Van Hall Larenstein University of Applied Sciences (HVHL) and EQAD have launched a collaboration to develop a new hybrid, AI-powered judging model for equestrian sport. The initiative aims to strengthen objectivity and transparency in evaluation while explicitly preserving the role of professional expertise and practical applicability. Rather than replacing judges, the model is designed to support decision-making by combining data-driven analyses with established evaluation principles and real competition practice. The project builds on several years of applied research in dressage judging decision-making, biomechanics, and welfare-related indicators in equestrian sport.
“Judging in equestrian sport has always been an interplay of clear principles and expert interpretation. Our goal is not to automate judging, but to make the underlying criteria more consistent, transparent, and comprehensible – especially where objectivity matters most.”
A deliberately hybrid approach
At the heart of the collaboration is a hybrid architecture. Artificial intelligence is deployed where it provides clear added value, such as automated recognition of anatomical markers and rapid biomechanical pre-analyses. This allows judges to focus more on interpretation rather than manually capturing data.
This approach is a deliberate decision to bridge the gap between technological innovation and evaluation practice. By embedding the model into existing evaluation systems, the partners aim to avoid black-box solutions and ensure interpretability, explainability, and trust.
“Technology should serve the sport, not redefine it from the outside. By working closely with researchers and practitioners, we're developing a system that aligns with how evaluation actually works – in the arena, under real conditions.”
Focus on consistency, welfare, and transparency
The collaboration also addresses broader discussions in equestrian sport about inter-judge consistency, criterion clarity, and horse welfare protection. By translating qualitative descriptions into measurable, interpretable parameters, the hybrid model aims to support fairer evaluations and clearer feedback without losing practical relevance.
Importantly, the final decision authority always remains with humans. The system is explicitly designed as decision support, not as an autonomous judging system.
Outlook
The partners are currently working on further development and validation of the model in close dialogue with practice, taking steps toward a more transparent, evidence-based future of evaluation in equestrian sport.