How every amateur match ends up looking like a Champions League final

By Linda Bak

Last year, lots of sports matches were played without an audience due to the corona measures. In order for loyal supporters to be able to enjoy the game, a lot of sports clubs used livestreams. Often this was an integral picture where the entire field could be seen. “That’s not so exciting to watch,” says Paul Valk, founder of Studio Automated, a spin-off from the Netherland Organisation for Applied Scientific Research (TNO). In such an all-encompassing picture, you often can’t see very well where the action is exactly. That’s why during TV broadcasts the camera operators and directors make sure that the viewer sees the right kinds of shots at the right time. This can also be done fully automatically now, thanks to AI and computer vision technology.

Analyzing images

“Through the use of AI, we are able to predict how the game will play out in the next few seconds. This allows us to know exactly what part of the field we need to zoom in on to properly capture the action,” Valk explains. Football club Ajax, for example, used this technology during the corona pandemic to enable parents of young players to watch them during matches. The trainer of professional teams can also use the technology to analyze the playing patterns of different players. For example, areas for improvement can be identified this way. Sports and media are key markets for Studio Automated.

”In addition to this, the company is also developing security-related applications in collaboration with TNO, among others. “AI and computer vision technology can hep to make society safer,” says Christophe Hoegaerts, Research Manager Intelligent Imaging at TNO. For instance, the technology can detect a commotion even before people actually get into fistfights. This enables law enforcement officers and police officers to intervene before anything goes wrong and helps to keep the peace on the streets. “The AI software learns which situations deviate from normal. As a result, the system can help improve the work effeciency. So operators no longer have to spend hours scanning camera images to keep an eye on the public order,” Valk says. “That’s also better for their health and helps increase overall safety.”

Johan Cruijff Arena

“The various application areas are culminating in pilots at the Johan Cruijff Arena and Ajax. With our technology, we can convert relevant content from matches into a fully automated broadcast of a match. We can also analyze visitor flows to see whether it’s getting too busy around the bar and the toilets, for instance,” Valk adds. Such insights into visitor flows are very interesting at large events. “By seeing to it that there are clear walking directions and short waiting times, we are able to improve the experience for visitors.”

A leap of faith

The technology developed by Studio Automated offers endless possibilities for a variety of sectors. Researchers within TNO who discovered and developed the technology a few years ago were well aware of this at the time. Paul Valk and his partner Paul van den Haak were both working at TNO at the time. “We spotted various applications and markets for this technology. Except that it was still difficult to transfer knowledge about the algorithms to existing companies,” says Valk. The technology is still relatively new and undiscovered, which brings a lot of risk with it for companies.

That is why Valk jumped into the deep end, along with his business partner. They founded a spin-off to bring the technology to the market themselves. Valk: “In a spin-off, we can combine the best of two worlds. At Studio Automated, we have the concrete commercial knowledge and skills in house to develop a product. A nd at TNO there are several experts who have a good understanding of high-end technology such as AI and computer vision. So, you learn a lot on both counts.” The spin-off and the knowledge institute are still working closely together to further develop and roll out the technology.

Reconsidering

Shortly after Studio Automated was founded in February 2020, the corona crisis broke out. “Our first customer indicated that because of the corona crisis that was erupting, he didn’t know if he could continue to meet his financial obligations in the short term. That was a bit nerve-wracking for a while,” he recalls. “You can do two things in crisis situations. You can sit idly by and wait for it to pass, or you can turn it around and look at the situation from a different angle.” Studio Automated has chosen to do the latter. “With our partners and their customers, we accelerated our investment in the technology. This allowed us to start developing our AI model for other sports as well.”

That turned out to be a wise move. In the meantime, Studio Automated is building a good partnership and customer network in the Netherlands and the rest of Europe. “Due to the accelerated growth of our partners and the scalability of the application, we will become the market leader in the media domain in Europe next year with our technology,” he states. In addition to a good team and some courage, a bit of luck also plays a role in the development of a company. “Ajax came to us about using our software because parents could no longer watch their children’s matches ‘live’. That was also an important moment for the company.”

Unique

Studio Automated is not the only company working on AI and computer vision. “A lot of companies are able to use algorithms to develop a product where automatic recognition is involved. They then often succeed in recognizing and estimating correctly up to eighty percent of cases,” says Valk. “But that isn’t enough. It only becomes interesting when you have a system that is more than 95 percent accurate. Then you are much better able to really tell a story with the footage. In this respect, Studio Automated is unique. “We offer a total package for the customer, so we can provide both the software for coaches and the software for fans, along with key partners who can also provide a complete setup (such as cameras and PCs).”

Loads of potential

With their approach and the latest technology, Studio Automated aims to make AI accessible, scalable and relevant for society and businesses. “AI has loads of potential, it can change the world in many different areas,” Valk contends. Christophe Hoegaerts from TNO sees plenty of opportunities for collaboration in the future, mainly in the area of security. “Ensuring a safer society is one of the important issues that TNO is working on. We can make great strides through cooperation between knowledge institutes and young companies,” he says. That increases the chances of success for new products. “AI can lead to fewer pickpocketing, drug deals, muggings and fights on the streets because we can recognize these kinds of situations before they happen. This enables the police to anticipate situations more effectively and act faster,” he continues.

In other areas such as smart cities and logistics, too, there is still plenty of room for improvement with Studio Automated’s technology. Valk: “The impact of AI is becoming increasingly visible. We want to make sure that AI improves the lives of as many people and companies as possible.”