Inovia’s AI solution gave new life to Lapland hydroelectric plant

Start where you are and build on what you have. The AI expert Inovia applied this philosophy to make the hydroelectric power plant at the tiny community of Saxnäs in Sweden’s far north efficient and profitable by connecting old but well-functioning technology to the internet.

The Swedish power company Vattenfall saw no profitability in the far-off hydroelectric power plant at Saxnäs in the inland far north of Sweden. The plant’s technology dated from 1968 and everything had to be done manually and on site. As a result, the plant was purchased by an entrepreneur who was interested in new technological solutions and had a vision of achieving efficiency and profitability. The AI company Inovia was brought in and today, less than two years later, production has increased by 13 percent and profitability by 43 percent.

“Many companies start major integration projects by digitalising the operations but we’ve worked in a different way here,” explains Robert Berg, business developer with the Inovia Group. “The existing equipment was indeed from the ’60s but worked well mechanically. We used it as our starting point and complemented it with internet-connected cameras and sensors.”

Large amounts of data

The sensors monitor vital processes and measure performance, the turbine speed in RPM, the heat of the turbine bearings, the precipitation, the depth of water in the lake upstream and so on. Analogue meters are monitored by Al-based image recognition. The data is then transmitted to Inovia’s data lake Insight, where the analysis is done. In order to further increase profitability, external data such as weather forecasts are added so that rainwater levels can be predicted and therefore how much water should be added behind the dam. In addition, data from the energy market is connected up in order to get indications of how much power should be produced and when is the best time to sell it. The AI system automatically adjusts the water flow once a minute based on the turbine’s status, current prices etc.

“It’s a matter of collecting a huge number of different types of data,” Robert Berg says. “Of course, it is a challenge to process it all and be able to do analyses in real time so as to optimise the process. It was important to create a user-friendly tool for the company that would give a rapid return on investment, and we succeeded in doing that.”

Saxnäs/Inovia is one of many cases at IoT Arena, where visitors can be inspired by real-life successful IoT cases from various segments of the industry.