科技部新聞稿

科研助攻運動產業—智聯網羽球場打造AI個人化貼身教練

日期：109年6月23日

發稿單位：人文及社會科學研究發展司

聯絡人：吳淑真

電話：(02)2737-7443

  E-mail：jenywu @nstc.gov.tw

您是否曾夢想和「羽球天后」戴姿頴、「臺灣一哥」周天成成為您的球友，甚至成為您的貼身教練，教您舉拍或是讓對手招架不住的完美殺球？有了人工智能助攻，選手在運動時就能分析揮拍軌跡和步法，還能偵測疲勞指數，不僅擁有科學且精準的運動訓練，還有全方位無所不在的教練紀錄完整的歷程！

科技部自107年起推動「精準運動科學研究專案計畫」，其中由國立成功大學體育健康與休閒研究所蔡佳良特聘教授所組成的跨校跨域研究團隊，針對羽球運動項目進行「AIOT智能羽球訓練模式」研發，執行團隊成員包括成功大學電機系與人工智能數位轉型研究中心王振興特聘教授、物理治療學系蔡一如教授，以及逢甲大學自動控制工程學系許煜亮教授等相關專長領域學者共同執行研究。

「AIOT智能羽球訓練模式」主要針對羽球運動項目的訓練歷程進行創新技術研究及產品化開發，整合各種新穎科技：智聯網羽球拍、智能手錶、AI影像辨識技術、智聯網技術與雲端大數據分析平台，開發全臺首創「智聯網羽球場整體解決方案」。此方案能協助球隊訓練數位轉型，羽球拍內建感測器能偵測揮拍動作訊號，精準分析各種球路歷程，並即時捕捉記錄球員在場上的肢體運動軌跡，透過AI雲端收集數據與分析演算，可提供教練與選手適性化的訓練建議與完整的運動表現指標分析！

本項研發成果已顛覆傳統球隊僅能依靠經驗值的訓練模式，以智聯網羽球場導入AI人工智能與物聯網，完整記錄運動員的訓練歷程、生活型態、心理素質及競賽表現，數據收集後由AI運算引擎再將分析結果回饋給球員，球員使用「UCoach」運動教練APP就能觀看運動指標分析結果，並為自己訂製客製化訓練課程。球員還可以搭配智能手錶，記錄睡眠時間數據、慢性疲勞累積程度等，達到運動科學化與精準化的訓練模式。

為持續優化智聯網羽球場數據分析，計畫執行團隊已與臺南市新豐高中密切合作並建置實驗場域，利用高精準光學系統同步作業，做為球拍參數、選手場上運動三維姿態分析的黃金驗證標準，並使用認知神經科學心理與生理訊號，做為選手訓練疲勞分析的黃金驗證標準。未來，智聯網羽球場預計於今(109)年底於國家訓練中心正式啟用，期望在訓練科學化與適性化基礎上，以更有效率的訓練方式幫助我國選手登峰造極，在奧運比賽再創佳績。

執行團隊自主研發之技術已進入商品化進程，與「勝利體育事業股份有限公司」簽訂合作備忘錄，並已成功將智聯網羽球拍感測模組整合進其高階球拍系列產品中(如附圖)。此外，並將智聯網羽球拍捐贈給新豐高中羽球隊所有球員，進行場域訓練歷程收案與實證研究。

未來，計畫團隊也將積極與國際體育及醫療團隊合作，除將研發之成果推廣至國內國手培育的場域，精進AI精準運動科技應用，提升運動訓練表現外，並將以關鍵技術轉化產品進行商轉，或技轉國內外品牌公司，藉此達到技術研發落實應用於產業及創造營運商機。

研究成果聯絡人：

蔡佳良特聘教授 

國立成功大學體育健康與休閒研究所

電話：06-2757575#81809 

Email：andytsai@mail.ncku.edu.tw

Press Release

June 23rd, 2020

Scientific Research Assists Sports Industry:

AIoT creates individual AI coach

Have you ever dreamed about playing badminton with Tzu-ying Tai, the international badminton queen, and Tien-Chen Chou, the badminton prince in Taiwan, or having them as your private coaches to teach you how to perform perfect and admirable skills? The dream may sound unachievable. However, with the help of AI (artificial intelligence)-assisted system, the dream will come true by high-tech analysis of swing trajectory and foodworks and the monitoring of fatigue indices when you are playing the sport. Such AI-based technologies and devices will help you improve your  badminton skills scientifically and precisely, and become your “ubiqutious coach” via recording the whole training process.

The Department of Humanities and Social Sciences of the Ministry of Science and Technology in Taiwan promoted the project of “Precise Sports Science” in 2018. Dr. Chia-Liang Tsai, a principal investigator and a distinguished professor at National Cheng Kung University (NCKU), invited three professors, Dr. Jeen-Shing Wang at Department of Electrical Engineering of NCKU, Dr. Yi-Ju Tsai at Department of Physical Therapy of NCKU, and Dr. Yu-Liang Hsu at Department of Automatic Control Engineering of Feng Chia University, to organize an interdisciplinary and cross-school research group aiming to develop professional AIoT badminton training model. The innovative technologies and product development in this project focus on recording the badminton training process, which integrates diverse cutting edge technologies, including the intelligent badminton racket, the smart watch, the AI image recognition and AIoT technologies, and the cloud-based big data platform, to develop the first “AIoT badminton court overall solution” in Taiwan. The solution can improve the badminton training with digital transformation in which the built-in sensors can monitor the signals of swing motion, analyze every badminton techniques precisely, and catch the motion trajectory of the players’ upper and lower limbs when they enter into the badminton court. The data uploaded into a cloud-based big data platform will be calculated and analyzed through an AI technology, providing very essential sport performance index and training suggestions for coaches and players.

The new products generated by the research group are an important breakthrough, very different from the traditional experience-based training modes. With the aid of these high-tech products, the players are able to know the individual’s sport index via UCoach APP and to order a customized training program. In addition, coupled with a smart watch, the players can also have the data of their sleep quality and the levels of  chronic fatigue. Through the above-mentioned precious information, players can obtain scientific and precise exercise training modes.

The research group is also collaborated with National Hsin-Feng Senior High School to build an experimental field to optimize the data collected from the AIoT badminton court. The golden standards of the badminton racket parameters, 3-D posture analysis, and players’ fatigue levels are set with high-precise optical systems and psychophysiological signals, respectively. At present, the experimental participants are the players from the badminton team at the National Hsin-Feng Senior High School. In order to optimize the data, the coaches are also invited to be members of the advisory committee. In the future, the AIoT badminton court system will be established in the National Sports Training Center at the end of the year to help the professional players with a more efficient training.

The commercialization of the self-developed technology has been working in progress with Victor Badminton Company, one of the two most famous badminton brands in the world, which has signed the Memorandum of Understanding with the NCKU research group. The sense module developed by the research group has been successfully integrated into the Victor high-class badminton racket (see Figure) and the patent application is working in progress. Moreover, in order to cultivate the players with badminton talents with the advanced deployment, the Victor Badminton Company will donate AIoT badminton rackets to the players in the badminton team of National Hsin-Feng Senior High School in June to enhance the experimental results gaining from the academic-industry cooperation.

In sum, the research group contributes to developing a AIoT and big-data collecting system for recording training/life history of badminton players continuously. The main purposes are to enhance the training quality and sport performances in badminton. Also, the core technologies developed and derived from the research will be transferred to the domestic industries in Taiwan to increase the competitive advantages of sports products in the international markets. In the future, the research group will also be proactive in collaborating with international sports teams to advance the application of the AI-based sports technology and further enhance the players’ sports performance. To commercialize the new products and transfer technology to the domestic or foreign corporations for creating the business opportunities are also the direction of the research group’s efforts.

Media Contact

Prof. Chia Liang Tsai

Distinguished Professor

Institute of Physical Education, Health and Leisure Studies, National Cheng Kung University.

+886-6-2757575#81809

Email: andytsai@mail.ncku.edu.tw

Jenny Wu

Department of Humanities and Social Sciences, MOST

+886-2737-7443

Email:jenywu@nstc.gov.tw