Today marks the start of the 109th edition of the Tour de France. Cycling fans have more and more digital opportunities to follow the Tour de France, such as live tracking of cyclists, see mountain stages in augmented reality and many Statistics†
Fifty-five data points per. cyclist is calculated by the IT company NTT in real time during the stages. For this it uses trackers under the cyclists’ saddles. For example, NTT uses clever algorithms to predict the top three of each phase. And TV viewers are regularly shown all kinds of data, such as the chances of an attack from the field succeeding.
From the tracker under the saddle, the signal goes very fast to a motorcycle or car nearby and then to the network. “Every second, we record each cyclist’s position and speed. This data is sent to our network as quickly as possible. We process it and display the data in our tools to the public, media and staff,” said Tim Wade, vice president. president of NTT, which has been an IT supplier to the Tour organization since 2015.
The company combines the data from the sensors under the saddles with, among other things, 3D information about the track, weather information and historical data from cycling races. “We know whether a rider is riding against the wind or up a mountain and what the gradient is. We use all this data for our machine learning models. For example, we can predict whether a new leading group will be ahead. And we do a sentiment analysis of the herd, where we predict if anything will happen in the race, Wade says.
“Last year, we had predicted exactly 74 percent of the stage winners. And in 2019, we also predicted that Egal Bernal would win the Tour, while it was something of a surprise.” According to Wade, there is still one problem with the trackers: the transmitters are only mounted under the bikes at the start of the stage. “If the rider changes bike along the way, we lose data.”
There is a limit
This year, NTT predicts as Tour winner Tadej Pogacar, followed by Primoz Roglic. But there is a limit to what technology can predict, Wade acknowledges. “Much more accurate than we do now becomes very difficult. Because so much can happen that one cannot predict, simply because those things cannot be measured, and therefore there is no data about it. Think, for example. a fan who wants to take a picture and hit a rider’s handlebars, or a puddle of oil on the road. All of this can have a big impact on the race. “
NTT’s data and forecasts can not only be seen in the TV broadcasts, but also on the social media account LeTourData and in the Tour’s online fantasy game. During the stages, the audience often gets more data about the riders than the team leaders in the cars, Wade says. “Teams will receive a separate version. There are restrictions on what data teams can use for tactical decisions during the race. This is in line with the rules of the Cycling Union UCI.”
This year, data analyzes will also be made for the first time at the Tour de France for women, taking place at the end of July. According to Wade, this is a little more complex because there is less data available on women’s cycling. “We’re going to track the women in the same way as the men and use the data the same way.”
Internet at the top of the mountain
Previously, Tour staff had to keep in touch over the radio and relied on whiteboards that someone held up on the back of a motorcycle in the meantime. Now they are always and everywhere online and everything can be checked live. For the connections, NTT works with the French telecommunications company Orange.
“The challenge is to analyze all the data we collect from the stage as quickly as possible and send it back to the employees. It’s getting better and better. It’s pretty unbelievably easy to reach 100 megabits per second on top of a high mountain. download and upload speeds, “says Wade.
NTT has built a so-called ‘digital twin’ for the employees in the cycling round: an exact digital copy of the entire Tour organization. This allows them to see the current situation in any place, such as where all the support vehicles are located, or where the trucks are at take-off and arrival or the stalls and toilets in the fan village. Thanks to sensors and smart cameras, they can also see exactly the number of spectators in locations.
According to Wade, the goal is to help employees make faster and better decisions. “It makes it easy for employees to adjust things that are immediately visible to everyone. For example, there are hundreds of trucks and other vehicles. They can be seen in one 3D model because they have all been tracked. Everything can be checked in real time, really cool. “
Every day, the Tour caravan is a different place. “Although this tool saves an employee five minutes in figuring out where a particular truck is, it is very valuable. Anyone can do their job a little faster and easier.”
Tonie van Ringelestine