Type of thesis
Development of control strategies for robotic rehabilitation and assistance in daily-life scenarios
Scientific disciplinary sector
INGEGNERIA - Biorobotics
Relatore Prof. VITIELLO, NICOLA
- Nessuna parola chiave trovata
Exam session start date
In recent years, robotic exoskeletons have been developed for several purposes, including augmenting human strength, rehabilitating people with neurological disorders, or assisting individuals in performing activities of daily living. In order to achieve the desired performance, proper physical and cognitive human-machine interfaces must be developed, matching human kinematics to transfer torque and forces safely, and operating accordingly to the users’ intentions. <br>In this work, different robotic platforms developed for upper-limb movement assistance and rehabilitation are presented. Position and torque controller are designed and characterized to implement both patient-in-charge and robot-in-charge control modalities: the robot can accomplish autonomously the task or assist users’ residual movements, depending on their residual movement capabilities.<br>On the high-level control layer, different interfaces have been integrated in order to detect the user’s intention. A finite-state machine has been designed to perform several tasks typical of activities of daily living, including drinking from a glass, pouring into a glass, eating and reaching-grasping-moving objects. The user can initiate the movement of an upper-limb exoskeleton and continuously control grasping motion of a hand exoskeleton by means of a hybrid electrooculography-electroencephalography (EOG-EEG) interface. Experimental tests on healthy subjects and patients with different level of upper-limb impairment (from post-stroke to spinal-cord injury) have been performed and feasibility and safety of such approach are discussed.<br>To be used as an alternative interface, an algorithm for the detection of upper-limb movement onset using EMG signal for upper-limb exoskeletons in reaching tasks is presented. Offline performances of the algorithm are shown and discussed and preliminary test in real-time applications are presented.<br>Finally, a behavioral study exploring eye-head coordination by using a commercial gaze tracker and a neck exoskeleton is presented. Obtained results can be used to design a control strategy allowing people with sever movement limitations to actively control the neck exoskeleton by means of eye movement.