Key takeaways
Short answer: One-piece flow moves a single unit between steps with no batching — the lean ideal for lead time and fast defect detection. Small-batch moves a handful of units, a pragmatic compromise when changeover times or process constraints make true one-piece flow infeasible. Both deliver most of the benefit over large batches; the question is rarely flow-or-batch but how small you can get the batch, and what is stopping you going smaller. See also batch vs flow production.
One-piece flow moves a single unit through the sequence — make one, pass one. WIP between steps is essentially one unit, lead time collapses to near the sum of the processing times, and a defect is caught at the very next operation. It is the theoretical ideal of lean flow, but it demands very short changeovers and well-balanced steps to be viable.
Small-batch moves a few units at a time — five, ten — rather than one. It captures most of the lead-time and defect-detection benefit of one-piece flow while tolerating longer changeovers and minor step imbalance. For many real lines it is the pragmatic destination, not a failure to reach one-piece flow.
A line runs batches of 200 with a nine-day lead time and finds defects only at final inspection. Moving to one-piece flow would be ideal, but the heat-treat step processes a fixed tray of twelve, so true single-piece flow is physically impossible there. The team settles on small batches of twelve — the tray size — everywhere. Lead time drops from nine days to under a day, and a defect is now caught within twelve units instead of 200. They got most of the one-piece-flow benefit by matching the batch to the real constraint, rather than chasing an unreachable ideal.
The smaller the batch, the lower the WIP, the shorter the lead time, and the faster defects surface. The benefit is steep at first — going from 200 to 20 is transformative — and flattens as you approach one. That is why small-batch often captures 80-90% of the gain: the last step from a small batch to a single piece adds less than the first big reductions did, and may cost disproportionate changeover effort.
1. Treating it as flow-or-batch. The real question is how small, and what is blocking smaller.
2. Chasing one-piece flow past the point of diminishing returns. Small-batch may capture nearly all the benefit at far less cost.
3. Reducing batch size without cutting changeover. More setups overwhelm capacity.
4. Ignoring fixed-quantity process steps. Match the batch to the binding constraint.
Smaller batches expose losses fast — a stop at one step quickly starves the next instead of being hidden behind WIP — so OEE problems surface immediately. The trade-off is more changeovers, which is why batch reduction and changeover reduction (SMED) go hand in hand to protect OEE Availability.
See how Fabrico captures this automatically on your lines — explore OEE for manufacturing or book a demo.
It is the ideal, but small-batch often captures most of the benefit when constraints make one-piece flow infeasible.
As small as your changeover times and process constraints allow — often matched to a fixed-quantity step.
Long changeovers, fixed-quantity processes, and handling designed for large loads.
More changeovers cost Availability, so pair batch reduction with SMED.
The benefit flattens near one, while changeover cost can rise sharply — small-batch is often the sweet spot.
Programați o întâlnire individuală cu experții noștri sau înscrieți-vă direct în planul nostru gratuit.
Nu este nevoie de card de credit!
Login
Blog
