A Map Is No Longer Enough. Without High-Quality Data, AI Won't Deliver.
28.7.2026 | Reading time: 13 minutes
- Energy utilities, airports, and water companies need more than maps today. They need data—and the ability to predict what comes next.
- Digital twins make it possible to simulate crisis scenarios and plan billion-dollar investments with greater confidence.
- Artificial intelligence alone cannot solve these challenges without high-quality data, says Ivo Růžička of Unicorn.
Companies today are talking primarily about artificial intelligence. Yet you argue that data is even more important than AI itself. Why do you frame it this way?
AI may be the most visible technology today, but the most significant transformation of recent years has been the digitalization of data. Artificial intelligence is built on that foundation. If the data is inaccurate or incomplete, its outputs will be unreliable as well. Many organizations focus on choosing the right AI solution before addressing their underlying processes and data. The real priority should be establishing a single source of truth across the entire organization. For example, when an energy company plans a scheduled outage or a city undertakes a road reconstruction project, everyone involved cannot be working from different datasets. They all need to rely on the same reflection of reality. AI is therefore not the starting point of digital transformation—it is its natural next step.
Hasn't the boom around ChatGPT and generative AI overshadowed other technologies that may ultimately prove more useful?
Absolutely. Generative AI is excellent for producing text and writing code, but in industry and critical infrastructure, the greatest economic benefits over the coming years will come from other AI applications—such as computer vision, predictive analytics, classification, and change detection. A good example is field technicians. They can simply take a photo of a piece of equipment with a mobile phone, and AI instantly identifies the type of damage or any deviation from normal conditions. It doesn't replace an experienced technician, but it saves hours of routine work. The real value of AI isn't created when the system recognizes something—it's created when that insight enables people to make better, faster decisions.
You've worked in geoinformatics for many years. What does that term mean today? Is it still just about sophisticated maps?
Fifteen years ago, maps primarily served as asset inventories. Today, they are simply the user interface. The real value lies in a digital model of reality that connects information about infrastructure, its condition, capacity, and interdependencies. In the past, we asked where a utility pole stood or where a cable was located. Today, we ask what happens if we connect a new power source, close a valve, or begin construction work on part of an airport. The purpose of a digital model is no longer just to visualize the world—it is to support better decision-making.
How far do investments in modernizing infrastructure really need to go? What scale of spending are we talking about?
The figures are enormous. According to guidance issued by the European Commission last June, Europe will need to invest around €477 billion in transmission networks and another €730 billion in distribution grids by 2040. But before that trillion euros is invested in actual cables, pylons, and substations, one fundamental question must be answered: where? Where is the grid already constrained? Where will demand increase? Where should new infrastructure be built? This is precisely where geoinformatics comes in. It provides answers to these spatial questions. It is not a map hanging on a wall, but a digital model of the entire network on which these massive investment decisions are planned and managed.
Can savings really be achieved on projects of this scale?
Absolutely—and they can be substantial. For example, Eurelectric's Grids for Speed study shows that smarter grid planning and management could reduce annual distribution investment requirements from €67 billion to €55 billion. That's a saving of roughly €12 billion every year. Those savings won't come from buying cheaper concrete for utility poles. They'll come from better planning and making more efficient use of existing infrastructure. And every one of those decisions is driven by data. A high-quality digital model of the network is quite literally the tool that makes those savings possible.
So we're no longer talking about a static picture of the network. What does a digital model enable in practice today?
Above all, it enables informed decisions before problems occur. Today, organizations are interested not just in where something is located, but in the consequences of every decision. When an energy company connects a large data center to the grid, it must simulate whether sufficient capacity exists, whether a new substation will be required, and whether nearby customers could be affected. The same applies in the water sector. If a major pipeline fails, it isn't enough to locate the break. The system must immediately calculate how many people will lose water supply, determine which valves should be closed, identify how flows should be rerouted, and dispatch field crews to the right locations. That is no longer digital mapping—it's active risk management.
What you're describing sounds very much like a digital twin. Is that what it is, or is there more to the concept?
You're right, but many people still think of a digital twin as simply an attractive 3D model of a building or a city. That's a misconception. A digital twin is not a static representation—it's a living system. It is continuously updated and capable of simulating future developments. The key isn't what things look like, but how they behave. Building these living models is at the heart of what we do. But our role extends beyond creating the model itself. We help customers connect their data, processes, and people so they can manage operations and investments effectively using that digital twin.
Grid connection approvals are particularly relevant in the energy sector today. Distribution operators are facing an unprecedented wave of applications because of the green transition. Is this still manageable?
We're already close to the limit. Leonhard Birnbaum, CEO of E.ON, recently said that by 2030 Europe will need to process approximately one new grid connection every seven seconds during the working day. Eurelectric's data also shows that more than 450,000 renewable energy connection applications were submitted across Europe in 2024—a 133% increase compared with 2021. At a pace of one new connection every seven seconds, automation based on high-quality data is no longer optional. If an engineer has to manually search paper maps to determine whether the grid can accommodate a new solar power plant, the process takes weeks. With an up-to-date digital model, the answer can be calculated almost instantly.
Yet companies still wait months or even years for grid connections. Network operators themselves often don't know where capacity is available.
Exactly. That's why the European Commission now requires network operators to publish capacity maps—publicly accessible overviews showing where new connections are possible and where the network is already full. Creating those maps depends entirely on having a high-quality digital model of the network. They are generated directly from that model, and without it, no capacity map can be considered reliable.
Building new infrastructure in Europe is often slowed by bureaucracy. Can technology help?
This is one of the industry's biggest challenges. According to European studies, constructing new network infrastructure can take ten years or more, and more than half of that time is spent on permitting rather than construction itself. Technology cannot shorten statutory approval periods, but it can eliminate many unnecessary delays between individual stages. The permitting process is largely a spatial data challenge. You need to identify which land parcels will be crossed, which protected zones will be affected, and where conflicts with environmental regulations or existing development may arise. When all of this information is brought together in a single digital model, alternative routes can be evaluated in days rather than months, and authorities can receive complete, accurate documentation from the outset.
This sounds highly relevant for energy and transport. Does it have applications beyond those industries?
Absolutely. The underlying principle is universal. We build digital twins for virtually any domain. For the Czech National Development Bank, for example, we're creating a digital model of loans and guarantees. For the Ministry of Labour and Social Affairs, we're digitizing administrative processes. In every case, the goal is the same: to connect data, processes, and people so that better decisions can be made more quickly.
How are these systems developed? Do you build everything from scratch for each customer?
That would be both inefficient and expensive. We've developed our own platform, Unicorn Universe, which is essentially a digital construction kit built from proven components. Solution architects assemble these components into systems tailored to each customer's needs. The building blocks already incorporate cybersecurity, architecture, and compatibility. Instead of starting from scratch every time, we're assembling trusted, secure components, making development dramatically faster.
One of your flagship projects is the digital transformation of Prague Airport. What exactly changed there?
An airport is one of the most complex operational environments in the world. Within a relatively small area, it combines transportation, energy infrastructure, stringent security requirements, construction activity, and international regulation. Managing such an environment without a digital twin is no longer realistic. Our role, however, was not simply to introduce new technology. It was to help drive a fundamental change in the way the airport operates and makes decisions.
One of your flagship projects is the digital transformation of Prague Airport. What exactly changed there?
An airport is one of the most complex operational environments in the world. Within a relatively small area, transportation, energy infrastructure, stringent security requirements, construction activity, and international regulations all intersect. Managing an environment like that without a digital twin is no longer realistic. But our role wasn't simply to replace technology—it was to help transform the way people think and work.
In what way?
Previously, only a small group of specialists in the office had access to spatial data. If a field technician needed to know what cables were buried beneath a particular location, they had to call the office, wait for drawings to be sent over, and only then could they proceed.Today, they simply open a tablet—whether they're standing on the runway or anywhere else on the airport site—and have immediate access to all the information they need.The same data is available simultaneously to designers, operations controllers, field crews, and management. We call this the democratization of data. Decision-making has moved closer to the people actually doing the work, enabling them to act faster, more safely, and with greater confidence.
So the biggest benefit wasn't the new software—it was changing the way people work.
Exactly. For me, the greatest success is that many airport employees no longer even think about which software they're using. They simply see it as a natural part of their daily work. Building on that success, we developed Aerospan. We realized that virtually every airport in the world faces the same operational challenges.
So you're now taking Aerospan beyond the Czech market?
Yes. We're currently in discussions with around twenty airports across Europe. Procurement processes in this sector typically take years, but the feedback has been extremely positive. We've presented the solution at international conferences, where it has received several industry awards. Representatives from Athens International Airport even travelled to Prague to see firsthand how the system works in a live operational environment.
But if thousands of people suddenly gain access to such sensitive airport infrastructure data, doesn't that create a major security risk?
Quite the opposite. A digital twin actually enhances security. The system includes highly granular access control, ensuring that each user sees only the information necessary for their specific role. A technician working on the runway doesn't need access to terminal security systems—and vice versa. It's no different from a bank. Digitizing banking data doesn't mean every employee can view your account balance.
After five years of working with Prague Airport, do you have hard evidence of how much value the system has delivered?
We've presented the results at numerous industry forums, but in this field the benefits are difficult to quantify. How do you calculate the value of a major incident that never happened because a simulation identified the risk in advance? What is the cost of avoiding an hour-long airport shutdown? With critical infrastructure, the greatest value often lies in events that never occur thanks to better planning and decision-making. For me, the strongest proof of success is that Prague Airport continues to expand the system and that other major international operators are now pursuing the same approach.
Do you offer similar specialized products for industries beyond aviation?
Yes. One example is Perima, which automates the entire process of verifying the existence of underground utility networks. This is one of the biggest administrative bottlenecks in construction planning. By digitizing the process, it can be completed much faster while significantly reducing manual work.
Who are these construction solutions designed for? Developers?
Not just developers. Virtually every operator of large-scale infrastructure faces the same challenge: how to prepare investments more quickly while reducing administrative overhead. We work with organizations that own, operate, expand, or modernize extensive infrastructure assets. Our customers include CETIN, mobile network operators, and major energy companies. The objective is always the same: reduce bureaucracy, accelerate approval processes, and save millions by minimizing delays on construction projects.
Looking ahead, do you think every family home will eventually have its own digital twin? Could a house warn its owner about an impending boiler failure or suggest ways to reduce heating costs?
Technology is clearly moving in that direction. But the same principle applies to a family home as it does to an airport: without accurate data, it simply won't work. If the system doesn't know where the pipes are, what condition the insulation is in, or how energy actually flows through the building, AI cannot provide meaningful advice. Once it has an accurate digital model of the house, however, it can detect anomalies, recommend preventive maintenance, and predict when appliances are approaching the end of their service life.That is very much the direction smart homes will take over the coming years.
Europe has recently been discussing the growing risk of large-scale blackouts. Can digital twins help prevent them?
No software in the world can guarantee that a blackout will never happen. These events are usually the result of an extraordinary combination of unfortunate circumstances. What a digital twin can do is help us understand those risks much earlier, respond far more quickly, and minimize the consequences. It significantly reduces the time needed to identify the problem, coordinate the response, and restore normal operations. The real revolution is that data enables us to move away from simply reacting to emergencies and toward actively preventing them.
These systems must require substantial investment. Are they really worthwhile for smaller organizations?
It always depends on the scale. For a small regional water utility, we're typically talking about investments of just a few million Czech crowns. For major energy or transport operators, the investment can reach tens of millions of crowns annually. The important point is that this isn't a one-time software purchase. A digital model must be continuously maintained and updated—just as physical roads, pipelines, and other infrastructure require ongoing maintenance.
