開門與穿越門對於本研究所開發之服務型機器人是一個必須的動作。本研究的展
示情境是將機器人引導至實驗室的門口附近，透過標籤輔助執行開門動作，移動到桌
子上擷取有標籤的茶杯，再開門離開實驗室，沿走道移動到飲水機裝水，再將盛好水
之茶杯送回實驗室原來放杯子位置。在移動途中利用雙目視覺攝影機及超音波感測障
礙物，判斷是否必須閃避並觀察是否能通過。
本論文服務型機器人所使用的技術有 Aruco 標籤輔助定位、視覺深度學習及物體
辨識，物體辨識之方法為 YOLO 辨識，底部移動載具採用麥克納姆輪並透過 CAN bus
通訊協定傳輸使得機器人能夠全方位移動。
執行系統使用筆記型電腦作為系統核心，機器人上搭載一隻六軸機器手臂，手臂
前端裝有一顆攝影機並透過攝影機尋找標籤，機器人前方也架設一套自製的雙目深度
視覺攝影機，用以計算物體深度及辨識物體避開障礙，機器人四個邊分別有超音波感
測模組用來即時避開障礙。
本研究可以展示前述情境，並提出一套拉門策略，可適用於比較窄的空間，利用
機器人的直線運動來拉開無反彈能力的門，及開門後觀察及評估是否能進入，以及進
入後是否可以順利離開此空間的判斷方式，並實現完成。

Opening the door and passing through the door is a necessary action for the service robot developed in this research. The demonstration scenario of this study is to guide the robot near the entrance of the laboratory, open the door with the aid of a label, move to the table to get the labeled cup, open the door and leave the lab, move along the aisle to find the water dispenser, and fill the cup with water by the water dispenser, and return the cup with water to the original location in the lab. Binocular depth vision cameras and ultrasound are used to detect obstacles while on the move, to determine if it must dodge and to observe if it can pass.
The tools used in this research for the service robot are Aruco tag-assisted positioning,visual deep learning and object recognition. The method of object recognition is YOLO. The mobile carrier is a platform with four Mecanum wheels to support all-round mobility of the robot. The carrier’s controller supports CAN bus protocol and responses to commands from the host computer. .
A laptop computer serves as the core of the host system. The robot is equipped with a six-axis robot arm. The label is searched by the camera mounted on the arm. A 3D vision system calculates the depth of the detected object, identify the object, and check if the object is a to-avoid obstacle. The hardware of the 3D vision system is a homemade binocular depth vision camera in front of the robot. The robot has ultrasonic sensors on each of the four sides to avoid hitting obstacles in real time.
This study proposes a set of strategies for pulling a door open in a relatively narrow space, observing and evaluating the accessibility after opening the door and whether it is possible to leave this space after entering.

[1]International Federation of Robotics, "This is According to World Robotics 2021 – Service Robots,"2021.[online].Available:https://ifr.org/news/service-robots-hit-double-digit-growth-worldwide [Accessed: Dec.20, 2021].
[2]H.Gao et al., "Dynamic Modeling and Experimental Validation of Door-Opening Process by a Mobile Manipulator," in IEEE Access, vol. 7,2019, pp. 80916-80927, doi: 10.1109/ACCESS.2019.2919964.
[3]F. Endres, J. Trinkle and W. Burgard, "Learning the Dynamics of Doors for Robotic Manipulation," in IEEE/RSJ International Conference on Intelligen Intelligent Robots and Systems, 2013, pp. 3543-3549, doi:10.1109 10.1109/IROS.2013.6696861.
[4]陳政國，自動化奉茶機器人，國立成功大學工程科學系碩士論文，2015。
[5]Romero-Ramirez, Francisco J., Rafael Muñoz-Salinas, and Rafael Medina-Carnicer. "Speeded up detection of squared fiducial markers." Image and vision Computing 76 (2018): 38-47
[6]HONG CHIANG, "BellaBot,"[online]. Available: https://www.hong chiang.com.tw/zh-TW/category/Bella-/K01.html [Accessed: Mar.28, 2022].
[7]Toyota, "Human Support Robot,"[online]. Available: https://mag.toyota.co.uk/toyota-human-support-robot [Accessed: Mar.20, 2022].
[8]H. Su and K. Chen, "Design and Implementation of a Mobile Robot with Autonomous Door Opening Ability," in IEEE International Conference on Fuzzy Theory and Its Applications (iFUZZY), 2017, pp. 1-6, doi: 10.1109/iFUZZY.2017.8311808.
[9]S. Kobayashi et al., "Development of a Door Opening System on Rescue Robot for Search“UMRS-2007”," in IEEE SICE Annual Conference, 2008, pp. 2062-2065, doi: 10.1109/SICE.2008.4655001.
[10] B. Gaspers, J. Welle and D. Schulz, "Opening Doors with a Mobile Manipulator Without Force-torque Feedback," in IEEE International Symposium on Safety, Security, and Rescue Robotics (SSRR), 2013, pp. 1-6, doi: 10.1109/SSRR.2013.6719354.
[11] S. Ahmad, H. Zhang and G. Liu, "Multiple Working Mode Control of Door-Opening With a Mobile Modular and Reconfigurable Robot," in IEEE/ASME Transactions on Mechatronics, vol. 18, no. 3, pp. 833-844, 2013, doi: 10.1109/TMECH.2012.2191301.
[12] Boston Dynamics, "Spot Arm,"[online].Available: https://www.bostondynamics.com/products/spot/arm [Accessed: Mar.20, 2022].
[13] M. Stuede, K. Nuelle, S. Tappe and T. Ortmaier, "Door Opening and Traversal with an Industrial Cartesian Impedance Controlled Mobile Robot," in IEEE International Conference on Robotics and Automation (ICRA), 2019, pp. 966-972, doi: 10.1109/ICRA.2019.8793866.
[14] 洪瑋呈，麥卡倫輪型移動機器人之路徑追蹤控制，國立成功大學工程科學系碩士論文，2020。
[15] 企誠自動控制股份有限公司，"HS30光學編碼器"，[online]. Available: http://tw.honestsensor.com.tw/pp-02.htm [Accessed: Jan.28 2022].
[16] 維基百科，"CAN bus"，[online].Available: https://en.wikipedia.org/wiki/CAN_bus[Accessed: Jan.21, 2022].
[17] DFrobot,"CAN-BUS Shield V2,[online].Available:https://wiki.dfrobot.com/CAN-BUS_Shield_V2__SKU__DFR0370_#More [Accessed: Jan.6, 2021].
[18] 劉昇宏，六自由度機器手臂設計實作與軌跡規劃，國立成功大學工程科學系碩士在職專班碩士論文，2020。
[19]CEILING TSAI,"HC-SR04,"[online].Available: https://atceiling.blogspot.com/2017/03/arduino_28.html [Accessed: Oct.10, 2021].
[20] 維基百科，"3D列 印"，[online].Available: https://zh.wikipedia.org/wiki/3D%E6%89%93%E5%8D%B0 [Accessed: Feb.10, 2021].
[21] 林聖源，基於立體視覺及深度學習之人員跟隨及避障系統設計，國立成功大學工程科學系碩士在職專班碩士論文，2022.
[22] OpenCV,"Detection of ArUco Boards,"[online].Available: https://docs.opencv.org/4.x/db/da9/tutorial_aruco_board_detection.html [Accessed: Jan.21, 2022].
[23] Maxon,"539487+166933,"[online].Available:
https://www.maxongroup.com.tw/maxon/view/product/motor/ecmotor/
EC-i/539487[Accessed: Jan.21, 2022].
[24] CSDN, "Aruco tvec and rvec," [online].Available: https://blog.csdn.net/dgut_guangdian/article/details/108093643[Accessed:Feb.5, 2022].