological signals, such as electroencephalography and electromyography. The integration of these bioelectrical signals enables robots to interpret human intentions, thereby facilitating the performance of tasks that would otherwise require the constant assistance of an assistant, particularly in the context of assistive robotics applications. This work presents the implementation of brain-computer interfaces in shared control architectures, with a particular emphasis on electroencephalographic signals, which capture brain activity and facilitate user control of robotic systems through cognitive paradigms. The effectiveness of electroencephalography-based shared control systems has been demonstrated in teleoperation experiments involving mobile robots. These experiments have shown that such systems can enhance user autonomy and provide intelligent robotic assistance in tasks such as navigation and obstacle avoidance. They can also be used to control a robotic arm in manipulating objects in teleoperation scenarios. Furthermore, this thesis addresses the applications of shared control in human-robot collaboration, with a particular focus on scenarios that require physical interaction and co-manipulation tasks. In such contexts, shared control frameworks that are guided by electromyography play a pivotal role, translating muscle activity into control signals for robotic manipulators. This research presents the development and implementation of control systems based on myoelectric signals for robotic arms in co-manipulation tasks, wherein the human and robot work together to accomplish complex tasks necessitating precision and flexibility. The objective of shared control strategies is to adapt to the operator's intentions and adjust the robot's behavior in real time, thereby ensuring safe and efficient collaboration. Techniques such as admittance control, which adjusts the robot's stiffness and compliance based on human input, are employed to guarantee smooth and natural interactions. Experimental validation represents a significant component of the research process. A number of experiments have been conducted with the objective of evaluating the performance of proposed shared control systems in both assistive and collaborative robotic applications. For example, a comprehensive investigation of teleoperating a mobile robot via a brain-computer interface is presented, wherein users navigate challenging environments while the robot proactively assists in tasks such as path planning and obstacle avoidance. Furthermore, a scenario in which the user employs electroencephalographic signals to manipulate objects via the control of a robotic arm is also presented. Similarly, experiments involving electromyography-based control of robotic manipulators demonstrate the feasibility of using bioelectrical signals to perform co-management tasks, thereby providing insights into how shared control systems can enhance productivity and reduce user workload. The experimental results demonstrate that shared control techniques not only enhance the autonomy and efficiency of assistive robotic systems, but also improve the safety and usability of human-robot interactions in collaborative scenarios. The results indicate that the incorporation of bioelectric signals into shared control frameworks provides notable benefits in terms of adaptability and responsiveness to user requirements, thereby facilitating the customization of robotic assistance across a diverse range of applications. In conclusion, this thesis makes a significant contribution to the field of robotics by proposing innovative shared control architectures that bridge the gap between human capabilities and robotic assistance. The research enables more seamless and intuitive human-robot collaboration through the use of bioelectric signals, thereby opening up new avenues for assistive and collaborative robotic applications in medical and industrial settings.