透過您的圖書館登入
IP:3.237.24.82
  • 學位論文

擴增實境輔助之遠端吊車系統及其介面設計

Development of an AR-Enhanced Tele-Operation Crane and User Interface

指導教授 : 康仕仲

摘要


為增加吊裝機具的操作性,降低工程施工時產生安全危害的機率,並且延伸應用於特殊吊裝情境中,本研究發展改善吊裝行為及環境感知能力的遠端控制系統,其重點為整合吊裝系統硬體、遠端操作設備與環境的整合平台,並利用實際吊裝情境的使用性測試與驗證以證明其實際應用的可能。 依據研究中提出之吊車遠端操控架構,本研究整合工業用機械手臂(KUKA CR 16)做為吊裝機具的實驗機體,配合分別架設於手臂左右兩側、上方及遠處的攝影機(Camera),將吊裝實驗場地的影像傳回給遠端操作系統,該系統介面以多螢幕排列方式呈現回傳的影像,提供操作者多角度的吊裝資訊,並整合擴增實境(Augmented Reality)及運動規劃(Motion Planning),進行吊裝物定位且計算虛擬的吊裝路徑,並呈現於場地的擷取影像中,讓使用者可依這些資料進行遠端吊裝作業的操作。經過使用性測試證實多視角與AR資訊的整合有其效益,受測者多認為AR整合是值得發展的方向。 除了整合擴增實境與多攝影機視角而完成的系統硬體外,系統介面部分依據使用性測試做出改良,提出以操作者注意力為基礎之操作介面,將介面資訊分為三種形式:持續注意型(Focused View)、即時參照型(Ambient View)與警示型(Alert View),分別以不同方式來呈現,利用整合之擴增實境技術加以顯示,讓使用者可依這些資料進行遠端吊裝作業的操作,更有效地完成指定的模擬吊裝任務,該介面亦已於使用性測試中證實其有助於減低操作手發生碰撞錯誤的機率,並且證明本研究的資訊分類及顯示方式有其效果,可引導操作者完成吊裝任務。 本研究亦進行了實際吊車手的使用性測試,藉以了解系統導入的可行性及需要注意的實際考量,吊車手的測試完成時間普遍較一般受測者短且失誤發生較少,並給予正面評價。針對吊裝路徑資訊的正確性及顯示方式,亦提出修改建議,根據這些意見,本研究實作改良型的吊裝路徑演算法,除了安全(Safety)、效率(Efficiency)外,亦考量到可操作性(Operation Feasibility),針對吊車運動模式及吊裝習慣修改吊裝路徑的指導,共考量了三個面相:可視性(Visibility)、駕駛特性(Control Feature)與動態環境(Dynamic Environment),該演算法已整合於一虛擬吊裝環境內,並驗證其路徑產生具有即時性與可行性,在未來得以整合進本研究中的遠端操作系統中,達到AR輔助吊裝功能的最大效益。

並列摘要


This research focuses on the development of a tele-operated crane to improve the safety and efficiency of construction rigging operations, through multi-disciplinary research collaboration. The multi-disciplinary technologies integrated in the system involve the tele-operation of construction cranes, multiple views, augmented reality, attention-based user interface design, and automatic motion planning. Following the system architecture for a tele-operated crane proposed in this research, a laboratory crane is implemented, including both hardware and software components. An industrial robot arm is utilized as an experimental crane, named Zebra. By mounting cameras on the left, right, global, and top sides of this robot arm, environmental information around the arm can be retrieved and multiple viewing angles of rigging objects can be covered. In addition, to provide guidance to operators, this work utilizes Augmented Reality (AR) and Motion Planning technologies to generate a virtual path that is then superimposed on images from the cameras. With the guidance information and multiple display screens, participants attending a usability test for technology validation provided positive feedback and pointed out research directions that are worth pursuing. In addition to the AR and multiple views integrated in the Zebra system, a method is also provided for classifying crane operation information into three types of views: focused view, ambient view, and alert view, which are displayed in the user interface according to human attention features for the tele-operators. According to the results of the usability test conducted for this attention-based user interface, the possibility of collision during rigging activities can be decreased using the proposed user interface, and the information classification strategy used for assisting tele-operations can effectively reduce the information load on operators while maintaining operating efficiency. By conducting the usability test with actual crane operators, the feasibility of implementing the proposed interface in the field is demonstrated. Most crane operators completed the test tasks with fewer errors than students did. They also provided positive feedback and suggestions, especially on augmented rigging paths. Based on these suggestions, the algorithms for generating rigging paths can be improved by considering three feasibility factors: Visibility, Control Features, and Dynamic Environment. A virtual environment, named PathGuider, for evaluating the efficiency of the developed algorithms is developed. This proved that the developed algorithms can generate rigging paths in real-time for users to make efficient decisions.

參考文獻


Kuan, C. P. and Young K. Y. (2003). "Challenges in VR-Based Robot Teleoperation," Proceeding of IEEE International Conference on Robotics and Automation, Taipei, Taiwan, Sep. 14-19.
Miller, T. and Johnson, D. (2010). XNA Game Studio 4.0 Programming, Addison, Wesley, Boston.
Ali, M. S., Babu, N. R. and Varghese, K. (2005). "Collision Free Path Planning of Cooperative Crane Manipulators Using Genetic Algorithm," Journal of Computing in Civil Engineering, 19(2), pp. 182-193.
Al-Hussein, M., Alkass, S. and Moselhi, O. (2005). "Optimization Algorithm for Selection and on Site Location of Mobile Cranes," Journal of Construction Engineering and Management, 131(5), pp. 579-590.
Azuma, R. T. (1997). "A Survey of Augmented Reality," Teleoperators and Virtual Environments, 6(4), pp. 355-385.

延伸閱讀