Application of data mining techniques to monitor the network-controllable robot´s performance

This study presents the data mining techniques to monitor and classify the network-controllable robot´s performance. The robot is a part of the networked production system, which can be accessed and controlled via the Internet. In this study, the robot´s repeatability is the main focus of the performance variables, of which being gauged with the use of a networked vision system and a precision calibration grid. The collected data are analyzed and made available over the network. The ability to monitor and analyze the robot´s important performance variables presents many benefits. For example, the remote operator can decide whether the robot is suitable for a new assembly task. The data mining techniques utilized in this study effectively classify the repeatability data. The visual display of classification results is also presented.