| 摘要: | 個案公司之客戶為世界級電子大廠，對於出貨檢驗資料報告之需求非常嚴謹。其中主要的要求為全自動生產資料之蒐集，包括產品製造過程中的產品檢驗統計製程/品質控制報告 (SPC/SQC, Defect Data) ，機台生產參數 (Set Value, Process Value) 等相關設定，有些良率較差的品項甚至被要求提供全檢或即時之資料蒐集及監控/防呆機制。由於機台大部分為可程式自動化控制器 (PLC) 架構，PLC之裝置種類繁多，市面尚無較完整功能之套裝程式可供應用。於是個案公司客製設計及建置了不少機台連線及資料蒐集之架構。 由於連線項目眾多，客製所要求之配套眾多，若有一步驟不符合，資訊服務既面臨中斷。若未考慮斷線配套，復線時即喪失斷線時間內的機台資料蒐集，其影響最嚴重者會造成客戶將某料號之訂單轉給其他製造商。 因此，本研究主要目的是設計一套機制以改善此機台資料蒐集之流程，在有限之資源下，可以提供適當之資訊、通知對的人、很快找到真因、縮短系統回復資料蒐集之時間，亦可以提升客戶對公司產品品質控管之信任。 This study is a case study on a electronic manufacturer, whose customers are world class manufacturers having very stringent inspection report requirements for shipments of orders. A major requirement is that of automatic data collection during the manufacturing processes, including inspection reports on statistical process/ quality control (SPC/SQC, Defect Data), machining parameters (Set Value, Process Value) and other relevant settings. For parts with relatively poor yields, the reporting requirements even include total inspection data, real-time data collecting and monitoring/control mechanism. Since the majority of machines are PLC-based, and there are many different types of PLC’s devices, there is currently no packaged solution in the market that can cover all the functions needed to perform the total data collection requirement. Thus, the case company has to custom design and implement an architecture to carry out inter-machine connection and data collection. Since there is a large number of connecting machines, coupled with very large number of customized matching features, any mismatch or irregularities may result in disruptions of the IT Services. If the recovery plan is not well laid out, there will be a loss of data during the network outage. In the worst case, this may result in a customer switching its order on this item to competitors. Thus, the aims of this study is to design a mechanism to improve the process of data collection, so that under severe resource constraints, the system can still provide appropriate information, inform the right people, quickly identify problems, reducing recovery time for data collection, and also enhance the trust of the customer on the company's product quality control. |
