Automatic line or two machines: when is each more profitable?

Why this choice is often judged wrong

The mistake usually starts with price. A manager sees that two separate machines cost less up front than a line, and the conclusion seems obvious. But what the shop buys is not metal and electronics — it buys part output per shift, operator load and predictability of the run.

If you count only purchase price, all the unpleasant things disappear from the calculation: the part must be removed after the first operation, inspected, carried, re-mounted and aligned, and then the size must be checked. On paper this takes a couple of minutes. In a real shift those minutes add up to a noticeable takt loss.

Another common error is taking the machine’s rated cycle time as the actual shop figure. It’s a convenient reference, but it works only in ideal conditions. In a shift there are always tool dressing pauses, first-part checks, fixture changes, small stops, operator waits and gaps between cycles. If you compare a line and two machines only by catalogue numbers, the cheaper option almost always looks better. For serial orders that conclusion quickly falls apart.

People often confuse a one-off job with a stable series. If a batch is small and the part mix changes constantly, two separate machines do offer more freedom: they’re easier to reconfigure and the risk of an expensive changeover is lower. But when the order repeats for months and volume is steady, the calculation must change. Then it matters not only how many minutes of cutting you get, but also the number of part transfers, operator loading, changeover time and downtime.

A simple example: the first operation takes 2 minutes, the second 2.5. In a meeting it looks like two machines will handle the task easily. Then you find out the operator spends an additional 40–60 seconds removing the part, carrying it, blowing it off, re-mounting and checking. At 300 parts per shift that’s several hours of non-cutting work.

A fair comparison doesn’t start with price or a datasheet. First calculate the real time a part spends in a shift: machining, handover between operations, inspection, waiting, changeover and human involvement. Only then will it be clear where two machines are truly advantageous and where a line pays off by smoothing output.

What to compare at the start

Comparison only makes sense under equal conditions. Use the same part, the same material, the same tolerance and the same production volume. If for the line you calculate a 10,000-piece batch while for two machines you look at a mixed flow of different parts, the numbers mean nothing.

The schedule must match too: the same number of shifts, the same shift length, the same breaks and the same maintenance plan. Otherwise one option will look better simply because of different assumptions.

A frequent omission is manual transfer between operations. For two separate machines this is a mistake. The operator removes the part, carries it, re-clamps, checks the datum and only then starts the next cycle. All that time must be fully counted.

A line follows a different logic. The model includes blank feeding, part transfer between operations and finished part removal. If the line performs these automatically, those seconds cannot be thrown out of the calculation. Often the difference in serial-order economics rests exactly on these seconds.

Next, look at four things: takt, changeover time, operator loading and downtime losses. Takt is better calculated from actual good-part output per shift rather than catalogue data. Changeover should be considered in full: tool change, adjustments, trial parts and size checks. For operators it’s important not only how many staff are present, but how much time they actually spend at the equipment.

Be strict about downtime. For two machines count how much time is lost waiting for an operator, re-clamping and small stops between cycles. For a line ask what happens if one unit, feeder or transporter stops. On serial orders even 10–15 minutes of such loss per shift noticeably changes the result.

If you calculate honestly for a single part, the picture clarifies fast. You’ll see where two machines provide needed flexibility and where a line removes unnecessary movements, reduces manual work and keeps output steadier.

How to calculate takt without complex math

First you don’t need a machine — you need a production plan. If a shift must produce 80 parts, calculate from that number, not from the machine’s rated speed.

Convert the shift into minutes and keep only the time when work actually happens. An eight-hour shift has 480 minutes, but some of that is setup, first-part checks, tool changes, cleaning and normal short pauses.

The logic is simple: take the planned output, calculate available minutes per shift, subtract all recurring losses, and divide the remaining time by required output.

The formula is simple:

Такт = доступное время за смену / нужный выпуск

A short example. The shift lasts 480 minutes. 35 minutes go to setup, 20 minutes to inspection and 25 minutes to pauses and small stops. That leaves 400 minutes of net time. If the plan is 80 parts, takt is 5 minutes per part.

Then don’t look at average time for the whole cell but at each operation separately. If one operation takes 6 minutes while your takt is 5, the plan already fails. It doesn’t matter whether you use a line or two machines — the bottleneck remains.

If the first operation is 3.8 minutes, the second 4.6 minutes, and transfer between them takes 20–30 seconds, output looks realistic. But if the second operation sometimes stretches to 5.5 minutes because of measurement or tool change, the buffer almost disappears. On paper the math looks good, in a shift you will fall behind.

It’s better to keep a small reserve. If calculation gives a 5-minute takt, it’s safer to target operation times around 4.2–4.5 minutes rather than 4.9. On long runs short stops quickly eat that margin.

This kind of calculation is sobering: it shows immediately where time is enough and where it isn’t.

Where two separate machines lose time

When comparing a line and two machines people often look only at price and rated cycle time. In practice hours are lost not in cutting metal but between operations. For one part these losses look trivial; on a series they add up to whole shifts.

The first loss comes at the handover. The operator removes the blank from the first machine, carries it to the second and sees the latter is still busy. The part waits. The operator waits or switches to another task and then comes back. On paper both machines appear loaded; in the real flow a pause has already appeared between them.

Inter-operation inspection breaks the flow further. After the first operation a part often needs measuring, datum verification, deburring or just checking dimensions. This is common practice, especially on serial orders with tight tolerances. Every such inspection rips the flow. While the check is happening the second machine receives no new part.

Changeovers hit harder than they seem. With two operations on two machines any batch change shifts the schedule at two points. The first machine may already be ready for the new part while the second is not — or vice versa. The result is more work-in-progress and a lost rhythm.

The queue between machines often hides the bottleneck. From the outside the flow may look fine: there are parts on a cart, people are busy, equipment runs. But if the queue grows one operation no longer keeps up with the other. Signs are simple: the second machine constantly has a pile of parts, the operator spends all their time carrying them, after a changeover one machine stands idle longer than the other, and size checks upset the usual shift rhythm.

For turning operations this is especially noticeable on recurring batches. If a series runs every month, hidden pauses stop being trivial — they become permanent losses in takt, operator load and delivery predictability.

When a line reduces operator load

When choosing equipment people often focus on price and forget about labor. Yet operators often determine whether a cell can maintain steady output or will constantly chase the plan.

On serial work an operator with two separate machines is rarely occupied only with cutting. They pick up a blank, carry the part between operations, wait for a cycle to finish, change containers, verify batches and re-clamp. Each action seems small. Across a shift they consume a lot of time and energy.

A line removes part of that manual work. One operator can oversee several stations: loading, machining, unloading and simple status checks. Instead of constant walking they act on signals and scheduled checks.

This changes the shift rhythm. People walk less with parts; output is steadier because the part’s route is fixed, and the supervisor notices a falling takt sooner.

In practice the difference is clearest on long runs. If a cell machines the same part for many days, a line keeps pace more calmly than two separate machines with manual transfers. Shift performance depends less on who’s faster carrying blanks and who frees the second machine earlier.

But a line has a condition: it reduces manual load only if the team acts decisively on faults and changeovers. If a sensor errors or a tool drifts, you can’t sort it out ad hoc. You need clear procedures: who stops the line, who checks the unit, who performs the first quality check after restart.

So a line doesn’t just save operator steps — it shifts some load from manual actions to shift discipline. For serial production this is usually a plus. For a shop where one day you run one part and the next something completely different, this format isn’t always suitable.

How downtime risk changes the calculation

Downtime risk can easily break a nice-looking economics on paper. If you only look at machine price and takt, you might pick an option that’s great in ideal conditions but starts losing money at the first serious failure.

With two separate machines one unit failing doesn’t always stop the entire cell. One machine can finish its current operation, process already started parts or at least keep people busy. Output drops but not always to zero.

A line is tougher. If a feeder, conveyor, loader or another flow-critical unit fails, the line often stops completely. On serial orders this is felt strongly: even a few hours of stoppage quickly wreck the shift plan.

Don’t count one failure, count its cost in lost parts. Answer four questions: how often does each option stop, how long do stoppages typically last, how many parts per hour do you lose, and can you continue partial work?

If one machine stops four times a month for an hour while the other keeps working, losses will be noticeably lower than one line failure of four hours that stops the entire flow. Frequency matters, but stoppage duration often hits harder.

There is a flip side. A well-built line with good service can have fewer unplanned stops than two machines that alternately demand attention. Then its advantage grows. So compare real statistics for the nodes that fail most often, not horror stories about total shutdown.

Service changes the outcome strongly. If the supplier responds fast, does commissioning, trains staff and pre-agrees a list of critical spare parts, downtime drops from days to hours. For such projects this is part of the economics, not a minor detail.

EAST CNC covers this with a full cycle: from selection and supply to commissioning and service. For serial cells it’s useful to discuss this before purchase because recovery speed can matter more than attractive catalog numbers.

A simple example: a line gives 18% more output than two separate machines. But if it stops twice a quarter for half a day while the two-machine cell loses only a third of a shift with similar issues, the advantage quickly disappears. Over a long run what counts is not rated performance but the number of parts you actually ship per month.

Example of a serial order

Take a typical order for a supplier in automotive: 3,000 identical parts per month, each requiring two consecutive turning operations. A similar scenario appears in components for construction machinery. On paper, two separate machines look straightforward: the first does operation one, the second finishes the size.

With two machines the operator removes the blank after the first operation, carries it to the second, waits for a suitable moment to load it and re-clamps. Machining itself might take, say, 48 and 42 seconds. But to those numbers you almost always add 15–20 seconds for removal, carrying and re-clamping. A few more seconds go to small pauses: the second machine is busy, the operator is distracted by a measurement, a queue has built between operations. In the end a finished part comes out not every 48 seconds but closer to 70.

On a line the same two operations run in one flow. Transfer between positions happens without manual carrying and the pause between operations almost vanishes. If operation one is 48 seconds and operation two 42, the line’s real takt is usually close to the longer operation plus a short service pause — for example 52–55 seconds. For a 3,000-piece batch the difference is visible: about 58 hours versus 46 hours of pure cutting time. The line doesn’t cut faster by itself; it removes walking, waiting and extra touches.

There’s a second effect. On two machines the operator constantly moves parts. On a line they monitor blank feeding, tool wear and equipment signals. The same volume means less manual routine and more even operator loading.

But this example changes fast if the batch is small. If instead of one 3,000-piece run the shop receives ten 300-piece batches with different sizes, the line’s advantage may disappear. Every changeover stops the whole flow. With two machines you can set up operations in sequence and sometimes avoid a full stop. So on a stable long series a line usually wins on takt; with frequent part changes two machines are often more convenient and cheaper to operate.

Errors that distort the calculation

The line vs two-machines debate is often decided by catalogue numbers. That’s convenient but too weak for a real choice. You don’t buy spindle rpm or axis speeds — you buy parts per shift.

Two similar machines can have near-identical catalogue specs while the eight-hour result differs. One option loses time on loading, measuring and part transfer; the other holds a steady takt and produces more finished parts despite humbler datasheet numbers.

People often don’t count time to the first good part. That’s not trivial. The setter installs a tool, runs a trial, measures the part, makes adjustments and re-checks. If the run isn’t long that 30–60 minutes matters.

Another mistake is ignoring small losses that feel like background noise. In reality they eat a shift in pieces: tool change, bringing blanks, removing finished parts, waiting for containers, the operator moving between machines. On paper a cycle might be 90 seconds, but in practice every twentieth part faces an extra pause.

A poor calculation is easy to spot. It takes the best day when everything is tuned and the tool is fresh, ignores time to the first good part, doesn’t count operator minutes spent on feeding and inspection, and forgets planned service stops.

A best day says little. You need an average week, or better, data across several shifts. Then you see how often people stopped the process, how long changeovers took and how often the rhythm broke.

Service is the same story. A planned 15-minute stop seems trivial until you multiply it by a month. On serial orders downtime is counted in parts, not abstract percentages. If you remove these stops from the model both options will look better than reality.

A proper calculation is straightforward: average output per shift, time to first good part, real stoppages, operator work minutes and internal logistics losses. That’s enough to make a calm decision without self-deception.

Quick checklist before deciding

If a series lasts only a month and then the cell switches to other parts, a line and two machines will give different economics. A line likes repeatability; two machines tolerate frequent part changes better.

First look not at catalogue performance but at the real rhythm of orders. If the same volume holds for at least six months the calculation usually becomes fairer. Over that time a line can show its advantage in steady takt and lower operator involvement.

Before deciding answer a few direct questions: what monthly volume can you sustain without dips for six months in a row? How often per week do you retool the cell for another part? Who actually works on the cell each shift, not only who’s on the roster? What happens to shipments if a unit stops for two hours? And how much spare capacity remains if the order suddenly grows by 10–15%?

There’s a nuance with changeovers. On paper two machines look more flexible. In practice the operator loses time for tool fitting, adjustment, first good part and repeated checks. If such shifts happen several times a week the small losses accumulate quickly.

Staffing also breaks many calculations. By day a cell may be carried by a skilled operator while at night a novice or one person runs several machines. In that case a line with a clear sequence sometimes yields steadier output even if it’s more expensive.

Also check a two-hour stoppage scenario. If one machine in a pair stops the other often continues making semifinished parts and builds a queue. A line may stop entirely but recovers takt faster after restart. There’s no universal answer — look at which setup loses fewer parts per shift.

A good sign for a line: the series is long, changeovers are rare, shift staffing is light and output targets are strict. If volume fluctuates, parts change frequently and the shop relies on flexibility, two machines are usually safer.

What to do next

Put the calculation into a single table. Enter parts per shift, takt, changeover time, number of operators, manual steps, downtime and scrap losses. Then collect the same numbers for a month. You’ll quickly see where the difference between two machines and a line exists in reality and where it exists only on paper.

Don’t count only the ideal shift. Take two scenarios: an average day and a bad day. On a bad day the operator may be distracted, a tool may need earlier replacement and a changeover can easily stretch by 15–20 minutes. This comparison often reveals which option is preferable.

Then check four things: how many parts the cell actually produces per shift (not from catalogue); how many times per month people stop work for changeovers; how many operators are needed in normal and peak loads; and how much output falls if one unit stops for two hours.

If the supplier talks only about performance, the picture will be incomplete. Ask about commissioning, training, service, consumables and recovery time after a failure. Sometimes a more expensive option wins not because of speed but because it suffers smaller losses when something goes wrong.

One more often-missed point: today the order may be 800 parts per month but in six months it may nearly double. At that point two separate machines often start losing time on manual work and changeovers, while a line keeps output steadier.

For serial projects in Kazakhstan and CIS countries this calculation is useful to discuss with those who handle not just supply but also start-up. EAST CNC is the official representative of Taizhou Eastern CNC Technology Co., Ltd. in Kazakhstan. The company supplies CNC lathes, machining centers and automatic production lines and helps with selection, commissioning and service. When you compare not only purchase price but also long-term output stability, this approach usually gives a more sober picture.

If numbers for both options are nearly equal, don’t choose from the catalogue. Pick the scenario that’s easier to start, easier to maintain and simpler to keep running month after month.