農機產業在全球市場環境競爭激烈、產品生命週期日漸縮短，逐漸走向顧客導向，對企業的快速回應與存貨壓力面臨很大的挑戰。而提高預測準確度，降低存貨成本，並加強企業內部與外部夥伴間資訊雙向的即時整合，以提高供應鏈流程掌握及企業應變能力。以往業者進行產銷預測時，僅考慮公司內部相關影響因子，而忽略供應鏈上合作廠商所造成的影響，且農機產業其上游關鍵零組件前置期長、獨占性強，往往造成需求預測落差，產生庫存過剩或不足的壓力，甚至無法即時滿足顧客需求。本研究探討協同預測作業、農機產業特性與供應鏈管理，並藉由實務訪談歸納出影響農機產業產銷協同預測六項主因子與十四項次因子。再結合層級程序分析法，找出因子間相對重要性。結果顯示歷史銷售資料之「農林機械銷售記錄」、產能之「產能利用率」、存貨之「存貨週轉天數」依次為影響農機產業產銷協同預測之主要因子。 此外，本研究建構出進行協同預測的作業流程架構，並以此架構深入訪談農機業者，推論協同預測模式可用於農機產業，但合作夥伴間信賴程度需要加強，以提高成功合作的可能性。而本研究發展的協同預測作業流程適合該產業，可供農機業者導入時參考。 The industry of the agricultural machinery is competitive in the global market environment as shortened products life cycle and customer-oriented marketing nowadays. A well designed collaborative forecasting process will reduce the inventory, shorten the production lead time and improve the cooperation among the participating partners. The critical components in the industry are typically expensive and with long order lead time. Moreover, the customers usually ask for short delivery time. This forces the manufacturer to put large quantities of inventory at hand, resulting in costly and less efficient supply chain. It’s of importance to develop an adequate collaborative forecasting process for the partners in the supply chain. In this study, literature was investigated related to collaborative forecasting. The characteristics of agricultural machine industry and supply chain management was analyzed, and an in-depth interview of personnel in six companies was conducted. As a result, we constructed a hierarchy of factors that was crucial to the forecasting process in the agricultural machine industry. The hierarchy of factors consists of six main forecasting factors and 14 sub-factors. We finally applied the Analytical Hierarchical Process to rank the importance of these factors. The result showed that historical sale amount, capacity utilization rate and inventory turn-over days are three major factors for collaborative forecasting process in the agricultural machine industry. In addition, this paper developed a structure of collaborative forecasting process for agricultural machine industry, and verified this structure by interviewing managers in agricultural machine enterprises. To be brief, the structure of collaborative forecasting process developed in this study is suitable for agricultural machine industry, and could be a reference model for further researches.
