2008-08-012024-05-18https://scholars.lib.ntu.edu.tw/handle/123456789/706994摘要：台灣地區每年需進行養護及維修（maintenance and rehabilitation, M&R）之道路面積平均超過一億七千萬平方公尺，為能維持既有鋪面的服務能力（serviceability），各道路主管機關均需針對道路鋪面的現況進行定期檢測、監測與評估，適時施予適當的M&R策略，以延長道路鋪面的使用壽年。然而，大量檢測的需求會耗費相當人力與物力，然大部分的施測作業皆是一成不變的慣常程序，若能以自動化的方式進行，除能減少人物力資源的耗費外，透過自動控制之變動性較人為操作低的特性，也會無形間增加檢測結果的可靠度。本整合型計畫即配合國家營建自動化政策的發展方向，結合現今全球蓬勃發展的機器人技術，針對所有公共工程中幅原最廣大、最不具規則特性且維護管理不易的鋪面工程，發展自主式機器人於鋪面「縱向平坦度（roughness）」與「表面破損（distress）」兩項最重要特性的智慧化檢、監測技術，以提昇國內鋪面工程的水準。 整體計畫主要包含三個子計畫：（1）機器人模擬與資訊系統整合：目標是發展機器人與鋪面檢測系統間的整合平台，並發展道路檢測用機器人之行為模擬與控制平台；（2）鋪面檢測自動化暨機器人檢測模式發展與驗證：針對目前檢測作業進行檢測自動化之規劃研究，並以人機介面（Human-Computer Interaction, HCI）研析現有國內外鋪面管理系統（Pavement Management System, PMS），並設計發展適於台灣地區環境特性之PMS，最終發展以機器人為考量之鋪面檢測模式並進行驗證與比較，及檢測資料探勘分析；（3）道路檢測用輪型機器人雛型載具及演算法研發：主要進行機器人載具機構的設計與研發，並開發機器人智慧型路徑規劃演算法。 整體計畫將以學術研究機構間團隊合作及跨領域整合方式，透過道路工程、電腦輔助工程、人工智慧、機器人學等不同領域專長之研究人員的整合，達到本整合型計畫發展機器人智慧化檢測系統之目標，並藉此整合高科技技術研發與傳統營建產業，除可促使產業升級，亦可增進營建自動化人才培訓及應用推廣。<br> Abstract: The pavement areas need to be maintained and rehabilitated in Taiwan, including the surface of highways, bridges and local routes, reaches more than 170 million square meters. Each year, a great amount of resources is consumed on regular inspections, monitoring and evaluations, in order to properly maintain serviceability of the roads. However, most of the inspection and maintaining processes are routine work, requiring immense man-hours. Therefore, to have a fast and effective automated method on road inspection has become highly important. In this research, we propose to develop a robot-based pavement inspection method. By introducing robotics, sensors, and artificial intelligence, an autonomous robot will be created for pavement roughness test and distress test, two of most important indices in road serviceability. Three subprojects will be included in this research. One of the subprojects focuses on integrating multiple sensors and actuators to develop an intelligent robotic platform. Another subproject focuses on the technical problems on adapting existing equipment for pavement inspection to the robot. The other subproject will research the integration issues between the robot platform and inspection equipment. The expected research results include an autonomous robot with path planning intelligence to perform pavement inspections. Multiple simulation tests and a field test will be conducted to verify the whether the developed methods can be applied for engineering purposes.營建自動化鋪面檢測自主式機器人路徑規劃人工智慧construction automationpavement inspectionautonomous robotpath planningartificial intelligence