The microfactory: a compact automated plant with a short capital runway

Designing a plant around a small footprint

The format forces choices: compact cells rather than long lines, machines selected for footprint and modest power supply, materials handled in totes instead of pallets, and services a leased industrial unit already provides. Capital is lower in absolute terms and, more usefully, arrives in small increments — a cell at a time rather than a building. Commissioning is quick because civil works are minimal. The matching limitation is that nothing scales gracefully inside the box; more output means another cell, another unit, or another site altogether. The constraint disciplines the design as well, since anything needing a permanent foundation or a specialist utility connection usually signals that the format is wrong for the process.

Which product families fit inside one

Suitable products are small, assembled or fabricated from bought-in materials, and valuable enough that modest output supports the overhead. Additive processes, computer-controlled machining, electronics assembly, personalised or configured goods and short-life items that suffer from long shipping all sit comfortably here. What does not fit is bulk material handling, heavy process utilities, long curing or drying stages, or a controlled-air envelope over a substantial area. Demand suiting the format is variable and local: many small orders, frequent variants, short promised lead times. Value density matters too: a product cheap enough that freight from a large plant is trivial removes most of the reason to make it locally at all.

Automation replaces the volume argument

A conventional plant justifies automation through throughput; a microfactory justifies it through headcount. With few people on site, machines must run under light supervision, tolerate unattended periods and recover from minor faults without an operator. That pushes investment toward reliable handling, vision-based verification and remote monitoring rather than raw speed. Growth is horizontal — clone the unit elsewhere — and the economics only hold if the cell design is genuinely repeatable rather than a bespoke arrangement the founders assembled once and never documented. Documenting the cell as a replicable package — layout, equipment list, programmes, procedures — is what turns a successful first site into a credible second one.

Small teams, wide skills, thin cover

Everyone does several jobs, which is efficient until somebody leaves. Quality, maintenance, planning and safety are held by people who also run production, so the documented system carries more weight than usual: written procedures, calibration records and change control are what survive turnover. Characteristic failures are key-person dependency, deferred maintenance because nobody has an uninterrupted day for it, and a quality function with no independence from output pressure. External contracts for calibration, servicing and specialist audit usually cost less than the equivalent headcount. Cross-training deliberately, and rotating who signs off what, is cheaper insurance than learning during a resignation that only one person could release product.

Supply and systems for a plant with no warehouse

Storage is measured in square metres, so material arrives often and in small quantities, and a delivery failure halts production within a day. That requires suppliers willing to hold stock on your behalf or ship frequently, and it favours standard, widely available inputs over bespoke items with long lead times. Purchasing operates at small scale and pays for it, though aggregating across sites recovers some ground. Systems should be light and hosted rather than heavy and local: order intake, a simple production record, machine monitoring visible remotely, and traceability adequate for the sector, all running without a dedicated administrator.

Frequently asked questions

Is a microfactory cheaper per unit than a conventional plant?

Rarely on direct conversion cost, and that is not usually the argument. Unit cost runs higher because fixed overhead spreads over less output and purchasing volumes are small. The gains sit elsewhere: less capital at risk, quicker commissioning, lower distribution cost when the site is near customers, faster response to shifts in demand, and the option to stop without stranding a large asset. Compare total delivered cost and the cost of being wrong, not conversion cost alone.

How do we cover maintenance with only a few staff?

Design for it rather than staffing for it. Choose equipment with accessible service points and well-supported spares, keep a documented preventive schedule with tasks short enough to fit between runs, hold critical spares on site rather than relying on next-day availability, and contract specialist servicing for anything beyond routine. Remote diagnostics help, provided the network access is secured properly. The situation to avoid is a plant where every machine differs and no supplier can attend quickly.

Can a small automated unit satisfy a regulated customer's audit?

Yes, though the burden per unit of output is heavy. Auditors look for a defined quality system, competent people, calibrated equipment, controlled documents and traceable records — none of which depends on floor area. What small sites struggle with is independence of roles and depth of cover during absence. Naming deputies, using an external auditor for internal audit, and keeping records genuinely current are the practical answers to those particular questions.