From opponent analysis, tactical insights, officiating and improving athlete performance, artificial intelligence and machine learning have already been incorporated into sport. The NBA is using generative AI for personalized user content. Two-time Olympic marathoner Des Linden has a digital twin heart to simulate her heart rate, blood flow and oxygen levels, to fine tune training and improve performance. The Mercedes-Benz Stadium, home to Atlanta Falcons, has implemented facial recognition ticketing.
In spite of these impressive applications, reports indicate a modest influence of the sports industry on the global AI market. From June 2020 to June 2023, 363 AI-related patents were granted in the sports sector worldwide, compared to 4,155 in the automotive industry. Here we explore some potential AI applications that could support environmental sustainability and address climate change-related needs for the sports sector.
Athlete injury prevention
Climate change is exacerbating risks to athlete health. Training and competing in extreme heat, poor air quality due to pollution or wildfire smoke or on surfaces hardened by drought are just some of the ways we’re already seeing this manifest. These conditions can contribute to increased risk of injury through dehydration, pulmonary illness, fatigue-induced strain/pull and repetitive strain.
AI offers real-time biomechanics precision monitoring and predictive analysis for athlete form, posture and movement. If indicators of stress or exhaustion are detected, coaches, managers and players could make more informed decisions on intervention, which may reduce the risk of injury in some cases. Wearable devices equipped with sensors and AI offer instant feedback.
Former Australian Diamonds netball player, Amy Steel, suffered a severe heat illness during competition in 2016 that ended her netball career and resulted in potentially lifelong health difficulties. When asked if she thinks AI applications might have helped prevent this, Amy responded, “We can see from my experience that it’s no longer adequate to take a heat and humidity reading before a game, and then assume it will be safe for the remainder of the match.”
She went on, “We know that people have individual responses to heat, and we can see that in my case. I had broken the club fitness testing record only a couple weeks before my heat stroke, and yet I was the player who experienced the most severe heat illness. Perhaps wearables may be able to help understand these individual differences and help to keep players safe. Ideally in the long term we would hope that these type of technologies can translate into better heat management strategies that can enable grassroots sport to continue safely without the need to rely on expensive equipment.”
Predictive modelling for weather
AI can now reportedly outperform conventional weather forecasting, predicting severe weather such as extreme heat and the path of cyclones faster and more accurately. The Financial Times reported that it could be 1,000 times cheaper in terms of energy consumption too.
As instances of extreme weather events or inhospitable conditions increase due to a changing climate, this application could assist sports organisers, especially organisers of outdoor sports, touring sports or competitions, to be more accurately informed further in advance. This could enable them to take necessary actions to adapt or postpone events to ensure athlete, staff and spectator safety, or provide welfare advice earlier to participants.
Predictive modelling for fan behaviour
In the aftermath of COVID-19 when fans were beginning to go back into stadiums, technology firms used AI to work with teams on ticket pricing, to better forecast what fans would be willing to spend on an event. LaLiga has created a machine learning solution that maximizes TV audiences and stadium attendance when scheduling matches. Sports organisations have begun to use predictive modelling for delivering targeted content and promotions also.
A potential environmental benefit of predictive modelling of fan behaviour in stadiums is waste reduction. With improved understanding of attendance trends, fan demographics and food, beverage and merchandise preferences, sports organizations could reduce unnecessary waste by leveraging this data to find efficiencies. Promotions could be tailored according to fan preferences and habits, further reducing potential waste streams.
In addition to waste reduction, some organizations are utilizing AI-based solutions to encourage and gamify waste separation, rewarding sports fans for proper sorting that enables recycling. This technology can also offer venues more data on waste types, levels and separation for reporting purposes.
Energy efficiency
Sports stadiums and venues are energy intensive when they are in use for games and events, and sometimes have staff offices and conference spaces that mean smaller parts of the venue are used more frequently. Building management systems are used to control and monitor various systems and functions of a building, and create a lot of data.
Implementing AI in smart buildings, that have sensors and Internet of Things technology, can maximise energy efficiency. AI algorithms can analyse and interpret building management systems data to make real-time adjustments to heating, cooling and lighting systems. This can optimize energy consumption and reduce overall environmental impact.
Emissions prediction and strategy development
A recent research paper presents a novel approach to developing AI-powered carbon emission strategies for sports event management. In brief, a model for analyzing the influence of population, wealth, and technology on carbon emissions in sports events is used, alongside a neural network that predicts future emissions trends. This is then enhanced with transfer learning, creating a comprehensive approach for carbon emissions analysis in sports event management.
Whilst there are limitations in the study, it shares that ‘significance of this research lies in its potential to empower sports event managers with a data-driven approach to carbon emissions management.’ The researchers posit that sport could transition towards greater sustainability by leveraging AI to predict carbon emissions in sports events accurately and to develop effective carbon neutrality strategies.
Generative AI for fan travel data
Collecting exhaustive data for fan travel to games for greenhouse gas emissions reporting is difficult for sports organisers. Often a small data sample is collected and extrapolated to give indications of fan travel emissions to sports events or competitions.
Generative AI can employ machine learning techniques to create synthetic data that closely resembles real data, simulates various scenarios, generates hypothetical datasets, and fills gaps in existing data. Whilst a potentially interesting application, there are risks relating to creating data in this way. It is crucial that this kind of generative AI extrapolation is tested, reviewed and validated by humans.
The opportunity
Some of these AI applications for sport are already in place, others will need much more exploration and regulation before they are robust enough to be considered. We also can’t forget the environmental impact of AI, which the International Energy Agency says ‘uses more energy than other forms of computing – a crucial consideration as the world seeks to build a more efficient energy system. Training a single model uses more electricity than 100 US homes consume in an entire year.’
The global sports industry is already being reshaped by the application of AI and other emerging technologies that have formed the Fourth Industrial Revolution. However we already stand on the brink of the Fifth Industrial Revolution or the cognitive age, which will prioritise purpose, the protection of the environment and sustainability. As sports organisations create, influence or adopt AI and ML applications, it is vital they take advantage of the chance to emphasize sustainability, environmental consciousness, and inclusivity at every possible opportunity.
