In a cycling team meeting, the best team directeur sportifs will explain the intricacies of the courses, they know where the peloton will shatter, where the winning attacks will be placed and which riders will be contenders. They will rarely be wrong. Their knowledge, which becomes innate, will be based on years of experience, having watched and raced thousands of races. They are students of the sport who will think about the racing more often than not. They are deeply passionate. They will have modelled race outcomes in their heads, they will know when the gradient of a climb will catch the peloton off guard, on what side of the road the cobbles are smoother, and they will know which riders are struggling or poised to win based off of their past performances on similar courses. They will see, and understand, the countless variables that determine outcomes that will go unnoticed to others.
The team directeurs who have profound tactical acumen are few. As artificial intelligence advances, their jobs, and their knowledge, will be enhanced which will influence race speeds and outcomes. Ultimately, machine learning will eclipse the directeur’s lifetime of experience. Their big asset to a team in the future will be their rapport with the riders and their ability to synthesize the data to motivate, galvanize and mobilize the team.
As in the battlefield, a chess match, or football game, Ai is revolutionizing most facets of life, and professional cycling is no exception. The integration of AI into competitive cycling has the potential to significantly influence the outcome of tactical decision-making. From optimizing race strategies to enhancing rider performance, AI will transform how teams approach the sport.
By analyzing large sets of data, computers will find blind spots in the current approach to race tactics, which will lead to novel tactics, and perhaps, less formulaic and predictable racing that is more interesting to the spectator.
Cycling has historically been a sport which is slow to adapt to post race tactical analysis. Racers and staff use Veloviewer and other models to preview courses but in comparison to field games, like football, baseball or soccer, where every play is analysed and dissected, cycling lags behind. NFL players are given iPads before and after the game, with game tape of plays to analyze and memorize. Rarely do cyclists or coaches review or study race film of themselves, their competition or the courses. As Ai models become more potent and prescient, riders and teams will be able to analyze every aspect of future courses, their competition and past races.
In a sport with thousands of variables, one of the most profound impacts of AI in professional cycling is its ability to analyze vast amounts of data quickly and accurately. Teams collect data on various factors, including rider performance, weather conditions, and course profiles. AI algorithms can process this data to identify patterns and predict outcomes, helping teams to formulate effective race strategies.
For instance, AI can analyze past performances of riders and opponents to suggest optimal pacing strategies. By understanding the strengths and weaknesses of competitors, teams can make informed decisions on when to attack or conserve energy.
Over the last fifteen years, professional cyclists have made their lives increasingly public through social media and training platforms. That data can be scraped by their competition and analyzed to gain a race day advantage. As Ai analysis advances privacy online will be an asset on the race course.
Knowing course conditions, and a rider’s physical abilities as well as the competitions’, AI can suggest the best moments for a rider to break away from the peloton or when to bridge across to a group. It can also advise on the ideal time to refuel based on the rider’s energy levels and the race profile. Real-time insights can give teams a competitive edge by enabling them to react swiftly to changing race dynamics.’
Race speeds have been increasing year over year as technology becomes more pointed. Ai will continue to accelerate the pace of the peloton yet it will also help predict where pinch points are in races, and where crashes may occur. This can cause chaos in the peloton as every team will be racing into position to avoid the crash; this already occurs with race radios but will only become more extreme. However, to increase safety and avoid panicking the peloton, race organizers can also use data analysis to create safer courses to avoid high speed crashes.
The financial disparity between professional cycling teams is vast which is evident in the race results. Teams with resources will increasingly hire analysts and develop Ai models which will further grow the performance gap. At the 2024 Tour de France, Team Visma Lease a Bike had a van following the race which was dedicated to race analysis. Seeing it as an unfair advantage, the ASO, the Tour de France organiser, banned it from being close to the course while the UCI, the sports governing body, said they would look at how the van was being used.
The advances in machine learning are occurring rapidly. The UCI needs to be considering how it will change the sport, where they may want to limit its use and how they can use it to improve safety. The debate over whether race radios (the two way pocket radios that riders wear to communicate with each other and their team directeurs who follow the race in the team car) should be banned has been ongoing for over a decade as they are a tactical aid to the rider making races more predictable. The counter argument is they can make the racing safer so should be permitted. Direct radio communication between the team car, which will have access to AI tactical modelling, and the riders will further affect race outcomes making an already uneven playing field less even as teams with greater resources will be a step ahead.
AI will revolutionize tactical decision-making in professional cycling, changing the team directeur’s role within the team and how races are ridden. Teams will increasingly rely on computer modelling to compete. The directeur will remain the orchestra conductor within the team; the information Ai produces will only be useful if properly communicated to the riders to bring them together to perform as one.
Just another tool in the quiver in my opinion. I am pretty sure that Sports Directors will still play a huge role in race craft. One area that AI would be useful is in designing safe parcours. I think that the UCI and Race Organizers should spend more time and effort in addressing rider safety. The recent example at the World Championships where there are timing chips and multiple cameras on course and rider crashed and was not found for an extended period of time which may contributed to her passing is just on example.