Researchers at the Sharif University of Technology have developed a method for 'wearable' robots (aka Exoskeletons) to learn wearer's movement and reactions accordingly, IRNA reported.
One of the major challenges facing wearable robots is controlling them. They either have joy sticks or handles or the robot needs to recognize the wearer's action by itself. The better the recognition, the more in control is the user.
Learns to Behave
Dr. Mehran Jahed, head of the research team, explained that movement behaviors such as walking, picking things up, and grabbing objects are among characteristics exclusive to each individual. In existing control methods, movements are pre-programmed and performed alike for all users. However, Sharif researchers have developed a technique where the robot learns how to behave according to users' movements.
Helps Physicians
Mohammad Ghassemi, project member, elaborating on the new method said human body follows specific rules controlled by the brain. The most challenging part of the project was identifying these rules and copying them.
Using commanding nervous signals that are recognizable from the skin is the principle behind the new control system. This allows the robot to directly receive the command from the brain the way a real muscle does, he explained.
Ghassemi added: "This is a physiologic model with meaningful parameters in medical science which not only facilitate control for users but allows physicians to monitor the patient's condition."
