There's a line we hear often on plant floors: "Our line runs at over 90% utilization." Yet the same plant misses delivery dates and fills the gap with overtime. Why is a 90%-utilization plant this busy?
The answer is simple: the "utilization" we usually talk about and a machine's real capability are two different things.
The illusion of 90% utilization
Utilization typically measures "time the machine was on." But even while a machine is on:
- It's actually stopped — waiting for material, changing dies, adjusting
- When it runs, it often runs below design speed
- And some of what it produces is scrapped as defects
OEE (Overall Equipment Effectiveness) multiplies these three views together — Availability × Performance × Quality. World-class plants hit 85%; most plants measuring for the first time land around 60%. A plant that believed it ran at 90% is really performing at 60%. Put the other way: that gap is production capacity you can recover without adding a single machine.
Why now — and why Korean manufacturing in particular
OEE applies to any factory, but several forces make it urgent for Korean manufacturers right now.
1. Solving with more people is no longer an option. Hiring shortages and rising labor costs are structural. When output must grow, adding shifts and headcount is increasingly off the table. Two options remain: add equipment, or recover the hidden capacity in what you have. OEE is the map for the second.
2. High-mix production multiplies losses. Korea's core industries — components, materials, food and beverage, electronics assembly — keep moving toward more frequent product changeovers. Frequent changeovers mean die changes, cleaning and adjustment become a major cost axis. What feels like "we're just busy" is a loss structure OEE decomposes into numbers.
3. Paper logs and spreadsheets can't see the losses. Many plants still write production reports by hand at shift end. Accuracy is not the only problem: a five-minute micro-stop repeated dozens of times a day never makes it into the report, and speed losses aren't even a field on the form. Most of OEE's six big losses — breakdowns, setup and adjustment, minor stops, reduced speed, process defects, startup yield — are invisible without automatic collection.
4. Recovery is cheaper than expansion. Raising OEE from 60% to 70% adds roughly 17% capacity on the same equipment. Compare that to the capital cost of a new line. The tighter the margin pressure, the stronger the case for recovering hidden capacity first.
Starting is lighter than you think
If you frame OEE as an enterprise MES project, it never starts. The minimum signals are just three:
- Run/stop — is the machine actually running?
- Production count — how many did it make?
- Good/defect — how many of those are usable?
These signals already exist in most PLCs and sensors. The key is collecting them automatically, without human hands. OEE that depends on manual entry rarely survives three months — it degenerates into paperwork that manufactures numbers.
A realistic sequence
- Automate run/stop, count and defect signals on one critical line.
- Put real-time OEE and a loss Pareto on a shop-floor dashboard — metrics only managers see never change behavior.
- Pick the single largest loss and run the improve → measure → verify cycle.
- Extend the proven pattern to the next line.
Closing
OEE is not a new idea. But with labor shortages, high-mix production and cost pressure converging, for Korean manufacturing today it is not a nice-to-have metric — it is almost the only way to gain capacity without capital expansion.
PlantPulse covers this path on one platform: signal acquisition, real-time OEE calculation and loss-analysis dashboards — proven in production deployments including the real-time OEE system at Lotte Chilsung Beverage. If you're curious what your line's real capability is, start with three signals.