Researchers have developed a deterministic IFC-based workflow that evaluates whole-building material choices for both embodied carbon and circularity. In a reference building, 150 wall-floor-roof configurations were reduced to 12 non-dominated candidate scenarios, with stable results across weighting, threshold and uncertainty checks.
Choosing a building material is rarely a one-dimensional decision. A material with a low carbon footprint may be difficult to recover at end of life, while a highly recyclable material can carry a much larger manufacturing footprint. Designers therefore need ways to compare these trade-offs early, before specifications become difficult to change.
A study published in Smart Construction presents a reproducible workflow that uses open Industry Foundation Classes (IFC) data to connect building information with material carbon and circularity assessment. Rather than relying on a proprietary BIM application, the prototype reads the exchanged IFC file directly, extracts component quantities and materials, checks data quality, maps the information to a prototype material database, and evaluates whole-building material combinations.
The researchers tested the workflow on a buildingSMART reference IFC model. Eight target elements were extracted; seven contained sufficient material and quantity information to continue into the calculation, while one incomplete roof aggregate was excluded with the reason recorded. The retained set consisted of four walls, one floor slab and two roof slabs.
The material decision space was deliberately small enough to evaluate exactly. Six wall materials, five floor materials and five roof materials created 150 whole-building configurations. Every configuration was scored for cradle-to-gate embodied carbon and a prototype Building Circularity Score, and all 150 were compared to identify Pareto-optimal solutions - configurations for which neither objective could be improved without worsening the other.
Twelve configurations were non-dominated. Six were low-carbon-oriented and were dominated by timber-roof systems. The other six were circularity-oriented and used aluminium roofs. No configuration met the study's balanced classification thresholds. The contrast illustrates a practical feature of multi-objective material choice: the design space may contain clear clusters rather than a single obvious "best" answer.
The roof choice had a particularly strong effect in this reference case because the two roof slabs accounted for 53.5% of the extracted material volume. For one paired comparison, replacing a timber roof with an aluminium roof raised embodied carbon from 591.8 to 21,706.5 kgCO2e while increasing the circularity score from 0.712 to 0.865. The study treats this as a case-specific result rather than a general rule for buildings.
The workflow also adds a controlled semantic query layer. Twelve predefined query types can retrieve low-carbon or high-circularity options, identify component-level material drivers, compare Pareto candidates, or explain a selected option. Because these queries operate through fixed rules, the same data and query produce the same output. The system can also package the results into structured prompt text for optional downstream interpretation, but the study did not call or evaluate a large language model.
Several checks were used to test whether the reference-case results were sensitive to modelling choices. The same 12 candidate scenarios were retained under five different weighting schemes for the circularity score. The classification remained 0 balanced, 6 low-carbon-oriented and 6 circularity-oriented across nine threshold settings. In 1,000 scenario-based Monte Carlo runs, every baseline candidate remained non-dominated in at least 83.2% of the runs under the stated uncertainty assumptions.
The full 150-configuration case took about 0.79 seconds, including about 0.023 seconds for pairwise Pareto filtering. Runtime tests with larger synthetic sets showed that exhaustive pairwise filtering does not scale indefinitely, so more efficient Pareto-search strategies would be needed for much larger design spaces.
The authors emphasise that the work is a proof of concept. It uses one reference IFC model and a prototype material database, and it does not yet include structural performance, fire resistance, moisture, cost, procurement, regulation, operational energy, or product-specific verified EPD coverage. Testing across multiple authoring platforms and real projects will be required before field deployment can be assessed.
The study nevertheless shows how open BIM data can be used as more than a visual or documentation layer. By keeping component quantities, material information, carbon calculations, circularity scores and candidate explanations connected in one deterministic process, the approach provides a traceable basis for exploring early-stage material trade-offs in low-carbon and circular construction.
The work was carried out by researchers from Loughborough University, the University of Sheffield, Newcastle University, and Beijing University of Technology.
Publication details
This paper was published in Smart Construction:
Meng Y, Sun Y, Huang H, Zhang C. Industry Foundation Classes (IFC)-enabled whole-building semantic optimisation and query-driven decision-support workflow for low-carbon and circular smart construction. Smart Constr. 2026(3):0019. https://doi.org/10.55092/sc20260019
DOI: 10.55092/sc20260019
