How AI and machine learning are helping athletes reach peak performance

Data and machine learning within sports and sports technology have been transformative. The integration of data and machine learning in sports has evolved from basic statistical analysis to advanced predictive models, supporting data-driven decision-making.  

This evolution impacts how athletes train, how teams strategise, and how fans engage with their favourite sports through personalised content and experiences.  

For example, using computer vision and wearable sensors, we can track football players’ positions, monitor their movements for quantitative performance analysis, and analyse game strategy and tactics.  

Tracking data on player loads can help prevent overuse injuries. AI-driven technology also provides fans with personalised content, including statistics, recommendations and automatically extracted highlights.  

Additionally, we work with the football industry to carry out data analysis that helps identify talent based on performance metrics and potential.

Wearable devices can track various metrics such as speed, acceleration, distance and heart rate. This detailed data helps in understanding an athlete’s physical performance.   

AI-powered deep neural networks can analyse players' postures, movements, game events, and interactions on the field from videos, providing a comprehensive view and insights into their performance.   

The data-driven approach provides quantitative feedback, analyses an athlete’s strengths and weaknesses, and helps design tailored training programmes to target specific areas, maximising their potential. 

Using data is more accurate, real-time, comprehensive, and objective compared to traditional human observation and judgment, which can vary based on expertise.   

Data-driven insights and technology provide objective metrics and consistent criteria across various performance aspects, minimising bias and reducing reliance on subjective evaluations.   

The widespread understanding and awareness of technology in sports, the sports industry, and among the public are crucial for transforming and empowering traditional sports. Data, although often fragmented and underutilised, holds significant potential when fully developed and leveraged for its intrinsic value.  

While specialised analytics companies and technology providers dedicated to sports analytics have played an increasing role, substantial advancements in knowledge and state-of-the-art AI technology are also embedded within research institutions.   

Collaboration between academia, soccer organisations, and the sports industry is pivotal for the early adoption of AI and the advancement of data analytics in the sport.  

Academic research not only provides a theoretical foundation and methodologies but also facilitates practical application through knowledge transfer projects, education and dedicated postdoctoral training programmes. 

My research expertise focuses on computer vision and machine learning, particularly state-of-the-art deep learning technology. We are also interested in combining vision technology with large language models.   

Loughborough works extensively with national sports organisations and the sports industry. My AI research in sports includes motion analysis, data-driven  player performance analysis, automated event/highlight extraction, injury prevention and talent identification. The research involves human motion tracking and posture analysis and can also be applied to other sports such as cricket and basketball.   

Human motion analysis and behaviour understanding is fundamental in AI technology. Leveraging AI, I also conduct research in biomechanics, robotics, human-robot interaction, and healthcare applications related to human behaviours.  

For instance, we have developed a robot system for sandwich production to improve productivity and hygiene. Additionally, our work includes AI-powered digital twins for sustainable agriculture and reducing greenhouse gas emissions.  

Furthermore, using AI, we are involved in projects aiming to provide real-time passenger load information in train carriages to enhance railway management and improve customer experiences. 

These applications highlight the diverse impact of AI across various domains, from industrial automation to environmental sustainability and transportation efficiency.