Gustav There is a bronze statue in a small town square somewhere in Scandinavia. It has stood there for over a hundred years. The detail in the casting - the fabric folds, the facial features, the texture of the base - was sharp when it was made. Today, decades of acid rain, frost cycling, and bird activity have softened everything. The sculptor's intent is still visible, but the precision is gone.
Nobody scanned it ten years ago, when it was in better condition. Nobody has scanned it today. At the rate it is going, the detail that survives now will be partially gone in another decade, and nobody will have a record of what was there.
This is not a dramatic or unusual story. It is the ordinary, slow loss happening to thousands of objects across the world, in parallel, right now.
The scale of the problem
Heritage loss is not primarily a story about dramatic disasters - fires, floods, bombs, deliberate destruction - although those happen too. It is mostly a story about attrition. The gradual softening of carved detail. The slow flaking of pigment from frescoes. The wood that dries out and cracks in a storage room with inadequate climate control. The stone figure whose face has been smoothed by weathering to the point where the expression is ambiguous where it was once clear.
Each of these changes is small. Collectively, across thousands of objects over decades, the loss is enormous. And it is mostly silent - nobody announces it, nobody writes the press release, it simply happens while everyone is dealing with other things.
Why photography is not enough
There is more photography of historical objects today than at any point in history. Every museum has a photographic archive. Every tourist with a phone photographs the things they visit. This is genuinely valuable - but it does not solve the problem that photogrammetry and 3D scanning solve.
A photograph records one viewpoint, at one moment, under one set of lighting conditions. It captures appearance but not geometry. You cannot measure a photograph. You cannot rotate it to examine an underside that was out of frame. You cannot use it as a template for restoration because it cannot tell you the exact curve of a surface or the depth of a carved line.
A 3D scan is a geometric record. It captures shape with a precision that no photograph can match - accurate to fractions of a millimetre in ideal conditions. It can be measured. It can be reproduced. It can be compared to future scans of the same object to track change over time. It is the closest thing we have to making a physical object permanent.
A photograph tells you what an object looked like. A 3D scan tells you what it was. The difference matters most when the object is no longer there to examine.
The tools that make this achievable
Photogrammetry - the process of reconstructing 3D geometry from overlapping photographs - has become significantly more accessible over the past decade. Software like Agisoft Metashape and RealityCapture can process several hundred photographs into a detailed, textured 3D model in a matter of hours on a capable workstation. The camera required is a high-quality DSLR, not a specialised scientific instrument.
This matters for heritage applications because it means the cost of a scan is no longer prohibitive. The equipment exists. The software exists. The skill required to produce a useful result is learnable. What has been missing, in many cases, is the awareness that this is possible and the initiative to do it before the window closes.
The window closes in different ways for different objects. For an outdoor sculpture in a coastal climate, the window might be measured in decades. For a fragile organic object in deteriorating storage conditions, it might be years. For something in an active conflict zone, it might be months or less.
What a digital scan actually preserves
A well-executed photogrammetry scan preserves three things: the geometry of an object, the surface texture and colour at the time of scanning, and the spatial relationships between different parts. That is enough to do most of what matters in preservation work.
Researchers can study the object without handling it - reducing wear from examination. Educators can create interactive versions that students anywhere in the world can explore. Exhibition designers can create virtual displays of objects that are too fragile to show physically. Restoration teams have a reference point if damage occurs. Future generations have something to study even if the original is gone.
The data can also feed into immersive experiences and VR exhibitions - making heritage objects accessible to people who could never travel to see them in person. A scan of a Scandinavian runestone that lives in a regional museum in Denmark can become part of an exhibition experienced by someone in Tokyo or São Paulo. That is a different kind of preservation: not just of the object, but of its ability to reach people.
The argument for doing this proactively
The instinct in heritage institutions is often to prioritise the objects that are already in crisis - things that need immediate intervention, objects with active damage, pieces that are about to be moved or loaned. This makes sense as triage but not as strategy.
The most valuable scans are the ones made before anything goes wrong. A scan of an outdoor sculpture in good condition today is a better record than a scan made after five more years of acid rain weathering. A scan of a collection before a museum undergoes renovation is a better safety net than a scan attempted in the middle of it.
The cost of a proactive scan is small. The cost of not having scanned something when the object is later damaged, lost, or destroyed is permanent. That asymmetry is the clearest argument for treating digital preservation as a standard part of collection stewardship rather than a last resort.
What this looks like in practice
My own photogrammetry preservation work has focused on sculptures and statues - objects with complex geometry, surface detail worth capturing at full resolution, and exposure conditions that mean the current state will not last indefinitely. The workflow involves a systematic capture session with a high-resolution DSLR, processing in RealityCapture and Agisoft Metashape, and delivery of a clean, textured 3D model in whatever format the client needs.
The result is a file. But what the file represents is something that will outlast the object it was made from - a record that does not age, does not weather, and does not depend on any single institution or storage location to survive.
That seems like a worthwhile thing to make.