PolyU Science Research Impact 2022 EP04 - Improving Manufacturing Efficiency

Abstract

由於估算不準確，紡織品製造商經常因產量不足或生產過剩而未能達到生產目標，導致材料及能源浪費等問題。假如能準確地預測原料數量和產品合格率，將有助於減省成本及減少浪費。 理大理學院蔣濱雁博士及研究團隊建立了一個混合統計模型，能準確預測合格產品的生產機率，並按原料狀況調整機器輸出數值。該模型成功幫助一家為全球提供紡織品的製造商提高生產效益，每年減少生產成本達250 萬港元，同時亦減低對環境造成的污染，包括每年減少排放300 噸溫室氣體以及 2,155 噸廢水。 Textile manufacturers often fail to meet their production target due to insufficient production or over production. This leads to wastage of resources and energy due to the inaccurate estimation of raw material needs and production planning. An accurate and effective manufacturing forecast would help mitigate such loss and wastage. Dr Jiang Binyan and a team of researchers from PolyU Science have developed a hybrid statistical model to accurately predict the product qualification rate and adjust the machine output value based on the condition of the raw materials. The model has helped a global textile manufacturer improve production efficiency and reduce production costs by HK$2.5m annually. In addition, it has also minimized the environmental footprint by eliminating 300 tons of greenhouse gas emission and 2,155 tons of sewage annually after implementation.