Ask three suppliers to integrate a programmable logic controller and an operator panel with the same machine and the same duty, and the quotations will not look alike. One arrives with a short line and a low figure. Another runs to several pages and costs several times more. At the moment of signature the two look like the same job with a different margin attached, so the buyer compares totals, and the totals are the one thing in the comparison that has any real meaning on its own. The useful comparison sits underneath the number: what each quotation has already decided, what it has left for a later invoice, and which of those decisions you would have made differently.
This article assumes an automation supplier is being asked to deliver the control layer together with the equipment it drives, which is the arrangement most buyers of factory equipment are actually quoted. Read it as a budgeting method: how to break one lump figure into parts you can price, and how to tell a narrow scope from a thin scope. It is not a price guide, and it does not quote figures from any supplier, machine or project.
First, what you are buying when you buy integration
A PLC and an HMI are two devices, and it is tempting to budget for them as line items on a bill of materials, because device list prices are easy to find and feel like an anchor. Spell the same pair out as a programmable logic controller plus an operator panel at the enquiry stage and nothing changes except how many words land on the page. Integration is not the devices. It is the work that makes them behave as one designed package: deciding which sensor or drive each signal belongs to, what the panel shows an operator at each step, what the system is forbidden to do while another part of the process is running, what happens when a signal disappears in the middle of a cycle, and what record the machine keeps afterwards.
That work is what the money goes on, which is why two quotations carrying identical hardware can differ so much. A narrow quotation supplies a controller, a panel, wiring and a program written from the supplier's own habits. A wider quotation has already answered the questions above and written the answers down, so the equipment arrives with a defined start, stop, hold, alarm and hand-back behaviour. Neither of those is wrong, but a narrow scope priced against a wide scope is how a buyer ends up paying for the same decision twice.
Break the lump figure into the parts that actually cost money
Write the following blocks on one sheet and put each quotation against it, so that the difference between two suppliers becomes visible block by block instead of one number against another.
The engineering block covers the input and output list, the signal names, who drives each shared line, what the system does when a device is disconnected, and the sequence the machine is expected to follow. This is usually where the largest hidden difference lies, because it is the block suppliers assume and buyers discover late.
The panel block covers the screens themselves: start and stop, mode selection, the parameters an operator may change, the ones locked away, the alarms in words an operator can act on, and the help screen that tells them what to do next. Screens are cheap to draw and expensive to redraw, so a quotation that shows you a screen list before you sign is telling you something about the rest of the scope.
The record block covers anything the machine is expected to keep: cycle counts, alarm history, a parameter set that can be recalled, or a per-unit record. Buyers routinely leave this out of a budget, then add it after go-live, when it costs more because the program structure already exists without it.
The site block covers cabling to the boundary, power supply, air or other utilities the control layer depends on, network access if anything talks to a plant system, and the physical position of the cabinet and panel relative to the machine and the operator. The supplier cannot do most of this and will not always say so; the budget has to say it.
The commissioning block covers what happens after delivery: the tests that prove the behaviour was built, the training, the documentation left behind, and the acceptance signature. Commissioning is where a thin scope is discovered, and discovery on site is paid for in lost production time rather than in line items.
The change block is the one almost no quotation prices and almost every project needs: the cost and lead time of a change after the program is written. Ask how a change is raised, who quotes it, and what a small change costs in that supplier's terms. A comparison that skips this block is comparing the price of the first version only.
Compare the assumptions each supplier has quietly made
Every low figure rests on assumptions that were never written down, and the way to find them is to ask the same handful of questions in the same words to every supplier and read the answers as scope.
Who owns the machine's interaction with the rest of the line? If a conveyor, a robot, an upstream feeder or a plant system has to agree before a cycle starts, somebody has to write that agreement. If the quotation does not name that somebody, the assumption is that it is you, or that the question will never arise.
What happens when a signal fails? A control layer can handle a missing sensor in three ways: stop, hold and warn, or carry on with the last known value. Each is legitimate for a different risk, and each costs a different amount to build and test. A quotation that does not mention failure is not free of the decision, it has just made the cheap one without telling you.
What is forbidden, and who enforces it? Interlocks are the part of integration that protects people, product and equipment, and they are also the part most easily dropped from a scope because they are invisible until something goes wrong. Ask for the forbidden conditions in words, and ask how each one is tested before acceptance.
Where does the supplier's responsibility end and the building's begin? Cabinet position, the power feed, the network drop, earthing, and the run of cable between the panel and the machine all sit on a boundary. The cheapest quotations are often the ones whose boundary sits furthest from the machine.
Who can change the machine after handover? If your own engineers are expected to edit programs or screens later, ask what access, what documentation and what training the quotation includes. If none is included, the supplier has assumed they remain the only person who can touch the machine, and that assumption has a price attached to it later.
Put the costs that arrive after go-live into the same budget
A control layer is not finished when it works; it is finished when the people who run it can keep it running. Several costs are easy to leave out of the initial number precisely because they surface months later and are then booked as something other than the project.
Training is the first. An operator who cannot read an alarm message correctly will clear the machine in a way the program was not written for. Ask how many people the quotation trains, on which machine, for how long, and whether the training happens on the delivered machine or on a demonstration unit that behaves differently.
Supportability is the second. When the machine stops in the middle of the night shift, the cost is not the fault, it is the waiting: finding the person who understands the program, explaining the symptom, and getting a response. Ask what documentation is handed over, whether program and screen sources are available, and what the supplier's route is when the line is down at the weekend. Then compare answers, not promises.
Spare parts for the control layer are the third. Controllers, panels, drives and power supplies fail, and the replacement you need is often the one thing nobody ordered. Ask which parts the supplier recommends holding, how long the current model is expected to remain purchasable, and what happens to the program if a component is replaced with a newer revision.
The first changes are the fourth. Once the machine is in production, somebody will want a new mode, an extra alarm, a different sequence or a record that was not in the original list. Put a deliberate allowance in the budget for the changes you can already foresee, and ask each supplier how an unforeseen one is handled.
Ask how integration is quoted before you read the figure
Two commercial habits shape these quotations, and both distort a comparison of totals.
The first is bundling. A supplier who delivers the machine and the control layer as one number may be carrying the integration honestly and simply not showing it, or may be carrying an assumption you disagree with. Either way you cannot see it, so ask them to split the machine, the hardware and the control work into three numbers and to state the assumptions behind the control number. A supplier who can do this on request is usually easier to work with afterwards as well.
The second is time-and-materials for work nobody has scoped yet. Some suppliers will not fix a control price until they have seen the machine and the process. That is reasonable, but the gap should be handled on purpose: ask for the rate, the estimate range, and the point at which the estimate becomes a commitment, and hold that figure next to the bundled totals rather than leaving it out of the comparison.
When a plant controller is part of the conversation
If the equipment has to join an existing line rather than stand alone, the budget grows in a specific direction and it is worth naming that direction in advance. Handshakes between machines, shared safety conditions, a common alarm view and a record that the line can read are all work that neither the machine supplier nor the line owner may have priced. Where a demonstration line or a reference installation is offered as evidence, ask what that line actually records and which conditions it was built for; a card with a title on it proves nothing about your process.
This is also the point where the boundary between catalogue equipment and designed equipment matters. A machine built to your process, which is what non-standard automation equipment covers, carries its control layer as part of the build, so the integration questions above belong inside one quotation. Ask the same questions in writing in either case; the wording does not change because the scope was bundled.
What to ask for in writing before you sign
Put these on one page and send the same page to every supplier. The replies are the comparison.
One: the input and output list, with the device each signal belongs to and the failure behaviour of each one. Two: the sequence, in words, from cold start through a normal cycle to a stopped state, including what is forbidden. Three: the screen list, with what an operator may change and what is locked. Four: what the machine records, and in what form you receive it. Five: the boundary, in words, of where supplier wiring ends and building work begins. Six: the acceptance test for each of the above, and who signs it off. Seven: the change process, with a rate and an example of a small change. Eight: handover package: documentation, program and screen sources, spares advice, training attendance and duration.
Where the equipment itself is being specified as well as controlled, a written brief is what keeps the two estimates honest. Our customized steam cleaning equipment enquiry page is built around exactly that: what is cleaned, on what surface, at what cadence, with what utilities on site. The same discipline applied to the control layer is what turns three unlike quotations into one readable comparison.
Where an engineering partner changes the arithmetic
Some buyers carry the integration questions with their own engineers; others bring a partner in early to write the requirement before quotations go out. The second route has a cost of its own and removes a different cost: the ambiguity that every supplier resolves in their own favour. If the requirement is written by somebody who has specified control layers before, the quotations you receive will be closer to one another, and a closer spread is itself a saving.
That is the reason we list integrated engineering team services separately from equipment: a buyer who needs the specification, the testing plan and the acceptance checks as a defined piece of work can commission them as such rather than hoping they arrive free inside a machine price. Whatever route you choose, the number to compare is not the total; it is the number of decisions each quotation has already taken off your plate and written down.
A closing note on what this budget is for
A control layer that has been budgeted carefully is not the one that cost least at signature. It is the one that is still being used properly long after handover: alarms that mean something, a sequence your operators follow rather than work around, a record you can read when a problem appears, and a change you can buy without replacing the machine. Those outcomes are cheap to ask for before the quotation is signed and expensive to add afterwards, which is the entire argument for comparing scope, failure behaviour, boundary and change cost rather than comparing totals.
