The whole-life cost of an automated cell: the equipment quotation is the smaller half

Getting it into the building and connected

Delivered equipment cost excludes nearly everything needed to make the cell work in your building. Floor preparation and load capacity, foundations for anything that vibrates, a supply of the right rating with switchgear to feed it, compressed air of the right quality and volume, extraction where the process demands it, network provision and the security work attached to it, guarding and its validation, plus craneage, rigging and occasionally removing part of a wall. Several of these need permits and lead times longer than the machine. Survey and cost them at quotation stage, since finding the air supply inadequate after delivery is expensive.

Internal engineering time is real money and never invoiced

Automation consumes internal engineering across a far longer span than the installation window. Somebody writes the requirement, evaluates suppliers, sits through design reviews, runs acceptance tests, coordinates the services work, retrains operators, updates routings and work instructions, then supports the cell through the period where it stops for reasons nobody yet understands. That effort comes from a small pool of people also running the rest of the plant, and treating it as free is why projects overrun on time rather than on money. Estimate it, name who supplies it, and say what they will stop doing.

Change parts, tooling and the cost of adding a product

Every product variant passing through the cell brings its own tooling: gripper fingers, fixtures, guides, format parts, programs and somewhere to store all of it. That cost recurs whenever a product is added and is the line most often forgotten when the commercial team accepts new business. Wear items accumulate alongside: seals, filters, belts, flexing cables, vacuum cups, tips and nozzles. Establish the cost of introducing a new product early and make it visible to whoever agrees new orders, because a cell that looks cheap per part turns expensive once the product range grows faster than anyone forecast.

Licences, support contracts and the control refresh nobody provides for

Software carries recurring cost. Programming environments, sensing packages, monitoring tools and safety configuration software all involve licences, and some need maintaining simply to stay able to open your own files. Industrial computers reach end of support long before the mechanics wear out and have to be refreshed under the plant's own change rules. Controllers eventually go end of life too. A cell expected to run for many years therefore needs a provision for a control refresh part-way through, and including that in the original financial case is far easier than requesting it later against equipment management believes is fully paid for.

The end of life that gets left out of the appraisal

Bespoke automation has almost no resale value. When a product ends, the cell that made it is worth its robot, its drives and its scrap metal, and even that recovers little once removal is costed. Redeployment sounds attractive and rarely survives contact: fixtures suit one part, the layout suits one building, and re-engineering approaches the cost of buying new. Budget disconnection, removal, disposal of anything hazardous, making the floor good and terminating support agreements. Raise it at approval, because a cell with a defined product life and no residual value should be depreciated to match.

Frequently asked questions

What is most often missing from an automation budget?

Services and building work, internal engineering time, and change parts for future variants. The first surfaces during installation, the second gets absorbed silently by an engineering team that then delivers nothing else that year, and the third arrives when a new product is accepted and nobody costed the tooling. A useful discipline is requiring the sponsor to state the installed cost including services, plus the cost of adding one further product variant, before approval is granted.

Do automated cells have residual value?

Very little where the cell is bespoke. Standard machines and robots retain some value, though removal, reconditioning and reprogramming absorb much of it and a buyer has to want that exact configuration. Cells built around specific fixtures in a specific layout are effectively worth their component parts. Treat residual value as close to nothing when appraising an investment, and where a finance case leans materially on a resale assumption, ask what evidence supports it beyond the supplier's optimism.

How should we budget for spares on a new cell?

Split the question by lead time rather than by failure frequency. Items obtainable quickly can be bought when they fail. Long-lead items, custom items and anything whose absence stops the cell should be held, and the integrator should supply that list as a contract deliverable with manufacturers' references rather than internal codes. Add wear parts on a consumption estimate, and revisit the list after the first period of running, because the real failure pattern seldom matches the predicted one.