ENGLISH TEXTS
types of arrhythmias, atrial fibrillation, etc. With all these parameters, we can analyze the horse’ s level of exertion and identify potential pain through abnormal heart rate variations. We also incorporate external data to support the analysis.
G. How does AI make your tool more powerful?
V. R. At Arioneo, we use three types of AI. First, predictive AI to analyze the horse’ s performance and health. It allows us to generate forecasts for performance, identify trends in physiological and locomotion data, estimate a horse’ s optimal form and create predictive models. This helps us anticipate injury risks, for instance. We also use AI to detect arrhythmias in ECGs, and even classify them- this saves valuable time for veterinarians. It’ s part of a research project we’ re conducting in the United States, particularly focused on sudden death in racehorses, where we’ re trying to identify early indicators that could help prevent such incidents. We’ re also working on fracture detection and early warning signs during training or racing. Then there’ s generative AI, which uses this data to automatically generate analysis sentences in reports, to pre-interpret the data for users and offer recommendations, for example. We recently launched Arion Intelligence, which generates the PDF reports sent to trainers, with the goal of saving them time. Lastly, we use machine learning, sensor-based technology that can automatically detect whether the horse is walking, trotting, cantering, or galloping.
G. How was your tool received in the beginning?
V. R. The perception varied widely depending on the country, and it still does today. As with any innovation, there are early adopters, and others who prefer to stick with traditional methods. Australia, for example, with its strong sports and tech culture, was very receptive- today, sensors are fully integrated into racehorse training, and almost all trainers use them. Some even hire dedicated teams just for data analysis and processing. We’ ve seen real evolution in how the product is used from the beginning to today. Some trainers waited to see how the tool performed with others before adopting it themselves.
� PAGE 54
BACK TO THE ARTICLE �
By Olivier Villepreux
AI MAKES ITS WAY INTO ALL SPORTS, START- ING WITH RUGBY
AI specialists have started appearing within coaching staffs. In rugby, for instance, its use ranges from play- ers’ physical preparation to assisting coaches in decision-making.
Data has always been an intrinsic part of sports. It adds an extra layer of interest beyond the result of a race or competition, both for the public and for the participants. If we consider just the notion of the“ record”( which first appeared in men’ s athletics in 1914), it helped create standards( qualifying times, performance thresholds), and made it possible to study in more detail how and why one athlete might outperform another under certain conditions, or over varying timeframes. The work of data collection, therefore, is nothing new, but the way that data is interpreted has varied widely over time. This depends on whether it’ s being used to shape strategy, especially in team sports, or to break down a performance in hopes of replicating it through training inspired by another athlete’ s technique or preparation. Despite the abundance of information sources, sensors, images, statistics, medical data, human beings still remain at the heart of athletic or team preparation, even if AI has entered the sports world as quickly as it responds to a question. But caution is advised, as Saad Drissi, head of data for the Stade Toulousain rugby team, reminds us:“ Context determines how algorithmic processes are used.”
146