Key takeaways
Short answer: Batch production runs a large lot at one step, then moves the whole lot to the next — simple, but it piles up WIP and delays defect discovery. Flow moves one piece (or a small lot) through the steps continuously, slashing lead time and surfacing quality problems immediately. Flow is the lean ideal, but batch still suits processes with long setups or genuine economies of scale. See also oee for manufacturing.
Batch production processes a large quantity at one step before moving the entire lot onward. It minimises changeovers per unit and feels efficient at each station, but the work spends most of its life waiting in queues between steps.
Flow moves one piece, or a very small lot, continuously through the sequence. WIP stays low, lead time collapses, and a defect is caught at the very next step instead of buried inside a finished lot.
A shop runs three steps in batches of 500. A drilling error starts at unit 50 of a batch but is not discovered until inspection after step three — by then 450 defective parts exist. Convert to one-piece flow and the same error is caught at step two on the very next unit, so the defect count is two, not 450. The lead time also drops from days (waiting in queues) to hours. Same machines, same people — the batch size alone was hiding the defect and stretching the lead time.
Flow exposes problems while they are cheap to fix and collapses the time between making and discovering a defect. It also compresses lead time dramatically, which is the core of lean and the reason most improvement programs push toward smaller and smaller batches.
Processes with very long setups, heat or chemical treatments that are inherently batch, or real economies of scale may justify batching. The goal is the smallest batch the process and economics genuinely allow — not batch for habit's sake.
1. Big batches by default. Habit, not economics, drives most oversized lots.
2. Ignoring the WIP cost. Cash and lead time tied up in queues is invisible on a station efficiency report.
3. Flow without changeover reduction. Small batches need short setups (SMED) to be viable.
4. Measuring station efficiency, not flow. Local efficiency can rise while lead time gets worse.
Flow makes losses visible fast — a stop at one step starves the next, so OEE problems surface immediately instead of being buffered out of sight by WIP. Batch can mask losses behind inventory, which is comfortable and expensive.
Fabrico exposes the starving and blocking that flow reveals and batch hides, so you can see the true cost of your batch size. Book a demo to see flow losses across your line.
Usually for lead time and quality, but not for every process — long setups and true scale economies can justify batch.
Long setups, inherently batch processes, and unbalanced lines.
Yes — a whole lot can be defective before anyone inspects it.
It exposes losses immediately rather than buffering them behind WIP.
Changeover reduction — small batches are only viable with short setups.
Login
Blog
