What the Detecon Energy & Data Event reveals about the digital energy transition
An industry under tension
There are moments when a single sentence sums up an entire industry. At this year’s Detecon Energy & Data Event in Zurich, one such sentence dropped almost in passing. Modern power grid operation now demands “the mind of a trader at the speed of a machine.” What used to count as an operational routine task, keeping the grid balanced between production and consumption, has become real-time risk management across several markets at once.
Andrea Tribelhorn, Partner and member of the Executive Board at Detecon Switzerland, opened the event with a diagnosis that surprised no one in the room and yet occupied everyone. The energy industry is under multidimensional transformation pressure. Markets and prices are shifting rapidly. Regulation keeps evolving. And technological innovation accelerates the change even further. Three forces that reinforce one another, and that no energy provider can afford to view in isolation anymore.
Three fronts of transformation
Tribelhorn’s view of the Swiss energy market can be organized along three fields of tension.
Markets and prices
Volatility on the spot markets is extreme. Unpredictable production swings in solar and wind power put margins at risk. And the phenomenon of price inversion, low prices under strong sunlight and high prices under weak, makes planning procurement and sales portfolios increasingly difficult. Those who cannot react flexibly end up paying for it. The answer lies in the ability to flexibilize: AI-based forecasting models, algorithmic trading optimization, the expansion of storage capacity, and dynamic tariff models that pass wholesale market price signals directly through to consumers.
Regulation
Above it all hangs the uncertainty around the EU electricity trading agreement. And the striking part is that both scenarios lead to substantial yet differently shaped consequences. A yes would force new business models through market liberalization, with the end of the monopoly position held by local energy providers and the arrival of open competition. A look at Germany shows where that can lead: non-industry aggregators, “discount electricity tariffs” from Lidl or Aldi, and municipal utilities losing market share to purely digital sales companies. A no, on the other hand, raises the structural risks in the power system. Without legal integration into the European internal electricity market, access to cross-border capacity, balancing energy, and coordination mechanisms remain restricted. Over the medium term, this can strain security of supply and stable grid operation, and it calls for additional national safeguards. Regardless of how the vote turns out, the importance of resilience against cyber risks in the energy infrastructure also grows.
Technology
The energy system is turning around, from central and fossil toward decentralized and renewable. Consumers are becoming “prosumers.” Coordinating heat pumps, EV charging stations, and local energy communities all at once pushes existing grid and IT infrastructures to their limits. Distribution networks that grew over decades were simply never designed for these load peaks. At the same time, the generation costs for renewable energy and battery storage keep falling, which makes decentralization not only technically possible but economically attractive.
From principle to practice: Three stories
The optimizer: When the machine outpaces the operator
The first story is about a paradox. The operational “DNA” of a grid operator is built for stability, not for the multi-market optimization the future demands. As the number of grid imbalances above 500 MW rose dramatically year after year, an operational task turned into a financial risk in real time.
The solution was a shift from reaction to anticipation: a machine learning system that predicts grid imbalances before they occur. The development of this model architecture is itself a lesson. From a first single ML model, through an ensemble approach, to a hierarchical system of more than 50 models with MLOps, forecasting performance improved step by step, rising 37 percent above the baseline.
The results speak for themselves. 98 percent of balancing energy activations are now carried out by the Optimizer, the forecasting error fell by 37 percent, and projected annual costs dropped by 23 percent.
Yet perhaps the most honest slide of the entire event showed something else. The system’s adoption curve, plotted on the Kübler-Ross change curve, that model of shock, denial, frustration, and finally integration that usually describes grief. A “turbulent go-live” in May 2024, followed by months of skepticism, until a performance report in December revealed the enormous economic value and turned the tide. Only in May 2025 did the actual autopilot go live. The message: the biggest hurdle was never the technology. It was the people.
Yuon: Heating networks that think along
The second story came from Dr. Sebastian Hersberger, CEO of Yuon, and moved the focus from the power market to the heat market. Thermal networks today cover around 10 percent of Swiss heating demand, with roughly 1,200 networks and a strongly rising trend. The city of Zurich alone plans to invest around 2.3 billion Swiss francs in expanding district heating by 2040. District heating is increasingly becoming the “infrastructure of the heat transition.”
Yuon’s approach sounds almost too simple to be true: get more output from existing networks without a single additional hardware investment. Through a digital twin of the network, ML-based detection of system parameters, and model predictive control, load peaks are broken, temperatures are optimized, and thermal storage is used in a forward-looking way.
The implementation examples showed what that means in practice. At a district heating center, morning load peaks fell by 27 percent, with cost savings of more than 15 percent. In one district heating network, fossil oil consumption was cut by around 23 percent. And at a shared connection in Zurich, load peaks were reduced by 33 percent, lowering the required connection capacity by around 26 percent. In total, Yuon promises up to 30 percent more network capacity and 20 percent lower operating costs, through IT-based optimization alone.
Uniper: COODE turns business ideas into scalable use cases
The third story centered on COODE, COO Digital Evolution, presented by Raquel and Sergej from Detecon, who supported the program at Uniper. The starting point was a familiar challenge. Good ideas exist, but they often stay isolated or fail to scale.
COODE creates a structured foundation for turning business ideas into scalable digital use cases. At its core is an integrated data governance architecture that connects critical data. More than 100 data sources form the basis for over 300 digital solutions, built on reusable data capabilities.
One example is CO₂ reporting. The goal was a single source of truth for Scope 1 emissions data. The starting situation was fragmented, with data spread across systems, Excel, and manual processes, plus complex calculation and approval logic.
The solution: central data consolidation, standardized and automated CO₂ calculation, and delivery through a scalable architecture. Today, users mainly verify the calculated values. Reporting became far more reliable, transparent, and reusable. At the same time, manual effort dropped considerably and auditability improved significantly.
The effect reaches beyond individual use cases. A scalable ecosystem for data-driven innovation emerges, with 72 percent process improvement, 59 percent time savings, and 43 percent better data access.
COODE shows how ideas turn into reproducible digital value, with scalability designed in from the very start.
The real lesson: It is not about technology
Anyone who reduces the event to its metrics misses the core. Across all the talks, one insight crystallized, the one Andrea Tribelhorn captured in her notes. The greatest challenge is usually not the technology but bringing people along the way.
Data projects create measurable value. Uniper reported four euros of value for every euro invested. But this value only emerges under certain conditions: fast realization of value instead of years of lead time, a focus on the business side instead of pure IT topics, and above all long-term, sustainable initiatives instead of isolated one-off projects.
The event’s central insights can be grouped into five bundles.
1) First, every scaling effort needs a robust data foundation. Data is increasingly becoming a bottleneck, not a given.
2) Second, a platform and technology shift is underway, moving away from physical infrastructure toward integrated, digital systems with agile, product-oriented ways of working.
3) Third, artificial intelligence acts as a fundamental lever of transformation, comparable to industrialization, while the traceability of the models and a “human in the loop” remain indispensable for trust and governance.
4) Fourth, new systems are often unstable at first in operation. Only iterative work and consistent use raise performance and acceptance.
5) And fifth, the common thread running through the entire day: the biggest hurdle is not the technology, but bringing employees along on this rapid, multidimensional transformation.
Conclusion
The Optimizer story with its grief curve and the Yuon examples with their sober percentages ultimately tell the same thing. The transformation of the energy industry is less a technology project than a holistic shift in business models, organization, and decision logic. The tools are here. The models work. The savings are real.
What decides between success and failure is the ability to integrate data, AI, and governance consistently, and to take the entire organization, with all its people, on the journey. As Sebastian Hersberger put it for heating network optimization, the same holds for the whole industry: the biggest step is the first one. Optimization begins with data, not with perfection.
The Energy & Data Event 2026 took place on 26 May in Zurich. It brought representatives of the Swiss energy industry together with experts from Detecon and partner companies such as Yuon.
