AI solution put to the test during the construction of the Hertsö swimming pool

Ensuring that the construction site is safe is one of the site manager’s most important and demanding responsibilities. Using industrial AI and robotics, researchers at Luleå University of Technology have automated parts of this safety monitoring process. The technology was tested in a real-world project during the construction of the new swimming pool at Hertsön in Luleå, in collaboration with NCC.

“Safety on a construction site is paramount. The site manager needs to know that the site is safe, and much of the monitoring is done manually. In our AI project, we were able to alleviate that burden while simultaneously improving the quality of safety,” says Ramin Karim, professor and department representative in Operations and Maintenance Engineering at Luleå University of Technology and project manager.

The AI project has now been recognized by SBUF, the Swedish Construction Industry Development Fund, through a newly established award for innovative thinking—an award that highlights innovations with the potential to transform the construction industry.

Collects data when the workplace is at a standstill

The researchers’ AI solution utilized industrial AI in several stages: to control robots and drones, to analyze the data, and to visualize the results in 3D. In the project for the construction of the Hertsö swimming pool, the robot Spot and drones equipped with sensors were used to document the construction site. Data collection took place automatically, as a recurring patrol, and was performed when production was not underway. The data was then analyzed using industrial AI and compiled into a report for the next workday.

“When the site manager arrived in the morning, there was already a report to work from. You could see if anything was out of place, if anything needed to be fixed, and if safety was at the required level,” says Ramin Karim.

A unique opportunity for the industry

The collected data was processed and visualized in a 3D environment where the construction site could be digitally explored, for example using VR glasses. This provided a comprehensive overview of the project’s status and made it possible to identify risks and deficiencies without having to be on-site at all times.

“Many industries are now looking at what AI can do for them. Here, it was about linking AI to workplace safety issues and daily operations. The site manager bears responsibility for the work environment and needs to ensure that everything is in order. At the same time, we see that it’s difficult to recruit site managers—the role is perceived as demanding, and many hesitate in the face of the responsibility. That’s why tools are needed that provide support in everyday work and make it easier to make the right decisions. With this model, it becomes visually clear and if something isn’t right, it’s flagged immediately. This makes it easier to quickly identify risks and take action. This is the future, even for us, with a focus on the work environment and the site manager’s day-to-day work,” says Josefin Hermansson, Production Manager at NCC.

Part of AI Factory – From Data to Decisions

The project was carried out within the AI Factory at Luleå University of Technology, where industrial AI solutions are developed in close collaboration with industry. The core idea is to connect the entire workflow—from data collection to analysis and visualization—into a functional decision support system.

“That’s exactly how we work at AI Factory. We connect collection, analysis, and visualization into a cohesive whole that works in practice,” says Ramin Karim.

Award in fierce competition

The cutting-edge AI project at Hertsö Bathhouse was funded by SBUF, where research is conducted in close collaboration with the construction industry. Every year, hundreds of project applications are processed, and only a few are nominated for the award.

“Being nominated was a big deal. The fact that we also won in the innovation category shows that our way of working is relevant to and appreciated by the industry. This is the result of several years of research and hard work,” says Ramin Karim.

The AI project did not aim to develop a finished product, but the results demonstrated how construction sites can be monitored in a more systematic way.

“We have demonstrated that it is possible to automate the entire workflow—from data collection to analysis and visualization. This opens up new ways of working with safety in construction projects,” says Ramin Karim.