Elite cyclists are using artificial intelligence to get precise estimates of how many calories they will need for each stage of endurance races such as the Tour de France.
Riders burn around during the Tour and food is so important to success that most teams employ numerous chefs and nutritionists.
at Maastricht University in the Netherlands and his colleagues worked with the Dutch professional cycling outfit Team Jumbo-Visma, using machine learning and mathematical methods to improve the way they planned their riders’ diets.
Advertisement
“The way they were predicting the energy intake was… well, not very efficient,” he says. “How they computed it was only based on experience. There wasn’t a real reasoning behind how it was done. You have to have it spot on – any improvement helps, of course, to get the win.”
The researchers took data from previous races, including the body measurements and power output of each rider, the route and elevation of the stages, the weather and the wind direction. They used machine learning to analyse this data and create a statistical model, which could then be used to estimate calorie demands for any rider on any new stage route.
In experiments, the researchers asked coaches to estimate riders’ calorie requirements for previous stages in the Tour de France and Giro d’Italia from 2019 and compared them to estimates from the statistical model. The results were compared in another statistical analysis that provided a score between zero and one, with higher values being more accurate. The coaches scored 0.55, while the model achieved 0.82.
The results were so accurate that Team Jumbo-Visma have since been using the model for dietary planning, including in last year’s Tour de France, in which their rider Jonas Vingegaard on a climb and went on to win the whole race – something van Kuijk puts down to a nutritional mistake causing Pogačar to run out of energy.
“We know that, since then, other teams have picked up on using AI in cycling,” says van Kuijk. “Now it’s become the big hype in the cycling world.”
But the AI tool will be of limited use in planning diets for non-athletes because people don’t generate a wealth of data in their day-to-day lives like professional cyclists do, he says. “Of course, you can simplify it, but then the accuracy drops.”
arXiv
