以優勢約略集合、形式概念分析與流量圖探勘影響主機板維修成本之因素

Abstract

主機板為個人電腦系統之核心，近年來，全球個人電腦主機板產業已逐漸成熟，為因應全球市場成長趨緩、競爭激烈、且產品毛利率逐年下滑之趨勢，各廠商極力降低成本，以求提昇競爭力。雖然分析造成主機板故障原因，並且進一步提昇毛利，為重要議題，但相關研究甚少，為跨越研究缺口，本研究將探討影響售後維修成本之因素。為推衍影響維修成本之因素，本研究將導入優勢約略集合法 (Dominance-based rough set approach，DRSA) 進行資料探勘，並以全球前三大某主機板製造商過去十年之售後服務維修紀錄，歸納出造成主機板故障，並影響維修成本之決策規則，再使用形式概念分析 (Formal concepts analysis) 與流量圖 (Flow Graph)，討影響維修成本之因素，並加以視覺化。實證研究以產品設計、維修用料、維修結果、返修天數等因素為條件屬性，維修成本為決策屬性，推衍出影響維修費用之決策規則與因素，進而有訂定降低維修成本之策略。研究結果顯示，造成較高維修成本的主要因素，分別為使用後返修的天數、產品銷售價格、電路板材大小、最終修復成功率。當產品使用一定的時間，且主機板的價格與電路板材屬高階時，產品不良返修容易使整體維修成本提升。若最終無法順利修復，更造成產品報廢，必須賠償客戶新品，成本極高。本研究方法可作為電子業售後維修成本分析之用，並可分析設計或品質不良，降低維修成本之用，分析架構與研究結果亦可與其他產業提升產品品質之用。Motherboards are critical components of personal computer systems. In recent years, the global motherboard industry for personal computers has matured. Given challenges such as slowing global market growth, fierce competition, and declining gross margins, manufacturers have intensified their efforts to reduce costs and thus, enhance their competitiveness. Understanding the reasons behind motherboard malfunctions and their impact on profit margins is essential, but there is a need for more academic research in this field. Thus, this study aims to cross the research gap by examining the factors that influence post-purchase repair costs.This research employs the Dominance-Based Rough Set Approach (DRSA) for data mining to identify the factors influencing repair costs. By analyzing a past five years of after-sales service and repair records from a leading motherboard manufacturer all over the world, we have identified patterns that contribute to motherboard malfunctions and impact repair costs. This data was further analyzed using formal concept analysis (FCA) and flow graph techniques, providing a more precise understanding and visual representation of the factors that impact repair costs.For the empirical studies, factors such as product design, materials used in repairs, repair outcomes, and the number of days for return repairs serve as conditional variables, while repair costs are considered the primary decision variable. This framework enables us to determine the primary factors contributing to repair expenses and propose strategies to mitigate these costs.The research results show that the main factors contributing to higher maintenance costs are the number of days after use before returning for repair, the product sales price, the size of the circuit board, and the final repair success rate. When a product has been used for a certain period of time and the price of the motherboard and the circuit board material are high-end, product defects and returns can easily increase the overall maintenancecost. If the product cannot be successfully repaired in the end, it will result in product scrapping, requiring compensation to customers with new products, which is extremely costly. This research method can be used for after-sales maintenance cost analysis in the electronics industry and can analyze design or quality defects to reduce maintenance costs. The analytical framework and research results can also be used to improve product quality in other industries.