Buildings are the primary venues for human activity, and precise numerical models can support many applications. At regional and national scales, however, different formalisms are used to represent building components, and building land use is frequently oversimplified—even when mixed-use buildings are considered. In the study, building land use is defined as an inventory of the distinct functional spaces hosted by a building. Manual measurement can produce highly realistic interior models, but it is labour-intensive and must be conducted separately for each case.

Colomb and Perret’s workflow integrates three complementary tasks. Use-oriented building object modelling creates building objects that capture land use. Activity census and conflation then aggregates and reconciles activity records drawn from datasets with different structures and levels of detail. Finally, logical attribute generation derives detailed attributes for housing units and activities in relation to the characteristics of their host building and other environmental factors. Together, these stages connect spatial representation, heterogeneous activity data and context-sensitive attributes within a fully automated pipeline.

The system is designed to accommodate varying levels of data availability and attribute richness. It also supports “recent-period historicisation,” defined in the study as reconstructing past states of the built environment. Automation allows the workflow to be deployed at large scale, and the researchers demonstrated its application across an entire country. The supplied study abstract does not identify the country, the number of buildings modelled or the types of buildings included.

The authors evaluated the method through multiple validation procedures, including comparisons with ground-truth measurements. Those comparisons indicated that the workflow reliably reproduced building land use. The reported results support its country-scale application and its capacity to combine substantial datasets with different levels of detail, but the abstract provides no numerical error rates, processing times, comparisons with named alternative models or performance results for individual building types. The demonstrated method is therefore an automated process for modelling fine-grained building use by linking use-oriented objects, reconciled activity records and attributes derived from building and environmental context.

Subject of Research: Automated country-scale modelling of building land use from heterogeneous datasets

Article Title: Systematic modelling of a large-scale, historicised, well-detailed building land use, with the fusion of massive and heterogeneous datasets

Article References: Colomb, M., & Perret, J. (2026). Systematic modelling of a large-scale, historicised, well-detailed building land use, with the fusion of massive and heterogeneous datasets. PLOS One, 21(10), e0359346. https://doi.org/10.1371/journal.pone.0359346

Image Credits: AI Generated

DOI: 10.1371/journal.pone.0359346

Keywords: building land use, automated modelling, country-scale mapping, mixed-use buildings, urban analytics, heterogeneous datasets, historicisation, housing units, ground-truth validation, Systematic, modelling, large-scale