Interview

How data accelerates the energy transition

As energy systems become more decentralized and dynamic, the role of data has never been more important. From planning to operation, access to high-quality, well-structured energy data is key to making smarter, faster, and more cost-effective decisions. But how can organizations harness this data effectively and what tools are available to help them do so? To shed light on these questions, we spoke with Mahsa Faraji Shoyari and Dr. Jessy Matar, researchers at the Architectures and Service for Critical Infrastructures competence field at fortiss. They share insights into how modern data infrastructures, automated processing, and digital twins are revolutionizing the way energy systems are planned, managed, and optimized.
How can modern data infrastructures improve energy planning and efficiency?

Jessy Matar: Efficient energy planning starts with strong foundations – and that means having the right data, in the right format, at the right time. At fortiss, we’re investigating how modern infrastructures can transform the way energy data is collected, organized, and applied.
We focus on data pipelining and orchestration frameworks that enable real-time, low-latency data flows from sensors, meters, and external sources to analytics platforms. Using application programming interfaces and semantic models, we ensure easy interoperability. This approach minimizes integration complexity and enables large-scale data exchange between various systems, such as data providers (e.g., geoportals, dataspaces) and consumers (e.g., planning tools, analytics, simulations).

Mahsa Faraji ShoyariThis structured approach empowers planners, operators, and policymakers to access a comprehensive, coherent view of the energy landscape. With high-quality, timely data, decision-making becomes faster and more reliable. This reduces uncertainty, accelerates project timelines, and allows for the optimization of energy systems before a single foundation is laid. That leads to lower investment risk, better long-term performance, and more sustainable energy outcomes. 

How does the integration of diverse data sources enhance planning accuracy?

Mahsa Faraji ShoyariEnergy systems don’t exist in isolation – they’re influenced by everything from weather conditions to building use patterns. Integrating these diverse data streams provides a more complete and contextual view, which is essential for accurate forecasting and resource planning.
By combining data from buildings, smart meters, and local weather models, we enable precise energy demand analysis. To structure and interpret this data more meaningfully, we use knowledge-based approaches such as semantic models, which describe what things mean and how they relate. We also use domain ontologies, which define the key concepts and relationships within a specific domain.

Jessy Matar: This not only improves interoperability across systems but also supports explainable and context-aware decision-making.
It’s about moving from isolated data points to an integrated, intelligent system that supports better decisions at every level.

What role do digital twins play in optimising energy infrastructures and reducing costs?

Mahsa Faraji ShoyariDigital twins are becoming a major pillar of modern energy planning. These virtual models replicate real-world energy infrastructures – down to the smallest detail – and allow us to simulate scenarios, detect inefficiencies, and test improvements before they’re rolled out in the real world.
We use data-driven simulations to guide economic and technical planning. Whether it’s evaluating the cost-effectiveness of different technologies or stress-testing a grid design against real-time conditions, digital twins provide a powerful way to reduce both risks and costs.
Ultimately, they help ensure that investments in energy infrastructure are future-proof, resilient, and aligned with long-term sustainability goals.

How do you make sure the AI models stay accurate when forecasting, especially when the weather or other conditions suddenly change?

Jessy Matar: We ensure AI models stay accurate by using adaptive models or ensembles that learn from the latest data to maintain performance. Combining machine learning with physics-based models can also add extra robustness during enexpected shifts. In addition, including real-time weather forecasts, satelite data, or policy-related indicatoars(e.g regulary changes, economic policies or environmental regulations) as input features can help the model anticipate changes.