ENGLISH TEXTS
G. So, for or against?
G. V. For surveillance, definitely. Against, personally, when it comes to measuring racing performance and planning matings. I’ m not the type to hand over my judgment to a machine. No way I’ m giving up my breeder’ s soul. Breeding is not mathematics. Sure, tech is in the air and we’ re testing the waters. I’ ve seen good ideas in these new tools, but you can’ t replace the human touch. We need to stay reasonable.
� PAGE 51
BACK TO THE ARTICLE �
IA IN RACING, IN THE SAME STRIDES AS OTHER SPORTS?
AI has found its place in training around the globe, mainly through data processing. But what about France? How is it making its way into a profession rooted in observation and intuition? A deep dive into the tracks and the digital world.
� PAGE 42
ARIONEO
By Mégane Martins
“ IT WAS SPORT THAT INSPIRED US”
Arioneo is one of the first companies in the French equine industry to specialize in data collection and equine data analysis. Founded in 2014 by Erwan Mellerio and Valentin Rapin, the startup successfully completed a first fundraising round of € 1.1 million eighteen months after its launch and is now part of the LIM Group. Arioneo has since maintained a strong innovation strategy and is now considered one of the leading players in its market worldwide, with a strong presence in North America, the Gulf countries and Australia. Valentin Rapin explains Equimètre, a tool that uses artificial intelligence to provide a range of data to analyze both the performance and health of horses, whether in racing or training.
Galorama. Equimetre positions itself as a pioneer in measuring racehorse performance. How did you come up with your tool?
Valentin Rapin. The idea for Arioneo was born in 2014. We were inspired by the world of sports, where tools had started to emerge to capture players’ physiological parameters, followed by platforms to display that data, and finally data analysts to interpret it for coaches.
G. What were the different development phases?
V. R. After engaging with industry professionals and institutions, we launched the prototyping phase, which involves collecting a large amount of data to later develop the algorithms, some of which use AI. The goal was to ensure that the data interpreted by the end user is reliable and accurate. We chose to keep our engineering team in-house, so we have all the expertise close at hand during product development. The prototype is then field-tested before being brought to market.
G. What data did you rely on to develop the tool?
V. R. We use GPS data to calculate distance, trajectory, speed, acceleration and split times- all the metrics that allow us to objectively assess a horse’ s acceleration, top speeds, speed endurance, etc. We also use data from IMUs- motion sensors that measure locomotion factors like stride length, cadence, symmetry, and regularity. And finally, we use heart data to monitor heart rate as well as electrocardiograms during exertion- when the horse is going over 60 km / h. For this, we developed our own electrodes, which are patented and scientifically validated for ECG acquisition during effort. This allows us to detect cardiac issues such as various
145