Algorithm Helps Rehab Robots to Move Naturally

Researchers at the Shibaura Institute of Technology in Japan have developed a control algorithm for rehabilitation robots that ensures that they move naturally during rehab sessions. The control system accounts for the angles the joints in a human arm naturally…

Researchers at the Shibaura Institute of Technology in Japan have developed a control algorithm for rehabilitation robots that ensures that they move naturally during rehab sessions. The control system accounts for the angles the joints in a human arm naturally make while performing various activities, and won’t let the robot attempt to bend the joints in an unnatural manner. The algorithm reduces the complexity and number of calculations required to ensure safe and acceptable movements for such robots.

Restoring movement and function after a stroke can be a challenge and may require extensive rehabilitation, which is expensive and time-consuming. Rehabilitation technology is advancing and an array of systems may soon be available to help physical therapists to perform rehabilitation more easily. One such device is a rehabilitation robot, which could guide and assist someone’s arm movements, for example, during rehabilitation. However, it is important that such robots operate predictably and safely. Moving an arm into an unnatural position could cause discomfort or even injury, particularly in someone with limited mobility to begin with. The arm has a wrist joint, an elbow joint, and a shoulder joint, and moving these joints through a series of specific angles describes a particular motion, such as picking up an object. At present, most control algorithms for such devices look at the final position of the arm once it has completed a specific task, and then work backwards to calculate the required joint angles to achieve this. The mathematics behind such calculations are highly complex, and to date such control algorithms do not adequately consider whether the joint angles they instruct a robot to make are feasible for human joints. This latest mathematical feat, dubbed “self-adaptive control parameters in Differential Evolution with search space improvement (Pro-ISADE)”, accounts for the feasibility of movements for the human arm. As the number of allowed movements is smaller, this means that algorithm calculations speeds are faster. So far, the Japanese researchers have trained the algorithm to conduct two activities that are essential to everyday living, drinking a glass of water and brushing your teeth. “Robot-assisted rehabilitation allows for higher intensity training, longer duration, and more repetition,” said Tam Bui, a researcher involved in the study. “Upper limb rehabilitation robots could help post-stroke patients quickly reintegrate into their daily lives.” Study in Artificial Intelligence Review: Using proposed optimization algorithm for solving inverse kinematics of human upper limb applying in rehabilitation robotic Conn Hastings Conn Hastings received a PhD from the Royal College of Surgeons in Ireland for his work in drug delivery, investigating the potential of injectable hydrogels to deliver cells, drugs and nanoparticles in the treatment of cancer and cardiovascular diseases. After achieving his PhD and completing a year of postdoctoral research, Conn pursued a career in academic publishing, before becoming a full-time science writer and editor, combining his experience within the biomedical sciences with his passion for written communication.