Before the peak: surveying a distribution center that cannot stop

A point cloud of your distribution center is a millimeter-level survey of what is actually built and stacked inside it: racking, columns, beams, services and real clear heights, captured with a laser scanner instead of inherited from a drawing. Before a demand peak it does one very specific job. It lets you decide on measurements that are true, rather than on a file that stopped matching the building several retrofits ago.

The calendar is tight. In Mexico, El Buen Fin 2026 runs from Friday 13 to Tuesday 17 November, landing days before Black Friday and ahead of the Christmas and Three Kings’ Day cycle. Whatever you intend to change in the layout has to be decided while the site is still running, not once the season is on top of you.

At Foundtech we do exactly that kind of survey on facilities that are operating. At Migros Millas we documented 11,000 m² of a working retail facility, and at the Werdhölzli treatment plant the team discovered, as it started planning the modernization, that it had no updated drawings or engineering for a facility with stacked systems. We have delivered more than 200 projects across Europe and the Americas and over 10 million m² modeled, and we are ISO 19650 certified. That is what separates a survey from a good-looking file: the measurement is taken off the building, not off what the project said would be built.

A 7-minute read. By the end you will know what a point cloud actually measures inside a full warehouse, what it cannot see, how it gets captured without stopping operations, and how much time you have to subtract from 13 November to make it.

If what you are after is what to optimize in the layout once you have the measurements, we cover that separately in how to optimize your logistics warehouse layout with Scan2BIM. This article is the step before: how you get the measurement in a building that cannot stop.

The drawing stopped being true and nobody raised a hand

No distribution center looks like its construction drawings. A rack level was added, a dock door was moved, a new duct was hung, the lighting was swapped, an aisle was closed off for a value-added area. None of those changes made it back into the archive, because each one was solved as an operational emergency rather than as a project.

The cost lands on the decision, not on the archive. When the moment comes to add racking, change the picking flow or install a system, somebody has to answer how much space there is. And the answer comes from an old drawing, a tape measure and the warehouse manager’s memory. The three sources contradict each other, and the contradiction surfaces once the equipment has already been bought. MHI, the U.S. material handling industry association, places warehouse layout and space utilization among the core levers of running a distribution center.

That gap between what the information says and what the building is has a measured cost. NIST put it at $15.8 billion a year in the U.S. capital facilities industry, with two-thirds absorbed by owners and operators during operations and maintenance. That is not a construction cost. It is the cost of operating blind.

What a point cloud measures inside a full warehouse

A point cloud is a set of millions of coordinates captured by laser sweep. In a distribution center it answers precisely the questions a drawing answers badly.

The real clear height under beams, ducts and sprinklers, which is almost never the one on the drawing and almost never the same across the building. The effective aisle width with the racking already installed and deformed by use. The floor flatness and level changes, which decide whether a taller reach truck can operate there. The exact column positions against the rack module you intend to install. And the dock geometry: heights, ramp slopes and maneuvering room.

Those five measurements are what a racking supplier needs in order to quote without assumptions. The Rack Manufacturers Institute, the body behind the U.S. structural standard for racking, works from the real conditions of the building to define the design and capacity of a storage installation. Declare those conditions wrong and the calculation is run against a warehouse that does not exist.

And what it does not measure, which also has to be said

The laser sees surfaces. Anything hidden behind stacked product is not captured: if a rack is full to the top, the structure behind the pallet does not make it into the cloud. It does not read the inside of a wall either, or the run of a pipe embedded in a slab.

A point cloud is also not a logistics diagnosis. It will not tell you how many lines per hour leave each zone, where your picking bottleneck sits, or whether your slotting is wrong. That lives in your warehouse management system and in your operation. The cloud gives you the exact container; the contents and the flow are yours. Our work does not cover construction management or executing the layout change either: we deliver the model and the information your team or your contractor decides from.

How it gets captured without stopping operations

This is the part that stalls the decision most often, and almost always because of a wrong expectation: people picture scanning as construction work occupying the building. It is not, but it is not invisible either. It needs coordination, and it is worth planning around these criteria.

It is scanned zone by zone, not in one pass. The survey is split into stretches that get closed and stitched together. That lets the crew work whichever zone is quietest in each shift and leave the one that is receiving for later.

The natural windows are shift changes, Sundays and scheduled maintenance stops. In a distribution center running continuously, those hours are worth more than any special permit.

Clear what you can, but nothing needs emptying. An aisle free of loose pallets and parked forklifts gets captured better and faster. Emptying racks is not a requirement: what we are after is the envelope and the structure, not the inventory.

It counts as a floor visit. Your warehouse rules apply: safety induction, vest, footwear, pedestrian routes and an escort. Better to settle that before the crew arrives than at the gate. The critical point is sharing space with moving forklifts, which OSHA treats as a specific workplace hazard with its own rules on traffic and visibility. A well-planned scan works around those routes, not the other way round.

What is hidden gets logged. Where product blocks the view, it is recorded as an uncaptured zone instead of being filled in with an assumption. A model that is honest about its gaps is more useful than one that is complete and invented.

The calendar, counting backwards from 13 November

This is where the conversation gets concrete. Our published timelines for Scan to BIM are about a week for small projects and three to five weeks for complex industrial bays, which is where most distribution centers land. On top of that you have to add your internal decision time and, above all, the physical execution of the change.

With the peak starting on 13 November, a survey that begins in September delivers the model in October and leaves October clear to quote, approve and execute. One that begins in October delivers on top of the peak, when nothing can be touched any more. Both routes scan equally well. The difference is that one lets you decide and the other only lets you document for next year.

On this kind of work we report up to 25% savings in project time and up to 35% fewer reworks, precisely because the model arrives coordinated and the measurements are not argued halfway through the job.

What you get, and who uses it

Delivery includes native editable files in Revit or Archicad (RVT, PLA), drawings generated from the model, and free viewers, so anyone on your team, whether that is the warehouse manager, the maintenance lead or the racking supplier, can open the project without needing a license. It does not lock you in with us.

That same material is what lets several suppliers quote against the same basis. At Migros Millas, with 11,000 m² documented, the value was not the model in itself: it was that everyone involved worked inside one digital environment instead of across each party’s own interpretation. In a distribution center, that difference shows up the day three racking quotes arrive and all three measured differently.

If the project stays alive after the peak, because you plan to instrument, monitor or scope the next expansion, the model can grow into a digital twin. If you only need to solve this winter, it does not have to. Pay for what you are going to use.

How to tell whether this conversation is yours

The signal is simple and needs no diagnosis. If on your last modification someone had to climb up with a tape measure to check something that was already on a drawing, if a supplier delivered equipment that did not fit, or if nobody in the company can tell you with certainty what the clear height under the beam is in the reserve area, what you have in front of you is not yet a layout problem. It is a measurement problem, and it gets solved first.

Tell us what you need to decide in your distribution center before the peak, and we will define together the scope of the survey and the level of detail you actually need. Book a technical session here.