If the Belgian track cyclists win a medal at the upcoming Olympic Games, it will not only be the result of top performances but also of top research. In the Flemish Cycling Centre Eddy Merckx, a stone’s throw from the Watersportbaan, a group of researchers from the IDLab of UGent-imec supports our track cyclists with the help of artificial intelligence (AI).
Professor Steven Verstockt leads the Sports Data Science team, as the research group is called, in the right direction. As an inveterate amateur cyclist, he eagerly looks forward to the Olympics and the performances of the Belgians, in which he also has a part.
Steven, your Sports Data Science team has been working for years to use technology to make the track riders of Cycling Vlaanderen and Belgian Cycling better?
Steven Verstockt: “Yes, a few years ago we built the Wireless Cycling Network (WCN) on the practice track for track cyclists of the Eddy Merckx Centre, a permanent set-up that we are constantly working on to improve the performance of athletes.”
How does that work exactly?
“The track is surrounded by receivers that record and measure all the data of each cyclist; the cyclists themselves are not burdened with additional equipment. They only wear their classic heart rate monitor and on their bikes are the average power and cadence meters.
We don’t even need access to their personal bike computers, but harvest all the sensor data about power, rhythm, and heart rate while the cyclists give it their all on the track. These data are processed at lightning speed, analyzed with AI, and continuously forwarded to the coaches so that they can make very targeted adjustments.
With the help of the video images, the driving style is also scrutinized. These results are so positive that we will also integrate this aspect into the permanent setup.”
The coaches are already directing the riders while they are still working?
“Track cycling involves minor adjustments that can have major consequences. An important discipline is the team pursuit. Four riders have to cover a certain distance as fast as possible in a team, with the riders relieving each other at the head of the train.
The speed and quality of those substitutions are decisive for the result. The more efficient and smoother, the better. The WCN records all rider changes to review them. Our AI model determines in the blink of an eye: this switch was good, this one was bad. Thanks to the WCN, we are refining the technique of track cycling, such as the timing and height of the switches.”
Text originally from: https://www.durfdenken.be/en/research-and-society/ghent-university-researchers-support-olympic-track-cyclists-optimal