本個案研究探討如何運用源自於日本豐田在製造業設立的典範-豐田式生產系統（Toyota Production System）的精實管理(lean Management )的方法論與工具，解決線性滑軌（Linear Guideway）生產製造過程中的品質與生產效率問題，俾便符合客戶要求。使用磨床進行加工時公差是很重要的品質要求，因為全檢成本太高，惟有靠管理手法解決問題，本研究個案公司依照客戶要求的公差±10μm來進行生產，每月約有退貨1千顆左右，客訴平均3~4件，因此減少製程變異以及如何改善使得生產效率更高，是本研究預計要解決的問題。應用的方法包括：問題特性要因圖、價值溪流圖、分析誤差種類、流程分析圖、動作分析圖、流程圖、工廠佈置等改善方法。本研究導入精實管理方法：確認誤差來源、分析誤差種類、校正系統誤差、消除錯誤，研磨機台趨勢控制常態誤差變異，確認常態誤差±0.75μm，製程能力要求±3.0μm，如生產抽驗±5.0μm，則品保人員及生產技術人員馬上要求停線確認消除誤差源。改善結果發現提升生產效率40%，改善後隨學習曲線進入成長期，預期產量提升，且切削液過濾後使刮傷趨近於零。

This case study employs methodology and tools of lean management, derived from Toyota's paradigm - the Toyota Production System, to address quality and productivity issues in the production process of linear guideways. The goal is to meet customer requirements. Maintaining precision tolerance during grinding processes is a critical quality requirement. Full inspection is expensive, so problem-solving relies on lean management techniques. In this case, the company produces with a tolerance of ±10μm as specified by customers, resulting in approximately 1,000 returns per month and an average of 3-4 customer complaints. Therefore, the study aims to resolve the problems of reducing process variations and improving production efficiency. The methods employed include problem characteristic cause-and-effect diagrams, value stream mapping, analysis of error types, process analysis diagrams, motion analysis diagrams, process diagrams, and facility layout improvements. The study incorporates lean management methods to identify error sources, analyze error types, rectify system errors, and eliminate mistakes. Trend control of normal error variation on grinding machines confirms a normal error of ±0.75μm. The process capability requirement is ±3.0μm, and for production sampling inspection of ±5.0μm, quality assurance personnel and production technicians immediately request a production halt to confirm the elimination of error sources. The improvement results show a 40% increase in production efficiency. After the improvement, production enters a growth phase along the learning curve, leading to expected production increases. Additionally, filtering the cutting fluid reduces scratching to near zero.