Measuring OEE in practice

What is OEE?

OEE (Overall Equipment Effectiveness) expresses in a single number how efficiently a manufacturer uses its production assets. It is the product of three factors:

OEE = Availability × Performance × Quality

• Availability: what fraction of scheduled production time was the machine actually ready to run (not stopped by breakdown, changeover, or material shortage).
• Performance: actual throughput rate versus the rated throughput (slow running, minor stops).
• Quality: fraction of output that was good first time (no scrap or rework).

Example

If a machine has availability of 90%, performance of 85%, and quality of 95%:

OEE = 0.90 × 0.85 × 0.95 = 72.7%

This means 72.7% of the theoretically available capacity is actually realised as good product.

What counts as good?

Industry benchmarks:

| OEE | Rating | |-----|--------| | Above 85% | World-class | | 65–85% | Acceptable, improvable | | Below 65% | Significant losses, intervention needed |

Most plants measure 40–60% OEE on first measurement — this is not company-specific weakness, it is the industry average. Reaching the 85% "world-class" level takes sustained Lean/TPM effort.

How is OEE measured in practice?

1. Manual measurement (entry level)

The operator records on paper or in Excel:

• Shift start and end
• Downtime events with reason (breakdown, changeover, material shortage, etc.)
• Units produced and scrap count

Advantage: cheap, quick to start. Disadvantage: data entry errors, late feedback (you find out at end of shift).

2. Automatic data collection (MES level)

The machine PLC automatically signals states (running, stopped, fault). OEE is visible in real time on a dashboard. The operator only records the reason for downtime on a tablet.

This is the level where OEE measurement actually enables interventions — because you see immediately when a machine stops, not 8 hours later in the daily report.

How to improve OEE

Improving OEE is not a quick fix — you need to work on all three factors:

Improving availability:

• Schedule preventive maintenance
• Reduce changeover time (SMED methodology)
• Fast fault-finding process (escalation path, spare parts stock)

Improving performance:

• Review rated cycle times (they may be set too pessimistically)
• Identify causes of slowdowns (worn tooling, bad material, setup error)
• Analyse micro-stops (stoppages under 30 seconds that add up)

Improving quality:

• Categorise scrap causes and run Pareto analysis
• Build in process controls (poka-yoke)
• Feed back results to setup engineers and process technologists

When do you need an MES for OEE measurement?

If you're measuring OEE on one or two machines and the data volume is manageable, Excel can work. But if you need to:

• track 5+ machines in parallel,
• see shift-by-shift breakdowns in real time,
• feed data into your ERP,

…then a MES system pays for itself almost automatically.

Related articles:

• What is an MES and when does it make sense?
• Manufacturing support