Robots that support the daily lives of elderly people with weakened muscles are being developed.Robots that support each movement are the mainstream, but since standing, walking, and sitting are a series of movements, it is desirable to support with one robot.In addition, a system that is small and has few sensors is desired for home use.If you rely on the robot for all the movements, you will be dissatisfied with the weakening of the muscles and the decrease in the sense of subjectivity of the movement. Developed.

　The research team focused on the position of the center of gravity to estimate the state of the robot user.Since it is not possible to obtain an accurate center of gravity position with a small number of sensors, we have developed a method to calculate the range in which candidates for the center of gravity position exist, taking into consideration the range of motion of joints.Using this center of gravity candidate, we made a robot that operates according to the user's state estimation and abnormality detection by machine learning called a support vector machine.

　This makes it possible to know the state of the user trying to stand or sit, and to provide the necessary support accordingly.It also has a function to prevent accidents by detecting abnormalities such as a state in which it is likely to fall.

　The research team believes that the human condition estimation using the center of gravity candidate can be applied to welfare robots of various shapes and uses.He also wants to clarify the actions that robots should take and how to communicate with humans in order to make the robot easier to use for the elderly.

Paper information:[IEEE Access] Standing, Walking, and Sitting Support Robot Based on User State Optimization Using a Small Number of Sensors