½ñÈÕºÚÁÏ

The University of Malta presented research on using Artificial Intelligence (AI) to support swimming performance at the - Swimming, held in Maia, Portugal, on 18 and 19 September 2026.

Associate Professor Vanessa Camilleri, from the Department of Artificial Intelligence within the Faculty of ICT, delivered a dedicated European Aquatics Innovation Hub session introducing SWIM-360, a University of Malta-led research project investigating how multimodal and explainable AI can give coaches richer, more accessible information about swimmer performance.

The one-hour session, “SWIM-360 project: an AI-driven coaching support platform”, formed part of an international programme bringing together expertise in sport psychology, environmental physiology, swimming talent development, research and elite performance.

A central message of the intervention was that AI should support, not replace, the coach. SWIM-360 is being developed around this principle, bringing together sources of information that can be difficult to capture and interpret simultaneously through conventional poolside observation. These include wrist-mounted inertial sensors, muscle oxygenation measurements, and dual-perspective above- and underwater video.

The session took coaches through the research pipeline, from data collection and dataset development to AI-based pose estimation, automated video analysis, and the project's developing coaching dashboard. The computer vision component explores state-of-the-art pose estimation approaches to analyse swimming movement, while ongoing work is investigating how video and sensor information can be synchronised and interpreted within a unified multimodal system.

Particular attention was given to explainable AI (XAI). Rather than presenting coaches with unexplained algorithmic outputs, the research aims to make AI-generated insights interpretable, allowing coaches to understand why particular aspects of a lap or stroke have been highlighted and to consider this information alongside their own expertise and knowledge of the athlete.

The presentation generated substantive discussion with coaches working in elite and Olympic-level swimming environments around the role of technology in contemporary performance analysis. The exchange highlighted the increasingly sophisticated use of computer vision, data analytics, and performance technologies at the highest levels of the sport, where extremely fine margins can be significant.

These discussions also provided valuable perspectives for the continuing development of SWIM-360. While high-performance environments may have access to specialised technologies, expertise, and extensive performance data, the exchange raised important questions about how research in AI and sports analytics can be translated into accessible, meaningful tools for coaches working with different levels of technological and specialist support.

Coaches attending the session were also invited to explore the SWIM-360 dashboard and provide feedback on the metrics, information, and forms of AI assistance that would be most useful in coaching practice, creating a direct link between ongoing research and practitioner needs.

The European Aquatics conference therefore provided an opportunity not only to showcase AI research being undertaken at the University of Malta, but also to engage directly with the international high-performance coaching community and to reflect on how future research can respond to the diverse technological and practical realities of swimming environments.

SWIM-360 is financed by and the Malta Digital Innovation Authority (MDIA) through the R&I Thematic Programmes: Digital Technologies Programme.