With the opening 2024 ceremonies starting today, the Olympic Games are one of the last bastions of pure human versus human competition—or so we would like to believe. The reality is that technology is woven so thoroughly throughout the fabric of the games that performances cannot be measured or viewed without millions of dollars worth of technology.
Intel aims to make Olympic arenas accessible for people of all abilities using a handheld LiDAR to scan environments and machine learning platforms powered by Intel Xeon processors. Image used courtesy Intel
The 2024 Olympics in Paris, France, will leverage the best that technology has to offer. From advanced timing by Omega to artificial intelligence (AI) by Intel, technology factors into the preparation, transportation, crowd management, conducting, and judging of the events. In this article, we take a look at three key Olympic technology areas, starting with the humble timekeeper.
A Millisecond Here, a Millisecond There, and Pretty Soon You’re Talking Real Time
Formal timing came to the Olympics in 1932 when a single watchmaker from the Swiss company Omega traveled to Los Angeles with 30 calibrated stopwatches in a briefcase. Each watch had been certified to be accurate to 1/10 of a second. Omega was chosen in large part because it was the only watchmaking company that could provide 30 certified stopwatches. These were handed out to timekeepers to be used for all events.
Electronic timing first entered the scene 16 years later at the 1948 Winter Olympics in St. Moritz, Switzerland. Those games saw the first use of the “magic eye”, a photoelectric cell-based finish line light sensor. The first “photo finish” followed at the Summer Olympics in London that same year where the magic eye was coupled to a slit camera to capture the finish. Still, the manual timing era was not yet over.
American runner Harrison Dillard is seen winning the 100-meter gold in the Olympics’ first photo finish at the 1948 London Games. Screenshot used courtesy courtesy of Monochrome
In the 1960 Rome Olympics, swimmers were still timed manually. Three judges covered each lane with mechanical stopwatches, which were averaged to determine the swimmer’s final time. In the 100-meter freestyle final, two swimmers received the exact same average time. In a controversial move, the head judge gave one of the swimmers the Olympic gold medal and recognized the other as the Olympic record holder.
Omega starting blocks with pressure sensors to detect false starts. Image used courtesy of Omega
Two Olympics later, in 1968, Omega responded to the controversy with the first electronic start-and-stop system. The clock started with an electronic attachment to the starting gun, and each electronic swim lane timer stopped with a swimmer-activated touchpad at the finish line. The same system is still used today, just with greater precision. 2024 speed rock climbing competitors, for example, use a similar system to stop their clocks with a touchpad at the top of the climb.
The Start Is as Important as the Finish
For track and field events, electronic starting “guns” today don’t rely on the sound from a single blank pistol round. The speed of sound delay between the inside and outside lanes can give competitors closer to the starting gun an unfair advantage. Now, electronic starter pistols trigger a speaker behind each competitor, so they all hear the start signal at exactly the same time.
Omega is still the designated timekeeper. For 2024, the company is deploying 450 tons of equipment for the timekeeping tasks in Paris. They will utilize 200 kilometers of wire and optical fiber cable, 350 sport-specific scoreboards, and 85 public scoreboards, all managed by 900 volunteers and 550 on-site timekeepers. Today’s electronic timers have a resolution of one microsecond, and 2024’s “Scan ‘O’ Vision” photo finish cameras take 40,000 images per second. Omega is also using single- or multi-comer computer vision technology coupled with AI models trained for each sport.
The First Olympic 8K Digital Television Viewing Experience
Closed circuit or local area television coverage started with the 1936 Olympics. The 1960 winter games were the first to be broadcast in North America. It was not, however, live TV. Television satellites did not yet exist, and undersea cables didn’t have the bandwidth. Tapes were edited, copied, and sent by airplane to the United States for broadcast. Judges in the men’s slalom race that year asked CBS TV crews if they could review a videotape to verify if a skier had missed a gate or not. This milestone is thought to be the birth of the instant replay. The first “live via satellite” Olympic coverage came in 1964.
For 2024, Intel is collaborating with the Olympic Broadcast System (OBS) to use AI tools for live 8K (7,680 x 4,320) resolution broadcasts and viewing on PC-based 8K platforms. The highest current commercial broadcast resolution is 4K (3,840 x 2,160). Throughout the games, OBS will have a number of 8K cameras. Intel has developed AI-optimized broadcast servers to encode and compress the 8K video for delivery over the Internet. Those who have licensed the 8K stream and have a suitable PC on the viewing end will have access to the 8K live video.
The AI video compression runs on servers powered by 4th and 5th Gen Intel Xeon processors and with Intel’s AI accelerator (AMX) and Deep Learning Boost technology. The compression system is fed video data at 48 gigabits per second and produces a 40–60 megabit per second output stream. The 1,000 X compression takes place in near real-time with just a 400-millisecond delay.
Artificial Intelligence Enters the Olympic Games
2024 isn’t the first Olympic Games to have some AI component, but it may reach a tipping point this year. In addition to the AI-enabled 8K video, AI is being used in a number of other arenas. Omega, the official timekeeper, also has AI-based performance analysis video equipment to help athletes analyze their performance. Participants and coaches can see data and imagery on when, how, and where to work to improve performance.
AI-based athletic performance analysis. Image used courtesy of AiScout
The UK-based startup AiScout has produced an athlete evaluation platform that works as, essentially, an AI talent scout. The app starts with smartphone videos taken in the field. The server-based component then uses AI to analyze the athlete’s performance, as shown in the video. The hope is that athletes with potential can be scouted from around the world and may be given opportunities for international competition, like the Olympics, that would not have otherwise been possible.
AI and the Future of the Olympics
The International Olympic Committee has deemed AI to be significant enough that they have created an Olympic AI agenda. AI can be used for scouting, training, performance judging, and a host of other areas. Soon, AI may be replacing qualitative judging with quantitative analysis. Such a change would deliver impartial scoring for rated events like gymnastics and diving. AI is already helping to develop better sporting technology for every event that has associated equipment. It’s possible that Olympic coverage in years to come will not be possible without AI.
