Artificial intelligence to make Olympic Games more inclusive | Computer Weekly

pigprox - Fotolia

In just a few weeks , the Olympic flame will reach Paris , and some 10,500 athletes from 206 countries, 20,000 journalists and more than 15 million tourists will flood the French capital.

The International Olympic Committee (IOC) will employ AI technology to make sports more inclusive, competitive and engaging in the upcoming Games. This innovative approach, championed by IOC president Thomas Bach, is not just about embracing AI, but about charting new sports and Games paths. “With this, we are embarking on a thrilling new era of sports,” he said.

In a strategic partnership with Intel, which has integrated AI into all of its products as part of its AI Everywhere strategy, the IOC has harnes se d the power of AI to ensure the Games are fair and inclusive.

This includes talent identification, personali se d training methods, and the enhancement of sports equipment and programmes. AI also plays a pivotal role in promoting fairness by enabling accurate judging, optimising event organisation, transforming sports broadcasts, enhancing the audience experience and making it more personali se d .

The IOC’s most impactful use of AI is through digital twins , said Ilario Corna , chief technology officer (CTO) of the sports organisation. “These digital models have been instrumental for event planning, staff training and, most importantly, improving accessibility, demonstrating our commitment to inclusivity in these Games. ”

The IOC leverages digital twinning to optimi se various aspects of planning and execution. “A digital model of an Olympic venue, for example, facilitates collaboration, improves access to information and supports decision-making,” he said. “This reduces planning errors and provides accurate, up-to-date information about and during events. ”

A practical example is the 3D models created for various facilities in Paris. Combined with AI and computer vision, the se models enable indoor and speech navigation through smartphone applications. “By using real-time data and detailed 3D environmental models, visually impaired u se rs can receive accurate directions and information about their surroundings, improving their mobility and independence. ”

The Olympic Broadcasting Se rvices (OBS) produces all live TV, radio and digital content during the Games. “In Paris, we will be recording in 8K resolution for the first time,” said Sotiris Salamouris , CTO at OBS.

“This is se en as a significant step forward in video quality, although it demands higher storage and processing capacity,” he said. “The current distribution standard is 4K UHD HDR, but 8K production will also occur for certain key events such as athletics, urban sports, and the opening and closing ceremonies . ”

Intel is crucial in enabling this 8K production by providing advanced compression technologies and processing with their latest generation of processors. “This ensures that the massive amounts of data associated with 8K can be effectively managed and distributed worldwide without requiring excessive energy and resources,” said Salamouris .

OBS provides neutral, impartial footage to broadcasting organisations worldwide, which can then create their own stories about national athletes and teams. “We offer AI services based on Intel technology that allows those broadcasters to quickly and easily extract specific content or individuals from the material they receive,” said Guillermo Jimenez Navarro , director of broadcast engineering at OBS.

“AI can help editors in different countries quickly create daily highlights of specific sports or athletes,” he said. “We will produce about 11,000 hours of original footage this summer. You can imagine how much time and money it would take if all that material had to be manually sorted and reviewed by humans. ”

The talent identification platform is another AI application tested by the IOC in collaboration with Intel. “This AI platform helps analyse massive amounts of data to discover potential Olympic talent, even in remote areas,” said Sarah Vickers , head of Intel’s Olympic and Paralympic programme. “Together, we have tested thousands of young athletes and analysed the results using millions of data points and metrics to identify exceptional talent. ”

For Corna, this is a special project. “It can potentially change the future of many young athletes. ”

AI can also be a valuable addition to scoring and judging. Using advanced imaging techniques, AI systems can perform real-time analy se s of athletes’ performances, which is very u se ful in sports such as gymnastics and diving, where performances are not always easy to judge with the naked eye. AI can accurately compare athletes’ movements with pre se t criteria, resulting in more consistent and objective scores, and reducing human bias.

The biggest challenge for the International Olympic Committee (IOC) regarding AI at the upcoming Olympic Games is balancing the technology’s benefits with managing the associated risks and ethical considerations.

“Within our organisation, we have a team dedicated to responsible AI,” said Corna. “They work on developing and understanding AI systems to eliminate as much bias as possible in the results. ”

A vital aspect of this is ensuring the privacy and se curity of athletes, spectators and other stakeholders. AI systems collect and analy se large amounts of data, posing risks to data se curity and privacy. It’s crucial that the se systems are free from bia se s, and that decisions and analy se s are fair and transparent. “This remains an ongoing challenge,” he said.

But Corna doesn’t se e significant risks regarding AI. “Which is not to say they don’t exist,” he added. “We take the time and do our best to identify risks thoroughly . We have found some u se ca se s that we believe do not po se significant risks, but we continue to look for potentially greater risks. ”

To mitigate potential risks, the IOC has developed an internal framework for all of the se u se ca se s. “It is an internal framework, with the primary component being integrity, ensuring that there is no bias in the algorithms,” said Corna. “We take this very se riously. ”

For him, AI systems do not replace the human factor behind the scenes. “Our vision is that AI i nstead support s them, enabling us to innovate the Olympic and Paralympic Games continually, ” concluded Corna.

A source: www.computerweekly.com/news/366588823/Artificial-intelligence-to-make-Olympic-Games-more-inclusive