Can a plant run a cell and a general shop side by side?

Yes, and it is usually the realistic outcome. The stable family moves into a cell while the irregular tail stays in the functional area, which preserves the ability to quote unusual work. Two things need deciding at the boundary: which work is allowed into the cell, and who arbitrates when a job could go either way. Without a rule, the cell absorbs awkward jobs because it has a reputation for delivering, and its flow advantage disappears within a few months.

What if one expensive machine cannot be dedicated to a single family?

That is the usual constraint, and there are three common responses. The cell can be built around the operations that can be dedicated, with the shared machine treated as an outside operation with a scheduled slot. The shared machine can be scheduled to serve the cell at fixed times so waiting becomes predictable rather than open-ended. Or the operation can be reviewed to see whether a smaller, cheaper machine inside the cell would cover the family's requirement adequately.

How do I know whether a layout change is worth attempting?

Measure how much of your current flow time is actual processing and how much is waiting or moving. In a functional layout the processing share is often a small fraction, and that gap is the size of the prize. Then check whether a family exists that is large and durable enough to justify committing machines to it. If the flow time is already close to processing time, or no stable family emerges from the routing analysis, the disruption will not repay itself.

Job shop or cells: laying out a plant around machines or around parts

Two plants can hold identical equipment and behave nothing alike, because layout decides how far work travels and how long it waits. Grouping machines by process keeps every routing possible and lets expensive equipment serve everything, at the price of queues between departments. Grouping machines around a part family gives short predictable flow, at the price of committing equipment to a family that must actually exist. The decision follows the stability of your part mix.

Comparison criteria

Criteria are stated explicitly and neither option is declared a winner: which one fits depends on the constraint that binds hardest in your operation.

Criterion	Job shop: machines grouped by process	Cells: machines grouped around a part family
How work moves	Between departments in transfer batches, joining a queue at each one, so distance travelled and time waiting both accumulate across the routing.	From station to station within a defined area, often one or a few pieces at a time, so movement is short and the next operation is visible.
Routing flexibility	Almost unlimited. Any part can visit any machine in any order, which is what lets the shop quote work it has never seen before.	Constrained to what the cell contains. Parts needing an operation the cell lacks must leave it, and every excursion erodes the benefit.
Work in progress and flow time	High, because each queue holds work that has been started but is waiting, and total time in the shop is dominated by waiting rather than cutting.	Low, because a cell holds only what it is working on and a little around it, so flow time approaches the sum of the operations.
Predictability of a delivery date	Weak. Time through the shop depends on what else is queued, so the same part can take very different times in different weeks.	Strong. A cell with a known load produces repeatable flow times, which makes promising and expediting much less of an art.
Setup behaviour	Every job is a fresh setup on whichever machine is free, and similar jobs are only grouped when a planner notices the opportunity.	Setups within a family resemble each other, so tooling stays close, fixtures repeat, and setup reduction work compounds across the parts in the cell.
What it asks of operators	Depth on one process, which suits apprenticeship structures and lets a skilled operator handle unusual work on demanding equipment.	Breadth across several machines, plus responsibility for the cell's output as a whole, which takes training time and a pay structure that recognises it.
Utilisation of expensive equipment	High, because one costly machine can serve the entire plant and is kept loaded by work from everywhere.	Lower, since dedicating a machine to one family accepts idle time in exchange for flow, and a shared costly machine may have to sit outside the cell.
Scheduling effort	Considerable. Every job needs sequencing at every department, and the interactions between departments defeat simple rules.	Modest. The cell is scheduled as a unit against its own load, and internal sequencing can often be left to the people in it.

Choose Job shop: machines grouped by process when

Incoming work is one-off or unpredictable and arrives as enquiries rather than schedules
A small number of costly machines must be available to every part you make
Routings differ so much between jobs that no family accounts for a meaningful share of hours
Your commercial proposition is the ability to take on work nobody else will quote

Choose Cells: machines grouped around a part family when

A recognisable family of parts accounts for a large share of your production hours
Parts within that family follow similar routings through similar equipment
You are losing orders on delivery reliability rather than on price
Operators can be trained and paid to run several machines within one area

Finding the family is the analytical work that decides the outcome

Cells fail when the family is asserted rather than discovered. The honest method starts from routings and volumes: group parts by the sequence of operations they actually require, count the hours each group consumes, and see whether any group is large and stable enough to justify dedicated equipment. Frequently the answer is that one family covers a large share of hours while a long tail covers the rest, which points to a cell alongside a retained general shop rather than a full conversion. The analysis is unglamorous and repeatedly skipped, and the resulting cell then spends its life sending work out to other departments.

Utilisation and flow time pull in opposite directions

A functional layout maximises machine loading, which looks efficient on a utilisation report and produces long, variable queues. A cell accepts idle machines in order to keep work moving, which looks wasteful on the same report while delivering shorter and steadier flow. Neither measure is dishonest; they answer different questions. The trap is running a cell and then judging it on utilisation, because the cell will be dismantled by its own reporting. Decide before the layout changes what the plant is being optimised for — throughput of a stable family, or the ability to absorb any job — and then measure the thing you chose.

Converting a layout is a people project wearing overalls

Moving machines is the visible part and usually the easiest. The difficult part is that a departmental structure carries supervision, skills grading, pay progression and identity, and cells cut across all of it. An operator who has spent a career on one machine type is being asked to run three, take responsibility for output rather than for a workstation, and report to someone who used to run a different department. Plants that treat the change as a materials-handling exercise get the machines moved and the behaviour unchanged. Deal with supervision structure, training time and pay before the first machine is unbolted.

Frequently asked questions

Can a plant run a cell and a general shop side by side?: Yes, and it is usually the realistic outcome. The stable family moves into a cell while the irregular tail stays in the functional area, which preserves the ability to quote unusual work. Two things need deciding at the boundary: which work is allowed into the cell, and who arbitrates when a job could go either way. Without a rule, the cell absorbs awkward jobs because it has a reputation for delivering, and its flow advantage disappears within a few months.
What if one expensive machine cannot be dedicated to a single family?: That is the usual constraint, and there are three common responses. The cell can be built around the operations that can be dedicated, with the shared machine treated as an outside operation with a scheduled slot. The shared machine can be scheduled to serve the cell at fixed times so waiting becomes predictable rather than open-ended. Or the operation can be reviewed to see whether a smaller, cheaper machine inside the cell would cover the family's requirement adequately.
How do I know whether a layout change is worth attempting?: Measure how much of your current flow time is actual processing and how much is waiting or moving. In a functional layout the processing share is often a small fraction, and that gap is the size of the prize. Then check whether a family exists that is large and durable enough to justify committing machines to it. If the flow time is already close to processing time, or no stable family emerges from the routing analysis, the disruption will not repay itself.

Data limitations

Manufacturing figures are operator-supplied inputs, not market data. GeoBusinessIQ holds no factory costs, production volumes, yields, cycle times, tooling prices or capacity data and does not estimate them — every result reflects only the figures you enter.
No manufacturer, supplier, vendor or factory is recommended, rated or ranked anywhere in this cluster, and no directory of them is published. Selection material describes how to run your own assessment; the assessment itself remains yours.

Explore the graph

Across GeoBusinessIQ

Sources

NIST Manufacturing Extension Partnership — NIST MEP (accessed 2026-08-30)
Covers: A public programme supporting small and medium manufacturers with operational, quality and technology adoption practice.
Does not cover: Results attributable to any specific manufacturer, or improvement figures transferable to another plant.
Why it matters: Cited for the operational practice it publishes for smaller manufacturers, not for benchmarks or outcome claims.
Review cadence: annual
United Nations Industrial Development Organization — UNIDO (accessed 2026-08-30)
Covers: Industrial development analysis, industrial statistics methodology, and manufacturing capability programmes across member states.
Does not cover: Company-level data, factory costs, supplier information, or real-time production statistics.
Why it matters: The United Nations agency for industrial development; used for structural framing of how manufacturing sectors develop, never for point figures.
Review cadence: annual
International Labour Organization — ILO (accessed 2026-08-30)
Covers: International labour standards, occupational safety and health conventions, and working-conditions research.
Does not cover: National enforcement practice, wage data for a given plant, or employment terms in a specific contract.
Why it matters: The UN agency setting international labour standards; cited for the framework behind factory labour and safety obligations.
Review cadence: annual

Educational and operational information only — not legal, engineering, safety, customs, tax, or financial advice. Requirements vary by jurisdiction, product, process, and contract; confirm with the relevant authority or a qualified professional before acting.

Last updated: 2026-08-30