What does it take to make a building “smart”? For Stathis Stamatopoulos, researcher at the Institute of Communication and Computer Systems (ICCS/NTUA), it comes down to turning data into decisions; forecasts that lower energy bills, simulations that improve comfort, and tools that help design better buildings before construction even begins.
In this interview, Stathis tells us about his work leading the development of AI-based operation tools and an early design tool for DTERBIM, and why he believes digital twins should one day be as common as central heating.
Could you briefly introduce yourself? What is your background, and what kind of work do you focus on currently?
I am Stathis Stamatopoulos, researcher at the Institute of Communication and Computer Systems (ICCS), the research arm of the School of Electrical and Computer Engineers of the National Technical University of Athens.
I hold an Engineer Master’s degree in Electrical and Computer Engineering, and I am currently a doctoral candidate in decision support systems utilising machine learning techniques in buildings. Today, I work mainly on Horizon Europe projects, managing technical project teams but also supporting research and technical developments.
In DTERBIM, I lead Work Package 5 (WP5), “Digital tools for optimal Operation & Maintenance and Deconstruction.” WP5 is, in a sense, where DTERBIM meets the real world: it covers everything that happens after a building is designed and built.
On one side, we develop the tools that keep a building running well day to day, while on the other, WP5 also looks at the end of a building’s life. ICCS leads WP5 as a whole and is directly responsible for the AI-based energy profiling and the indoor environmental quality simulation and optimisation.
What do you enjoy most about your work? What keeps you motivated?
I enjoy the moment when something abstract (a model, a forecast, an algorithm) becomes a decision that improves how a real building runs. Energy bills drop, comfort improves, and a maintenance issue is caught before it becomes a failure. That tangibility is rare in research, and it keeps the work grounded. I’m also motivated by the dynamic environment of European research and innovation: every year brings new challenges to tackle, new partners to work with, and new ideas to test, and that constant movement is what keeps the work exciting.
“I enjoy the moment when something abstract (a model, a forecast, an algorithm) becomes a decision that improves how a real building runs.”
Now let’s zoom in on your role within DTERBIM. What is your team responsible for, and what challenge are you addressing?
In WP5, the ICCS team directly leads two of the most data- and AI-intensive tasks. Our team is building the scalable workflows and AI models that turn raw energy data into operational intelligence.
We work with deep learning architectures alongside gradient-boosting trees to produce high-resolution energy profiles and forecasts at device, room, and whole-building levels.
We also address the fact that “comfort” is multidimensional: it includes thermal conditions, as well as acoustics, lighting, and visual comfort, and these dimensions interact. We are building a simulation and optimisation framework that fuses static BIM data with real-time sensor streams and evaluates operational plans against established references.
We are also developing dynamic thermal simulation models that predict indoor temperature and humidity, and predictive algorithms that model occupant behaviour and interactions with the spaces they use. In collaboration with the IEQ optimisation service, our aim is to achieve adaptive comfort: HVAC (Heating, Ventilation, and Air Conditioning) setpoints that adjust automatically to actual preferences and usage patterns rather than a fixed schedule.
Beyond operational methodologies and tools, ICCS also leads the development of the “Early design tool supporting efficiency, circularity and inclusiveness”, which sits at the opposite end of the building lifecycle. ICCS develops a decision-support tool that helps designers, owners, and investors compare design and renovation scenarios on energy, cost, environmental, sustainability, circularity, and accessibility indicators before a building is built or refurbished.
“We are developing a decision-support tool that helps designers, owners, and investors compare design and renovation scenarios on energy, cost, environmental, sustainability, circularity, and accessibility indicators before a building is built or refurbished.”
What has your team achieved so far, and what are your main goals for the next phase of the project?
In this first year, our most visible progress as ICCS has been on the early design tool. It is one of the first tasks to deliver tangible technical output in DTERBIM. It started early in Month 2 and, as a result, has already moved beyond specifications into a working interface and a first set of functionalities.
We have designed the user interface, implemented the first functionalities that allow designers, owners, and investors to compose and compare early designs, and put those early versions in front of relevant stakeholders to gather structured feedback.
For the next phase, our priorities are clear. First, we will deliver the early design tool, refined based on stakeholder feedback and ready to support the design-stage release of the DTERBIM Toolkit. Second, we push our other activities from their methodological foundations into running implementations.
In your view, what is the most significant impact DTERBIM could have on the construction sector? What changes would you like to see as a result of your work on the project?
From the work that ICCS leads in DTERBIM, two concrete shifts matter most to me. First, better decisions earlier: design tools that put energy, cost, environmental, circularity, and accessibility indicators in front of designers and owners when changing course is still cheap. Second, buildings that are operated proactively rather than reactively: AI services that anticipate inefficiencies, comfort issues, or failures before they happen, and that close the long-standing trade-off between energy performance and occupant comfort.
The change I want to see is cultural: digital twins moving from a technology buzzword to a piece of everyday infrastructure that owners, facility managers, ESCOs (Energy Service Companies), designers, and even occupants rely on without thinking about it.
If five years from now, a mid-sized building owner in Europe expects to operate their building with a digital twin the same way they expect to have heating, then DTERBIM will have made a real difference.
“If five years from now a mid-sized building owner in Europe expects to operate their building with a digital twin the same way they expect to have heating, then DTERBIM will have made a real difference.”
This is the second in a series of interviews with DTERBIM partners, highlighting the people, expertise, and innovations driving the project forward.
Want to meet more of the experts behind DTERBIM? Follow us on LinkedIn, Bluesky, X, and YouTube, and subscribe to our newsletter for future interviews and project updates.