Tour de France gets added dimension with data technology boost

Tour de France gets added dimension with data technology boost
Christopher Froome of Great Britain riding for Team Sky in the leader's jersey rides past the Arc de Triomphe during stage 21 of the 2017 Le Tour de France, a 103km stage from Montgreon to the Paris Champs-Élysées on July 23, 2017 in Paris, France. (Photo by Chris Graythen/Getty Images)
Predictive analytics, machine learning incorporated into race to boost viewer experience
Published: 7:45 PM, August 13, 2017

SINGAPORE – At this year’s Tour de France (TDF), fans could challenge the predictions of each stage’s winners by Dimension Data, an information technology services firm.

The company would tweet its picks at its dedicated TDF handle (@letourdata), before the start of each stage and users would be asked for their opinions. On average, Dimension Data’s predictions were spot on around 60 per cent of the time in the first week.

These predictions were made possible by the use of predictive analytics, as part of a new pilot this year between TDF organisers Amaury Sport Organisation (ASO) and Dimension.

The partnership saw machine learning technologies, where computers learn without being explicitly programmed, being introduced to the iconic cycling race, to provide deeper levels of insights as the races unfolded.

The technologies helped to revolutionise the viewing experience for fans, whose demand for real-time statistics has increased in recent years, according to Dimension Data’s Chief Executive Officer for Asia-Pacific, John Lombard.

“We actually have a transponder that sits under every seat of every rider, it’s basically collecting information of that bike, like its speed and positioning, and all of it gets processed at a data centre, which we build every day to support the tour,” he told TODAY in a recent interview at the company’ Singapore office, noting that three billion data points were created and analysed.

“A lot of the information collected is matched with things like weather information, temperature, gradient of the course… (and) we make sure we provide that to the broadcasters and general public.

“We call it the second-screen experience, where you’re watching it on television and (at the same time), you have your iPad or mobile and you’re watching the statistics. Most people who watch the Tour – we are talking about 17 million people – are accessing these online sites for live statistics and predictive analytics.”

The use of such technology is not new to sports.

Major European football clubs use the likes of Prozone and Opta to track players’ performances in-game, right down to every pass, sprint and tackle.

In tennis, the technology to review calls has long been employed, while there are also programmes that predict winners of matches.

In Formula One, teams can make real-time adjustments as cars send back a constant stream of data to pit personnel.

While Dimension’s focus at the moment is on enhancing the fan experience, Lombard admitted the data they collect at the TDF – won by Chris Froome this year – can be also used to help the cyclists improve their performance.

“In real time, we are taking all this live information, using what we call Internet of things… and marrying that with historical information,” he explained.

“We are not just doing it at one static point at the start or end of the race, it’s happening constantly.

“One of the things we can do is the concept of rider comparison, such as taking historical performances on different stages at different events, we can match that (for one rider) against another rider... how, if I was a professional cyclist, I should expect to be performing at this particular stage of the tour, and looking at how I could improve.”

Lombard added that such technologies will not only become more commonplace in sport, but beyond that, as machines’ abilities to process huge amounts of data have broad implications in society, manufacturing, healthcare and government.

However, because of the vast amount of information, cyber-security has now become a big risk.

“Last year, we had just over 1.4 million unauthorised attempts at access (to the TDF data), which we protected,” he said.

“That’s why it’s so important that in designing a solution.

“It’s not just about putting some information on a computer and running some algorithms, we are talking about data flowing across a radio network, information flying across the internet… and at every single stage of the journey we need to make sure the information is secure.

“Anybody who’s in the business of providing that information needs to make sure they secure it, because in future, there may be elements to that information that should not be in the public domain.”