以機械手臂實現自動化取放任務及人機協作應用於模擬/現實環境

Automatic Pick-and-Place and Human-Robot Collaboration with Robotic Arm in Simulation/Real World

林哲民(Che-Ming Lin)

指導教授 ： 王學誠

國立交通大學/電機學院/電機與控制工程系所/碩士(2018年)

摘要

現今在一般傳統產業已經開始以機器逐漸取代人力之傾向，且在工廠中之生態逐漸講求統一工作效率以達到整體效率之提升，故藉由機器人之固定工作頻率來控制整體效率。已經有許多產業相關研究正在進行機器人產業之應用，然而卻經常侷限於實驗場域之大小而無法將其研究實際在工廠生產線中進行測試。本論文以此作為基礎分為三部分，依序為模擬生產線取放任務實現、現實取放任務實現、人機協作之應用。

透過模擬環境建立工廠之生產線場景並藉此場域進行整個產線規模之工件取放任務及搬運實驗，其中包含系統狀態流程設計、UR10機械手臂之操作、物件資訊監控及夾取姿態計算。由模擬之測試所得結果，本論文將其規模縮小並延伸實現於現實環境中，在實現取放及搬運任務添加以AprilTag辨識輔助物件之姿態偵測、UR5機械手臂及夾爪控制、AGV自動追跡，並將模擬測試中之系統流程簡化應用於現實中。在最後的實驗以現實機械手臂取放任務為基礎添加人機協作之安全應用，在人與機器共用空間時做出機械手臂安全停止機制後找尋新的路徑避開人而繼續運作。

關鍵字

機械手臂 ； 自動化 ； 模擬 ； 拾取

並列摘要

Nowadays, the tendency to replace the labor with machines has begun in the traditional industry, and the ecology in the factory gradually emphasizes unifying work efficiency to improve the overall efficiency. Therefore, the overall efficiency could be controlled by the fixed operating frequency of the robot. Since there've been many researches about industry being applied in robotics industry, but they are often limited to the size of the experimental field and couldn’t be tested in a real factory. In this thesis, it divided into three parts: Production line Pick-and-Place task in simulation, Pick-and-Place task realization in real world, and Human-Robot cooperation.

The factory's production line environment is established to used for Pick-and-Place task experiments through simulation system, which include: System state machine design, Manipulation of UR10 robotic arm , Object Information monitoring, and Grasp pose estimation. In this thesis, we reduce the scale of the results in simulation and extend it into the real environment. In Pick-and-Place and transport tasks, we add the detection of the AprilTag for object pose detection, UR5 robotic arm and gripper control, AGV automatic trail following, and also simplify the state machine of simulation for real world. In the final experiment, we add the safety application of human-robot cooperation based on the Pick-and-Place task in real world. When human and the robot share the same space, the robotic arm should immediately stop safely and then start planning a new path to avoid human and continue operating.

並列關鍵字

Robotic Arm ； Pick-and-Place ； Simulation

參考文獻

[1] W. Qian, Z. Xia, J. Xiong, Y. Gan, Y. Guo, S. Weng, H. Deng, Y. Hu, and J. Zhang, “Ma- nipulation task simulation using ros and gazebo,” in Robotics and Biomimetics (ROBIO), 2014 IEEE International Conference on. IEEE, 2014, pp. 2594–2598.

[2] S. Jotawar, M. Soni, and S. Kumar, “Motion planning for an automated pick and place robot in a retail warehouse,” in Proceedings of the Advances in Robotics. ACM, 2017, p. 40.

[3] U. Ozguner, K. Redmill, S. Biddlestone, M. F. Hsieh, A. Yazici, and C. Toth, “Simula- tion and testing environments for the darpa urban challenge,” in Vehicular Electronics and Safety, 2008. ICVES 2008. IEEE International Conference on. IEEE, 2008, pp. 222–226.

[4] T. Laue, K. Spiess, and T. Röfer, “Simrobot–a general physical robot simulator and its application in robocup,” in Robot Soccer World Cup. Springer, 2005, pp. 173–183.

[5] Y. Zhang, L. Li, M. Ripperger, J. Nicho, M. Veeraraghavan, and A. Fumagalli, “Gilbreth: A conveyor-belt based pick-and-sort industrial robotics application,” in 2018 Second IEEE International Conference on Robotic Computing (IRC). IEEE, 2018, pp. 17–24.