How Fortum Uses Artificial Intelligence to Detect Issues at Power Plants and Entire District Heating Networks – From the Press

At the core of modern industrial operations is the need for reliability, efficiency, and safety. mi Solutions’ Prexello, an AI-powered software platform, is revolutionizing how industrial sites predict failures before they happen, allowing for proactive maintenance and significant cost savings. One of the standout implementations of Prexello is in Finland, where it is coupled with Caverion’s industry experts to provide Caverion Intelligence AD services. This service is used by Fortum, a leading Finnish heat and power company, to ensure smooth operations in their district heating power plants. Read the original story at lansivayla.fi.

The Espoo Success Story

In a recent feature in Länsiväylä by Miia Saari, the power of artificial intelligence at Fortum’s Suomenoja district heating plant in Espoo, near Helsinki, was brought to light. The story highlights how Prexello, through Caverion Intelligence, helps Fortum predict potential equipment failures, ensuring more efficient operations.

Fortum’s digitalization director Viki Kaasinen shared that Fortum is still at the beginning of its AI journey, but as “one of Fortum’s great successes in the use of artificial intelligence, Kaasinen names a system that you have to go and marvel at on the spot, at the Suomenoja district heating plant in Espoo”

Here “the reporter witnessed how artificial intelligence alerted the Fortum district heating power plant of a danger threatening it – this is how companies in Espoo use artificial intelligence.” This real-time insight, enabled by Prexello, has become a critical tool for Fortum’s maintenance teams, allowing them to respond to issues before they escalate.

Inside the Control Room

Janne Metsomäki, Fortum’s operations master, demonstrated the software’s capabilities, showcasing how the system monitors key equipment in real time. “Fortum’s operations master Janne Metsomäki opens the Caverion Intelligence software on the big screen. Soon two curves, green and blue, will appear on the wall”, Metsomäki explained. “The blue curve shows the vibration intensity of the condensate pump motor bearing located in the turbine hall of the power plant in real time, in millimeters per second. The green curve below it is nothing more and nothing less than artificial intelligence”.

The difference between these curves is critical. When Prexello’s green line deviates from the blue real-time performance curve, it triggers a warning—alerting engineers like Metsomäki to investigate and take preventive action. As Metsomäki explains, “Artificial intelligence tells the users of the system how strongly the engine, pump or other device would vibrate under different process conditions and operating situations, and also takes weather conditions into account.” This level of precision, taking into account both operational and environmental variables, makes Prexello indispensable for a dynamic industry like energy.

Prexello’s easy to use dashboard effectively guides users to relevant issues – even at very large sites such as district heating networks.

Actual measurements in blue vs. Prexello’s predictions in green. Deviations are signs of looming failures or equipment issues.

Years of Learning

Caverion Intelligence using Prexello doesn’t just monitor systems in isolation—it learns from historical data. “Artificial intelligence has been measuring the vibrations and other process parameters belonging to the power plant’s processes since 2017 and has learned which vibration levels, temperatures, and pressures at different power levels and seasons are normal, and which are not,” Metsomäki explains. This long-term learning means the predictions become more accurate over time, allowing Fortum to detect even the subtlest deviations that could signal future failures.

Impact on Operations

The practical impact of this AI-powered approach cannot be overstated. The Caverion Intelligence service, powered by Prexello, helps optimize maintenance schedules, reduce unscheduled downtime, and ultimately, save costs.

In addition to cost savings, the use of Prexello enables companies to improve service reliability and quality. As highlighted by the article the district heating network that supplies both Espoo and Kirkkonummi is kept running smoothly thanks to AI’s constant monitoring.

Conclusion

Prexello’s AI technology is a game-changer, enabling companies to proactively manage their critical assets, improve service quality, and reduce costs. The collaboration between Prexello, Caverion Intelligence, and in this case Fortum exemplifies the future of industrial maintenance—where artificial intelligence not only predicts failures but actively helps to prevent them, ensuring safe and reliable energy delivery to communities.