Helsinki – European utilities are facing a historic turning point: increasingly complex grids, extreme weather events and customers demanding reliable and sustainable electricity are exerting pressure on infrastructure designed and built in a different era. At the 2026 edition of Eurelectric’s Power Summit, held this year in Finland, Renewable Matter interviewed Shubbhronil Roy, VP of Strategy, Marketing & Transformation at Schneider Electric, a world leader in energy management technologies.

Roy offers a privileged perspective on how to overcome data fragmentation, unleash hidden capacity within the grid and use artificial intelligence (AI) not as an “autopilot” but as a co-pilot serving operators and planners, tracing a clear path towards Europe’s “grid intelligence” by 2030.

Utilities today operate through fragmented systems, each with its own data silo. What is the real cost of this fragmentation for decision-making and AI adoption in the grid, and how do you build a shared data foundation without discarding existing investments?

There is One Grid, many memories, but no single operational truth. Utilities have SCADA systems, ADMS, GIS, OMS, metering, and asset planning tools that are somehow integrated. At one point, we even used to refer to integration as the holy grail of this industry. Utilities have spent huge sums on integration that never delivered a single operational truth. The industry has suffered a lot from this. It is part of our legacy, and it is very structural, because many of these silos come from the way utilities themselves are organised. I recently saw a utility where the SCADA department was separate from the ADMS department. That shows the level of fragmentation. Each department wants its own single “source of truth”, which, of course, makes sense from their perspective. But what becomes truly compelling is when you cut across those silos and think about the customer – the utility customer – and about the utility as one enterprise that connects planning, operations, asset management and all the other functions together. AI inherits fragmentation and doesn’t cure it. AI needs trusted, interoperable and governed data, as well as the context required for operational decision-making. This is where Schneider Electric’s One Digital Grid Platform, for instance, was invented; it is connecting ADMS, DERMS, GIS, planning, asset and flexibility capabilities into one governed foundation, rather than replacing any of these systems.

Utilities have long operated reactively: an outage happens, then you restore. Is there a shift toward predicting and preventing grid risk?

This shift has clearly begun, and it is accelerating. It is being driven by ageing assets and by increasingly extreme weather events, floods, storms, wildfires, heat waves. I just came back from the US, where we held our Innovation Days, and there you can see how serious the situation has gotten. In some cases, a utility can go bankrupt because of wildfire liability laws if it is proven that the fire originated from utility operations. So the shift is not only about restoring faster after an outage. It is about predicting risk earlier, preparing better, and reducing the impact before events escalate.

Which data sources are proving most valuable to anticipate outages? How mature is this approach in Europe compared to more advanced markets?

The value really comes from combining three different layers of data. First, grid data: SCADA, ADMS, OMS, smart meters, sensors – everything happening across the network. Second, asset data: age, condition, maintenance history, failure records, inspection reports. Third, external data: weather forecasts, vegetation information, satellite data, LIDAR imagery, wildfire and flood risk indicators. The breakthrough does not come from one data source alone. It comes from bringing these layers together into one decision environment so utilities can understand not only what is happening, but what could happen next. We also work extensively with partners. One of our key partners is AiDash. Together, we focus a lot on major event mitigation, not necessarily on avoiding events altogether, but rather on helping utilities prepare much better for them, reducing losses by as much as 50% through better management. I would say the US is slightly ahead, mainly because of the liability and weather challenges I mentioned. In some parts of the country, weather disruption is extremely severe, and that is why utilities have adopted these technologies promptly. Europe is moving in the right direction, but progress is uneven. Many DSOs still lack sufficient low-voltage observability; for many of them, the last mile of the grid remains something of a blind spot. In the end, AI’s role is to support decision-making, not replace humans. As we often say, AI should not be an autopilot, but a co-pilot: helping operators, planners and maintenance teams make better decisions. That is where we see the acceleration happening.

Europe needs massive grid investment to support electrification, but permitting and construction timelines are slow. How can digital visibility and planning help unlock more capacity from the existing grid in the meantime?

I will share a story very close to my heart, in which I was directly involved. The story is not set in Europe, but it is highly replicable here. It comes from California. Tesla approached the DSO there, Pacific Gas and Electric, with the intention of building a high-power charging station in Southern California. When PG&E ran the standard calculations, they concluded that Tesla would need to wait three years, as a new substation and power lines were required. Then, a proof of concept began using our EcoStruxure ADMS and DERMS platform. Through collaboration with the customer, within two to three months we found – through simulations, situational awareness, and analysis of available capacity – that we could deliver the megawatts Tesla needed for 340 days out of 365. In the past, planning often worked on conservative multipliers, i.e., if you needed 100, you planned for 300, but that approach no longer holds. To conclude, the substation was then built within the following three months. We went from a three-year timeline to roughly six months. This is a concrete example of what digital technologies can achieve, and we see similar cases in the Nordics, where better visibility and partnerships with charging operators deliver comparable results. More broadly, it highlights the importance of integrated planning. Planning can no longer remain siloed: it must become more integrated and closer to real time. The key framing is that integrated planning does not replace grid investment; it makes grid investment smarter, faster and more targeted. It changes the question from “what do we need to build?” to “what can the grid safely support today, and where do we truly need reinforcement?”

With rooftop solar, EVs, batteries and heat pumps, distribution networks are becoming bi-directional and more volatile. DERMS (Distributed Energy Resource Management Systems) is often seen as part of the answer, but who really has the visibility and authority to orchestrate flexibility in real time, and how must the DSO (Distribution System Operator) role evolve?

One crucial issue is visibility and how orchestration will work. In Europe, DSOs arguably have the best understanding of local constraints, but they do not own all assets. This reflects a broader shift: the traditional definition of the grid – ending at the customer meter – is no longer valid in a world of EVs, PVs, heat pumps, and prosumers. Much of today’s regulation was designed for that legacy system, while the grid has already evolved. Regulation now needs to enable DSOs to orchestrate this new reality. Whether a single actor can do this alone is unclear; it will likely require an ecosystem approach involving DSOs, TSOs, retailers, prosumers, and technology providers like us. Regulation should encourage this collaboration and incentivise DSOs to lead it. There is also significant hidden capacity in the grid: in the last 12–18 months, over €7 billion of renewable energy in Europe was curtailed. That energy was produced but not used because parts of the network lacked sufficient flexibility or perceived reliability. Better orchestration could recover a substantial share of this value. So the role of the DSO is evolving: from operating the network to orchestrating flexibility across the local energy system.

You mentioned incentives, and we know there are many forms. In your experience, which are the readiest to deploy and effective?

The UK has some of the most advanced regulatory mechanisms, such as dynamic pricing and active network management or “envelope” schemes. There are, of course, penalties when performance moves in the wrong direction, but overall, the frameworks have been applied constructively. Another strong example is Germany. The challenge is that approaches remain fragmented: each country implements these ideas differently, which still creates friction. We can also incentivise DSOs by compensating them for non-wire alternatives. Instead of reinforcing the network – since today they are primarily rewarded for the physical assets they build – we should reward them for non-wire solutions. This means going further in using flexibility and treating traditional reinforcement as a last resort, including solutions such as storage. This would better incentivise DSOs and help unlock underused capacity in the grid. In simple terms, regulation should reward the capacity unlocked, not only the assets built. That can mean faster connections, lower system costs, less curtailment and better use of existing infrastructure.

The power sector is facing a “silver tsunami” of retiring engineers and operators. Can AI help operators and planners to preserve institutional knowledge rather than just automating tasks?

It is about augmenting people, not replacing their judgement. We will see a significant number of new hires, a more digitally savvy generation for whom AI will feel natural. AI will also help bring forward the domain knowledge held by people with thirty years of experience. If you think about someone joining today, after five years they cannot have that same depth of experience, but with AI, they can progress much faster along that learning curve. In that sense, AI can play a major role in preserving and transferring institutional knowledge, capturing how experienced engineers diagnose faults, read network behaviour and make trade-offs, then embedding it in digital assistants, playbooks, training and decision-support tools. So the value is not only automation. It is also helping utilities turn decades of operator knowledge into structured guidance for the next generation.

“Grid intelligence” and AI are widely discussed, but adoption is the hard part. How do you move a traditionally conservative DSO from decades-old operating models to trusted, AI-assisted or even partially autonomous grid operations?

AI is already here and advancing at a pace we have never seen before. However, I do not think we will see a big-bang transformation in the market. Its adoption and the development of grid intelligence should be gradual and carefully managed; it must be trustworthy. Trust is critical in grid operations and for grid customers, because these are mission-critical systems. Grid intelligence will therefore be driven by AI for visibility, AI for recommendations, and AI to enhance operator actions. Only step by step will autonomy increase, confidence build, and the approach extend across more use cases. Europe’s distinctive stance is trustworthy, human-centric AI. The EU AI Act’s risk-based approach and the Commission’s link between energy digitalisation and renewables, cybersecurity, privacy and sovereignty, all push this way.

Finally, where do you see EU utilities in 2030?

I believe that, while Europe may not be moving the fastest today compared to the US or Australia, it will ultimately be the most reliable and trustworthy in orchestrating this transition. On flexibility, for example, the UK is clearly ahead of continental Europe, so there is still catching up to do. I would say Europe may not be the fastest, but when it moves, it will do so in a robust and beneficial way, like in the Nordics or some projects in Italy with major customers. We are already seeing this holistic orchestration emerge through partnerships, technology, and collaboration between DSOs and retailers. It is not about rushing ahead and then stumbling, but it is about moving at the right pace to address the major challenges we face. For these reasons, I am more optimistic than some reports would suggest.

Cover: Shubbhronil Roy