What we can save: additive manufacturing and the future of architectural heritage

The Grande Poste d'Alger was completed in 1910, designed by Jules Voinot and Marius Toudoire in the Moorish-revival style that French colonial architecture occasionally produced at its most generous — a style that attempted, however imperfectly, to synthesize European structural rationalism with North African ornamental tradition. The building is extraordinary: the central hall with its horseshoe arches, the geometric tilework that covers the lower walls, the delicate foliated capitals that sit atop the interior columns, the exterior cartouches that frame the windows with Arabic calligraphic motifs rendered in stucco. It is also slowly disappearing. The stucco ornament, applied over a lime mortar base that is no longer maintained with the skills that once existed in abundance, is fragmenting. Individual cartouches are gone. The tile patterns are interrupted by repairs that do not match. The building is still magnificent. It is less magnificent than it was, and it will be less magnificent still in ten years if nothing changes.

My thesis at EPAU asked a question that seemed simple but turned out to be deep: what if we could print the ornament that is falling off? Not as a theoretical exercise, but as a practical intervention — using photogrammetric scanning to capture the surviving ornament in three-dimensional detail, processing that scan into a printable model, and using concrete extrusion or polymer printing to produce replacement elements that match the originals to sub-millimeter tolerance. The technology is available. The workflows are established. The cost is a fraction of what traditional stonecutting or stucco casting would require. The question is not whether the technology works. The question is whether we are allowed to use it, and whether we want to.

The ethics of reproduction

Heritage conservation has a complicated relationship with reproduction. The Venice Charter of 1964 — the foundational document of modern conservation practice — is explicit that distinguishability is required: any addition to a historic structure must be identifiable as an addition, readable by a trained eye as not-original. This principle was developed in response to the heavy-handed "completion" projects of the 19th century, where Viollet-le-Duc and his peers essentially invented missing medieval elements from imagination and installed them as if authentic. The fear is reasonable: a reproduction that looks too good is a lie told in stone. But the Venice Charter was written before high-resolution photogrammetry and before concrete extrusion. A reproduction printed from a scan of the surviving original is not invented. It is not imagined. It is copied, at precision levels that medieval craftsmen could not have achieved and that modern observers cannot distinguish from original without chemical analysis. The ethical category is different from Viollet-le-Duc's fantasies.

The Ottoman tilework tradition that once covered the interiors of Algerian mosques and palaces — and that survives in beautiful fragments in the Musee des Arts et Traditions Populaires and in the interiors of a handful of preserved medina buildings — presents a different challenge. The tile patterns are geometric, governed by the mathematics of Islamic ornament: repeating units based on six-fold or eight-fold symmetry, which can be parameterized perfectly and reproduced without any original surviving element, purely from the underlying geometric logic. Here the printing technology enables something beyond repair: it enables reconstruction from mathematical first principles, producing geometrically authentic ornament even where no surviving original exists. Roman mosaic floors that have been damaged by centuries of foot traffic and settlement present a similar case — the pattern can be reconstructed from documentation, the color palette matched to surviving fragments, and the missing sections printed in matched stone or polymer and inserted with documented joints that identify them as interventions.

Three categories of heritage elements that additive manufacturing can save

The argument I am making is not that additive manufacturing makes conservation easy, or that it resolves the philosophical tensions at the heart of the discipline. It does not. What it does is change the economic calculus in a way that makes comprehensive rather than selective intervention possible. The Grande Poste currently receives the level of ornamental restoration that its budget permits — which is partial, inconsistent, and insufficient. If the cost of reproducing a damaged cartouche drops from $4,000 to $400, the scope of what is economically viable changes entirely. The entire ornamental program could be completed, consistently, at the cost that currently allows only a few isolated elements to be addressed. The building would be more fully itself. That matters — not just aesthetically, but as a civic and cultural fact. The Grande Poste is one of the most visible buildings in Algeria. Its condition is a statement about what Algeria values. We should value it more.