Key takeaways
Buy it before handover, and use it to receive the plant rather than to maintain it. The first job of the system on a greenfield site is to be the container that the equipment supplier's data lands in: the asset register from the equipment list, the manuals and drawings attached per machine, the manufacturer's service intervals as preventive plans, the recommended spares as a catalogue, and the warranty end date as a field. Every one of those is easy to get while the supplier still wants your acceptance signature, and expensive to get afterwards.
The second job is measurement. Availability, performance and quality calculated from the PLC give you an independent record of how the line actually behaves through commissioning and ramp-up. That record is what you take to a supplier when the contracted rate is not being met. Fabrico calculates availability times performance times quality from PLC data, with IoT sensors or AI cameras where a machine offers no usable signal, holds documents and history per machine, and supports bulk import with a quoted 3 days of Fabrico-side setup for the CMMS layer. Machine connection is a separate exercise paced by line access and by the commissioning schedule.
On an existing plant, the hard part is the data you already have: a register full of duplicates, thirty years of habit, and technicians who know things the system does not. On a greenfield, the hard part is the data you do not have.
There is no failure history, so MTBF is undefined, top-problem-machine reports are empty, and criticality cannot be inferred from what has hurt you before. There is no local knowledge, because the team may be new too. There is a supplier relationship that is at its strongest right now and will weaken permanently at handover. And there is a warranty clock running, typically twelve to twenty-four months, during which your record keeping is the difference between a claim and an argument.
That combination inverts the usual advice. On a brownfield site you start with reactive work and add structure later, because the structure has to be discovered. On a greenfield you can and should start with the structure, because it arrives in a crate with the machine.
This is the highest leverage paragraph in the article, and it belongs to procurement rather than to maintenance, which is exactly why it is usually missed. Before the equipment contract is signed, specify:
1. The asset and component list, in a spreadsheet, with a defined column set. Tag number, description, manufacturer, model, serial number, parent equipment, and where relevant the motor, gearbox and drive details. Ask for it in your naming convention. Suppliers will supply their own convention if you do not specify yours, and reconciling two conventions later is a job nobody has time for.
2. Manufacturer's maintenance schedules, as data and not only as a PDF manual. Task, interval, and whether the interval is calendar based or runtime based. This becomes your first preventive plan directly.
3. The recommended spare parts list with lead times and, separately, criticality. Lead time is the number that matters and it is the one most often omitted. A cheap part with a twenty-six week lead time is a bigger risk than an expensive part available in three days.
4. Documentation, indexed to the tag number. Manuals, electrical drawings, P&IDs, lubrication charts, certificates. Indexed to the tag, not delivered as a folder of scans named by document number.
5. The performance guarantee and how it will be measured. If the contract says a line will run at a stated rate with a stated availability, the contract must also say who measures it, from what source, and over what window. Agreeing to measure from the machine's own control system, into your system, removes the argument before it starts.
You cannot rank by failure frequency because there have been none. Rank by consequence and by recovery instead, which is available on day one from the design.
Walk the process flow and mark every asset that has no bypass and no redundancy: if it stops, the line stops. Then overlay replacement lead time, which you now have from the spares list. Then overlay safety and regulatory consequence, meaning anything whose failure creates a hazard or a compliance breach. An asset that scores on all three is critical regardless of how reliable the supplier says it is.
This is a deliberately crude method and it is the right one for month zero. Refine it after twelve months of real data using the approaches in our guide to asset criticality analysis. The mistake to avoid is waiting for perfect data before assigning any criticality at all, because criticality is what decides which preventive plans get built first, and building them in tag-number order is how a new plant ends up with beautiful schedules on the conveyors and nothing on the single-point-of-failure compressor.
The usual sequence is to get the plant running, stabilise it, and add monitoring once there is something worth monitoring. On a greenfield that sequence costs money in two specific ways.
You lose the ramp-up curve. The first months of a new line are the richest source of information you will ever have about it: which stations starve, which changeovers are slow, which faults recur. Measured, that becomes a debottlenecking list. Unmeasured, it becomes folklore, and by the time monitoring is installed the line has been tuned around problems nobody documented.
You lose the contractual leverage. A rate guarantee is enforceable while acceptance is still open and much harder afterwards. If you can show, from the machine's own data, that the line achieved a given performance over a defined window against a contracted figure, that is a conversation with an outcome. A dispute conducted from shift reports and recollection is a conversation without one.
There is a practical caveat worth planning around. Connecting machines is physical work: PLC addressing, tag mapping, and in some cases retrofit sensors or camera mounting. On a greenfield it is far cheaper during commissioning, when the panels are open and the integrator is on site, than six months later. Get it into the commissioning scope rather than treating it as a follow-on project. See the OEE for manufacturing guide for how the calculation is built, and OEE monitoring without a PLC for the machines that arrive without a usable signal.
Question 3 is small and pays disproportionately. A warranty expiry field with a task attached turns a date nobody remembers into an inspection three months before it matters, which is the last chance to claim on anything marginal.
For a greenfield the relevant capabilities are the ones that let you build a system before you have operational data. Fabrico provides a machine registry with files, documents and history per asset, bulk import with support included, an annual preventive maintenance plan with recurring templates and conditional tasks, custom fields for things like warranty expiry and criticality, inventory covering catalogue, minimum and maximum levels, deliveries and consumption tied to work orders, and approval workflows for the acceptance and handover steps.
On the measurement side, availability, performance and quality are calculated from PLC data, with IoT sensors and AI cameras for machines with no usable signal, a real-time OEE dashboard, micro-stop detection, and downtime, MTTR and MTBF analytics that begin accumulating from the first shift. Data moves through Excel import and export, a REST API, webhooks, and a bidirectional SAP PM sync including S/4HANA where the new site is joining an existing ERP landscape.
Implementation on the CMMS side is quoted as 3 days of Fabrico-side setup covering configuration, users, roles and bulk import, with live and on-site training available. Machine connection for OEE is separate and paced by the commissioning schedule rather than by software, so build it into the integrator's scope. The platform is ISO 27001, ISO 9001 and ISO/IEC 20000-1 certified, hosted in an AWS EU region, with a 4 hour recovery time and recovery point objective, and the interface ships in English, Bulgarian, German, French and Polish.
One scope note so it is not discovered late: Fabrico does not include a production planning module, so scheduling of production orders remains with your ERP or MES.
Take a packaging line bought with a contracted output of 120 units per minute at 92 percent availability, measured over a continuous 72 hour acceptance run, with a twelve month warranty.
Done badly, acceptance is run with a supplier laptop logging to a spreadsheet, the plant signs off in the room, and four months later the line averages 104 units per minute. The conversation with the supplier now hinges on whether the shortfall is the machine, the material or the operators, and nobody has data that separates them. The line stays at 104 and the loss is permanent.
Done well, the OEE connection is made during commissioning as part of the integrator's scope. The 72 hour acceptance run is measured from the machine's own signals into your system, giving availability, performance and quality separately rather than as one blended number. When output drops later, the split is already visible: if performance is intact and availability fell, it is stoppages, and the stoppage reasons are recorded; if performance fell with availability intact, it is speed, which is a machine or material question.
The value here is not the dashboard. It is that separating availability from performance turns an unwinnable argument into a specific one, and that separation has to exist from the acceptance run onwards or it never exists at all. A greenfield gives you exactly one chance to install it at the right moment.
No, and waiting is the more expensive error. The register, manuals, spares list and manufacturer schedules all arrive during the project, and if there is no system to receive them they land in email and shared drives, from which they are reassembled later at real cost. Buying early also means the maintenance team learns the tool during a quiet period rather than during a ramp-up.
Use it as the catalogue, not as the purchase order. The list tells you what exists, what it is called and, if you asked for it, what the lead time is, which is genuinely valuable. What it does not do is optimise your working capital, because it is written to minimise the supplier's exposure to downtime disputes. Stock against criticality and lead time: long lead time plus no redundancy justifies stock, short lead time usually does not, however cheap the part.
Start with the manufacturer's intervals, because they are the only defensible starting point and they also protect the warranty. Expect them to be conservative and generic, since the manufacturer does not know your duty cycle or environment. Plan to review them after twelve months against real failure and inspection data, and record the reasoning when you change one, so the change survives a personnel change.
If a group standard exists, join it, and negotiate only on the taxonomy. A new site is the easiest place to adopt a group naming convention because there is no legacy to convert, and being consistent with the rest of the group is worth more than any local preference. Where the site is in a different country, check the data residency and any employee representation requirements before assuming the existing setup transfers, which is covered in our European multi site rollout playbook.
Availability and its stoppage reasons, plus preventive plan completion. Not MTBF, which needs failures you have not had yet, and not cost per unit, which is dominated by ramp-up noise. Stoppage reasons in the first six months are the debottlenecking list for the following year, and preventive completion tells you whether the regime you designed on paper is one the team can actually execute.
If you are still in the equipment procurement phase, the cheapest thing you can do today is fix the data deliverable in the purchase order. Book a demo and ask for the import specification so you can attach it.
Last updated: 7 August 2026.
Login
Blog