Digitizing a Hand-Sculpted Craft
For a luxury residential project in New York City, Kemperle Industries provided digital fabrication support for a series of custom architectural ornamentation elements, including corbels, frieze, and repeat molding profiles designed for the space. The work paired hand-sculpted reference pieces with a digital production pipeline built around 3D scanning, CAD modeling, and 3D printed master patterns, giving the design team a way to move from concept to mold-ready geometry without losing the character of the original sculptural work. The goal throughout was to keep the finished plasterwork indistinguishable from a fully hand-sculpted piece, while gaining the accuracy and repeatability that only a digital production step can provide.
Rather than treating fabrication as a separate, later step, the project connected every stage of development from the first reference through final cast. That alignment made the resulting molds more accurate, the production timeline more predictable, and the finished plaster elements ready for installation with minimal on-site adjustment.
Residential ornamental plaster carries its own set of constraints. Unlike a public or institutional installation, there is little room to revisit a piece once it is set in place, and every element has to read correctly at close range in a finished living space. Building the digital pipeline in from the start gave the design and production teams a way to confirm geometry, proportion, and repeat-pattern accuracy before committing to a mold, rather than discovering a mismatch after casting.
The project began with hand-sculpted reference pieces and supporting architectural drawings that captured the intended ornamental detail down to individual repeat motifs. These references set the geometric and stylistic baseline for everything that followed.
Kemperle Industries scanned the reference elements to capture precise surface geometry, including the fine undercuts and relief detail that define ornamental plasterwork.
Scan data was refined into clean CAD geometry built for production, correcting for mold draft angles, wall-thickness allowances, and panel breaks while preserving the original ornamental detail. Some pieces were designed entirely in CAD.
The finalized CAD models were 3D printed as master patterns, giving the mold-making process a precise, dimensionally stable physical reference.
Molds were produced directly from the 3D printed masters, and final plaster elements were cast for production and installation.
Ornamental plaster work is traditionally a hand process, which makes it difficult to reproduce complex detail consistently across multiple elements or coordinate tightly with a construction schedule. This project needed a way to preserve the character of hand-sculpted reference pieces while making the geometry repeatable enough for confident mold-making, predictable lead times, and an installation that would fit correctly on the first attempt in a finished residential interior. There was little margin for the kind of trial-and-error adjustment that a hand-only workflow can require, especially with a full team of trades already scheduled around a fixed install date.
The individual elements also had to work as a set. Corbels needed to read as load-bearing and consistent from piece to piece, and the frieze had to repeat cleanly along its run without a visible seam or drift in the pattern. Achieving that kind of consistency by eye, across every hand-finished piece, is difficult even for an experienced plasterer, and any small variation becomes obvious once the elements are installed side by side in a finished room.
Kemperle Industries combined hand-sculpted craftsmanship with a digital production pipeline built for accuracy and repeatability.
Reference elements were 3D scanned to capture their exact surface geometry, then modeled in CAD to prepare that geometry for mold-making, correcting draft angles and panel breaks without losing the sculpted detail that made the design worth reproducing in the first place. In some cases, elements were designed directly in CAD based on sketches. From there, the finalized models were 3D printed as master patterns, giving the mold-making stage a precise, dimensionally consistent starting point rather than a one-off hand original. Molds were built from those printed masters, and the final plaster elements were cast for production and installation.
Since every stage fed directly into the next, changes could be made early and cost-effectively in CAD rather than late and more expensively in cast plaster, and the finished elements matched the original design intent with far less rework than a fully manual process would allow. The digital record also meant that if additional elements were needed later, or if a piece was damaged during installation, the same master patterns and mold data could be used again without re-sculpting from scratch.
Before committing to a full production run, a single test piece was produced from the 3D printed master and checked against the original reference for fit, proportion, and surface quality. That step caught any adjustments early. That same test-and-verify approach carried through mold making and casting, with each cast checked against the source geometry before completion, so any variation was caught before handoff rather than discovered after installation.
The finished ornamental plaster elements matched the original design intent with a level of geometric consistency that would be difficult to achieve through hand-sculpting alone. Moving the workflow through scanning, CAD, and 3D printed master patterns gave the mold-making and casting stages a reliable, repeatable foundation, which kept the production timeline on track and reduced on-site adjustment during installation. Repeat elements, like matching corbels or a run of identical molding profile, came out of the mold consistent with one another. Just as importantly, the approach gave the design team confidence in the geometry before any material was committed to a mold, removing much of the risk that comes with translating a hand-sculpted concept into a finished architectural element, and leaving a reusable digital record the project can draw on if future phases call for it. For a residential installation where every surface is on close, permanent display, that combination of craftsmanship and repeatable accuracy is the difference between a good result and one that will hold up to scrutiny at arm’s length.
The finished result also demonstrates a capability that carries forward. Because the corbel, frieze, and molding geometry now exist as reviewed CAD files rather than as one-off hand-sculpted originals, they can be revisited, adapted for a different profile, or extended into an adjoining room without starting the sculpting process over. For a residential client considering additional custom plasterwork elsewhere in the home, that groundwork already exists.
