| 摘要: | 本三年期單一整合型計畫擬與中華電信及啟碁公司合作(已簽訂合作意願書)，於中央大學之FESTO智慧製造系統及非傳統加工智慧工廠建置5G終端模組，與各連網機器設備及相關感測模組介接，達成5G網路通訊，並建置具備資訊安全防護之邊緣計算模組，結合自主移動式機器人(AMR)及人機協同相關技術，實現5G智慧自動化工廠，並以藍寶石基板加工為例，驗證5G於邊緣計算模組應用之優越性。本計畫將於所建置之5G智慧工廠利用電化學放電加工機、雷射加工機、自動化光學檢測系統執行藍寶石基板加工，並於邊緣計算模組即時蒐集製程參數及檢測數據，進行數據前處理與分析，執行各式人工智慧演算法，以建立一Data-driven模型，進行尺寸、表面粗糙度、幾何位置、品質預測等功能，且透過智慧型穿戴裝置即時監控機台運作，提升製程良率及品質。本計畫亦將應用光達及深度攝影機，結合AMR，透過三點定位通訊、光達飛時測距、自走車定位校正、虛擬場景、機械手臂三維環境感知等技術，達成即時避障及人機協同目標。預期目標為第一年：5G智慧自動化工廠基礎設施建置；第二年：5G智慧自動化工廠產線加工優化；第三年：5G智慧自動化工廠人機協同產線建置，實現5G智慧自動化工廠，以期對國內工業智慧自動化之發展有所助益。 ;The purpose of this proposal is to establish 5G terminal modules at the nontraditional machining smart factory and the FESTO smart manufacturing system located at the Department of Mechanical Engineering, National Central University, in cooperation with the Chunghwa Telecom Co. and Wistron NeWeb Co. The 5G terminal modules will connect manufacturing equipment and various sensors through an edge computing module to achieve 5G network communication. The edge computing module will be featured with information security protection. In order to establish the targeted smart factory with 5G mobile networks, the edge computing module will be integrated with the autonomous mobile robot (AMR) and human-robot collaboration (HRC). The superiority of the 5G smart factory with edge computing technique will be verified by conducting sapphire substrate machining as an example. A laser processing machine, an electro-chemical discharge machine (ECDM) and an automated optical inspection (AOI) system built in the 5G smart factory will be used to perform sapphire substrate machining. The edge computing module will collect real-time processing data and execute data preprocessing process for artificial intelligent algorithms to build a data-driven model. The model will be applied to predicting surface roughness, size and location of the workpiece, product quality, etc. Real-time machine monitoring will be developed through an intelligent wearable device to improve product yield and quality. An AMR system cooperated with LIDAR sensors and KINECT modules will be built to develop several techniques, such as ultra-wideband communication, time of flight, positioning correction, virtual scene establishment, and 3D-environment perception, for establishing collision avoidance, HRC, and automated logistics management. The first-year goal is to set up the fundamental infrastructure required for the 5G smart factory. In the second year, relevant techniques will be developed to optimize the processing parameters and improve the machining quality through on-line, real-time control. The third-year, final goal is to establish a production line with heterogeneous equipment integrated with AMR and HRC in the planned smart factory. It is hoped the implement of this project will be beneficial to the development of 5G intelligent automation industry in Taiwan. |
