BIM Coordination for Stone Packages: A Fabricator's Guide

Building information modelling has moved from a novelty on marquee projects to a routine requirement on mid-size commercial work, and stone fabricators are increasingly asked to participate in a process that was not designed with them in mind. The typical request arrives as a line in a specification: provide the stone package modelled to a stated level of development, coordinated in the project model, with clash detection complete before fabrication. For a shop whose digital workflow runs from a laser template straight into a computer-aided machining program, that requirement can look either trivial or impossible depending on how it is interpreted.

The good news is that stone fabricators already work in three dimensions with tighter tolerances than most trades in the model. The difficulty is that the model and the shop use different software, different coordinate systems and different definitions of what a finished element means. Understanding the vocabulary of levels of development, and where the stone scope genuinely benefits from coordination, turns the requirement from a compliance burden into a tool that prevents the site problems fabricators complain about most.

Levels of Development and What They Ask For

The framework used across the industry defines six levels describing how completely and reliably a model element has been developed. The base definitions come from the American Institute of Architects, with one intermediate level added later by an industry working group specifically to support trade coordination. The levels are not a measure of graphical detail; they are a statement about how much the project team may rely on the element for design, coordination, estimating, fabrication and operations.

At the first level an element may be no more than a symbol, a mass or a space reservation, and any quantity, size or location taken from it should be treated as approximate. The second level represents approximate geometry and design intent. The third provides accurate size, location and geometry suitable for detailed design. The intermediate level adds the connections and interfaces needed for clash detection between disciplines, sitting above the third level in demand but below the fourth.

The fourth level is the one that matters most to a fabricator, because it supports fabrication and installation and carries manufacturer-specific data. A stone element modelled at that level should be dimensionally correct as it will be produced, with its edge profile, thickness, joint positions and anchor or support features represented. The final level describes the element as actually constructed, verified in the field, and exists to support operations and maintenance rather than production.

Reading a specification correctly therefore means finding which level is required and at what stage. A requirement for the third level during design development and the fourth before fabrication is a reasonable and common structure. A requirement for the fourth level at bid stage is not reasonable, because the shop has not templated anything yet, and it is worth raising as a question rather than pricing an impossibility.

Practical Coordination for a Stone Package

Agree the coordinate system first

Nothing wastes more time than models that do not align. Establish the project base point, the survey point and the units before exchanging anything, and confirm that the shop's export lands in the right place when the project team opens it. A stone package that appears half a kilometre from the building because of a shared coordinates mismatch is a fifteen-minute fix that regularly consumes a fortnight of email.

Model what matters, not everything

A model element that represents a countertop as a correctly dimensioned solid with the right thickness, edge profile and cutout positions is enormously useful. A model that also represents every polishing step and every adhesive joint is noise. Decide with the coordinating party which attributes carry information the project actually uses, typically material, thickness, finish, weight, support requirement and fire or slip rating, and populate those consistently.

Use clash detection where stone genuinely clashes

Stone conflicts with a predictable set of things: plumbing and waste under sinks, electrical outlets and data boxes in splashes and islands, structural supports under long overhangs, mechanical services above wall cladding, and door swings against projecting counters. Running clash detection against those systems catches the problems that otherwise appear on install day. Running it against every element in the model produces thousands of trivial hits that nobody reviews.

Keep the fabrication model separate

The coordination model and the machining program are different artefacts with different purposes. Trying to drive the saw and the machining centre directly from a design model usually fails, because design models carry approximations that fabrication cannot tolerate. The workable pattern is to coordinate in the project model, template on site, then produce the fabrication geometry from the template and reconcile any differences back into the model as a revision.

Scanning, Verification and Closing the Loop

Three-dimensional scanning has become affordable enough that verifying a site against the model before fabrication is now realistic on commercial work. A scan produces a point cloud that can be overlaid on the design model to reveal where the building as built differs from the building as drawn, and on stone work those differences matter because walls are rarely as plumb or as square as the model believes.

The most valuable use of a scan is not producing the stone geometry, which a good laser template still does more efficiently for typical work. It is quantifying the discrepancy early enough to do something about it. A scan that shows a wall out of plumb by a significant amount three weeks before fabrication allows the scribe allowance to be planned; the same discovery on install day is a return visit.

Where a scan is used, agree who owns the registration and the accuracy statement. A point cloud without a stated accuracy is an impression rather than a measurement, and a fabricator who cuts to it and then finds the pieces do not fit has no recourse. Register to the project control network and record the residuals.

Feeding as-built information back into the model closes the loop and is what the highest level of development actually describes. For a stone package that usually means updating the model with the final piece layout, seam positions, material lot references and product data. It costs a modest amount of time at closeout and it produces a genuinely useful record for the facilities team, who otherwise inherit a building whose finishes are documented only in a photograph.

It also produces a commercial benefit for the shop. A fabricator holding an accurate as-built model of a large installation is the obvious party to call for replacement pieces, alterations and future phases, because nobody else can produce a matching piece as quickly.

Making It Work in a Real Shop

Start smaller than the specification suggests. A shop new to this should aim to receive, open and check the project model competently before it attempts to author elements into it. Being able to measure a wall length in the model, check a service position and raise a coordination question already puts a fabricator ahead of most trades on a mid-size job.

Nominate one person and give them time. The skill is not difficult but it is unfamiliar, and expecting the estimator to acquire it in the gaps between quotations guarantees it will not happen. A few days of structured training and a licence for the appropriate software is the realistic investment, and the payback comes from avoided site problems rather than from a line item on an invoice.

Price the work honestly. Model authoring, coordination meeting attendance and clash review are labour, and on a large package they are meaningful labour. Where a specification requires participation in a coordination process, include an allowance for it in the bid rather than absorbing it, and state what is included so a request for additional meetings can be handled as a change.

Set expectations about what the model can and cannot resolve. Coordination catches geometric conflict; it does not catch a wall that will be built out of tolerance, a supplier who ships a different lot, or a design decision that has not been made. Fabricators who describe those limits clearly are more useful to a project team than those who imply the model has solved everything.

Over a few projects the process starts paying for itself in a way that is easy to measure: fewer return visits, fewer field modifications, fewer arguments about whose responsibility a conflict was. Those are the costs that quietly erode margin on commercial work, and they are precisely what coordination is designed to remove.

File management deserves a rule rather than a habit. Coordination models are reissued frequently, sometimes weekly during an intense phase, and a shop working from a superseded version will produce beautifully coordinated work against a layout that no longer exists. Keep a single current folder, archive superseded issues rather than deleting them, record the issue date and revision on every file, and check the revision before any fabrication decision is taken from the model.

Weight and structural data are among the most valuable attributes a fabricator can contribute. Structural engineers frequently size supports from generic allowances because nobody told them the actual thickness, density and span of the stone element. Populating the model with real weights lets the support design be checked properly, and it occasionally reveals that a proposed cantilever or a long unsupported run was never going to work, at a stage where the design can still change.

Fire, slip and acoustic attributes matter on commercial projects and are routinely missing from stone elements in a project model. A flamed granite floor and a polished marble floor perform very differently underfoot, and a specification that assumed one and received the other creates a problem that nobody notices until a facilities manager reads a slip test report. Adding the relevant performance data to the element is a small effort with a disproportionate benefit.

Interoperability between the coordination model and shop software is improving but remains imperfect. Geometry usually survives an exchange; attribute data frequently does not, and complex surfaces can be simplified in ways that matter at stone tolerances. Test the round trip early on a representative element rather than discovering the limitation on a deadline, and agree with the coordinating party which format will govern if the two disagree.

Coordination meetings are worth attending in person or with the model open rather than treating them as status calls. The questions that save money are the ones asked while the model is on screen and someone can rotate the view: where does that waste pipe actually run, is that bracket in the way of the sink, does the door clear the counter. Those conversations rarely happen by email and almost never happen after fabrication has started.

Whatever the model says, the pieces still have to be cut and finished accurately. Explore precision cutting tooling and browse the Dynamic Stone Tools catalogue for the equipment that turns coordinated geometry into parts that fit the first time.