AI at the Paris 2024 Olympics: A Q&A with PAI Partner Intel

Thalia Khan: Intel is utilizing AI technology to enhance the performance and experience of the Olympics for athletes and fans, can you tell us how you are doing this?

Sarah Vickers: We are working with a number of stakeholders in the Olympic ecosystem, including the International Olympic Committee (IOC), Olympic Broadcasting Services (OBS) and the Paris 2024 Organizing Committee to utilize artificial intelligence platforms to help solve problems. That could be anywhere from addressing the operational challenges of delivering such a complex event, enhancing the fan experience and meeting the demands of the viewers at home. We’re really thinking about how AI can enhance the experience operationally from an environmental and legacy perspective— for the fan experience both on the ground and at home.

TK: How does Intel ensure the safety and security of its AI applications in a high stakes environment like the Olympics?

SV: We know the industry is evolving quickly and Intel is very disciplined in how we approach our solutions with AI. We follow a set of responsible AI practices and consistently collaborate internally to ensure that we are enabling AI responsibly. This applies to our entire approach to AI, but in regard to the Games, ensuring that the data sets are gathered securely, that comparisons are done in the right way, and that we remove as much bias as possible are paramount in our approach.

TK: I read that Intel used AI and digital twinning technology for venue planning for the Olympics. How does this contribute to the operational efficiency of the event and sustainability?

SV: From a sustainability perspective, there are a couple of ways it helps. One is that if you can provide people with digital twins, there’s a reduced need for travel because they can see the space, experience the space, and test things out virtually. Traditionally, people would have to travel to the venues to see and experience them, sometimes even multiple times. When you think about operational efficiency, the ways that digital twins can help are really vast. There are so many different ways that they can help you plan for providing onsite fans with the best experience while simultaneously ensuring that the broadcast is the best it can be.

As an example, many of the venues for these Olympic Games are temporary. If you think about beach volleyball at the Eiffel Tower, where you can’t necessarily plan for a physical venue a year in advance, the digital twin can help you understand its exit routes, people flow, broadcast cameras, or what the impact of the sun will be on the attendees.

Another area where we’ve found that digital twins are useful is helping the transition from the Olympic Games to Paralympic Games because many venues will be used for both. As an example, in venues where you know there will be athletes navigating in wheelchairs, you can use the digital twin of the space to identify where ramps can be installed. The digital twin enables you to identify opportunities or challenges ahead of time, so you’re not dealing with them when they happen.

TK: One of the ways Intel is supporting the event is by enhancing accessibility for visually impaired people. Could you describe how AI is being used to improve their experience?

SV: We’re partnering with Team USA to deploy a wayfinding solution at their training facility just outside of Paris, where many athletes visit to prepare ahead of the Games. The tool that we’re using was originally designed to support the visually impaired, but it actually has multiple applications.

For example, if you are in a wheelchair or have mobility issues, you can use the tool to identify a step free option. It can also help people with hearing impairments move through the venue.

The tool provides people step-by-step directions on how to navigate the space. For someone who is visually impaired and visiting a space for the first time, finding areas like the restroom or where to eat traditionally required the help of a guide. This technology gives people more independence to navigate those spaces more freely.

What we would love to see with the application of this technology in the future—and we’re already seeing it more and more—is that it will be available everywhere for fans in the stands or at train stations and airports to really help people navigate spaces more independently.

TK: How does Intel’s AI technology help in creating dynamic broadcasting snippets that capture key moments of the events?

SV: We’re working with OBS to bring AI-generated highlights to Paris 2024. Traditionally broadcast highlights have been done manually through individually tagging footage and the amount of highlights was limited by the number of people who could do that. Because there’s so much content from the Olympics and Paralympics, this meant there were a limited number of highlights that could be produced.

If you think about the most popular sports or the largest countries, they tend to get lots of highlights. But for the less mainstream sports or countries that have a smaller viewing audience, they often couldn’t produce all of the highlight content that viewers want.

With AI and automation, OBS and Media Rights Holders (MRHs) can create as many highlights as they’d like. It’s really endless. This results in a more dynamic experience for the viewer at home. For example, if you’re a fan of archery and there is somebody representing your home country at the Games, but you happen to live in a country with a smaller viewing audience, historically it would have been really hard to find the footage of them competing. But with the help of AI automated highlights, it’s easier to watch your favorite athlete.

TK: As a partner with Partnership on AI, Intel is committed to responsible AI practices. How has this approach to developing and deploying responsible AI positively influenced the way you support the Olympics?

SV: We’ve worked closely with the IOC on their AI Agenda, which is their long-term strategy on how they’re approaching AI. By working across teams and understanding our responsible AI practices, we’ve been able to ensure that our knowledge and awareness of what we’ve learned is being passed on throughout the ecosystem.