A small team at the EU's Joint Research Centre is building the most detailed map of Europe's buildings ever assembled. Their open source software project, the Digital Building Stock Model (DBSM), matters for energy, housing, and the climate.
It’s clear to all governments in the European Union: renovation of the building stock is an essential frontier for energy and climate action. Buildings are Europe’s largest energy consumer: 85% of buildings in the EU have been built before 2000 and 75% have poor energy performance.
Yet policymakers driving legislation like the Energy Performance of Buildings Directive and the Renovation Wave have long been forced to make decisions without granular data. At EU scale, nobody can yet say with precision where the buildings are, how old they are, or how much energy they consume. Statistics exist at the level of countries or large regions. Below that, the picture quickly fades.
The member states often lack complete records, or fail to make all records publicly available. Some countries have detailed cadastral databases; others do not yet share any building footprint data, or not in a usable form. The data formats vary so much between countries that combining them is a heavy task. “The map is still patchy,” says Ana Martínez, the DBSM team lead at the Joint Research Centre (JRC) in Ispra in Italy. "Especially for building attributes such as age of construction, or energy demand."
"We can’t rely only on authoritative data sets,” she says. “It will 'take forever' to combine, and gaps remain likely.” Instead, Ms Martínez and her colleagues link all that is available and then use machine learning and satellite data to fill the gaps. Their model, a consistent, building-by-building dataset covering all of Europe, will let us finally understand the challenge.
Building the puzzle
Launched in 2023, the DBSM project is the work of a small team of part-time researchers and developers. Rather than patching national datasets together region by region, the team examines each individual building and selects its data from whichever source ranks highest in a defined hierarchy. (This approach is called conflation.) Authoritative national sources come first, then OpenStreetMap, then AI-generated building footprints derived from satellite imagery. A building’s geometry might come from a national cadastre, while its height or roof type is sourced elsewhere. Each element is selected, validated and assembled into a single, consistent record.
The model goes well beyond location. It includes building footprints, heights, construction period, function, and compactness - an indicator of energy efficiency. “Complete coverage is possible thanks to satellite-based raster maps provided by the Global Human Settlement Layer”, says Ms Martínez.
Assembling this data is anything but linear. “It’s a dance: if something goes wrong, we have to take two steps back to recalculate,” says developer Olympia Gounari, describing the constant iteration required to process and align the datasets.
The team publishes the software on code.europa.eu, the open source software repository for all European Union institutions. Each release comes with peer-reviewed papers and reports. The logic is simple: public money funded the work, and sharing the results is the natural extension of that. “Releasing incrementally, rather than waiting for a perfect product, means other researchers can already build on it and point out what needs fixing,” says Ms Martínez.
The missing piece
One key dataset remains difficult to access: energy performance certificates. These certificates, familiar to anyone who has bought or rented a home in Europe, contain detailed information about insulation, heating systems and renovations. They are critical for understanding how buildings actually perform. Yet across Europe, access to these registries is uneven.
“Only a handful of Member States make these registries openly available,” Ms Martínez notes.
Combined, even without detailed energy audits, the attributes start to tell a meaningful story. The DBSM is also proving useful beyond its original scope. Researchers are using it to estimate rooftop solar potential (at EU scale), to explore links between building conditions and socioeconomic vulnerability, and to estimate the risks of floods and other disasters.
“What the team has built is, at its core, an aggregator: a project that finds useful data wherever it exists and combines it in a principled way,” concludes Ms Martínez. “We are making it public as open source so anyone can use and improve it.”
More information:
Digital Building Stock Model on code.europa.eu
DBSM documentation
Energy Performance of Buildings Directive
Renovation Wave
Identification of essential variables for the estimation of energy demand in buildings at European scale
Mapping Europe’s rooftop photovoltaic potential with a building-level database