Alliander N.V.
AI is emerging as one of the most consequential technologies of our time. GenAI has the potential to boost AI/ML applications already in place, exploiting the power of this disruptive approach and the latest developments build on it, such as AI agents. GenAI-based foundation models have the potential to revolutionise and replace several traditional AI/ML applications. The integration of AI in the energy sector can yield significant advantages, such as enhanced grid flexibility, optimised congestion management, predictive asset maintenance, and advanced data analytics.
A workshop on 31 March 2025 organised by ADRA together with representative of DG CONNECT and DG ENERGY, gathered more than 250 online participants to discuss the transformative potential of Generative AI/ AI in the energy sector analysing quick-win applications that deliver immediate value to smart grids. The energy industry can harness the full benefits of this powerful technology to pave the way for a more resilient, efficient, and sustainable energy future
While further electrification creates challenges already today, such as grid congestions as often referenced to the congested situation in the Netherlands (see map below), AI solutions might have a pivotal role in accelerating the adoption of intermittent renewable energy sources like solar and wind, managing demand-side dynamics with advanced forecasting and optimisation while targeting high utilisation of existing grid capacity. As grid operators like Tennet and Alliander further reaching better understanding of the grid situation, it is the AI capability of forecasting of energy flows and possible grid saturation. Accurate demand forecasting is crucial for balancing electricity supply and demand in smart grids. AI improves forecasting by analysing large datasets, identifying patterns, and adapting to real-time changes.
The potential of foundation models to enhance grid operations was highlighted, demonstrating their versatility in performing a broad range of tasks, including state estimation, forecasting, and contingency analysis. By leveraging intensive self-supervised training on large datasets, these models can be fine-tuned to excel in various applications, streamlining the process and eliminating the need for manual feature engineering and task-specific model development.
The workshop gave some insights into dynamic grid reconfigurations that could not be imagined previously. It introduced innovation at substation level (e.g. virtualisation and edge computing), where new features have been shown (for example, a substation switching and routing of imminent energy demands or supply). Those features, when combined with AI applications (such as reinforced learning) and agentic AI, are instrumental for topology reconfiguration and agile grid management. Industrial operational platforms are likely to play a key role, and leading European system integrators SchneiderElectric have the potential to bring GenAI capabilities on premise for more efficient and agile grid operations for key business in areas like buildings, data centers, or grid system operations.
The combination of AI and the Internet of Things (IoT) may be the basis for transforming demand-side flexibility (DSF) by enabling real-time monitoring, automation, and optimisation of energy consumption. Standardised technologies of the Web of Things (WoT) are reported to coordinate energy flows across smart edge devices and make energy applications convenient in a way similar to application development on the Web.
The Commission supports the diffusion and uptake of AI through its strategy on the Competitiveness Compass hinting at the forthcoming Apply AI strategy, as well as by mobilising €200 billion into AI support through its programmes, such as Horizon Europe and InvestAI. More specifically, the Commission is tackling the integration of Generative AI across different edge devices at industrial or consumer end points through its Digital Spine concepts which will be supported under Horizon Europe WP 2025 under Cluster 5. It may help consumers and grid operators efficiently adjust energy usage based on grid conditions, price signals, and renewable energy availability. More to details will be communicated during the planned InfoDay on 6 May 2025. Find further information on Horizon Europe Info Days – Cluster 5
Map of the grid congestion in the Netherlands
Courtesy of Netbeheer Nederland