臺灣地區2.8萬座橋梁之檢測方式以目視檢測方式為主。近年來，國內外應用無人飛行載具(Unmanned Aerial Vehicle, UAV)於現地勘災或近遠景拍攝等高空作業，從未使用自動飛行之UAV進行橋梁檢測。因此，本研究嘗試以自動飛行之UAV代替檢測人員至危險地域，或不易到達處進行橋梁檢測，並比較其與傳統橋梁檢測之優劣點。 本研究先由文獻回顧及專家訪談，了解目前UAV發展現況與應用範圍後，進行UAV之組裝、測試其自動化飛行與應用上之特性。其中，UAV自動化飛行能力係由各零件所附之程式撰寫而成。本研究發現以自動飛行之UAV進行現地橋梁檢測時，除適合無經驗的橋檢人員使用外，在時間、人力、成本、安全性、便利性與影像品質更勝於或等同於傳統工具之檢測結果；其缺點僅為定期檢測21個構件中僅可擷取19個構件之影像。另外，使用此自動化UAV進行橋梁檢測時，則須避免橋面版寬大於60公尺、橋墩高小於5公尺、以及飛行高程超過100公尺或負20公尺之橋梁。 本研究依前述條件自臺灣地區橋梁管理資訊系統中，統計出適合使用UAV進行橋梁檢測之公路橋梁數高達4,807座。本研究之成果，可供各橋梁管理單位於辦理橋梁檢測工作時之參考，以提升橋梁檢測成果之品質。 ;There are twenty eight thousands of bridges in Taiwan which are mainly inspected visually with hand-held devices. Recently, popular unmanned aerial vehicle (UAV) have been used in many places such as elevated shooting of images for disastrous areas. However, UAV with automated route-flying capabilities have not yet been used in bridge inspection. Thus, this research explores the feasibility using such UAV equipped with cameras in dangerous or hard to reach areas where bridges are located, and compares the quality differences of photos that are taken by inspectors.
This research firstly obtained relevant knowledge regarding current status of UAV through literature review and expert interviews. A few UAV were then assembled using various commercially available parts or hardware to test their abilities in automated route-flying for bridge inspection. The automated route-fly capabilities of UAV were specially programmed using available computer codes accompanied with the hardware. It is found that such UAV are suitable for inexperienced bridge inspectors to obtain images of bridge components with many advantages; better or no less than that of traditional inspection, in time, labor, cost, and safety aspects. The only deficiency is that the UAV only can capture 19 out of 21 bridge components. Notably, using such UAV for bridge inspection should avoid bridges wider than 60 meters, piers shorter than 5 meters, or 100 meters above or 20 meters below the bridge to maintain normal operation of the UAV.
Based on the aforementioned criteria, there are 4,807 bridges in the Taiwan Bridge Management System (TBMS) that are suitable for using such automated UAV for bridge inspection. Results of this research can serve as useful references for bridge management agencies in Taiwan in planning bridge inspection tasks so as to improve the quality of bridge inspection.
