Prognostics and Health Management of a Wafer Handling Robot Arm

• Main Authors: Wu,Bo-Han, 吳柏翰
• Other Authors: Chung, Kuan-Jung
• Format: Others
• Language: zh-TW
• Published: 2017
• Online Access: http://ndltd.ncl.edu.tw/handle/rqmd3z

碩士 === 國立彰化師範大學 === 機電工程學系所 === 105 === In recent years, the applications of robot arm significantly have been increased due to the industry 4.0 requirement. It is an important equipment of automatic manufacture line for productivity increase and costs reduce. Therefore, when the robot is unexpectedly shutdown, it may loss of yield and increase downtime of the equipment in the process station. With the rapid development of Industry 4.0, the application of the fabrication field and the environment will be more complex, the in-line lifetime prediction of the robot arm system is even more critical and hard to deliver the precise results. In order to minimize this issue, manufacturers are developing new technology for in-line robot diagnosis, prediction failure time and then optimize the yield and equipment maintenance named Prognosis and Health Management, PHM. This study is focus on the developing the system which enabling sends the failure information when the robot arm in operation at a semiconductor manufacturing process. The automated optical inspection (AOI) technology is applied to assembly a basic/advaced visual identification system including one high-level CCD camera and image-handling software. Thee data-driven prognostic model, it contains a Wiener stochastic process and a Monte Carlo simulation based on the particle filter, was applied to predict the lifetimes of the robot arm. Furthermore, the residual useful lifetimes were updated continually to predict the maintenance time of the robot arm during the operation period.