Finnish energy company uses AI to predict grid faults and reduce outages

A Finnish electricity network operator has deployed an artificial intelligence system to monitor grid conditions and prioritise maintenance, aiming to prevent power outages before they occur.

Järvi-Suomen Energia, a network company in the South Savo region, developed its own AI model after finding no suitable commercial solutions, the public broadcaster Yle reports. The system analyses real-time data from sensors and drones, cross-referencing it with historical network records to identify high-risk areas—such as overgrown vegetation threatening power lines.

“We trained the AI by feeding it our past maintenance decisions,” said development manager Tomi Öster. “It now proposes exactly the kind of interventions we’d choose ourselves.”

The company’s 28,000-kilometre network—one of Finland’s largest—spans heavily forested and rural areas where maintenance is costly. Annual upkeep runs €8–10 million, with an additional €30–60 million invested yearly in weatherproof underground cabling.

CEO Arto Nieminen said AI-driven efficiency won’t immediately lower customer transmission fees, as major infrastructure upgrades will continue through 2036. “After that, we shift to maintenance mode, and price reductions may become possible,” he noted. “For now, we’re allocating funds where they deliver the most benefit.”

Järvi-Suomen Energia’s network serves some of Finland’s longest average line lengths per connection, contributing to its position among the country’s highest transmission tariffs. The AI system is expected to reduce outages by targeting resources more precisely than the previous schedule-based approach.