從飛航組員自願報告探究疲勞發生的潛在原因

Abstract

做好疲勞管理是航空業界提升飛航安全的重要趨勢，專家、學者對駕艙飛航組員的疲勞管理相關研究亦已不勝枚舉。惟其研究資料來源多為問卷、訪談及生理監測儀器資料等。因資料提供者都是在已知參與研究的狀態下，故較具主觀成份。本研究的資料來自樣本航空公司飛航組員的自願報告資料庫，資料蒐集自2010至2014共五年期間。因報告提報性質屬自發與無償，且提報者事先不知道報告內容會成為未來的研究資料，故較具客觀性。在研究方法上，本研究應用資料庫知識發掘流程與決策樹模組演算法來分析資料。最後探勘出十二項樣本航空公司各機隊飛航任務模式潛在疲勞發生原因的分類規則，並據以提出管理意涵。本研究結果可讓樣本航空公司在規劃未來飛航組員任務時，消弭或減緩飛航組員疲勞發生。

Well implementation of the fatigue management has become the important trend of flight safety improvement in the aviation industry. Experts and scholars have completed numerous of researches related to fatigue management of cockpit flight crewmembers. However, the data sources of their researches are questionnaires, interviews and records of physiological monitoring equipment mostly. Besides, the data has more subjective elements for those sample providers knowing that they are being part of the research. In this research, data source is the database of flight crewmembers’ voluntary reports of sample airlines with 2010-2014, 5 years period of time. Those data are more objective as the reports are voluntary and gratuitous. Furthermore, the reporters are never known that their reports will be the future research data. Regarding the research methods, this study applies the knowledge discovery in database processes and decision tree algorithm modules. Eventually explored twelve classification rules of potential fatigue causes in all fleets type flight patterns of sample airlines and recommend six management significations accordingly. The results of this study enables sample airlines could reduce or eliminate occurrence rate of fatigue when design future flight duty patterns.