When I saw that the European Commission had launched a consultation to shape a Strategic Roadmap for AI in the energy sector, I was very pleased. This is a timely and important step. Artificial Intelligence has the potential to accelerate digitalisation, improve efficiency, and support Europe’s decarbonisation goals. But as I read further, I noticed what was missing: the role of AI at the edge.
The conversation often focuses on the cloud. And for good reason. Cloud AI allows us to aggregate and analyse vast amounts of data, forecast demand, and optimise entire networks. It has changed the way we manage energy. But the cloud also has limits. Every byte of data we send upstream consumes bandwidth, introduces latency, and relies on energy-hungry data centers. When connectivity drops, devices often fall silent, waiting for instructions. And when sensitive data is constantly in transit, it increases exposure to cyber risks and raises questions about privacy and sovereignty.
That is why I believe edge AI is the missing piece. By embedding intelligence directly in devices such as smart meters, gateways, and distributed assets, we give them the ability to act locally and autonomously. Data can be processed instantly without relying on a distant server. Devices keep working even if the network goes down. Privacy is easier to protect because sensitive information stays within the device. And when combined with cloud capabilities, we get hybrid systems that are both powerful and efficient: global optimisation from the cloud, local autonomy from the edge.
This approach is fully aligned with the European agenda and the Commission’s consultation signals growing awareness of AI’s role in energy. One report from the Commission already highlights how digitalisation and AI can make our systems smarter, greener, and more inclusive, while respecting consumer rights. At the same time, Europe has reached more than 63 percent penetration of smart meters for electricity consumers, with coverage expected to keep rising. These devices form a solid foundation for intelligence at the edge—but only if we stop treating them as static endpoints and start seeing them as adaptable platforms.
So how do we get there? The answer lies in secure execution environments. At MicroEJ, we developed VEE, a lightweight, AI-ready container designed specifically for constrained devices. VEE runs applications in isolation, optimises the use of on-device compute resources (including neural network accelerators), ensures portability across hardware families, and enables remote updates.
For utilities, this means they can dynamically add new features and services without replacing millions of meters. A device that once performed only basic measurements can evolve over time to support advanced metrology, DER disaggregation, predictive maintenance, or demand response services.
This approach also creates space for innovation ecosystems. With secure containers, SMEs, start-ups, and utilities themselves can develop applications that run safely on fleets of devices. Instead of a closed, hardware-bound model, we can enable a dynamic environment where services evolve continuously.
And this is not a theory. It is already happening. Leaders such as Sensus, a Xylem brand, and Landis+Gyr embed MICROEJ VEE in their devices through our solution VEE Energy. In the United States, utilities are deploying programmable meters based on this architecture. They are updating them with new capabilities such as load disaggregation and predictive maintenance, all without changing hardware. These real-world examples prove that edge intelligence in secure containers is not a future bet. It is a proven, scalable, and cost-effective way to modernise the grid.
For Europe, the opportunity is significant. By embracing edge AI, we can make our grids more resilient, accelerate innovation, and build systems that are both efficient and sovereign. We can reduce dependence on data centers, address privacy concerns more effectively, and foster a diverse ecosystem of developers and service providers. Most importantly, we can ensure that the massive investment in smart meters across the continent delivers ongoing value rather than becoming locked into static functionality.
As ESMIG members, we are not just suppliers of devices. We are shaping the digital backbone of Europe’s energy transition. That requires us to push for standards and interoperability, so innovation can move freely across markets. It means investing in developer ecosystems that allow new services to emerge. And above all, it means changing our perspective: devices are not endpoints, they are platforms.
The cloud has brought us far. Intelligence at the edge will take us further. The Commission’s consultation is a chance to ensure Europe recognises this. It is up to us, as industry leaders, to make sure edge AI becomes part of the roadmap for the future of energy.
Find out more about MicroEJ.
—
Disclaimer: This is a blog space for debate where ESMIG members share their thought leadership. All opinions expressed are the author’s. The content of this article is not an official position paper endorsed by the association.