A groundbreaking research project funded by Lero, the Research Ireland Centre for Software, and the IRFU is deploying artificial intelligence (AI) to analyse the tackle event in rugby to enhance player welfare and performance, resulting in a more exciting and dynamic game.
The project team, led by Professor Anthony Ventresque, Director of the Complex Software Lab at the School of Computer Science and Statistics at Trinity College Dublin, has the potential to provide coaches, players and referees with incredible insights into tackle technique, identify areas for improvement, and ultimately reduce the risk of injury.
“Our research is focused on developing AI that can understand the complexities of rugby tackles. By analysing large amounts of video data, we can identify patterns and trends that may not be apparent to the human eye. This information can be used to develop targeted training programs to improve tackle technique and player safety.”
PhD researchers Will Connors and Caoilfhionn Ní Dheoráin, are teaming up with Dr Kathryn Dane to harness the power of AI in the world of rugby. This collaborative research project aims to develop AI models capable of automatically identifying and analysing the tackle event, with the potential of improving training techniques.
Will Connors, who has represented Ireland at senior, U20, and sevens levels, said that as a rugby player with a Computer Science background, he is fascinated by AI’s potential to analyse and optimise tackle technique.
“I believe this research can help players at all levels improve their tackling skills and contribute to a more exciting and dynamic game.”
Dr Kathryn Dane, who has also represented Ireland at senior international level, said this project highlights the crucial link between technique and safety in rugby at all levels of the game.
“By using AI to analyse a large number of tackles, we can identify specific areas where technique can be improved to enhance both performance and player welfare.”
Computer Scientist Caoilfhionn Ní Dheoráin said she is excited by the challenge of applying Machine Learning at scale to analyse rugby tackles in the domestic club and school game.
“This project offers a unique opportunity to push the boundaries of AI and contribute to a deeper understanding of this complex and dynamic sport.”
The IRFU’s Medical Manager, Dr Caithríona Yeomans, who holds a PhD in Sports Sciences, said this research will be hugely helpful to enhancing player welfare in rugby
“By understanding the mechanics of tackles and identifying areas for improvement, we can help players develop safer and more effective techniques. The collaboration with Lero and the Complex Software Lab at Trinity College Dublin is invaluable in our ongoing efforts to make rugby a safer sport for all.”
This collaboration stems from the IRFU’s decision to lower the tackle height in the domestic game. The insights gained from the video analysis will help identify the trial’s impact on player welfare and the overall game.
The IRFU’s National Rugby Development Manager and Tackle Trial project lead, Colm Finnegan, says: “We are excited to work with Lero to be at the forefront of innovation in such an important area of Rugby, which reaffirms our aim of making the sport as safe as possible whilst also being an enjoyable game for all.”
One of the project’s recent publications, ‘Frisbees and Dogs: Domain Adaptation For Object Detection with Limited Labels in Rugby Data’, explores how AI can be trained to accurately detect essential elements in rugby videos, even with limited training data. This is a significant breakthrough in this area.
Another recent publication, ‘Are we tackle ready? Cross-sectional video analysis of match tackle characteristics in elite women’s Rugby Union’ examines tackle techniques in the women’s game. This study found that many tackles lacked full completion of World Rugby’s ‘Tackle Ready’ recommended techniques, highlighting the need for targeted interventions to reduce injury risk and improve performance.
The project’s findings are being presented at prestigious international conferences and published in leading academic journals. The research team continues developing their AI models and plans to make their findings available to coaches and players at all levels of the game.
See more stories here.
