Robot Patient for Nursing Self-training in Transferring Patient from Bed to Wheel Chair

In this paper, we proposed a robot patient for the nursing training in patient transfer. The robot patient was developed to reproduce the performance of the patients who are suffering from mobility problems. We targeted on the reproduction of movement of the patient’s limbs (arms and legs) with the consideration of physical and voice interaction between the patient and nurse. The robot patient had 15 joints including 2 active joints installed with motors, 4 passive joints installed with electric brakes and 9 passive joints without any actuators. To realize the physical interaction, potentiometer type angle sensors was utilized to detect the rotation angle of the joints of shoulders, elbows and knees. In addition, follow-up control approach was applied to the shoulder joint. By this way the robot could react accordingly when the trainees moved its limbs. A voice recognition module was applied to enable the robot to interact with the trainee by voice. An experiment was performed by a nursing teacher for examine the robot’s performance. The robot patient successfully reproduced the patient’s movement with physical and voice interaction, including embracing, keeping embracing, standing up, keeping standing and sitting down.