Lean Fundamentals · A Visual Explainer  

One Piece
at a Time

Two identical production lines get the same job: ten pieces, three steps. One works in batches. The other moves one piece at a time. The difference is bigger than it looks.

Scroll

Walk through almost any factory — or office — and you will find work moving in piles. Cut a hundred parts, then fold a hundred parts. Approve a stack of invoices, then enter a stack of invoices. Batching feels efficient: you set up once and crank.

Lean practitioners call the alternative one-piece flow: finish a piece completely and pass it on before starting the next. To see why it wins, we built two lines and gave them a race. Scroll to run it.

The setupHere's the job: ten pieces, three steps — cut, fold, pack. Every step takes one minute per piece, on both lines. The amber batch line goes first. Its rule: move work only when the whole batch is ready.
Batch · minutes 0–10Batching feels productive. The cutting station works flat out for ten straight minutes. But look downstream: nothing is getting any closer to the customer.
Batch · minutes 10–20Only when all ten are cut does folding begin. A pile of half-finished work piles up between every station — inventory you've paid for but can't ship.
Batch · minutes 20–30Packing finally starts at minute 20. The first sellable piece appears at minute 21. The last ships at minute 30.
Flow · minutes 0–4Now the blue flow line: same stations, same one minute per step. But each piece moves the moment a station finishes it. The first piece ships at minute 3 — eighteen minutes earlier.
Flow · minutes 4–12From minute 3 on, a finished piece comes off the line every single minute. All three stations work at once. The whole job is done in 12 minutes — with zero extra effort.
The race · minutes 0–12Head to head. By minute 12, the flow line has shipped all ten pieces. The batch line has shipped none.
The race · minutes 12–30The batch line ships its first piece at minute 21 and needs nine more minutes to finish. Same people. Same machines. Same work. The only thing that changed is how the work moves.

The scoreboard

Both lines did exactly 30 minutes of hands-on work. Only the movement changed.

ResultBatchFlow
First piece in the customer's hands21min3min7× sooner
Entire job complete30min12min60% less time
Half-finished pieces sitting in piles (peak)90no piles at all

Speed is only half the story. Now run the race again — but this time the fold station is set up wrong, and every fold it makes is defective. Nobody on either line knows yet. You, however, get to watch.

The hidden defectSame job, same lines — but the fold station is now miscalibrated. You can see the problem coming; the lines can't. There's no inspection until packing. The batch line goes first.
Batch · minutes 0–13Cutting goes perfectly — the broken station hasn't touched anything yet. Then folding begins at minute 10, and the first ruined pieces drop quietly into the pile.
Batch · minutes 13–21One after another, all ten pieces are folded wrong. Between fold and pack, bad work just accumulates — nobody downstream has touched a single piece yet.
Batch · minute 21Packing picks up the first piece — and finds the defect. Too late: the entire batch is ruined, and the cause is 11 minutes cold. Scrap all ten and start over.
Flow · minutes 0–3Now the flow line, same broken station. Piece 1 is folded wrong at minute 2 and lands at packing one minute later — where the defect is caught on the very first piece.
Flow · minutes 3–5The line stops — on purpose. The fold setup is corrected in two minutes, while the evidence is still warm. The piece caught mid-fold is simply redone.
Flow · minutes 5–14The line restarts and runs clean. Final tally: one defect instead of ten — and even after stopping to fix the problem, the job still finishes by minute 14.

Batch line

ALL 10 SCRAPPED · 0 SHIPPED

By the time the first bad piece reaches packing, all ten have already been folded wrong. The entire batch is scrap or rework — and the evidence of what went wrong is 11 minutes cold.

Flow line

1 SCRAPPED · 10 SHIPPED GOOD

The very first piece exposes the problem at minute 3, while the trail is fresh. One piece is affected; the station is fixed before piece two is folded, and the job still finishes by minute 14. Flow doesn't just move faster — it learns faster.

Piles don't just cost time — they cost square feet. Every buffer in the batch run needs racks to hold it, pallets to move it, and aisles wide enough for the forklift that feeds it. Shrink the batch, and the piles disappear; the stations can pull within arm's reach of each other — the compact cell layout lean factories are known for.

Try it yourself

Batch size10
1 · one-piece flow10 · full batch
Floor needed175sq ft
Space freed0sq ft
Pieces waiting20

Top-down view, 1 grid square = 1 sq ft. Drag the slider: as the batch shrinks, the storage racks and handling aisle disappear and the stations pull together.

Less floor space, less handling, less to track — and a layout where a problem at one step is an arm's length from the next.

Nothing about the work itself changed in this race — not the people, the machines, or the minutes of effort. What changed was the batch size, and with it four things every operation cares about.

Lead time collapses. Customers feel minute 3, not minute 21. In a real value stream, batching is why a product with 30 minutes of actual work can take six weeks to deliver.

Cash stops hiding in piles. Every half-finished piece in a buffer is money you've spent but can't sell yet. Flow keeps work-in-process — and the floor space, handling, and tracking it demands — near zero.

Problems surface immediately. A batch buries a defect under nine identical copies. Flow puts it in your hands within minutes, while the cause can still be found and fixed.

Space comes back. The buffers and aisles a batch needs are pure overhead. Pull the piles out and the work folds into a compact cell, freeing floor you can use for the next product instead of the last one's leftovers.

None of this means batching is always irrational. Long changeovers, ovens that cure 50 parts at once, shipments priced by the pallet — these push real operations toward batches. The lean response isn't to accept that, but to attack the reason: shrink changeover times, right-size the equipment, and keep driving batch sizes down toward one.

And the lesson travels well beyond the factory. Emails answered in a Friday pile, invoices approved monthly, software released quarterly — all batches, all with the same hidden cost. Wherever work waits in a stack, minute 21 is hiding inside it.

Good to know

Frequently asked

What is one-piece flow?
Moving one piece completely through each step before starting the next, instead of processing a whole batch at one step before the batch moves on.
Why is one-piece flow faster if the work is the same?
The hands-on time is identical; flow just overlaps the steps (all stations work at once) and stops parking finished work in buffers, so the first piece — and the whole job — finishes far sooner.
Doesn’t batching save on setups and changeovers?
It can, which is why long changeovers, shared ovens, and pallet-priced shipping push toward batches. The lean answer is to attack the cause (SMED/quick changeover, right-sized equipment) and drive batch size toward one, not to accept large batches.
How does flow improve quality?
A defect in a batch hides under every later copy and isn't caught until inspection; in flow the very next step catches it within minutes, while the cause can still be found and fixed.
MS
Matthew Savas

Founder of Kaizumi, an AI-powered Lean training platform. More about Matthew →

Updated June 9, 2026 · Drafted with AI assistance and reviewed by Matthew Savas for accuracy.