Almost nobody wants a rectangle of identical slats. The ceiling that gets approved has an island over the reception desk, one stretch turned to run with the corridor rather than across it, a couple of rows in the brand colour and light only above the desks. Every one of those decisions is easy to draw and expensive to count, which is why they usually arrive at the workshop as a sentence rather than as quantities. Module 01 of AxisMod, the ceiling baffle configurator, treats them as part of the configuration, so they land in the schedule at the moment they are made.
The standing caveat from the earlier posts in this series still applies: the module is a demonstrator. It computes end to end, but on a demo catalogue with indicative rates, and I have not yet quoted a live ceiling through it. The mechanism is the part worth writing about.
Where the uniform ceiling stops
A uniform ceiling is the one case the usual process handles well. Area, pitch, a multiplication, done. The job stops being uniform the moment somebody looks at the room, and from then on the ordinary workflow starts leaking.
- The feature gets described instead of specified. "Island over reception, accent rows as per architect" is a perfectly normal line on a drawing, and it means the person ordering material has to guess how many pieces are involved and in which finish.
- Accent pieces are a different product line hidden inside the same quantity. They are the same profile in another finish, so they disappear into the total and surface late, when the rest of the ceiling is already cut.
- The lit rows and the electrical schedule drift apart. Light above the desks and nowhere else changes the number of fittings, the run of the supply and the driver count. If that decision lives in a note, the electrician prices a different ceiling from the one the client approved.
- The suspension count is taken from the area. A turned stretch or a separate island is not just a visual choice: it is its own set of hanging points, and that item decides whether the install is a day or three.
None of this is difficult work. It is counting, done twice, by two people, from two slightly different versions of the same ceiling.
The layout is a choice, not a default
In the module the arrangement is one of the first decisions rather than a post-rationalisation. Runs can go in one direction across the room, turn through a right angle, cross in a grid, or collapse into a feature island sitting over one part of the floor while the rest of the ceiling stays open. The room keeps its dimensions through all of that; only the arrangement changes.
Service openings are drawn on the plan in the same pass. Luminaires, diffusers, sprinkler heads and speakers are marked as zones the ceiling has to leave alone, and the runs break around them with a clearance rather than being trimmed by hand afterwards. That matters here because an island is usually placed exactly where the services are busiest, over a reception desk or a bar.
Then it goes run by run
The second half is the part that normally never survives into production. With the plan on screen, individual runs and individual baffles can be touched: take this one out, put light on that one, make these two the accent finish. The ceiling stops being a formula and becomes the ceiling somebody actually designed.
Illumination works the same way. Light can sit on every run, every second or every third, it can throw down onto the floor, up onto the soffit or both, and any single run can override the pattern. "Lit above the desks, plain over the meeting rooms" is a configuration, not a conversation to have again later.
The important part is what happens underneath each of those clicks. Every edit re-runs the whole calculation rather than a part of it, so the plan, the section, the acoustic result, the quantities and the price always describe the same ceiling. There is no state in which the preview shows the island and the bill of materials still prices a plain grid.
What the client ends up holding
- A plan that shows which run is which, including the ones removed, the ones lit and the ones in the accent finish, so nothing on it has to be explained in a covering email.
- Quantities taken from the pieces on the plan rather than from the area, so the runs marked as accent are counted as what they are instead of being averaged into the total.
- A count of lit runs that the electrical side can work from, consistent with what the client approved on screen.
- Hanging points counted from the layout that was actually chosen, the island included, which is the number the install programme depends on.
- A cut list allocated to stock lengths with the waste visible, and the production documents generated from the same configuration.
- A price that already contains the feature. The honest version of "what does the island cost" is a number in the meeting, not a variation two weeks later.
Why it was hard: the plan had to behave like a drawing
The first version of the plan looked right and felt wrong. It was built as a picture that you could also click, and on a drawing tool that distinction shows up immediately: you would double-click the run you wanted to change and nothing would happen, because the click was being captured by the thing you were dragging a moment earlier. Clients forgive a plain-looking tool. They do not forgive one that ignores them, and the whole argument for editing the ceiling run by run collapses if the first attempt to touch a run does not register.
The fix was to let the drawing itself decide what was clicked rather than letting whichever element had the pointer answer for it. Dull work, and it was the difference between a plan view people use and a plan view they look at and then ask me to send a PDF.
What this is worth on a fit-out job
The test is not whether a configurator can draw a feature. It is whether the feature reaches the people who buy material and book the install as quantities. If the island exists on the render and as a sentence on the drawing, somebody is counting it by hand, and that count is the one that goes wrong.
The baffle module can be clicked through on screen, plan view included: AxisMod. Putting your own profiles, your own finishes and your own rates behind the same mechanism is exactly what a module is for. The shop-floor end of it, cut lists, nesting and CNC-ready files, is the part of my work this grew out of: production support.
