UAE explores the future of cycling with self-developed AI system 'Anna': "The capabilities are crazy"

The WorldTour is on the brink of a revolution in performance management. UAE Team Emirates, the most successful team of 2024, is developing an advanced artificial intelligence system that could improve riders' performance and enable new training methods. According to insiders, the 'Anna' system offers unique views and could dramatically change the future of cycling.

"We've been working for years with our partners in the UAE on our own data platform. It's now getting to the stages that are really phenomenal," Jeroen Swart, performance coordinator at the UAE Emirates, told Vélo. "The capabilities [of the AI] are crazy. It's in the process of being transformative."

Text continues below photo

Already the right answers, now to answer the question

With the help of 'Anna,' the team can analyze complex data sets and translate them into practical strategies. These can range from race tactics to personalized training schedules and recovery plans. Swart stresses, however, that the real challenge now lies in asking the AI the right questions. "The hardest part now is thinking about what to ask. We're getting the answers, but we need to know what the questions need to be, and unique questions that are not immediately obvious. So that's really what we have to do now. Because we have the capability," he explains.

Cycling has long been a sport in which data plays a major role. Riders and coaches rely on many metrics such as power, heart rate, aerodynamics, nutrition and sleep data. But as Dan Lorang, performance chief at Red Bull-Bora-Hansgrohe, points out, connecting all this data remains challenging.

"It would be a huge next step if we could bring together all this data we collect that lives in different places," Lorang says. "If we could understand how each variable influences and changes each other would change our entire approach. Being able to bring it together in one place and to get information and predictions out of it would be a massive breakthrough."

At the UAE Emirates they have already embraced this challenge. Their AI system can run simulations and create alternate realities where riders such as Tadej Pogačar can virtually race in simulated versions of races like the Tour de France.

Text continues below photo

Will Pogacar with AI-generated training schedules get even better? 

Experiments and guesswork will not be part of cycling in the future

According to performance experts such as Stephen Barrett, trainer at Decathlon-Ag2r La Mondiale, the future of cycling lies in micro-training. This means giving each rider a fully personalized training program based on how his or her body responds to specific stimuli.

"However, riders are all at different stages of development and progressing at different levels. You can't just use the same ideas for two riders because they can hit the same watts-per-kilo – especially now that everything is so marginal. Being able to better use what data we have on each individual could change that. We have to keep pushing forward with all this information that we have to figure out what works for each specific rider. And we need to be able to figure it out quickly in the athlete's career." Using artificial intelligence, trainers can largely avoid experimentation and guesswork. AI can provide certain insights and refine workouts faster.

Pogacar finished third at Milan-Sanremo in 2024

How does Pogacar win Milan-Sanremo?

The idea of integrating artificial intelligence into cycling is not limited to the UAE. Other teams, such as Visma-Lease a Bike and Red Bull-Bora-Hansgrohe, are also working on innovative technologies. Although they do not yet have programs similar to 'Anna', Swart emphasizes that this is only the beginning of what is possible.

UAE Emirates is ahead of its competitors thanks to its significant budget and technological developments in the UAE. The team uses alternate reality and 3D simulations to analyze performance and race insights. "They brought up a 3D profile of one of the stages of the Tour. And then we could ask 'Anna' to show us how the race unfolded using little miniature cyclists. It could tell us key stories, particularly focusing on Tadej and the data it has on him. 'Anna' could show us how his performance was and how that unfolded in the race."

Swart thinks AI will soon be widely used in cycling and immediately become an essential component. "This is something in future where teams have to put more money. I think it's an element that everybody has to try to invest in,' he believes. Because the system is not yet perfect. For example, 'Anna' cannot yet answer all the questions, such as: how does Pogacar win Milan-Sanremo? 'The most interesting answers from AI? I think we will find that out in the future," Swart concludes.

Read also

🎥 Just like with Pogacar, an amateur cyclist clings to Vingegaard's wheel: "Why don't you shave your legs in winter?"

UAE gem Del Toro cycles too fast for his friends, but not (yet) as fast as Pogacar: "Metrics much better"

Visma's biomechanist gave Jorgenson a reality check: "Like this you're one of the least efficient in the peloton"

Subscribe to our Newsletter

Get every week the latest news about CYCLING in your inbox!

Subscribe