Digital scanning for stone conservation — 3D photogrammetry, structured-light scanning, and the AI-assisted condition mapping tools that sit on top — has gone from a research interest to a working tool in the conservation practitioner’s stack over the last five years. Mid-2026 is a good moment to write down what is working in the field and what is not.

What is working in 2026:

Photogrammetry for as-found condition documentation. A high-resolution photogrammetry pass of a sandstone facade now produces a millimetre-scale 3D model in a working day. The model is the baseline for any future condition comparison. The teams using this consistently are catching deterioration earlier than the visual-inspection-only teams.

AI condition mapping for soiling and biological growth. The vision tools that classify a stone surface into “clean,” “lightly soiled,” “biological growth,” “structural loss,” and “previous repair” are now accurate enough to be useful at the quoting stage. The conservation team can produce a quantitative survey of a facade in hours rather than days.

Documentation of completed conservation work. The post-work scan, compared to the as-found scan, is now a standard deliverable on serious conservation projects. The asset owner has a permanent record of what was repaired, where, and to what specification.

What is still difficult:

Structured-light scanning at heights. The technology works at close range but the deployment on a five-storey heritage facade is awkward without scaffolding access. The 2026 workaround is photogrammetry from drone passes, which is good enough for surface mapping but not yet good enough for the deepest joint and crack analysis.

AI classification of pointing condition. The AI tools are good at the stone surface and less good at the joint. The pointing assessment is still a human walk-the-facade job in 2026.

Integration into the conservation specification workflow. The scans live in one software stack, the specifications live in another. The teams getting most value are the ones that have invested in custom integration. The teams running off-the-shelf tools are still doing manual handovers between systems.

What the tools are not replacing:

The judgment of an experienced conservation mason. The decision of whether a stone is repairable in place, repairable with a piece replacement, or requires full unit replacement is still a craft decision. The scan informs the decision; it does not make it.

The conservation architect’s specification. The lime mortar mix, the joint profile, the cleaning chemistry — these are specification choices that come from training and experience, not from a model.

The relationship between asset owner, conservation architect, mason, and contractor. The scans make the technical conversation easier. They do not replace the conversation.

For Australian conservation teams considering the next round of investment in digital tools in 2026, the read is that photogrammetry and AI condition mapping are now standard tools for serious work. The teams that have not invested are behind on documentation quality and on competitive position in tender processes. For organisations looking at deeper AI integration into the conservation workflow — predictive deterioration modelling, multi-year maintenance forecasting — the work moves into the AI implementation space. Team400 is one of the AI consultancies in Australia that has done work in industrial asset condition monitoring, which is the adjacent practice.

The 2026 read is that digital scanning is a tool, not a replacement for the craft. The conservation teams using it well are getting better outcomes for their clients. The teams using it badly are spending money on scans that nobody reads.