《From Broadcasting to Analysis… AI as a Multi-Player in Stadiums

The sports field, once dominated by the belief that only relentless practice leads to victory, is now experiencing a wave of artificial intelligence (AI). An era where AI designs tactics, broadcasts games, and edits highlights. A look into how AI has transformed the stadium landscape.》


It was a drizzly night. On the railing of the third-floor stands of Seoul World Cup Stadium in Mapo-gu, Seoul, three 'eyes' quietly watched over the Taegeuk Warriors. The illuminated grass, uniforms clinging with sweat, and the sliding ball were all recorded by these three eyes every second.

The AI cameras of the domestic company 'Bepro Company (Bepro)' meticulously analyzed the A-match between Korea and Brazil on the 10th. Without human intervention, the cameras continuously tracked the movements of the 22 players roaming the field. Key data such as player-specific shooting records and team possession collected by the cameras were transmitted in real-time and reappeared as numbers on the broadcast screen.

In the sports world, the phrase "effort never betrays" is perhaps the most frequently quoted. Athletes on the podium often express sentiments like "relentless practice is everything." Until recently, this was largely true.

However, as AI now dissects and analyzes games, such beliefs are being quietly reinterpreted. The extent to which accurate figures and data are secured directly correlates with performance improvement. Technological advancement is building 'evidence' to support an athlete's 'intuition.'

● Data Creates Tactics

"If a player's effort and passion are the default values, our role is to provide the coordinates of where to focus that effort."

Yang Jun-seon, head of Bepro's Asia-Pacific (APAC) division, said this during an interview at Bepro's office in Jongno-gu, Seoul, on the 14th. It signifies their aim to be a 'compass' that helps translate players' sweat into results.

The advent of AI is reshaping the landscape of the sports industry. Bepro is one of the companies leading this change. Bepro captures game and training scenes at 400 stadiums worldwide using AI cameras, processes players' movements into data, and provides it to clients.

Bepro counts all 12 clubs of Korea's K League 1, some teams from major European leagues such as the English Premier League (EPL) and the German Bundesliga, among its clients. During the Copa del Rey final in April, where Real Madrid faced FC Barcelona, the video and data captured by Bepro's AI cameras were delivered in real-time to both clubs and the Spanish Football Federation.

Bepro films the stadium by dividing it into left and right with three cameras. When the footage is stitched together in a panoramic format, a 'stadium at a glance' is created. This long video allows simultaneous viewing of movements from the goalkeeper to the forwards. Bepro provides this video to clubs for analysis.

An interesting aspect is that during video playback, it is possible to mark players of interest or measure distances between defenders. This is because the AI camera automatically recognizes and tracks each player during filming. This allows for the collection of not only play-related data such as pass success rates and attempts per player but also movement data, including top speed and instant acceleration.

Unlike traditional vest-type GPS equipment, which only recorded information for players wearing it, this system can also analyze the movements of opposing team players. This information is also conveyed to the coaching staff on the bench during games.

A particularly notable service is the 'bespoke data,' which provides customized data for each club. For example, the Pohang club receives data on the positions where the ball lands during corner kicks and the ratio of those that lead to shots or goals, which they use for training. A Pohang representative stated, "We use the data in various ways, such as analyzing the attacking patterns of the opposing team before a match to refine our defensive tactics."

The K League match statistics currently available on the Korea Professional Football League's 'Data Portal' also originate from these AI cameras. Under four main categories—attack, defense, passing, and goalkeeping—300 to 400 detailed indicators are accumulated. The Technical Study Group (TSG) under the league's technical committee also uses this data for researching and analyzing game strategy trends.

● Changing the Scout's Notebook

The use of AI data by professional clubs is not limited to game analysis and strategy operation. The prestigious EPL club Liverpool has been collaborating with Bepro since 2024 on a 'data-driven prospect scouting project.' They develop performance indicators that meet scouting criteria and request analysis of games they are monitoring.

For clubs that find it challenging for scouts to attend every game, this provides a new tool to save time and costs while enhancing competitiveness. Yang stated, "Data doesn't tell you everything, but it doesn't show what isn't there either," adding, "There's demand because it allows for an intuitive assessment of players' capabilities."

Bepro, which began its service targeting K League youth teams in 2016, now has branches in eight countries, including Germany, the United States, Spain, the United Kingdom, Italy, Japan, and France, providing services to 3,000 to 3,500 clubs and institutions worldwide. The business scope has also expanded to handball, futsal, and basketball.

Bepro's long-term goal is to create a 'football library' accessible to everyone. Yang stated, "In the past, sports had an asymmetrical structure where the amount of information held by coaching staff, players, parents, and fans varied," adding, "Our goal is to produce more objective and extensive information to reduce this gap." He continued, "What used to require a large workforce can now be achieved by AI. This change will serve as a foundation for the sports industry to grow to the next level."

AI is also utilized when recruiting players from other clubs. By analyzing a player's game data, it predicts how well the player will adapt to different leagues or team tactical systems. For instance, it assesses how a player reacts when they have less time on the ball or face faster opponents to determine if they can be competitive in the EPL, known as the league with the fastest transitions between attack and defense.

Large language models (LLM) like ChatGPT are also being used in the football transfer market. The Italian AI-based football data analysis company 'Soccerment' is integrating LLM into its data analysis platform. When asked, "If Arsenal (England) loses Bukayo Saka, who would be the most similar replacement?" AI identifies Saka as a player who creates threats with breakthroughs and forward passes, suggesting Barcelona's Raphinha as a replacement. This aligns with the reason Arsenal manager Mikel Arteta actually attempted to sign Raphinha in the past.

● AI Era in Recreational Sports

In sports, every player wants to review their performance, whether on the world stage or at a local field. In response to this demand, AI's influence is spreading to recreational sports and amateur scenes.

The 51st Korean Journalists Association Seoul Regional Football Tournament held last month was also broadcast live through the AI broadcasting platform 'Pochak.' A fellow journalist who scored a debut goal in this tournament said, "I was so satisfied that I replayed it more than 10 times."

Launched by KT Skylife in August, Pochak is based on the camera system of the Israeli AI broadcasting company 'Pixellot.' While Bepro's cameras track players, Pochak follows the ball to compose the optimal game screen. It focuses more on 'broadcasting' than analysis. Users can watch this through TV or mobile applications.

KT Skylife is responsible for broadcasting the 2025 Hwarangdae National Youth Football Tournament, the National Sports Festival for All, and the KT Futures League (second team) games in professional baseball. They also offer mobile camera services for groups without a dedicated home stadium. Schools like Suncheon Central Elementary have shown demand. Local governments are also interested. Gumi City in Gyeongbuk broadcasts social baseball team and club games through Pochak.

Interest in AI broadcasting platforms is also growing among parents with children aspiring to be athletes. Players who have not joined professional clubs often lack opportunities to have their games filmed or analyzed with professional equipment. AI platforms automate this process, allowing anyone to record and share game scenes at a relatively low cost. Kim Kwang-hyun (SSG), a representative left-handed pitcher in Korean professional baseball, also watches his son Min-jae's Little League games through Pochak.

A Pochak representative stated, "By filming games with unmanned cameras and automatically editing them for broadcast, costs can be reduced by up to 90% compared to traditional methods," adding, "It is particularly gaining attention in amateur games or unpopular sports where manpower and budget are limited."

Pixellot is currently operating over 30,000 AI broadcasting cameras in more than 70 countries. It has already become standard equipment in U.S. high school and college leagues. The global AI broadcasting market is estimated to be worth $700 million (approximately KRW 994.6 billion) as of 2024.

● Highlights at AI's Fingertips

When Son Heung-min, playing for Los Angeles (LA) FC in Major League Soccer (MLS), scored a 'lightning goal' just 52 seconds into the game against San Jose on the 14th of last month (4-2, LA FC win), what surprised fans was not just the scoring scene. It was the fact that the highlight video was uploaded to MLS's official Instagram just two minutes later. Behind this astonishing speed was AI. AI is evolving not only as a broadcaster but also as an editor and content producer.

The Israeli sports video tech company 'WSC Sports' uses generative AI to produce highlight videos. AI scores various elements such as crowd noise and commentator voice tone to extract key scenes. The company boasts, "Thanks to overwhelming data learning, the accuracy is high." As short-form content gains popularity on social media platforms like YouTube and Instagram, major sports organizations worldwide, including the NBA and UEFA, are producing real-time highlight videos with the help of WSC Sports.

Now, algorithms have permeated the tactical boards in training grounds and the notes of scouts. Data becomes a clue to change the next game, and AI becomes a new analyst that finds that hint faster than anyone else. Outside the stadium, it transforms into a producer offering opportunities to some and memories and entertainment to others. Wherever technology reaches, sports evolve into a more multidimensional form. AI is editing the future of sports with its unblinking eyes.

Popular News