以績效評估建構建築設備維護資料倉儲系統之研究

Abstract

近年來建築設施的維護管理逐漸受到社會重視，設備運轉效率也會反映在維護成本效益上，而建築設備的維護績效須兼顧經濟性與機能性等層面，但背後牽涉的影響因素往往過於複雜且衡量不易，考驗維護管理者的決策判斷，加上實務上的維護過程多未能有效記錄與探討，故本研究以績效評估為基礎來建構決策支援系統模式，應用資料倉儲（Data Warehouse, DW）達到決策查詢導向（Decision-oriented）的目的。 蒐集調查建築設備適用的維護績效指標(Performance indicators)與影響因子(Impact factors)，以及維護管理者關心的資料類型。建構日常維護作業的資料庫，也就是線上交易處理資料庫（On-Line Transaction Process, OLTP），如此可透過資料萃取轉換載入（Extraction Transformation Loading, ETL）等步驟，將線上交易處理資料庫內的資料載入到以星狀綱要（Star schema）及雪花狀綱要（Snowflake schema）為其設計架構的資料倉儲中，並建立多維度（Multidimension） 的資料方體（Data cube）， 再利用線上分析處理（On-Line Analytical Process, OLAP）技術，依決策需求快速彈性地取得所需的資訊。並可透過MS-Excel中的樞紐分析表功能，將資料利用前端工具瀏覽的視覺化（Visualization）方式呈現資訊，以提昇決策品質及時效。 維護管理者可依據所關心之事實(Fact)及觀看資料的角度(View) ，以多維的方式瀏覽資料，整理維護成本及設備效能這兩大類型績效指標，觀察各影響因子對維護績效指標值的變化，並對關鍵影響因子採取應對策略。透過本研究，與建築設備維護相關之績效評估模式及資料倉儲決策支援系統將被建立，此成果提供業界在系統設計時能有一參考模型，以縮短系統建置時間，並改善維護效率與經濟性。

Recently maintenance management of facility is more important to society; the equipment working affect cost-benefit about maintenance. But maintenance performance on building equipment must consider economic and capability. The impact factor is not only complex but also diffical to measure as a rule. It is a challenge for maintenance manager. And that most maintenance process is unable to record and analyze. The purpose of this research is to support “Decision-oriented” query using “Data Warehouse”, and build “Decision Support System” based on “Performance Evaluation”. Collection performance indicators apply to buiding equipment, as well as data type for decision makers care. Building database of day-to-day maintenance work. It is called “OLTP” (On-Line Transaction Process) that make data to “ETL”(Extraction Transformation Loading). Converting and loading data from database of OLTP to data cube of multidimension deciding on “Star schema” or “Snowflake schema”. Than we can Using “OLAP”(On-Line Analytical Process) technology to get real time “Decision-oriented” data flexibility. Finally user interface can make more visualization on data browsing, that using Pivot Table function of MS-Excel to advance quality and efficiency of decision. Maintenance decision maker can browse data based on Multidimension mode according as fact and view. Collating type of performance indicators for maintenance cost and equipment efficiency. Observing the affected variation for impact factors to maintenance performance. And than we can adopt action aimed the important impact factors. This research will building DSS of “Data Warehouse” for building equipment maintenance based on “Performance Evaluation” mode. The effort provide to design system for reference on actual situation. So that it can abridge time on building system, and improve on equipment economic and capability.