Can AI Revolutionize Dressage Judging?

Technology is changing at lightning speed and sports are adapting. But in the traditional world of dressage, can equestrian sport adapt and embrace new technologies? There are so many questions!

Can artificial intelligence (AI) revolutionize dressage and other subjective sports judging?
What is the potential of AI in assessing images, videos, and live events?
Does AI have the capacity to identify the beauty, artistry and harmony of dressage?
What is the feasibility of its implementation, public acceptance, and the challenges it may encounter from major stakeholders?
Could it be the future of dressage judging at major games?

Images, Video, and Live Events

Currently, most AI is focused on text and numerical data and information. But the progress in AI and computer vision technologies is moving at a rapid pace. Today’s algorithms can analyze images and videos, identify patterns, and even gauge human emotions. There is AI assisting other sports at top levels, such as with video replay and supporting referee decisions. In tennis, Hawk-Eye technology uses cameras and AI to track the trajectory of the ball and determine whether it is in or out of bounds.

Exactly how can it assist in dressage judging? AI can be ‘trained’ to detect certain aspects of dressage movements, such as rhythm and consistency of step, counting steps in changes or piaffe, whether the poll stays at the highest point, and many other details. The true challenge lies in replicating the nuanced understanding, emotional comprehension and artistry as seen by human judges.

Scroll to continue with content

The applications for future use in the sport are far-reaching, such as providing anyone in the world access to AI assessment: having your performances compared to each other to see how the horse is improving or struggling, detecting soundness issues early due to changes, and so on. AI could also introduce dressage to the public with in-depth information on how horses move, with data for livestream commentary.

AI and Beauty, Harmony, and Artistry

This is going to be a major challenge, since beauty and artistry, by nature, are subjective concepts that elude precise quantification. To assess these elements accurately, AI must utilize sophisticated algorithms capable of processing complex sensory inputs and recognizing subtle emotional cues in both horse and rider. This would require extensive training on vast amounts of high-quality data to gain a deep understanding of the intricate interactions between horse and rider, accounting for elements like footfalls, balance, and communication cues. Developing AI with such abilities represents a significant technical challenge.

While AI may excel at analyzing objective factors such as symmetry, precision, and synchronization, it may struggle to capture the intangible essence of artistic expression. The human element of subjective judgment, driven by emotions and experience, adds a unique dimension to equestrian sports and other art forms.

Perhaps judging may evolve into a combination of AI and human input. On one side, AI assesses the components of the performance it can objectively. On the other side, the human element provides the “je ne sais quoi” and artistry of a moving performance. In an interview with Jan Tönjes in The Horse, Jan mentions “…there’s no identically moving horse, we know that, but this [AI] could be … a computer to assist. It could all … help to support dressage.”

Challenges and Roadblocks

While AI judging holds great promise, there will be numerous challenges, with #1 being convincing stakeholders to make the change and accept AI’s credibility and fairness. This includes riders, trainers, and the broader equestrian community, whose sense of tradition and resistance to change will be factors. The replacement of human judges with AI may face resistance from traditionalists who cherish the emotional connection and expertise brought by experienced judges. Some may argue that AI lacks the intuition and profound understanding of equestrian artistry, potentially undermining the authenticity of the sport.

Canadian Olympian Shannon Dueck sees the use of AI already with the freestyle Degree of Difficulty program into which all riders input their freestyle choreography, and sees possibilities for it in the future of judging. “Not one hundred per cent, but maybe there is a way for it to help the judges? We still need their eye and feel [but] I think it is the future for sure.”

The new FEI Dressage director Ronan Murphy is open-minded to the possibility of AI assisting judges. In a Eurodressage article concluding the 2023 European Dressage Championships, he said “These technologies are there to assist, we use them in veterinary inspections, we use them in all sorts of ways, and I think if they add something to our sport, of course we will look to that. But at the moment we are completely relying on the knowledge and expertise of our fabulous judges.”

Japanese IT manufacturer Fujitsū has been using AI in gymnastics scoring since 2019. (Fujitsū photo)

Dressage icon Isabel Werth also added to Ronan’s comments at the same press conference “It’s very interesting to see how it improves and how we can use it as a tool and as an assistant for our judging system.” said Werth. “There’s always something to get better at, there’s always something to discuss. And I think whatever we can use to make it more objective, to make it more transparent, that helps and would be a good idea.”

Already in Play

A revolutionary AI-based scoring system made its debut at the 2019 World Artistic Gymnastic Championships in Stuttgart, Germany, solving many of the problems that plagued judging in that sport in the past. Developed by Fujitsū, the system comprises a scanning unit shining two million laser pulses per second on the gymnast, and a database of gymnastic routines for comparison. The system identifies elements subject to additions or deductions and displays the result as a score. While judges were concerned that AI would threaten their jobs, the benefits to gymnastics included more accurate and fairer scoring, which were hailed by the gymnasts themselves.

The potential of AI to revolutionize dressage judging and other subjective sports judging presents both excitement and challenges. While AI has made significant strides in assessing sports performances, capturing the emotional depth and artistic essence of equestrian sports remains. Change is inevitable – the question is, when?