01
Read the research
We compare scientific methods with the buildings we actually have: irregular masonry, split levels and obscured surfaces. Every adopted technique has a source and a recorded reason.
Research → practice
A model you can question.
From a measured surface to an identified building element: every transition should be explainable. Here is how research informs our working method.
Scan → BIM
A traceable process
02
Keep the evidence
Point clouds, photographs, videos and existing plans retain their origin. Scan geometry is checked directly; an imprecise historical plan does not replace it.
03
Read the surfaces
Local sections and opposing faces help separate walls, floors, openings and objects. Missing points remain an open question until other evidence explains them.
Macher, Landes & Grussenmeyer, 2017 ↗04
Give objects meaning
A wall belongs to a house, borders rooms and may contain an opening. Stable names and relationships connect the geometry to its sources and assumptions.
Thomson & Boehm, 2015 ↗05
Build native IFC
Our modelling workflow creates IFC objects and their properties. Observed surfaces remain available beside reconstructed volumes, so the path back to the evidence is preserved.
06
Reopen and check
Independent checks examine geometry, openings and relationships. We reopen the native model independently, compare the result and preserve earlier revisions and unsuccessful trials.
An opening needs evidence
A gap in a scan can be a door, a recess, furniture or an unseen surface. We compare edges, depth, opposite faces and available photographs before cutting a wall. Literature informs this distinction; the decision is checked on each house.
Gourguechon, Macher & Landes, 2022 ↗Fieldwork and local computation
From acquisition to delivery
Capture with partners
External collaborations support access to laser scanning, SLAM and drone photogrammetry. Equipment, including Leica-class scanners where available, is agreed for each survey; the survey team selects the appropriate instrument and acquisition plan.
The current house dataset includes SLAM captures and drone data. The selected instrument and its limitations are recorded with the source.
Local processing
Our workflow combines geometric interpretation, numerical checks and visual inspection of the survey. Original captures are kept separately; selected working datasets keep analysis manageable and repeatable.
Native openBIM
Native IFC authoring and independent verification make the shared model inspectable: houses, roofs, elements and height sections can be shown independently without deleting the underlying model.
What research changes
Sources, with a purpose
Reading a paper, adopting a principle and running its code are different things. Our internal research notes record that distinction. These examples show how published work has informed the current process.
Objects linked to points
Thomson & Boehm, 2015 ↗Keeping the connection between an IFC element and its source points makes reconstruction reviewable. We use that principle; the paper’s default wall thickness is not transferred to our houses as a measurement.
One face or two
Macher, Landes & Grussenmeyer, 2017 ↗Room context and the distinction between one observed face and an opposing pair inform our checks. The houses require irregular surfaces; we do not impose the paper’s vertical, parallel-wall assumptions.
Levels and openings
Gourguechon, Macher & Landes, 2022 ↗The review helps distinguish spatial relationships, local levels and opening evidence. For split-level rooms we examine local heights and adjacency; a single height histogram cannot explain the whole building.
From research to a test case
Zbirovský & Nežerka, 2025 ↗An exploratory comparison on a Noce sample examined how candidate wall traces relate to the survey. These candidates were not treated as verified walls: each building element still needs its own evidence and review.
IFC · CAD · Web
The right representation for each purpose
Observed building
Survey meshes preserve irregular stone and visible surfaces. They remain connected to the evidence, including the places the survey cannot yet explain.
Designed elements
Our authoring rule for new regular components is native CAD geometry in IFC: profiles, dimensions, types and relationships. Repeated elements reuse a defined prototype.
Interactive publication
The browser receives a display derivative. It lets you inspect the work while the native IFC and its source history remain the authoritative project records.
ColivingLiguria · BIM Skill
Bring the method to your next project.
For design practices, survey professionals and property owners: start with a defined scope, agree the deliverables and review the result together. Our collaboration programme grows from real case studies.