Traditionally, analyzing athletic performance required specialized equipment and trained professionals. This limited accessibility for athletes and coaches, especially at the grassroots level. However, a new smartphone application called aiScout® might be changing the game.
Researchers at Kingston University are investigating whether the aiScout® mobile application (APP) can provide accurate and reliable assessments to identify aspiring athletes with the help of artificial intelligence, offering a more accessible alternative to conventional methods.
aiScout Application
The aiScout application is the world’s first fully automated platform for discovering and developing talent from around the globe. It was created by the award-winning artificial intelligence company ai.io, which reached out to the sport and exercise science team at Kingston University to validate the app.
Designed to give players from anywhere in the world the chance to be discovered and impress clubs, the app allows users to upload skill-based videos where they participate in tests and exercises such as dribbling, push-ups, jumps, and 10-meter sprints. The videos are then analyzed by AI technology, and users receive scores and feedback. Professional clubs can view the footage and test results and, if deemed good enough, invite the user for in-person trials.
Using AI technology, the exercises assess users’ reaction times, speed, balance, agility, and fatigue levels. These data are then uploaded and stored in the app for analysis by professional football scouts.
Study Highlights the Potential of aiScout®
Ai.io contacted Associate Professor of Sports Biomechanics Dr. Simon Augustus to validate the data and ensure that the AI accurately scored the exercises, which would then be accredited by FIFA, the governing body of football. “Two of our alumni work for ai.io and approached us to conduct some tests. We used the high-tech equipment in our labs, including force plates and timing gates, as well as our Tolworth Court sports field. This allowed us to test all the exercises, collect the data we needed, and analyze them to test the accuracy of the AI model,” he said.
The study, recently published in the International Society of Biomechanics in Sports Proceedings Archive, explored the agreement between aiScout® and conventional methods for assessing athletic performance. The study involved 16 athletes who completed six different tests, including sprints, push-ups, jumps, and dribbling exercises. Both aiScout® and conventional methods were used to assess performance.
The results were promising. In most tests, including sprints, push-ups, lateral jumps, and dribbling exercises, aiScout® showed good to excellent agreement with conventional methods. This suggests that aiScout® can be a reliable tool for assessing performance in these areas.
Clearing the Hurdle: More Research Needed
However, the study also identified a limitation. aiScout® only showed moderate agreement for single-leg countermovement jumps, a test that assesses explosive power. This suggests that aiScout® may not be as accurate for complex jumping movements. More research is needed to understand why this discrepancy exists and how aiScout® can be improved for jump analysis.
aiScout®: A Promising Tool for Athletes and Coaches
Overall, this study highlights the potential of aiScout® as a valuable tool for athletes and coaches. The ability to use a smartphone app for performance assessment offers significant advantages in terms of accessibility and affordability. Athletes can track their progress and identify areas for improvement, while coaches can gain valuable insights into their athletes’ movement patterns.
Users can utilize the feedback they receive to develop and improve their game. “It’s very inclusive—users only need a soccer ball and a smartphone or tablet. There have already been success stories in countries across Africa, Sri Lanka, and India, where significant barriers exist to being discovered due to travel costs and logistics,” said Dr. Augustus.
For his part, Dr. Chris Howe, Director of Research and Innovation at ai.io, explained ai.io’s mission in designing the app in line with the values and skill-based perspective of Kingston University. “We want to level the playing field so that talent is discovered regardless of who they are or where they come from. The sport and exercise science team has done a great job using their expertise to validate the data and the technology’s accuracy so that we can provide an effective and useful tool to athletes worldwide; maybe we’ll even see a player discovered through the app playing in the Premier League in the not-too-distant future.”
The Future of aiScout®
While more research is needed to refine aiScout® for complex movements like jumping, the initial results are encouraging. As the technology continues to evolve, aiScout® has the potential to revolutionize how athletic performance is assessed, making it more accessible and efficient for athletes and coaches at all levels.
Currently, the app is being used by Premier League club Chelsea, Championship club Burnley, as well as teams in the United States’ Major League Soccer. This has already led to several amateur footballers receiving trial offers from professional clubs.
The app can be downloaded for free on any smartphone or tablet via the App Store and Google Play Store.
Reference (Open Access)
Augustus, Simon; Brouner, James; Howe, Chris; Kosciuk, Anna; and Felton-Thomas, Richard (2024) “AGREEMENT BETWEEN A SMARTPHONE-BASED MARKERLESS MOTION CAPTURE APPLICATION (AISCOUT ®) AND CONVENTIONAL ATHLETIC ASSESSMENT,” ISBS Proceedings Archive: Vol. 42: Iss. 1, Article 6. https://commons.nmu.edu/isbs/vol42/iss1/6
Note: Article based on information from the scientific study and the press release “Artificial intelligence helps football scouts unearth hidden talent through new innovative app.”
