Sports Data Science and AI Next Generation Graduates Program (NGGP)

We now have a huge array of data to inform us about athlete performance, whether at Olympic level or in community sport. Performance is also dependent on a large number of interacting physical, mental, environmental and social factors. This theme will use Complex Systems tools to integrate disparate data and learn about the holistic athlete: performance, health (injury and illness) and wellbeing. The aim is to support individualisation of athlete management and support.

Challenges:

• Layered AI for female patterns (complex individualisation - e.g. medication change - female hormone usage)
• Relationships between physical and psychological/emotional
• Profiling and clustering of data (e.g. understanding different female models or ‘types’ through AI links to genomic data - risk/opportunity analysis)
• Predictive weather modelling - integration of complex systems
• Can generative AI and LLMs provide automated feedback to athletes, coaches, other stakeholders from complex data analytics?

There has been an explosion of new gadgets, apps, software and tools for sports and athletes! How do we know what is beneficial? This theme will focus on proving the efficacy of innovative technologies in an athletic setting to provide significant health and performance outcomes for athletes and recreational exercisers.

Challenges:

How might we:

• Test, validate, and/or use new biomedical technologies to accelerate post-injury or training recovery? Also, what’s the optimal experimental design and minimum viable data to show benefits?
• Develop new tools, techniques, algorithms for managing large datasets in the biomedical disciplines (genomics, epigenetics, immunology, biochemistry, haematology) for research, clinical and field (sport) applications?

It is critical for the future of data science and AI in sport that athletes are confident the governance and the ethical framework for the use of data are based on solid consensual, legislative and human rights foundations.  This theme will create knowledge and tools to increase trust in sports data science & AI.

Challenges:

• How to design data collection/surveys to accurately represent a population (sport participation etc.)
• Integration of multi-modal data
• Data sovereignty and ownership (in relation to diverse data from different sources, including proprietary wearables and more)
• How to link data
• How to create value back to the athlete from the use of their data

How can we reduce injuries and illnesses in sport? In this theme, we aim to better understand the context of these adverse outcomes, including the effect of training and competition, differences between individual and team sports, and interactions between contextual factors. We will focus on developing models to predict outcomes, analyse trends and assist in individualised management of risk.

Challenges:

• What injuries/illnesses are preventable and predictable?
• Can we use biomedical data to predict injury (and prevent it)?
• What is the interaction between all contextual variables (e.g., athlete characteristics, training and competition info)?
• What is the training capability of individual athletes (i.e., when will they 'break')?
• How do we optimise return from illness or injury?
• Could we model return to training following pregnancy?
• Can generative AI and large language models be used to explain injury condition. time to return to play and rehabilitation plan/ protocol to players and coaches?
• Can we use AI to inform athletes and coaches about the role of menstrual status on performance and injury?

Current approaches provide a 'one size fits all' solution for athlete performance. This theme focuses on personalised optimal performance: finding the best solution for an individual athlete or team.

Challenges:

• Optimise individual athlete or team performance through the use of data, other information and a data science skill set
• What does it take to win for a particular individual or team? What are the determinants of success?
• What value (e.g. financial) can be derived from modelling performance (performance investment)?
• Can we take into account the female cycle hormonal phase to  optimise performance and manage risk?
• Can we better understand responders and non responders to different training modalities?
• Can we use other emerging technologies (e.g., genomics) to optimise individual performance?
• Can we develop ChatGPT type language tools to help athletes and coaches extract their own useful information from bodies of knowledge?

How can we measure the social impact of investments in sport (e.g., health economics)? How can we help athletes and sports build and monetise their brand, leveraging data science and AI? This theme focuses on 'the business of sport' through a data science lens.

Challenges:

How might we:

• Leverage data to better understand an organisation’s stakeholder network (e.g., volunteers), including network role, value contributed (e.g., in volunteer roles like  officiating, coaching, canteen, etc.), motivations to contribute (e.g., social, health, etc), and risk of (and potential reasons for) churn, to help inform strategies and automate tactics (e.g. personalised offers/rewards/recognition) to retain/improve engagement for the benefit of the stakeholder, sport, and its ecosystem. 
• How can we use data science to improve the fan experience?
• Are there opportunities with sports broadcasting, e-sports and fan engagement as part of the business of sport.?
• To have accessible & consumable strategies / technologies to support individuals in the community in their own performance.