Developing a real-time accuracy monitoring system for industrial robot arms based on lean principles

Abstract

In the Robot Laboratory of Mechatronics Department during practical classes of robot programming, the participants observed that the KUKA KR5 Arc robot arm’s practicing tool does not accurately follow the programmed path in every case. In accordance to Lean approaches, the observations of the participants were taken into consideration and the possible sources of error were identified. One of them is the robot arm itself, which is responsible for the execution of the programmed movements. The goal of this research is to develop a system which makes it possible to exclude the robot arm as a potential source of error from the error analysis. For this purpose, a monitoring system was developed which supports the Lean production by implementing one of the main pillars of the Toyota Production System. Jidoka is one of the main pillars, which means that the production process is automatically stopped if an error condition arises. The method was first implemented by Sakichi Toyoda in 1896, who developed a simple device for automatic looms, which stopped the machine if the thread broke. Using this principle, the autonomation – automation with human touch – of the robot arm is achieved.