When the Australian swimming team heads to Paris in July, computer vision – tracking their every stroke in the training pools – will be fueled by the same predictive insights that propelled them to the podium at the 2021 Tokyo Olympics.
Let’s dive deeper into the post-Tokyo data
With nine gold medals under its belt, Australia’s top swimming body returned from Japan confident that its investments in data analytics had paid off.
Jess Corones, General Manager of Performance Support and Olympic Campaign for Australian Swimming, said iTnews that from 2021, AWS is working to connect Swimming Australia’s data lake, Atlantis, with more data sources and create more predictive insight applications for coaches.
“After the success of the Tokyo Olympics and the work we’ve done with AWS improving the impact that data and data transformation and usability can have on performance, we can see the potential and have started to invest more,” she said.
In the run-up to the Games, each elite swimmer’s preparation – or ‘Dolphin’ – was guided by Atlantis’ predictions on optimal nutrition, training schedule, injury management plan, race plan and more.
It was also the first international swimming meet where both historical and real-time data on the performance of the Dolphins and competing teams throughout the day predicted the optimal combination and order of swimming relays.
Coaches accessed relay proposals in real time and made last-minute decisions according to the order; The Dolphins won medals in six of the seven relays they competed in, making them the most medal-winning relay team at the Tokyo Olympics.
Expanding computer vision
Since 2019 — when Swimming Australia released its customized “Swim Performance and Race Tactical Analysis” athletic improvement system (Sparta 2) — computer vision has followed the Dolphins, but only in competition.
Automatically recording each swimmer’s technique and performance during training and then feeding it into Atlantis’ performance insight models is one of Swimming Australia and AWS Professional Services’ current projects.
“We are currently in a current project with ProServ [AWS Professional Services] developing a new system called Training Insights,” Corones said during an interview with AWS re; invented in Las Vegas last December.
“The idea is that we have really rich competition data, but we want comparable data in the training pool.
“We’re developing a piece of software that can then use machine learning and image recognition to track athletes in the training pool to give us the same metrics we get in the competition pool.
“And that’s something we’ll be looking to release, hopefully, in the next few months.”
It’s much easier to track a swimmer’s distance per stroke, number of breaths, turn times and other statistics when every other variable is consistent, which it is in a competitive environment, Corones added.
“In competition you have eight swimmers, one per lane and they go up and down, but in a training pool there are so many more variables that computer vision has to take into account, pick up, [and] understand what to focus on and what to discard.
“In a training pool, you can have up to six swimmers in one lane, doing four different strokes … and not all of them do a full circuit – maybe they stop at 25 or 35 meters; maybe they do different segments. So how do you measure that in an automated way?”
Lane four
Swimming Australia engaged AWS in 2018 to turn data collected on athlete performance during training and competition into decision-making tools for coaches.
The data ranges from statistics from the International Association of Swimming Federations (FINA) to various athlete management systems and data from wearable technologies and Sparta 2.
The Professional Services team used the AWS Serverless Data Lake Framework to integrate data sources and establish a data lake environment.
Corones said that generating performance insights through Atlantis has allowed training to become “highly individualized to each athlete, each coach and each training program.
“What data lakes allow us to do is to have this really holistic view and understand each of the different variables and how they interact in performance.”
As well as connecting Atlantis to multiple data sources, another area of focus for Swimming Australia and AWS post-2021 is making insights more accessible to more coaches, Corones said.
“So once we create those insights, how can we put them in front of coaches in a nice visual way to allow them to use those insights to make decisions about athletes and training?”
In early 2023, Swimming Australia and AWS released an online portal that allows coaches to access their swimmers’ data and predictive insights from Atlantis through the AWS QuickSight dashboard.
“Lane Four” is “a one-stop shop that coaches can access on their iPad from the side of the pool to access everything they need to know about their athlete from strength to conditioning to nutrition,” Corones said.
“He can too [respond] to their natural language queries and find answers to the questions they want.
“So it could be, for example, that a coach could know who had the top three fastest times in the women’s 100m freestyle in the world since 2003. They can ask that question without having to search for filters and dates.
“This is so useful for coaches bbecause sometimes during training they need that information at the click of a finger; maybe they’re in the middle of training and need that benchmark to give the athlete a goal.”
Jeremy Nadel attended AWS re:Invent 2023 in Las Vegas as a guest of AWS.