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.