Design guidelines for power assist robots for lifting heavy objects considering weight perception, grasp differences and worst-cases

Abstract

This paper presents design guidelines for power assist robots for lifting heavy objects in industries. A 1-DOF power assist robot system was developed and human's weight perception was included in the dynamics and control. The robot system was simulated and subjects lifted objects of different sizes with the robot under three distinct lifting schemes – unimanual, bimanual and cooperative lift in usual working conditions. Then the subjects lifted objects with the robot in some worst-cases (unusual working conditions). Weight perception, load forces and object's motions in both usual and worst-cases were analysed and the findings were compared to each other. A novel control was applied to reduce excessive load forces in both usual and worst-cases that improved the system performances. The findings were then proposed as design guidelines to develop power assist robots for lifting heavy objects in industries such as manufacturing and assembly, mining, construction, forestry, agriculture, transport and logistics, etc.