Case study

    Staying steady in a volatile market: AI-powered forecasting at Stora Enso

    In industries such as renewable packaging, biomaterials, and wood products, planning has become a balancing act.

    Service

    Data and AI Technology and Engineering

    Industry

    Industry Manufacturing Energy and Resources

    Fluctuating energy prices, rapidly changing demand, and production challenging production decisions make accurate planning difficult. Manual forecasting often falls short, leading to gaps between projections and customers’ actual needs.

    Do unpredictable swings always signal uncertainty? AI-powered forecasting leverages real-time data and predictive insights to transform complexity into clarity, enabling Stora Enso to plan smarter, respond faster, and stay ahead rather than catch up.

    Stora Enso is a leading provider of renewable products in packaging, biomaterials, and wood construction, as well as one of the largest private forest owners in the world. Guided by its purpose, “Do good for people and the planet, and replace non-renewable materials with renewable products,” Stora Enso continually seeks to enhance operational efficiency and minimize its environmental footprint.

    Accurate forecasting enables better business

    A few years ago, Stora Enso recognized that accurate energy consumption forecasts are fundamental to one of the company’s core business priorities: optimizing value in a circular bioeconomy.

    “Most Stora Enso sites produce packaging and biomaterials, which requires electricity and steam for drying,” explains Christopher Tonk, Digital Solution Advisor at Stora Enso. The company produces parts of its operational energy needs in-house and purchases the  remaining from the market. This requires continual balancing of demand, supply, and prices, especially given the heightened volatility in energy markets since the outbreak of the war in Ukraine. 

    As Tonk explains: “We want to be able to accurately forecast the demand of steam and electricity, especially for the next 24 hours. Without the ability to forecast in near-real time, we could not react immediately to, e.g., planned and unplanned shutdowns. By improving and automating our forecasting, we would enable those near-real-time predictions.”

    A primary objective in developing energy consumption forecasting was to reduce Stora Enso’s energy costs. Additionally, the goal was to support the company’s sustainability agenda by decreasing energy use per ton of products sold. “Supporting our sustainability goals is very important for Stora Enso as a whole,” Tonk emphasizes.

     

    Co-creation brought together the process, the tool, and the people

    Stora Enso and Nortal co-created an energy demand forecasting solution. The solution was piloted in one of Stora Enso’s Finnish mills and consists of three elements:

    Process: The existing workflow was improved and optimized by implementing AI-based forecasting and integrating it seamlessly.

    Tool: A productized, AI-powered Energy Demand Prediction Tool (EDPT), running in the Azure cloud, is integrated into Stora Enso’s Energy Management System. Forecast visualizations are also available in Power BI.

    People: From the outset, the solution was designed to align with existing practices and enhance mill operators’ daily work with minimal disruption.

    To train the AI models, the system is provided with 1–2 years of historical process data. The training dataset also includes shutdown and production plans, as well as freshwater temperatures. Weather conditions and forecasts are also essential, given Finland’s significant seasonal temperature variations. Leveraging this intelligence, the solution automatically estimates the amount of electricity and steam required to produce a given product.

    Invisible technology with powerful impact

    “We have achieved excellent results under normal, steady operations,” explains Christopher Tonk, and continues: “With such accuracy, the solution will be an important lever in enabling us further to optimize our power plant and energy market operations.”

    Tonk also envisions additional impactful outcomes: “Now that we have that information, insights, and forecasts available, we are investigating other opportunities as well. Maybe we can find entirely new revenue streams from that intel.”

    Seamlessly integrated into existing processes and tools, the new, more accurate forecasting solution runs in the background, causing minimal disruption to site experts’ daily work. It is almost invisible – just as the best technology should be.

    Christopher Tonk, Digital Solution Advisor at Stora Enso

    “We have achieved excellent results under normal, steady operations. With such accuracy, the solution will be an important lever in enabling us further to optimize our power plant and energy market operations.”

    Listen and watch the full story

    Learn more about this project by watching the keynote by Christopher and Ergin, recorded at The Subcontracting Fair 2025 in Tampere, Finland.

    Get in touch

    Let us offer you a new perspective.