Artificial Intelligence
AWS re:Invent 2024: Data dives into the swimming pool
Data and AI gave Swimming Australia the edge in the relay events of the last two Olympic Games, writes ARTHUR GOLDSTUCK.
The margins separating victory from defeat in elite sports are measured in hundreds of seconds. At Swimming Australia, these razor-thin differences are being transformed by an invisible ally: data.
In the high-stakes world of Olympic swimming, technology is no longer a supporting player, but a strategic cornerstone. Jess Corones, general manager of performance support at Swimming Australia, told Gadget last week that that data and artificial intelligence were revolutionising the way athletes train and compete, creating a sophisticated ecosystem where technology and human potential intersect.
A relay application developed by Swimming Australia epitomises this data-driven strategy. By analysing historical relay performance data across international teams, coaches can now simulate and optimise relay team selection with unprecedented precision. At the Tokyo Olympics, this approach yielded remarkable results: Australia medaled in all seven relay events – a feat rarely achieved in Olympic history.
“Races are won and lost by ten-hundredths of a second,” said Corones, on the sidelines of the Amazon Web Services (AWS) re:Invent conference in Las Vegas, Nevada, last week. This microscopic margin underscores the critical nature of sophisticated data analysis. The relay app doesn’t replace human intuition but augments it, providing strategic insights that help coaches make more informed decisions about team composition and racing strategy.
The economic and technological implications extend far beyond the swimming pool. In an era where data has become a corporate currency, Swimming Australia’s approach offers a blueprint for organisational innovation across industries. Their methodology demonstrates how technological integration can create competitive advantages that are both measurable and transformative.
Since 2018, said Corones, a partnership with AWS meant tools like QuickSight and SageMaker has enabled coaches to interact with complex performance data in intuitive, personalised ways. The upcoming integration of Amazon Q will allow natural language querying of datasets, further breaking down traditional barriers between technical and non-technical team members.
A computer vision tool called Sparta represents a leap in performance tracking. Unlike traditional wearable technologies that can restrict movement, it is a non-invasive system that allows athletes to train naturally while collecting detailed performance metrics. The tool can track multiple swimmers in a single lane, capturing nuanced details about technique, speed, and efficiency across different swimming strokes.
“It’s a custom-built piece of software, and it uses machine learning and computer vision to track the swimmers in a race,” said Corones. “But (computer vision) is limited in the fact that it can only work with the constructs and confinements of an actual race environment. So we’ve developed a system that can get the same metrics, but in the training environment, which is a much more chaotic environment for computer vision, because in a training pool, you could have anywhere from two to six athletes in each lane. They’re swimming different strokes. They’re stopping halfway in the pool. They’re diving underwater. So tracking them in that environment’s a lot more chaotic for the computer vision.
“That’s one of the things we’ll be announcing (at re:Invent) and showing some insight into the architecture behind that system.”
Corones highlighted transformative principles the system could teach businesses. First, the critical importance of user-centric design: by involving coaches and athletes throughout the development process, they ensure that technological solutions are not just technically sophisticated, but practically applicable. Second, meticulous long-term planning mirrors strategic business cycles of preparation, execution, and continuous improvement.
“It’s about how you use the data,” said Corones. “Everyone’s trying to advance technology, but the real competitive advantage comes from understanding and applying that data effectively.”
The broader implications for organisational innovation are profound. By creating a culture that embraces technological transformation while keeping human performance at the center, Swimming Australia shows how data can be democratised, how complex technologies can be made accessible, and how human potential can be amplified rather than replaced.
* Arthur Goldstuck is CEO of World Wide Worx and editor-in-chief of Gadget.co.za. Follow him on Bluesky on @art2gee.bsky.social.
