The Six Big Losses in OEE — How to Cut Each One

By Daniel Crawford on May 30, 2026

six-big-losses-oee-cut-each-one

Overall Equipment Effectiveness is only as useful as your ability to act on it. A single OEE number tells you how much productive capacity you lost — the Six Big Losses framework tells you exactly where it went. Developed as part of Total Productive Maintenance (TPM) by Seiichi Nakajima, the framework gives every minute of lost production a category, a cause, and a path to reduction. This guide breaks down all six losses, shows what each one costs in real terms, and explains the specific interventions that cut them.

Auto-Categorized Loss Tracking

iFactory classifies every downtime and speed loss minute into the correct OEE category automatically — no manual logging, no guessing at the end of the shift.

How the Six Big Losses Map to OEE

OEE is calculated as Availability × Performance × Quality. Each of the Six Big Losses attacks exactly one of these three components — which is why the framework and the metric were designed together. You cannot improve OEE without reducing losses; you cannot reduce losses without knowing which category they belong to.

Availability
Planned Run Time − Downtime
Loss 1 — Unplanned Stops
Loss 2 — Planned Stops
×
Performance
Actual Output ÷ Theoretical Output
Loss 3 — Slow Cycles
Loss 4 — Small Stops
×
Quality
Good Parts ÷ Total Parts Produced
Loss 5 — Production Rejects
Loss 6 — Startup Rejects
85% World-class OEE target
60% Typical baseline for manufacturers new to OEE tracking
25 pts Gap between typical and world-class — entirely explained by the Six Big Losses

The Six Big Losses — Detailed Breakdown

Each loss has a precise definition, a real-world manufacturing example, a cost model, and a specific set of countermeasures. Treating all six with the same response is why most OEE improvement programs stall.

01
Availability Loss

Unplanned Equipment Failures

High Impact

Any unscheduled stoppage caused by equipment failure, mechanical breakdown, tooling failure, or process fault that requires intervention before production can resume. This is the most visible loss — and the most expensive per occurrence.

Real Example

CNC machining center hydraulic pump fails mid-shift. Repair takes 2.5 hours. At 480 parts/hr ideal rate, that is 1,200 parts of lost output. At $8.50/part margin, the single event costs $10,200 in unrecovered throughput — before labor and parts.

Overdue or skipped preventive maintenance
No condition monitoring on critical components
Reactive maintenance culture — fix on fail
Poor lubrication or contamination management
Aging equipment with no replacement plan
Planned preventive maintenance at OEM-defined intervals
Vibration and temperature condition monitoring on critical assets
Autonomous maintenance — operators perform daily equipment checks
MTBF tracking to identify chronic failure patterns
02
Availability Loss

Planned Stops & Setup / Changeover

Medium Impact

Production time lost to planned stoppages — changeovers between products, scheduled maintenance windows, tooling changes, and warm-up periods. These are known in advance, which makes them reducible through engineering and process discipline. Many plants accept these as unavoidable; world-class plants treat them as waste.

Real Example

An injection molding cell performs 4 mold changeovers per day averaging 45 minutes each. That is 3 hours of daily planned downtime. If SMED reduces changeover to 20 minutes, the plant recovers 100 minutes of productive time per day — roughly 42 additional shifts per year.

No standardized changeover procedure (SOP)
Internal tasks not converted to external (SMED principle)
Tools and materials not staged before production stop
Excessive adjustment time after setup — first-off not studied
SMED (Single-Minute Exchange of Die) analysis — eliminate internal setup steps
Changeover kits pre-staged at cell; shadow boards for tooling
Quick-release fixturing and standardized tool heights
Video-based changeover analysis to identify non-value-added steps
03
Performance Loss

Slow Cycles & Reduced Speed

High Impact

Any time equipment runs below its ideal or nameplate cycle time. This loss is the hardest to detect because the machine is running — it looks productive. A press running at 85% of ideal rate over an 8-hour shift loses more total output than a 30-minute breakdown, but almost never appears on a downtime report.

Real Example

Stamping press ideal cycle time: 4.2 seconds/part. Actual observed cycle time: 5.0 seconds (worn die, operators compensating). Loss per hour: 48 parts. Over a 2-shift day: 576 parts. At $3.20 margin: $1,843 per day in invisible throughput loss — $460,000 per year.

Operators intentionally running below speed for quality reasons — unaddressed process issue
Worn tooling increasing friction or resistance
Material variation causing process adjustments
Incorrect ideal cycle time established (too aggressive or outdated)
Drive or servo degradation not yet causing alarms
Establish and validate true ideal cycle time from machine capability studies, not estimates
Real-time cycle time monitoring with alarm when drift exceeds 5% of ideal
Tooling condition monitoring — replace on schedule, not on failure
Address root cause when operators reduce speed — do not accept speed reduction as normal
04
Performance Loss

Small Stops & Micro-Stoppages

Medium Impact

Stoppages typically under 5 minutes that are cleared by the operator without a maintenance work order — jams, misfeeds, sensor faults, part blocking a chute. Each event is small enough to be ignored on a shift report. Collectively, they can account for 15–25% of total production losses in high-speed automated lines.

Real Example

Packaging line experiences 18 jams per shift averaging 2.5 minutes each = 45 minutes of lost run time per shift. Because each event is cleared by the operator with no log entry, the shift report shows 95% availability. Actual availability: 84%. The 11-point gap is entirely invisible to management.

Worn guides, worn belts, or misaligned conveyor components
Material dimension variation causing feeding issues
Dirty or misadjusted sensors triggering false faults
No threshold or process for logging sub-5-minute events
Auto-detect all stops via machine signal — no operator logging required
Pareto analysis of micro-stop reason codes — target the top 3 causes
Set 3-strike rule: same micro-stop 3×/shift triggers a maintenance escalation
Incoming material inspection to catch dimensional variation before it reaches the line
05
Quality Loss

Production Rejects & Rework

High Impact

Defective parts or assemblies produced during steady-state production that require rework or are scrapped. Every rejected part consumed raw material, machine time, labor, and energy — then delivered zero revenue. Reworked parts cost the operation two full cycles. This loss directly reduces the Quality component of OEE.

Real Example

Welding cell with 0.8% scrap rate on 5,000 parts/shift = 40 scrapped parts. At $22 material and direct labor cost per part, daily scrap cost is $880. Across 250 production days: $220,000/year from a scrap rate that most supervisors consider acceptable.

Process drift not detected between scheduled inspection points
Incoming material non-conformance passed through receiving inspection
Operator error during complex assembly steps
Tooling wear producing gradual dimensional drift
Environmental variation (temperature, humidity) not controlled
In-process SPC monitoring with control limits — catch drift before rejection
Poka-yoke (error-proofing) at high-defect operations
Defect pareto by reason code — address top defect before moving to second
Tighter incoming inspection AQL for materials linked to top defect codes
06
Quality Loss

Startup & Yield Losses

Lower Impact

Defective parts produced from startup until the process reaches steady-state — after changeovers, shift starts, machine warm-ups, or after a breakdown restart. These parts are often written off as unavoidable startup scrap. In high-mix environments with frequent changeovers, startup losses accumulate into a significant quality drain.

Real Example

Plastic extrusion line discards the first 12 minutes of output after each material changeover (4 changeovers/day) as off-spec material. At 400 lbs/hr throughput and $1.80/lb material cost: 32 lbs scrapped per changeover × 4 × 250 days = 32,000 lbs/year = $57,600 in startup scrap alone.

No documented startup procedure — operators rely on experience
Process parameters not restored to last known-good settings after stop
Insufficient warm-up time before first-part inspection
No first-article check built into changeover procedure
Documented startup checklist with mandatory first-article inspection before production release
Parameter recipe management — restore exact settings from last successful run
Track startup scrap separately from production scrap to make the loss visible
Reduce time-to-stable through process warm-up automation where applicable

Calculating the Cost of Each Loss — A Worked Example

Most manufacturers know their OEE percentage. Few have translated it into a dollar figure by loss category. Here is a calculation for a mid-volume discrete manufacturing cell running a single shift to show what each loss actually costs.

Cell Parameters
Shift Length480 min
Ideal Cycle Time0.5 min/part
Ideal Output960 parts/shift
Part Margin$6.40/part
Current OEE64%
Actual Output614 parts/shift
Loss Category Time Lost / Shift Parts Lost Daily Cost Annual Cost (250 days) OEE Impact
Unplanned Failures 48 min 96 $614 $153,600 −10 pts Avail.
Planned Stops / Changeover 36 min 72 $461 $115,200 −7.5 pts Avail.
Slow Cycles — (speed) 58 $371 $92,800 −6 pts Perf.
Small Stops 24 min 48 $307 $76,800 −5 pts Perf.
Production Rejects — (quality) 38 $243 $60,800 −4 pts Quality
Startup Rejects — (quality) 34 $218 $54,400 −3.5 pts Quality
Total Loss 108 min 346 $2,214 $553,600 −36 pts OEE

Based on illustrative mid-volume cell parameters. Actual figures vary by industry, product mix, and margin structure.

See Your Six Big Losses — Right Now

iFactory connects to your machines, auto-classifies every stop and speed loss into the correct category, and shows you exactly which loss is costing you the most — updated in real time, every shift.

Priority Sequence: Which Loss to Attack First

Not all six losses are equal. The right sequencing depends on your current OEE profile — but the following decision matrix applies to most discrete manufacturing environments starting a Six Big Losses reduction program.

Loss
Typical Share of OEE Gap
Detection Difficulty
Reduction Speed
Start Here If…
1 Unplanned Failures

30–40%
Easy
Slow (3–12 mo)
Availability below 75%
3 Slow Cycles

20–30%
Hard
Fast (days)
Performance below 85%
4 Small Stops

15–25%
Hard
Medium (1–3 mo)
High-speed automated lines
2 Planned Stops

10–20%
Easy
Medium (1–2 mo)
High-mix, frequent changeovers
5 Production Rejects

5–15%
Easy
Medium
Quality below 95%
6 Startup Rejects

3–10%
Easy
Fast (weeks)
High-mix with frequent changeovers

How iFactory Tracks the Six Big Losses

The biggest barrier to Six Big Losses reduction is visibility. Most manufacturers cannot tell you how many minutes of Loss 4 (small stops) occurred on Line 3 last Tuesday. iFactory closes that gap with automatic loss categorization from machine signal data — no manual entry, no shift-end estimates.

1

Automatic Loss Classification

Machine signals are interpreted in real time and every stop, speed deviation, and defect count is assigned to the correct loss category — without operators logging a single event code. Classification rules are configurable per machine type and process.

2

Loss Waterfall Chart per Shift

Shift supervisors see an OEE waterfall — starting from 100% ideal, stepping down through each of the six losses in sequence, landing on actual OEE. The contribution of each loss is quantified in both minutes and percentage points.

3

Loss Pareto Across Time Periods

Quality and production managers can rank losses by time or cost across any date range — this week, this month, by line, by product. The top loss on any line is always visible at a glance, driving prioritization decisions without manual analysis.

4

Maintenance & Quality Workflow Integration

Loss 1 events (unplanned failures) automatically create a maintenance work order. Loss 5 events above a threshold create a non-conformance record. Both are linked back to the originating OEE event for full traceability from loss to corrective action.

5

Dollar-Denominated Loss Reporting

Every loss minute is converted to a dollar figure based on your configured part margin and ideal cycle time — so plant managers and operations leaders can make decisions in business terms, not just OEE percentages.

Frequently Asked Questions

Which of the Six Big Losses is typically the largest?
In most discrete manufacturing environments, unplanned equipment failures (Loss 1) account for the largest single share of OEE losses — typically 30 to 40% of the total OEE gap. However, in high-speed automated lines and packaging operations, small stops (Loss 4) and slow cycles (Loss 3) often collectively exceed the impact of breakdowns because they are chronic, invisible, and never trigger a maintenance response. The answer depends on your production profile — which is exactly why the Six Big Losses framework requires you to measure all six, not just downtime.
How is planned maintenance classified in the Six Big Losses — is it Loss 1 or Loss 2?
Planned maintenance windows are Loss 2 (Planned Stops) when they are scheduled in advance and the equipment is intentionally removed from the production schedule. Unplanned breakdowns that occur during scheduled production time — including failures that occur during a planned maintenance window's overrun — are classified as Loss 1 (Unplanned Failures). The distinction matters because the countermeasures are completely different: Loss 2 is reduced by doing more maintenance work in less time or during non-production windows; Loss 1 is reduced by preventing failures from occurring at all.
What is the difference between Loss 3 (Slow Cycles) and Loss 4 (Small Stops)?
Loss 3 is a continuous speed reduction — the machine is running but slower than its ideal rate, every cycle. Loss 4 is an intermittent complete stop — the machine fully halts for under 5 minutes before resuming. Both reduce Performance. The distinction matters for corrective action: Loss 3 root causes are typically process-related (worn tooling, material variation, intentional operator speed reduction), while Loss 4 root causes are typically mechanical (jams, misfeeds, sensor faults) and benefit from frequency-based pareto analysis.
Do I need to be connected to machine data to use the Six Big Losses framework?
No — you can start with operator-entered data on paper or in a spreadsheet. However, manual entry systematically under-reports micro-stoppages (Loss 4) and slow cycles (Loss 3), which means your OEE number will look better than it actually is, and your improvement efforts will target the wrong losses. Machine connectivity — even a basic digital counter or a relay-based run signal — dramatically improves Loss 3 and Loss 4 detection accuracy and is the fastest path to actionable data.
Stop Estimating. Start Measuring.

iFactory auto-classifies every loss minute across all six categories, converts them to dollars, and routes corrective actions automatically — so your team focuses on fixing losses, not logging them.


Share This Story, Choose Your Platform!