Artificial intelligence and big data analysis are boosting the performances of top athletes, but are they creating a sport-tech arms race?
Watching elite athletes run, leap and score, it’s hard to imagine there’s much room for improvement, but the Internet of Things, Big Data and virtual reality are shaving milliseconds from sprinters, extending the jumps of Olympians – and helping your favourite striker put the ball in the net.
That’s before bionics change sports forever, with predictions that the sprinters at the next summer Olympics could be outperformed by an athlete at the Paralympics. It all starts with data collection, which shouldn’t be a shock. The film Moneyball was based on the true story of Oakland Athletics’ team manager outwitting rivals with data science – and that was in 2002. Fast forward to 2018 and the combination of always-on sensors and connectivity takes the idea several leaps forward.
“The big initial change was testing work rate through GPS tracking,” said Rebecca Hopkins, CEO of the Sports Technology Awards. It started with companies such as Catapult Sports, an Australian startup that began tracking athletes with GPS in 2006. This took training research out of laboratories and onto the field – in this case, the fields of the Australian Football League, although the technology is now used by the NBA, NFL and many more.
Today, ever more accurate sensors are worn by athletes at all levels, from couch-to-5K amateurs to marathon runners. “This evolved into recording aspects of health and performance such as hydration, recovery, sleep levels, jump capability, hit force and so on,” noted Hopkins. “This is used to make decisions such as how to avoid injury, understand the best times to make substitutions and how to structure individual training.”
Training isn’t all about timing the body – it’s about teaching athletes the best way to react in specific situations as well as learn from their mistakes. Coaches have long shown players the video of a match to highlight mistakes or learn how an opponent operates. That’s now done in much more detail thanks to 360-degree video and VR.
“VR is already being used in training, both enabling athletes to prepare or using it to help them replay a game-time situation and understand what they could have done better,” said Hopkins. That includes letting a cyclist better visualise a route, bumps and all. “When Team GB cyclists go around a particular corner, [VR] lets them know where the kerb is and so on,” said futurist Matthew Griffin.
“While it’s not necessarily as good as being out there, it allows athletes to learn the different layouts of roads or pitches.” There’s also a psychological element: players can practise ignoring a stadium full of boos or cheers. Virtual reality can let athletes replay a failed moment by putting them back into a virtual version of the game as an avatar. But that’s only the beginning, noted Griffin. “Virtual reality lets you replay and learn from our mistakes, but where we’re starting to move to is being able to put professional sports men and women against artificially intelligent real-time environments.”
(Team GB during 2012 Olympics. Mitch Gunn/Shutterstock)
It’s like playing a computer game against a virtual character, with firms such as STRIVR offering immersive digital training environments, letting athletes practise without wearing out their bodies. It’s like a realistic video game to rehearse for the real thing.
“AIs are starting to ingest and learn knowledge about players, tactics and so on – there’s a huge database,” said Griffin. “There are millions of hours of NFL footage, of Premier League footage.” That means coaches and managers can analyse what works and doesn’t for their teams and opponents, tweaking strategy as required to win – that’s already being trialled with the NFL, thanks to researchers at Oregon State University.
Such an approach limits the AI to hunting out the best tactics from existing games, meaning it’s capped at what we humans can come up with. The next step is letting machines come up with their own ideas.
“The majority of modern AI will go back in time, look at a lot of footage, and see that Manchester United, for example, won this game because they used a specific strategy,” Griffin said. “That’s artificial intelligence that analyses human strategies and replays them. Where we’re starting to see some stunning advancements is where you give artificial intelligence the rules of the game and nothing else, and let it run its own simulations and let it create its own strategies.
“Even the coaches are like: ‘I’d never have come up with that strategy, it’s so random’”
“Some of the strategies are completely alien – things we wouldn’t have thought of before,” Griffin added. “When you start taking away human preconceptions and ideas, and start letting these alien AIs look at it through their own eyes to create their own strategies, some of those are just staggering. Even the coaches are like: ‘I’d never have come up with that strategy, it’s so random’. And when they play them, they work.” And if they do work, expect to see more AI-based surprises in sport soon.
Double-digit performance boosts from tech could lead to an uneven playing field, and sporting bodies have banned innovations before. Back in 2008, Speedo introduced LZR Racer swimsuits made from a hydrophobic material, which boosted swimmers’ performance by up to 2%. This led to over 130 world records falling in two years, and the swimsuits were banned as a “technological prothesis”, noted Griffin.
(Olympic swimmers Thiago Pereira and Ryan Lochte at the launch of Speedo’s Fastskin LZR Racer X swinsuit in 2015. Shutterstock)
But evaluating the performance impact of virtual reality and analytics isn’t as easy. “It’s one of the biggest problems that sports organisers have,” said Griffin.
Regardless of what organising bodies do to manage the growth of tech in sports, expect it to continue to shape competitions. “We’re increasingly getting to the point where technology is an advantage. [If you approached] every football team and Olympic team and told them you wanted to take away all their technology, they’ll kick you off the field,” said Griffin. “It’s difficult to argue that technology gives one country an unfair, technological advantage.” British cyclists may well be benefiting from VR, but there’s nothing to stop other teams from doing the same.
As is often the case with tech, regulators are years behind the curve, meaning that once they do take action, teams may well have already moved onto newer techniques. “We are in a technological arms race, and some countries lead the way in these technologies,” Griffin said. “The UK is one of them. Look at the performance of our cycling team at the last Olympics – there were countries saying you guys are cheating because you’re too good.” That was down to everything from new bikes, helmets and slick skinsuits, to VR simulators.
Where do we go from here?
The use of tech in sport will continue to grow – the more data coaches and trainers have, the further they can push athletes. “Analytics is likely to continue to develop, although it will be hard to see what gaps have been left to fill,” said Hopkins.
“The more data we get from athletes, the more we can feed back into AI systems”
More data offers more opportunities. Nanomaterials or reactive materials could change sports as much as the Speedo swimsuit impacted swimming. Such gear will heat up or cool down to help keep an athlete at the right temperature, for example, looking at indicators such as respiration. That’s helped by soft sensors that can be embedded into fabrics, analysing everything from how much salt or water an athlete loses to minutiae of posture and position. “The more data we get from athletes, the more we can feed back into AI systems,” said Griffin.
However, some of the advancements won’t be about pushing sports men and women to the next level, but protecting them from harm. With concussions and other physical damage making headlines within the NFL and other top sports, this is a key issue. “Athlete welfare is – and needs to be – huge,” said Hopkins. “Governing bodies have a moral and financial imperative to ensure athletes are well cared for and tech can play a very effective part in delivering this.”
The use of technology in sports could become yet more invasive. Gene profiling might reveal your potential before you even start training, while CRISPR DNA editing could be used to boost the effect of changes by hacking how your body works. Neurostimulation could take virtual reality training a step further, using a brain-machine interface to directly stimulate specific parts of the brain to boost performance – something Griffin said the American ski jumping team has already used to great effect using Halo Neuroscience’s headsets to boost muscle training.
“We’re not talking people who are eating pies and doughnuts all day and just make them a bit fitter,” said Griffin. “These were Olympic athletes that were already at the top of their game.” In other words, brain training could push sports to a new level. And then there’s bionics: the athletes already astounding us at the Invictus games and Paralympics may well dart ahead of their so-called “able-bodied” peers.
“It’s been claimed that the evolution of tech, such as blades for runners, means that whilst the Olympics will be about man’s sporting endeavours, the Paralympics will be about what man and tech can achieve on the track,” said Hopkins. “It is highly likely that in the next Games, we will see a Paralympic sprinter outperform the able- bodied Olympians. That has to be a very exciting prospect.”