MEP Fabrication Shop Software: Why Does the Shop Build Ahead of the Wrong Zone?

Prefabrication only pays if the right assembly is on the right truck

A sheet metal shop finishes a good week. The coil line ran at high yield, the nesting was efficient, 340 pieces of duct went through the plasma table, and the shop manager is pleased. On Thursday the field superintendent calls, angry. His crew on level 6 north has been waiting two days for eleven fittings, and meanwhile the yard is full of level 9 duct that cannot be installed for another month because the slab pour slipped.

Nobody did anything wrong by their own measure. The shop optimised for the machine, because that is what the shop is measured on. The field needs material in crane and install sequence, because that is what the field is measured on. There is no shared object between them that says what is needed when, so the two halves of a prefabrication strategy optimise against each other and the yard fills up with correctly made parts in the wrong order.

The information that would fix this exists. It is in a Revit model, a CAM package, a whiteboard in the shop, a spreadsheet the detailing manager keeps, a WhatsApp group between the foreman and the shop lead, and a stack of paper cut lists. What does not exist is a single item that knows what it is, which zone it belongs to, what state it is in, and when the field expects it.

Problem 1: an item has three identities and none of them is authoritative

The same elbow exists as a Revit element with a unique identifier, as a CAM item with a shop number, as a barcode label that got printed and possibly reprinted, and as a line on a cut list. When the field asks whether spool MP-06-114 is fabricated, someone has to translate between those identities to answer. Multiply by 8,000 items on a hospital project and the translation cost swallows the productivity gain that justified prefabricating in the first place.

The first thing a build must establish is a stable item master. One record per fabricated item, carrying the model element it came from, the shop item number, the spool or assembly it belongs to, the zone and system, and a label identity that survives reprinting. That sounds trivial in a specification and it is the hardest thing to retrofit later, because a numbering scheme that has drifted across three projects cannot be reconciled after the fact.

Autodesk Fabrication CADmep and Trimble SysQue live upstream of this, producing content and cut data from the model. They are not tracking systems and were never meant to be. STRATUS and MSUITE sit in exactly this gap and do it credibly, which is why either belongs on your shortlist before you consider building.

Problem 2: the shop schedules by machine, the field needs by sequence

This is the structural conflict and it is worth stating plainly. Nesting exists to maximise material yield from a coil or a sheet, which means grouping similar parts regardless of where they go. Pipe spool fabrication batches by size and schedule to reduce changeover. Both are rational. Both produce output in an order the field cannot use.

What a build introduces is a demand signal. The field publishes an install sequence by zone and date, usually derived from the project schedule and refined weekly by the superintendent who actually knows what is happening. The shop plan then runs against required on site dates rather than against a backlog list. Nesting still optimises, but it optimises within a window bounded by need date, which is a materially different instruction. The shop keeps most of its efficiency and stops building four weeks ahead of a zone that is not ready.

The dashboard that comes out of this is the one every operations director wants and almost nobody has: for each zone, what is required, what is fabricated, what is kitted, what is on a truck, what is received on site, and what is installed. Six states, one item, visible to both halves of the business.

Problem 3: kits ship incomplete and the crew waits anyway

A kit is not a list of spools. It is everything a crew needs to install a zone: the spools, the duct, the hangers and rod, the fittings and gaskets, the valves, the fasteners, and the drawing package. Shipping ninety five percent of a kit does not deliver ninety five percent of the value. A crew missing one box of threaded rod is a crew standing still.

So the kit needs to be a defined object with completeness rules, and the load list needs to refuse to close when a kit is short unless somebody with authority overrides it and the field is told what is missing before the truck leaves. Scan on at the shop, scan off at the site gate, and a receiving confirmation from the field, because material that arrives and cannot be found is functionally the same as material that never shipped. On a congested urban site, laydown location on receipt is worth capturing too.

Problem 4: nothing comes back from the field, so earned value is a guess

Shops typically know what they made. Almost none know what got installed. That gap matters for two reasons. First, percent complete and earned labour against installed quantities are the numbers your project managers report and they are currently estimated by walking the job. Second, without installed status you cannot tell the difference between an item that is missing and an item that is already in the ceiling.

The field scan has to be trivially easy or it will not happen. That means a phone, a QR code, one tap, and it has to work with no signal in a basement plant room and sync later. Anything heavier gets abandoned in week two, and the loop stays broken.

The same loop carries remakes. A spool that does not fit becomes a verbal request to the shop lead today, which means no cost code, no root cause and no measurement. In a build it is a remake request with a reason code, a photo, a link to the original item and a priority, and at the end of a project you can finally answer whether your remakes come from detailing errors, field dimension changes or shop workmanship. That answer is usually uncomfortable and always valuable.

Problem 5: traceability on pressure piping is a separate discipline

Mechanical contractors welding pressure piping carry obligations that a general tracking tool ignores. Weld maps by spool, welder identification against qualification records, procedure references, heat numbers from material test reports, and nondestructive examination results with their reports. When an inspector or a client's QA representative asks for the package on spool MP-06-114, it should be one export.

This is a genuine differentiator in a build, because it links the QC record to the same item identity the shop and the field use. Contractors who keep welding records in a separate binder end up assembling turnover packages for weeks at the end of a job, using people who should be on the next project.

What this costs and how long it takes

Across the 2,000 plus projects Digital Heroes has delivered, the shape here is as follows. A first release covering the item master linked to your model output, shop routing with station level status, barcode or QR tracking, and kit and load list creation by zone runs $70,000 to $150,000 and ships in 12 to 18 weeks. A full platform adding field receiving and install status with offline capability, remake workflow with reason codes, weld and material traceability, labour standards and earned value, and ERP integration runs $180,000 to $450,000 over 6 to 12 months.

What drives the number up in fabrication shops: machine and CAM integration, because posting cut data to a plasma table, coil line or pipe cutter is specific to the equipment on your floor. Label printing at industrial scale, which sounds trivial until you need durable labels that survive galvanising and a site winter. Multi trade shops, since sheet metal, pipe and plumbing have different work in progress models and pretending they are the same produces a system that suits none of them. ERP integration for material purchasing and job costing, where Viewpoint Vista, Sage and Acumatica are each their own project. And offline field capability, which is architecture rather than a feature.

What keeps it down: one trade, one project, and the six states from fabricated to installed. Resist adding weld traceability and earned value into release one, however much the quality manager pushes.

Build versus buy, and when buying is right

Buy if you run a single trade shop, model in Revit, and want conventional spool and duct tracking with kitting. STRATUS is well established with mechanical and sheet metal contractors and integrates with the Autodesk world your detailers already live in. MSUITE covers similar ground and is worth a look alongside it. If your process fits either, use it and put your energy into detailing standards, which is where prefabrication actually succeeds or fails.

Build when two or more of these are true. Your numbering and zoning methodology is genuinely proprietary and is the reason your install rates beat your competitors. One shop serves several trades and multi trade racks are a growing share of output. You need the demand signal to come from a field sequence that your superintendents control weekly. You require weld and material traceability tied to the same item identity as everything else. Or you run more than one shop and need to load balance between them, which no product handles well because it is a scheduling problem specific to your geography and your fleet.

How to choose a developer for fabrication software

Ask them to draw the item lifecycle before you sign anything. It should show a model element becoming a shop item, moving through routing steps at named stations, being assigned to a kit, loaded, received and installed, with a status change at each point and a barcode identity that survives reprinting. A developer who draws products and orders has built an ecommerce system and will discover work in progress the hard way.

Ask how the system behaves when the model changes after fabrication has started, because it will. The answer needs to cover superseded items, items already cut, and what the shop is told.

Ask about offline. Basements and plant rooms have no signal and that is where install scanning happens.

Ask what they have actually integrated on a shop floor, naming the machine and the controller, not integrations in general. And settle code ownership before kickoff: you should hold the repository, the infrastructure accounts and the right to bring in another firm. At Digital Heroes the client owns the code from the first commit.