The videos and analysis created by ai.io are stored in the cloud, where the AI models perform advanced human movement analysis that looks at skeletal tracking, inverse kinematics and biomechanics
In the competitive world of professional sports, scouts and coaches can have a tough time finding the right talent. And the most talented players can have a tough time getting noticed by the right teams. Worldwide, more than 300 million athletes play football (soccer to those in the U.S.). Aspiring players are desperate to be noticed, while scouts are overwhelmed with that pure volume of players.
Since 2021, British-born ai.io has been working with Intel to make the company’s results faster and more cost-efficient. ai.io began its work with Intel through 3DAT (3D Athlete Tracking), which enables camera data to be captured, rendering 3D data real time and provides 3D sports-biomechanics reporting. Recently, the company has integrated multiple Intel hardware and software solutions across its platform, including Intel Xeon and Intel Core processors, 5G communications technology, OpenVINO to improve efficiency and Intel Gaudi AI accelerators. The company’s custom computer vision models were trained on AWS DL1 instances powered by Gaudi accelerators, which reduces the AI training costs. On inference, its AI models are optimized with OpenVINO and run Intel Xeon Scalable processors.
ai.io came to life in 2017 when its founder, Darren Peries, identified the inequalities between grassroot sports clubs and the larger sports industry. Big sports clubs have plentiful resources, access to key data and contracts worth millions. There was no existing technology that allowed scouts access to reliable, benchmarked data to identify talent – a huge gap for the clubs and the players. Peries pulled together a team of experts focused on data, AI, biomechanics and motion capture and began working with leading Premier League football clubs to develop a solution. The result was the company’s first product: aiScout.
Partnering with pro teams, universities, leagues and federations, ai.io is changing the face of recruiting across sports. “We’ve been working with the ai.io team for the past three years,” said Jim Fraser, academy director for Chelsea FC. “Their ability to push the boundary of technology has already proved to be a game-changer in talent identification, performance analysis and data solutions.”
In India, ai.io is helping surface young football talent. Reliance Foundation Young Champs (RFYC) is a prestigious football academy in India offering scholarships for under-12s. Finding talent is challenging with accessibility issues and the sheer volume of players. During COVID-19 lockdowns, ai.io’s solution enabled the RFYC’s scholarship program to keep moving forward, allowing 7,000 people to participate in trials with the aiScout app.
The ai.io platform identified 400 players who were invited to participate in-person, with 19 signed to the program. Among those signed was a player living in a remote village who used a shared community phone and makeshift “cones” to perform drills. Three players who had never played organized football, but who had enough natural talent and ability recognized by ai.io, also secured spots. ai.io enabled the leveling of a competitive and often unfair playing field, identifying talent that would most likely have been missed.
The videos and analysis created by ai.io are stored in the cloud, where the AI models perform advanced human movement analysis that looks at skeletal tracking, inverse kinematics and biomechanics. Through this process, athletes are given scores helping them pinpoint specific areas to work on. And it allows clubs to benchmark the skillset of the players they have and find players to fill any gaps or areas they are lacking. In the past, analyzed data would need to be disaggregated and sit with different coaches, doctors and trainers. With this technology, it is centralized and tells the whole story.
Another innovative part of the ai.io platform is aiLabs. ai.io’s gold-standard equipment housed in a mobile lab provides sporting organizations with real-time access to data and insights from any location. Analysis helps players fine-tune their performance with specific real-time feedback, identifying areas to work on and pre-emptively spotting possible upcoming injuries.
Richard Felton-Thomas, COO and director of Sport Science at ai.io, joined Intel CEO Pat Gelsinger on stage during the Day 1 Intel Innovation keynote to discuss the way Intel technology has enabled the company to move fast, which is absolutely fundamental in such a quick-paced and competitive industry.
To run the heavy AI workloads needed, Felton-Thomas described how important the partnership with Intel has been, allowing ai.ao to continue to scale into more sports and positively affect the scouting process around the globe. “The critical thing for players and clubs is speed,” Felton-Thomas said. “The videos go into the cloud, where heavy AI workloads get run, and Intel was best in class for supporting us with that.”
