Joint Paper Published on Scan-to-IFC Workflows for BEM-Ready Building Models

A joint research paper titled “A Semantic Scan-to-IFC Pipeline for Automated Generation of BEM-Ready Building Models from Mobile Indoor Scanning Data” has been published.

The paper was authored by Federico Rossi, Hanwen Hu, Karsten Menzel, and Carlo Zanchetta, with contributions from the University of Padua and the Institute of Construction Informatics, Faculty of Civil Engineering, TU Dresden.

The study addresses a key challenge in building energy modelling: existing buildings often lack reliable as-built BIM models, while conventional BIM-to-BEM workflows require structured geometry, spaces, openings, and semantic relationships. The paper proposes a semantic Scan-to-IFC workflow that converts mobile indoor scanning data into simplified, BEM-ready IFC4x3 models.

Our institute’s contribution focused on the RoomPlan-to-IFC reconstruction pipeline. This included the development of a Python-based exporter that converts Apple RoomPlan JSON data into IFC4x3, supports multi-scan sessions through a scan manifest, reconstructs walls, slabs, windows, doors, and openings, and adds BEM-oriented completion layers such as inferred ceilings, analytical spaces, closure walls, missing-wall completion, and validation routines.

The workflow is designed for partial indoor scans of existing buildings, where the scanned data may be incomplete, noisy, or segmented across several scan sessions. By preserving scan metadata and clearly marking inferred elements, the approach keeps the generated IFC model transparent and suitable for downstream building energy modelling preprocessing.

Publication: Rossi, F.; Hu, H.; Menzel, K.; Zanchetta, C. A Semantic Scan-to-IFC Pipeline for Automated Generation of BEM-Ready Building Models from Mobile Indoor Scanning Data. Buildings, 2026.