本研究旨在整合六軸機械手臂進棒球訓練領域中，主要針對棒球投球與拋
球訓練開發系統的設計與效能評估。在投球訓練系統中，訂定了一個標準化的
選擇前端影片的流程，並為後端機械手臂的應用定義了明確的運動學參數，從
而促進了開發和修改流程的精簡化。最後對於達明機械手臂的實作進行詳盡的
實驗，評估了其優勢和不足之處。
相較於傳統的實時投球訓練使用的運動捕捉裝置，本研究所提出的系統利
用姿勢辨識簡化了開發流程，省略了昂貴的設備，並減少了數據收集和處理的
時間。通過分析投球影片中的關節點角度，系統實現了在機械手臂協同操作下
高效的投球訓練。
為了解決硬體角度限制，本研究採取了角度量化技術。這種優化方法考慮
了手臂運動範圍的限制，確保其位於手臂的安全操作範圍內。隨後的驗證階段
比較了原始角度配置和優化角度配置之間的位置錯誤率，兩者相比有著百分之
76.5 的明顯進步。
此外，本系統提供了第二種拋球訓練模式，可以根據不同身高區間的使用
者來調整發球機與使用者的站距，並提供外角、內角與中心三種訓練角度，來
強化訓練的多樣性與動作協調和準確性。隨後的驗證階段進行總共 60 次的重複
拋球，並計算每次拋球與預期落點的誤差，大約落在 1 到 2 公分之間。
為了提升出球速度，本研究設計了一個簡易的發射裝置安裝在末端控制器
上，這項裝置在水平方向速度上增加了顯著的百分之 38，來到了 2.16 公里每小
時。

This research integrates six axis robotic arms into baseball training, emphasizing the design and performance evaluation of a system for pitching and toss training.
The pitching system establishes a standardized process for video selection and defines
kinematic parameters for streamlined development. Detailed experiments assess the
strengths and weaknesses of the implemented Techman robotic arm.
The proposed system simplifies development compared to traditional motion capture in real-time pitching training. Utilizing pose recognition reduces data collection and
processing time, enabling efficient pitching training with the robotic arm’s cooperative
operation analyzing joint point angles from pitching videos.
To address hardware design constraints, this study employs angle quantization technology. This optimization method considers the limitations of the arm’s range of motion,
ensuring it stays within the safe operational range. The subsequent validation phase compares the positional error rates between the original and optimized angle configurations,
showing a significant improvement of 76.5%.
Additionally, the system offers toss training mode that adjusts the distance between
the ball machine and the user based on different height ranges. It provides three training
angles: outer corner, inner corner, and center, enhancing training diversity and movement coordination and accuracy. In the validation phase involving 60 repeated throws,
the system demonstrates an average error ranging from 1 to 2 centimeters from the expected feeding point.
The research introduces a simple ejection device mounted on the end effector to
increase ball speed. This device significantly increases the horizontal speed by 38%,
reaching 2.16 km/h.

[1] J. A. Potteiger, H. N. Williford Jr, D. L. Blessing, and J. Smidt, “Effect of two
training methods on improving baseball performance variables,” The Journal of
Strength & Conditioning Research, vol. 6, no. 1, pp. 2–6, 1992.
[2] S. D. Moore, T. L. Uhl, and W. B. Kibler, “Improvements in shoulder endurance
following a baseball-specific strengthening program in high school baseball players,” Sports Health, vol. 5, no. 3, pp. 233–238, 2013.
[3] S. Oyama, “Baseball pitching kinematics, joint loads, and injury prevention,” Journal of sport and health science, vol. 1, no. 2, pp. 80–91, 2012.
[4] C. J. Dillman, G. S. Fleisig, and J. R. Andrews, “Biomechanics of pitching with
emphasis upon shoulder kinematics,” Journal of Orthopaedic & Sports Physical
Therapy, vol. 18, no. 2, pp. 402–408, 1993.
[5] J. Zhao, B. Xie, and C. Song, “Generating human-like movements for robotic
arms,” Mechanism and Machine Theory, vol. 81, pp. 107–128, 2014.
[6] P. J. Fadde, “Interactive video training of perceptual decision-making in the sport
of baseball,” Technology, Instruction, Cognition and Learning, vol. 4, no. 3, pp.
265–285, 2006.
[7] B. L. Tripp, E. M. Yochem, and T. L. Uhl, “Functional fatigue and upper extremity sensorimotor system acuity in baseball athletes,” Journal of athletic training,
vol. 42, no. 1, p. 90, 2007.
73
[8] V. Patidar and R. Tiwari, “Survey of robotic arm and parameters,” in 2016 International conference on computer communication and informatics (ICCCI). IEEE,
2016, pp. 1–6.
[9] E. Van der Kruk and M. M. Reijne, “Accuracy of human motion capture systems
for sport applications; state-of-the-art review,” European journal of sport science,
vol. 18, no. 6, pp. 806–819, 2018.
[10] M. J. Axe, R. Wickham, and L. Snyder-Mackler, “Data-based interval throwing
programs for little league, high school, college, and professional baseball pitchers,”
Sports medicine and arthroscopy review, vol. 9, no. 1, pp. 24–34, 2001.
[11] H. Katsumata, K. Himi, T. Ino, K. Ogawa, and T. Matsumoto, “Coordination of hitting movement revealed in baseball tee-batting,” Journal of sports sciences, vol. 35,
no. 24, pp. 2468–2480, 2017.
[12] G. S. Fleisig, B. Bolt, D. Fortenbaugh, K. E. Wilk, and J. R. Andrews, “Biomechanical comparison of baseball pitching and long-toss: implications for training
and rehabilitation,” journal of orthopaedic & sports physical therapy, vol. 41, no. 5,
pp. 296–303, 2011.
[13] S. Liu, L. M. Ferris, S. Hilbig, E. Asamoa, J. L. LaRue, D. Lyon, K. Connolly,
N. Port, and L. G. Appelbaum, “Dynamic vision training transfers positively to
batting practice performance among collegiate baseball batters,” Psychology of
Sport and Exercise, vol. 51, p. 101759, 2020.
[14] R. F. Escamilla, G. S. Fleisig, S. W. Barrentine, N. Zheng, and J. R. Andrews,
“Kinematic comparisons of throwing different types of baseball pitches,” Journal
of Applied Biomechanics, vol. 14, no. 1, pp. 1–23, 1998.
[15] M. Lapinski, E. Berkson, T. Gill, M. Reinold, and J. A. Paradiso, “A distributed
wearable, wireless sensor system for evaluating professional baseball pitchers and
74
batters,” in 2009 International Symposium on Wearable Computers, 2009, pp. 131–
138.
[16] A. Kumar, P. S. Chavan, V. Sharatchandra, S. David, P. Kelly, and N. E. O’Connor,
“3d estimation and visualization of motion in a multicamera network for sports,”
in 2011 Irish Machine Vision and Image Processing Conference, 2011, pp. 15–19.
[17] J. Kim, M. Ra, H. Lee, J. Kim, and W.-Y. Kim, “Precise 3d baseball pitching trajectory estimation using multiple unsynchronized cameras,” IEEE Access, vol. 7,
pp. 166 463–166 475, 2019.
[18] A. Toshev and C. Szegedy, “Deeppose: Human pose estimation via deep neural
networks,” in Proceedings of the IEEE Conference on Computer Vision and Pattern
Recognition (CVPR), June 2014.
[19] Y.-C. Li, C.-T. Chang, C.-C. Cheng, and Y.-L. Huang, “Baseball swing pose estimation using openpose,” in 2021 IEEE International Conference on Robotics,
Automation and Artificial Intelligence (RAAI), 2021, pp. 6–9.
[20] T. Senoo, A. Namiki, and M. Ishikawa, “High-speed throwing motion based on
kinetic chain approach,” in 2008 IEEE/RSJ International Conference on Intelligent
Robots and Systems, 2008, pp. 3206–3211.
[21] U. Mettin, A. S. Shiriaev, L. B. Freidovich, and M. Sampei, “Optimal ball pitching with an underactuated model of a human arm,” in 2010 IEEE International
Conference on Robotics and Automation. IEEE, 2010, pp. 5009–5014.
[22] D. M. Lofaro, C. Sun, and P. Oh, “Humanoid pitching at a major league baseball
game: Challenges, approach, implementation and lessons learned,” in 2012 12th
IEEE-RAS International Conference on Humanoid Robots (Humanoids 2012),
2012, pp. 423–428.
75
[23] T. Senoo and I. Ishii, “Baseball robots based on sensory-motor integration,” in
2021 21st International Conference on Control, Automation and Systems (ICCAS),
2021, pp. 1772–1777.
[24] 阿 黑 棒 球 日 記, “Darvish pitching video.”
https://www.youtube.com/watch?v=MuUR2s3BaFI, 2023.
[25] T. R. INC., “Techman robot information.” https://www.tm-robot.com/zhhant/tm14/, 2023.
[26] 王清欉, “東, 西方職業棒球選手年齡體格之分析研究 (以美, 日, 台, 韓為例),”
政大體育研究, no. 16, pp. 97–112, 2004.