Benfica scores big with AI and spatial data

Sport Lisboa e Benfica has won Portugal’s domestic football league more than any other team¹ and regularly competes against Europe’s best in the UEFA Champion’s League. To sustain this success, the club has embraced new technologies and developed a focus on innovation, rather than resting on its laurels.

“To compete, we need to be smart and do things a little differently,” says Joao Copeto, chief information and technology officer at Benfica.²

Over the past 20 years, the organization has built a large sports data science program that invests in advanced technologies like player-motion-tracking computer vision and artificial intelligence, all with the goal of developing young, talented players into high-caliber professional athletes.

This isn’t only about wins and loss, but dollars and cents as well. Benfica operates junior-player development academies where it trains future professionals. Over the past 10 years, the team has generated some of the highest player-transfer deals in Europe. By leveraging data and AI, Benfica has turned player development into a profitable business.

At the junior level, the player academy collects a range of data on players, such as how they are developing their football skills, along with their nutritional, psychological, and social data. Coaches and teachers can use this data to develop personalized instruction to help players improve both on and off the pitch.

At the professional level, AI and spatial data are helping the team gain a competitive edge. For example, Benfica uses camera technology combined with AI and computer vision to track player movements on the field during games. These cameras track 2,000 individual points of each player’s body throughout the games to see where players position themselves, where they are looking, and how they are processing information.

All this data essentially creates a digital twin of each player. This allows the team to simulate player behavior in different positions or test new strategies. The team is now using these simulations to help contribute to tactical decision-making.

“There’s been a huge evolution in AI that’s pushing these models forward, and now we can use them in decision-making,” Copeto says.

AI is also helping team staff eliminate low-value work and focus on more valuable tasks. For example, after every game, the coaching staff wants to know how the team performed on set pieces like corner kicks and free kicks. In the past, an analyst had to manually review the footage and flag these set pieces before anyone could actually analyze them for insights. Now, with the help of AI computer vision models, this work is done automatically, allowing analysts to spend more time developing insights.

“What we want to achieve with technology is for the staff to use their time to analyze the game instead of using it to produce data,” says Copeto. “We want them to move to the next step. AI helps accelerate a lot of these processes.”

Copeto expects operations to continue getting more efficient as AI assumes a larger role within the team. Like most organizations today, Benfica produces a lot of data, and staff members’ day-to-day tasks involve generating, organizing, analyzing, and reporting on that data. But more and more, AI will perform all these tasks. Models are improving, and, importantly, organizations are finding more ways to use these models to improve upon previously manual tasks.

“What we can do as a team will improve now with large language models and multimodal AI tools,” Copeto says. “I think this will be groundbreaking.”