The Use of AI at the Olympic Games in Paris and Beyond

The Olympic spirit of continuous innovation and pursuit of excellence aligns perfectly with  modern technological advancements, so the use of artificial intelligence (AI) for the Olympic Games was only a matter of time.

 

The world has seen more AI and tech innovations at Paris 2024 than ever before, from AI support for athletes to revolutionary broadcasting.

Read on to learn how artificial intelligence is revolutionizing the Olympic Games experience for organizers, athletes, coaches, judges, spectators, tourists, broadcasters, and viewers worldwide. If you are contemplating a project that includes AI/ML (machine learning), please feel free to contact Onix.

 

We build and train ML models for various purposes, such as:

 

• image classification
• language identification
• news categorization
• social media content analysis
• social media sentiment analysis
• translation of graphic novels and webtoons into multiple languages
• comic strips animations
• conversational engines based on natural language processing & deep learning
• image denoising in info-communication systems
• product type and style recognition
• face recognition and manipulations in livestreams
• crowd video analysis and behavior pattern recognition
• mapping and geographic information systems
• green energy output prediction

The Olympic AI Agenda

Athletes, stakeholders, and the public had a chance to witness AI applications in the Olympic Games contexts before, for instance, at PyeongChang 2018 and Tokyo 2020. The digital revolution has been brewing since the International Olympic Committee (IOC) adopted the Olympic Agenda 2020+5 in 2021. 

 

However, only in April 2024 did the IOC present their ideas of what AI can deliver for sport in a formal AI agenda. It should streamline the growth of AI research and strategize the use of AI in the Olympics. The Paris 2024 Olympic and Paralympic Games were the first where this AI strategy was implemented.

 

Read also: Some Use Cases and Benefits of VR Sports Training

 

The Official Olympic Partners Alibaba, Deloitte, Intel, and OMEGA, and the Olympic Media Rights-Holders, such as Discovery, NBC, and Warner Bros., supported this initiative. For example, Intel’s AI platform technology was to be embedded throughout the Games experience in three key areas:

 

• creating an immersive sports experience for fans using innovative technology to measure athletic potential
• facilitating an end-to-end 8K broadcasting experience
• ensuring universal accessibility at selected Olympic and Paralympic facilities

 

Read also: How AI Is Used in Euro 2024 & 6 Ways You Can Use It Too

 

 

 

How AI Can Be Used to Enhance the Olympic Games Experience

Deloitte distinguished four impact areas of AI at the Olympic Games:

 

• Competition
• Consumption
• Management
• Monetization

 

Let’s follow this logic when reviewing the use cases of AI at the Paris 2024 Olympics.

 

Read also: How AI Can Transform Your Business

 

The Impact of AI Applications on the Athletes, Coaches, and Judges

 

AI enhances training and performance

For instance, Chinese AI tools reportedly assisted athletes with analyzing their performances. AI company SenseTime partnered with the China men’s basketball team, providing AI-driven sports data analysis and advice on strategy during the competition. 

 

China’s diving team used Baidu’s large language model (LLM) to supply accurate information and even score diving movements in real time for precise analysis and guidance.

 

Learn more: How to Build a GPT Model: Prerequisites and Essential Steps

 

AI assists judges

The IOC also expected AI tools to “enhance the fairness and accuracy of judging and refereeing through the provision of precise metrics” at Paris 2024.

 

AI’s data comprehension and analysis capabilities already improve officials’ judging accuracy by supplying verifiable information they can’t detect with the naked eye. For example, AI instruments integrated with cameras can track divers’ rotations or jump height.

 

For the Paris 2024 Olympics, the Official Timekeeper OMEGA provided several AI-powered solutions. Alain Zobrist, the chief executive of OMEGA Timing, said they intended to measure competition without disturbing athletes and to convince athletes to trust their methods.

 

“This is why Computervision and AI are so helpful,” he said. “It’s another way of how to measure time, how to measure time and explain performance.”

 

For example, two points are deducted from a diver’s score if their head passes dangerously close to the board. In Paris, a computerized camera (which Zobrist called “Computervision”) measured the distance between the head and the board to inform the judges’ decisions. 

 

 

AI guides athletes around

The IOC and Intel developed the AthleteGPT chatbot to assist thousands of athletes from around 200 countries participating in the Paris 2024 Olympics. This LLM was built using Intel’s Gaudi processors and an AI developed by Paris-based company Mistral AI. 

 

This AI chatbot, accessible through the Athlete365 mobile app, is designed to promptly answer FAQs regarding navigation within the Olympic Village, social media guidelines, anti-doping rules, and Rule 50 regulations. It can quickly scour through thousands of information pages to answer questions like how to get around Paris or where to eat 24/7.  

 

Read also: Hyper-Personalization in Hospitality: Reaching New Heights

 

Sarah Vickers, the leader of Intel’s Olympic and Paralympic Games Office, said: “We’ve heard really positive feedback that we’re taking away that pain for athletes to let them focus on their performance.”

 

Learn more: How to Create a Chatbot: The Ultimate Development Guide

 

AI protects athletes

In recent years, the Olympic movement has increasingly focused on athletes’ mental health and well-being.

 

Read also: Innovating Mental Health App Development: Features & Cost

Cyberbullying and online abuse are growing issues. Many high-profile athletes claim that they affect their mental health and performance at the games and call for measures to protect them.

 

The IOC expected the 2024 Summer Olympics and Paralympics to generate millions of social media posts, excluding the comments, about some 15,000 athletes with entourages and 2,000 officials. Cheering and congratulations are often mixed with hate, coordinated abuse, and even threats.

 

Although it’s possible to set up a filter detecting racial slurs and other noxious words, languages are more subtle than that. The IOC needed a tool to process an ocean of content and sort through meaning. LLMs can analyze the sentiment behind a piece of text even if it doesn’t contain tell-tale words. 

 

So, the IOC implemented Threat Matrix – Signify’s LLM-based monitoring system that scans social media platforms for abusive messages directed at athletes and flags harmful content for intervention by the partner platforms: Facebook, Instagram, TikTok, and X.

 

Learn more: How to Create a Machine Learning Model for Your Business

 

Threat Matrix can process text in 35 languages, discerning nuances varying across regions, read into sarcasm, and even understand how images and emojis shift a post’s meaning.

 

When it spots a problem in posts, a response team looks for context that the AI might have missed and, if needed, acts to mitigate problems: issue takedown requests for posts that violate social media policies, reach out to the victim to offer support, or even contact law enforcement about threats of violence. Often, they do it before the content has a chance to reach the target. 

 

Threat Matrix will also gather information to help the IOC understand online abuse and improve the protection of athletes and judges during future Olympics.

 

How AI Is Transforming the Olympic Experience for Spectators and Viewers Worldwide

 

AI enriches the fan experience on site

Intel’s stand at the Stade de France caused much buzz. Their interactive AI Platform Experience uses AI and computer vision to analyze volunteers’ athletic abilities and drills and match their profile to one of 12 Olympic sports. 

 

When a person enters a dark box with bright lights on the inside, a camera takes their body measurements. Then, they must run on the spot as fast as possible, squeeze devices as hard as possible, and touch flashing lights as quickly as possible. They conclude the tests with a final sprint and a superhero pose. When the system produces their result, it flashes brightly.

 

 

This AI-powered fan activation was trained on Intel Gaudi accelerators, running on Intel Xeon processors with built-in AI acceleration, and optimized with OpenVINO.

 

AI promotes accessibility and inclusivity

Intel is also using AI to make the Olympic and Paralympic Games in Paris more inclusive and accessible to visually impaired guests. Their AI solution was built on Intel Xeon to create 3D models of the Olympic facilities. A dedicated mobile app uses those models to provide indoor and voice navigation. 

AI enriches viewer experiences

AI is reshaping the Olympic experience for billions of viewers across linear television and digital platforms worldwide. 

 

In collaboration with the IOC’s commercial and broadcast partners, the Olympic Broadcasting Services (OBS) used AI to improve internal workflows and enrich the viewer experience with statistics, numbers, extra benchmarks for comparison, better storytelling, and personalized insights into sports events. 

 

Particularly, Chinese AI-driven technology provided advanced solutions for real-time live commentary, 360-degree live-streaming technology, and visual searches. 

 

For instance, Alibaba supplied a record number of multi-camera replay systems with AI-powered high-quality reconstruction in the cloud. This enabled OBS to create 3D models and additional viewpoint mapping to deliver more compelling replays from more camera angles for 21 sports and disciplines.

 

 

Alibaba’s Tongyi Qianwen LLM assisted the IOC’s official commentators.

 

The collaboration with OMEGA enabled OBS to deliver faster, more relevant, and insightful data. For example, intelligent stroboscopic analysis across diving, track and field, and artistic gymnastics improved the commentators’ and viewers’ understanding of the athletes’ movements and biomechanics. 

 

In the Olympic Aquatic Center, OMEGA set up four high-tech cameras to capture everything in the pool. These cameras and the computerized system operating them were trained to recognize and analyze specific movements. 

 

During a race, these cameras can calculate each swimmer’s stroke rate and the distance they are covering, compare that to the distance they have covered, the distance to go, and the other swimmers’ data, and figure out the tiny differences in acceleration and deceleration at critical moments in the race in real time.

 

 

In diving, OBS and OMEGA used AI to deliver enhanced graphics breaking down each athlete’s performance in the air and when entering the water.

 

During beach volleyball matches, smart cameras captured and tabulated each player’s movement, calculating the distance they covered in each match, the speed of the ball, and a data-driven understanding of tactics.

 

AI-based motion tracking technology also helped commentators and viewers monitor athletes’ positions during canoe sprints, marathons, race walks, road cycling, mountain biking, marathon swimming, rowing, sailing, and triathlon competitions.

 

Athletes pin a smart bib to their bodies during track and field events. A sensor inside constantly transmits data about the runner, just like the cameras collect data about swimmers.

 

A number of antennas send information to high-speed computers that calculate all the athletes’ positions on the track, their steps, stride rate, and which direction they are taking. The antennas send nearly 2K data points per second to the timekeeping room. This allows viewers to know who is gaining ground, who is losing it, and who is leading a 200 or 400-meter race in real time.

AI also helped create highlight videos in multiple formats and languages during the Paris Olympics. Traditionally, highlights were done manually through tagging and clipping. Now, AI automation enables broadcasters to speed up the delivery of more customized digital content.

 

The OBS Automatic Highlights Generation was used to capture key moments from 14 sports and disciplines, compile them into tailored highlight reels based on Media Rights Holders’ preferences, package tailored highlights, and distribute them to fans instantly, increasing efficiency across production and editing.

 

The automatic highlights inference engine is based on AI models. These were trained sport by sport with the help of Intel’s Geti computer-vision AI platform. The content they were ‘fed’ was retrieved from the Olympic sports video archive.

Personalized highlights were intended to engage more digital and social media audiences. With so much sporting action being recorded simultaneously, AI’s ability to pick out precisely what viewers want to see will be a game-changer. This ability may also benefit broadcasters from nations with limited production resources.

 

NBC covered the Olympic events using its streaming platform Peacock. Every day, it provided a 10-minute highlight package pulled from 5,000 hours of live coverage and tailored to each subscriber’s sporting preferences.

 

The voice recaps were delivered in the AI-generated voice of Al Michaels, one of America’s most respected sportscasters. The AI was trained on Michaels’ previous broadcast audios, but editors still must double-check the results, including name pronunciations and fact-checking.

 

AI revolutionizes broadcasting from the Olympic Games

AI also helped with clearer images and improved live streams. 

 

Read also: Live Video Streaming App Development Guide by Onix

 

For example, Alibaba’s cloud computing unit provided an AI-infused platform to support broadcasting from the Paris 2024 Olympics, overtaking the satellite broadcasting system – the primary distribution method for 60 years. 

 

Paris 2024 was the first to feature revolutionary 8K (4X the resolution of standard 4K broadcasts!) live over-the-top (OTT) broadcast workflow delivery. 

 

Broadcast servers powered by the latest Intel Xeon Scalable processors with Intel AI Deep Learning Boost technology encode and compress the signal produced by the OBS at 8K/60FPS/HDR “More/Faster/Better Pixels” (48 Gbps RAW to 40-60 Mbps distribution using VVC standard) in milliseconds. An 8K OTT signal is delivered to the latest Intel-based PCs and laptops connected to 8K TV worldwide in seconds.

 

Besides an unparalleled viewing experience, this collaboration with Intel provided a visible pathway for the future of low-latency, 8K resolution transmission over the Internet.

 

AI helps promote the Olympic legacy

Intel’s Emergent AI Lab developed neural object cloning to transform digital images of iconic Olympic moments into 3D models for display in the Olympic Museum in Lausanne.

 

This solution will bring some of the Museum’s artifacts into interactive digital environments, enabling visitors to explore pieces of Olympic history as if they were holding them in their hands.

 

Read also: Virtual Reality in Museums: How VR Improves Museum Experiences

The AI Role in the Olympic Games Management 

 

More efficient planning

The IOC, host cities, and partners started gathering various operational data in 2020 to learn how AI can be used to enhance the Olympic Games management.

 

Read also: The AI Life Cycle: Unlocking Business Value through Strategic Implementation

 

Ilario Corna, IOC’s Chief Information and Technology Officer, explained:

 

“For planning, we are now working with our Partner Intel using the concept of digital twinning, or digital representations of venues so we can foresee, for example, where we would need power, where we would need to place cameras, and if there could be any accessibility issues – all without needing to be on site every time. Using these digital twins of the Games venues, we can change the way we organize the Games.”

 

Read also: A Deep Dive into Product Digital Twin Development

 

After creating digital replicas of Olympic venues, organizers can simulate different scenarios to identify potential issues in advance and make informed planning decisions.

 

This approach saves time and resources and enhances the overall operational efficiency of the Games. The data collected from these simulations and during the Olympic events will also inform planning for the future Olympics.

Enhanced security

Ahead of the Olympic and Paralympic Games in Paris, France initiated the testing and deployment of AI-powered video surveillance technology to enhance security during the events. AI-enabled cameras had been previously tried at concerts.

The French company Wintics developed the Cityvision software – the core of this surveillance upgrade. It analyzes video surveillance images in real time and utilizes AI algorithms trained to identify suspicious activities or threats, such as weapons, abandoned packages, signs of fire, or individuals lying on the ground.

 

Monitoring crowd behavior, it can detect overcrowding, crowd surges, unauthorized entry, or movement against traffic direction rules. The system alerts the units in charge, who then review those risks and, if confirmed, act accordingly.

 

This advanced AI-assisted surveillance notably excludes facial recognition – a compromise balancing the need for advanced security measures with commitment to privacy and data protection. The legislation also strictly limited the use of these AI-powered surveillance capabilities to large-scale sports and cultural events where there’s a potential threat of terrorism.

 

 

Improved sustainability 

The IOC also pledged to halve the carbon footprint of Paris 2024 compared to previous Games. This determined another application of AI at the Paris 2024 Olympics and Paralympics – the monitoring and management of electricity consumption.

 

They utilized Alibaba Cloud’s Energy Expert platform, an AI-driven sustainability solution aiming to measure and analyze energy consumption across 35 venues around Paris during the Olympics and Paralympics and create venue-specific energy-use forecasts and recommendations. 

 

Energy Expert allows real-time tracking of all energy-related data, such as energy consumption, power demand contingency, venue capacity, competition-related information, and weather conditions.

 

Ilario Corna expects that the data-driven insight produced by Energy Expert will help organizers learn from each Games edition and apply that knowledge intelligently to make future events even more energy-efficient.

Final Thoughts

Like earlier innovations in sports science, communications, computing, and other fields, AI is transforming the Olympic experience in the 21st century.

 

The Olympic Games in Paris have provided many examples of using AI to make the Olympic Games fairer, safer, more inclusive and accessible, more sustainable, more fun for visitors, and more exciting and personalized for viewers worldwide. 

 

Behind the scenes, time and money saving, efficiency, and monetization are the areas where AI can improve performance.

 

As more stakeholders realize what benefits AI can deliver for sports, organizations, and communities, a better future is getting closer. 

 

Imagine a world where AI helps spot budding Olympic talent anywhere and tailor training programs to each athlete’s physical and mental abilities; where an infallible and impartial AI-based scoring system ensures that all selections and competitions are fair; where Olympic broadcasts transcend all barriers and fans can experience sports from an athlete’s perspective in virtual reality.

 

Read also: The Future of Chatbots: 10 Trends, Latest Stats & Market Size

 

The integration of AI in the Olympics requires a collective effort of athletes, International Federations, National Olympic Committees, IOC and the International Paralympic Committee, Worldwide Olympic Partners, and other organizations, as well as sports enthusiasts, innovators, entrepreneurs, philanthropists, environmentalists, and champions of social change.

 

If you wish to join the revolution in any of these roles, please consider Onix as your potential AI/ML development partner!

 

FAQ

 

​​How is AI being used in the training of Olympic athletes?

 

• Performance analysis: AI can analyze athletes’ performance data to identify strengths, weaknesses, and areas for improvement, providing personalized training plans.

 

• Decision-making: Coaches and athletes use AI to make informed decisions on strategies, training schedules, and recovery processes, enhancing overall performance.

 

• Technique optimization: AI-driven video analysis can break down athletes’ movements and offer real-time feedback and corrections during training sessions to optimize techniques.

 

• Virtual coaching: AI-powered virtual coaches simulate various training scenarios and opponents, allowing athletes to practice in diverse controlled environments.

 

• Injury prevention: AI can monitor biomechanics and training loads to predict and prevent potential injuries by suggesting adjustments in technique or intensity.

 

How is AI helping judges and referees at the Olympic Games?

AI systems analyze real-time data from multiple angles, enable instant replays, and offer precise measurements of athletes’ performances. For example, AI-powered video analysis can detect infractions or validate scores more accurately than the human eye.

 

In sports like gymnastics or diving, AI helps to assess the difficulty and execution of routines, ensuring consistent and unbiased scoring. AI also reduces the likelihood of human error and bias, ensuring more transparent and reliable judging.

 

How is AI being used to improve the spectator experience at the Olympic Games?

 

• AI-driven cameras provide dynamic angles and zooms in real-time, ensuring viewers see the most exciting moments from the best perspectives.
• AI offers real-time statistics, predictions, and insights during events, enhancing viewers’ understanding and engagement.
• AI curates tailored highlights, replays, and content based on individual viewer preferences and behavior.
• AI powers virtual assistants and chatbots that can answer questions, provide event information, and suggest content, creating a more interactive experience.
• AI can be integrated with augmented reality (AR) to overlay additional information like athlete stats or trajectory paths on live footage for a more immersive viewing experience.
• AI-driven translation services provide real-time subtitles and commentary in multiple languages, making the games accessible to a global audience.
• Organizers can implement AI for spectators’ entertainment, engagement, and communication with other fans located worldwide.

 

How can AI help advertisers during the Olympic Games?

 

• Programmatic ad buying: This AI-driven solution enables smaller brands to compete with major players for real-time ad placement during the Games.
• AI-powered creative analytics: This ad tech tool leverages ML to identify the most engaging media elements, helping craft the right textual and visual messages to enhance campaign performance.
• Real-time optimization: AI facilitates quick adjustments to ad creatives, maximizing budgets and enhancing campaign effectiveness.
• Targeted engagement: ML helps tailor ads to specific sports audiences, increasing relevance and impact.
• Cultural relevance: AI helps ensure ads respect cultural differences and values, maintaining effectiveness across diverse global audiences.
• Platform-specific advertising: AI optimizes content for various channels, ensuring consistent branding and engagement across platforms.
• Performance insights: AI analyzes ads to identify successful elements, guiding improvements for future campaigns.

 

How is AI used in Olympic sports between the Games?

 

• Personalized training: AI analyzes data from sensors, wearables, and smart equipment to provide athletes with valuable insights and tailored training, health, and fitness programs.
• Performance analysis: AI analyzes training data to optimize techniques, enhance performance, and give athletes a competitive edge.
• Injury prevention and recovery: AI-powered tools track biomechanics and health metrics to prevent injuries and speed up recovery.
• Olympic talent scouting: AI may help identify promising athletes even in remote areas, making Olympic training more accessible and equitable. Tools like Intel’s 3D Athlete Tracking analyze body movements and reactions to match athletes to sports.

 

Are there any challenges and concerns about using AI at Olympic Games?

Yes. Some of the concerns and hurdles are: 

 

• Unfair advantage: Wealthier countries using AI for athletes’ and coaches’ needs may gain an edge, potentially leading to inequalities that exclude developing nations.
• Lack of regulation: A framework is required to regulate the AI role in the Olympic Games, similar to rules regarding performance-enhancing drugs, to ensure fairness for athletes and fans.
• Privacy issues: The introduction of surveillance legislation has raised concerns about privacy violations, personal information protection, and the risk of discriminatory practices resulting from automated surveillance.
• Sport-specific challenges: AI integration into different sports may be hindered by limited funding in less popular sports or the complexity of applying AI to specific sports and disciplines.

 

What are the prospects for the use of AI in future Olympic Games?

The numerous applications of AI at the Olympic Games in Paris have set new standards for future Games and paved the way for further adoption of AI in the areas of athlete training, injury prevention and rehabilitation, scoring and judging, broadcasting, advertising, entertainment, online and offline safety, accessibility, event planning, operations management, and sustainability.