Manipulator handling device for assembly of large-size panels

摘要

Purpose - The purpose of this paper is to introduce a novel device to handle a robot manipulator which can grip large-size panels. This concept arises from questioning why the glazing task is always performed manually and it is assumed that if the panel is handled by worker's bare hands, the material is lifted by a robot system and can be assembled to a frame easily and intuitively. Design/methodology/approach - This study proposes the intuitive manipulator device (IMD) which can be attached on the panel directly and connected to it with the coordinate of robot end-effector based on a virtual coordinate of IMD. The virtual coordinate is defined by the detection of the location of the IMD from the robot end-effector using IR sensor scanning and origin point estimation method. In this study, the robot manipulator system is operated by a combination of the commands of two IMDs to perform the panel assembly test and its aspect of input commands is compared with the previous force-control based human-robot cooperative systems. Findings - The proposed system shows the better performance while reducing the frequent force reflection of robot system against an environment and simplifies the instant input source for robot control system. Those are caused by the intuitiveness of visual servoing performed by operators and the minimization of a force control strategy by utilizing the operator's own sensitivity. The proposed system shows the possibility of efficiency improvement and simple mechatronic system to realize the automation of panel assembly task. Originality/value - The proposed device alternates the expensive 6-axis F/T sensor system to handle the robot manipulator by using the two 3-axis load cell and those force/torque combinations. Also, the developed device is portable and can attach on the material anywhere. That is why this system could cover various sizes of materials. This system minimizes the computational load to control the robot system and improves the efficiency of an assembly task based on the human-robot cooperation strategy.