PolyU｜AI combines biomechanics to improve swimming skills. Sensors provide precise feedback to help improve training movements

[ad_1]

Contents

The stroke speed and head turning frequency are clear at a glance Electromyography analysis optimizes take-off and entry posture Support EJ Tech

originalPublished in Hong Kong Economic Times Financial NewsEJ Tech Innovation Lab“

The Paris Olympics in France will kick off at the end of this month. The Hong Kong swimming team has won a total of 9 tickets this year. Swimming is one of the Hong Kong team’s strengths, and with the latest technology to assist training, it is believed that it will help the athletes achieve better results. The Hong Kong Polytechnic UniversitySports Technology ResearchBased on various sensor data, the team developed a biomechanical training method that combines artificial BroadCast Unitedligence (AI) technology.cooperate,Co-constructionIn addition to the AI Swimming Technology Laboratory, the system has also been adopted by some Hong Kong team representatives.

Core member of PolyU Sports Science and Technology Research Institute, Assistant Professor of Department of Rehabilitation SciencesSu JunlongIn an exclusive interview with our newspaper, he revealed that he was a swimmer and also worked as a swimming coach. Looking back on his past training and teaching experience, he said that the biggest difficulty was that he could not give athletes accurate feedback, and it mainly depended on the other party’s understanding.

Su Junlong mentioned that when the team developed the AI system, they referred to the standards of Olympic athletes. (Photo by Huang Junyao)

The stroke speed and head turning frequency are clear at a glance

The system developed by Su Junlong’s team only requires a square sensor to be placed between the swimming cap and the back of the athlete’s head. The system algorithm can obtain various swimming data, including stroke length, stroke frequency, turning acceleration, breathing and head turning frequency, as well as segment speed, that is, the athlete’s speed distribution every 50 meters.

For athletes, a slight difference in swimming style can affect their performance. Su Junlong explained that in theory, the higher the stroke frequency, the better, but the longer the arm, the longer the path, and the longer the time consumed; in other words, the appropriate stroke range is far more important than the frequency. Su Junlong continued that in order to reduce resistance when swimming at high speed, the frequency of turning the head to breathe is also a certain consideration.

The sensor can display the athlete’s heart rate during swimming training. (Photo provided by the interviewee)

The “world” inside the swimming cap is not limited to this. An optical sensor will be placed at the athlete’s temple. Its principle and appearance are similar to the bottom component of the Apple Watch. It can display the heart rate during swimming training, reflecting whether the athlete has successfully entered the state, whether he has overtrained, etc. Su Junlong pointed out that the optical sensor is lightweight and is pressed against the athlete’s blood vessels by the swimming cap, so the accuracy is quite good.

The team attached the underwater surface electromyography acquisition system to different muscle parts to observe their movements. (Photo provided by PolyU)

Electromyography analysis optimizes take-off and entry posture

Su Junlong listed three “appropriate” factors, namely: appropriate time, appropriate angle, and appropriate muscles, indicating that it is not enough to rely on the coach to observe the athletes by the pool. The team attached the underwater surface electromyography (Surface Electromyography) acquisition system to different muscle parts of the athletes to observe their movements.

In addition, electromyography analysis is not only applicable to traditional sports, but also to lifesaving, which is a “Class B” elite sport. He explained that many athletes reported that some parts of their body were particularly sore when dragging a dummy. After electromyography analysis, it was found that certain muscles were used most frequently during exercise, providing a rigorous basis for coaches to train. If the program is combined with video analysis, it can also provide data such as take-off reaction, entry angle and speed in real time to help athletes improve their take-off posture.

Su Junlong shows the hand-worn force sensor. He believes that the Hong Kong-made system is suitable for the physique of Hong Kong people and the data analysis results are more accurate. (Photo by Huang Junyao)

As for another swimming force sensor worn on the hand, it can measure the athlete’s water resistance, hand movement trajectory, frequency and force, and observe the propulsion during exercise. Su Junlong revealed that the team is also launching a physiological study on oxygen consumption. The three aspects of swimming biomechanics, swimming physiology and swimming muscle activity analysis are crucial to understanding how to maintain the correct use of muscles during high-intensity training.

Interview and article by: Zhou Yongtong