The Six Big Losses in Manufacturing: What They Are, Why Most Go Undetected and How to Eliminate Them
The six big losses in manufacturing are the six categories of production loss identified in Seiichi Nakajima’s Total Productive Maintenance framework that collectively reduce OEE from its theoretical maximum of 100 percent. The framework is well known. What is less well understood is that without automatic, sensor-based monitoring, most manufacturers reliably detect only two of the six — leaving the remaining four to compound silently in the background of every production shift. This guide covers each loss category, which OEE component it affects, why manual systems miss it and how TEEPTRAK automatically detects all six from the first shift of deployment.
The Six Big Losses: A Complete Framework
Availability Loss 1 — Unplanned Downtime
What it is: unexpected equipment failures, mechanical breakdowns, tooling failures, material jams and any unscheduled stop that prevents the machine from running when it should be running.
Which OEE component it affects: Availability. Every minute of unplanned downtime reduces Availability directly.
Why it is usually detected: major stops are visible and disruptive. A machine that has been down for 45 minutes generates a maintenance call, a production escalation and a record in the shift report. This is the best-detected of the six losses in most plants.
What manual systems miss: the cause, precisely. The duration may be logged, but the root cause classification at shift end is reconstructed from memory and less accurate than real-time classification. The compounding effects — diagnosis time, parts retrieval, restart sequence — are often not logged separately from the repair time.
Availability Loss 2 — Planned Stops (Changeovers and Setups)
What it is: product changeovers, scheduled maintenance, operator breaks, cleaning cycles and any planned stop that is part of the production schedule.
Which OEE component it affects: Availability, when counted as downtime against planned production time. Properly configured as planned downtime, changeovers reduce the planned production window rather than appearing as unplanned loss.
Why it is partially detected: major changeovers are planned and recorded. The problem is accuracy: the actual changeover duration versus the standard changeover time is rarely tracked systematically. A changeover that should take 35 minutes but takes 52 minutes has generated 17 minutes of undetected Availability loss that appears as a longer-than-planned downtime event.
What TEEPTRAK adds: precise timestamps from sensor detection of machine state, not operator-reported start and end times. Every changeover duration is measured against the configured standard, and overruns are classified as the cause of the extended stop.
Performance Loss 3 — Minor Stops and Micro-Stops
What it is: brief machine interruptions lasting less than five minutes — material jams cleared by the operator, brief sensor faults, momentary blockages that the operator resolves and resumes. Individually trivial; collectively significant.
Which OEE component it affects: Performance. In OEE calculation, stops under a configurable threshold are classified as minor stops rather than unplanned downtime. They reduce Performance rather than Availability.
Why manual systems miss it: this is the most systematically under-detected of the six losses. A 90-second jam that the operator clears and resumes will not be logged at shift end — it was too brief to note at the time and too numerous to remember accurately. On a line experiencing 15 to 20 micro-stops per shift, manual logs capture approximately 3 to 4. The other 12 to 16 events are permanently absent from the production record.
What TEEPTRAK adds: automatic detection of every state change regardless of duration. A 45-second stop is captured with the same precision as a 45-minute stop. The micro-stop database reveals patterns that are invisible in manual systems — a specific machine, a specific time of day or a specific material lot driving the bulk of the accumulated micro-stop time.
Performance Loss 4 — Reduced Speed
What it is: a machine running below its nominal production rate. No stop occurs. The machine appears to be working. But the Performance component of OEE is degraded by the percentage gap between actual and nominal speed, for every minute the speed deviation persists.
Which OEE component it affects: Performance. A machine running at 85 percent of its nominal rate has a Performance of 85 percent for that period — silently, invisibly, generating throughput losses that accumulate with each shift.
Why manual systems miss it: no stop event occurs, so nothing triggers a log entry. An operator observing the machine from a distance sees it running. Without a system that compares actual cycle time against the configured nominal rate in real time, this loss category does not exist in the production record. It is the most invisible of the six losses.
What TEEPTRAK adds: continuous comparison of actual cycle time against the nominal rate, with automatic flagging of deviations beyond the configured threshold. Speed losses that would never appear in a manual system generate quantified Performance loss records from the first shift of deployment.
Quality Loss 5 — Startup Defects
What it is: scrap and rework generated during production startup and following changeovers, before the process stabilizes. Higher defect rates during the startup period are normal in many manufacturing processes — but if they are not tracked separately from steady-state defects, they are invisible as an improvement target.
Which OEE component it affects: Quality. Every defective part reduces the Quality component, regardless of when during the production run it was produced.
How it is typically tracked: startup defects are often absorbed into general scrap figures. The distinction between startup scrap and steady-state scrap is not made, so the opportunity to reduce startup defects through changeover improvement is not visible in the data.
Quality Loss 6 — Production Defects
What it is: scrap and rework during steady-state production. Parts that fail quality inspection during normal production runs — not during startup.
Which OEE component it affects: Quality. In OEE calculation, only good parts produced on the first pass count toward throughput. Every defective part represents machine time consumed without producing a good unit.
See how TEEPTRAK detects all six big losses automatically
Why Most Manufacturers Only Detect 2 of the 6
The detection rates for the six big losses in plants without automatic monitoring follow a consistent pattern: unplanned downtime and production defects are reliably detected because they are large, disruptive and leave records. Everything else is systematically underdetected.
Planned stop overruns are detected in aggregate but not individually. Minor stops and micro-stops are captured at roughly 20 to 25 percent of their actual frequency. Reduced speed is essentially undetected. Startup defects are merged into general scrap figures. The result is a calculated OEE that understates actual losses by 10 to 20 percentage points — leading to systematic underestimation of improvement potential.
When manufacturers deploy TEEPTRAK for the first time, the discovery that their true OEE is significantly lower than their manually calculated OEE is one of the most common and most important revelations of the first two weeks of data. The gap between perceived OEE and actual OEE is where the fastest improvement opportunities live.
TEEPTRAK: Automatic Detection of All Six Big Losses From 48 Hours
TEEPTRAK plug-and-play IoT sensors install on any machine in 48 hours without PLC modification or production stop, capturing every state change from the first shift of deployment. This means:
Unplanned downtime and planned stops: automatically detected and classified by operators in 30 seconds on a touchscreen interface.
Minor stops and micro-stops: captured automatically regardless of duration, building the micro-stop Pareto that manual systems have never recorded.
Reduced speed: detected by continuous comparison of actual cycle time against the configured nominal rate, quantifying Performance losses that are invisible without real-time monitoring.
Startup and production defects: logged through the operator quality interface, with timestamps that enable separate analysis of startup versus steady-state defect rates.
JEMBA AI: Identifying Which of the Six Is Costing You Most
Pareto analysis of two weeks of TEEPTRAK data identifies which of the six loss categories accounts for the largest share of OEE loss on each production line. JEMBA goes further: it identifies the specific root cause of the top loss categories by processing over 700 production variables simultaneously with 99.7 percent anomaly detection accuracy.
When micro-stops are the top Performance loss, JEMBA identifies whether the frequency correlates with a specific material batch characteristic, a time-of-day pattern or a machine parameter drift. When reduced speed is elevated, JEMBA identifies which process condition is causing the speed reduction. The CI team acts on a directed causal finding in hours rather than completing a manual investigation over weeks.
TEEPTRAK tells you what is happening on your shop floor. JEMBA tells you why it is happening and which of the six big losses to address first for maximum OEE impact.
Results
TEEPTRAK is deployed in more than 450 factories across 30+ countries. Customers average plus 29 OEE percentage points after deployment — driven by systematic elimination of all six loss categories, not just the two that manual systems detect. Hutchinson drove OEE from 42 percent to 75 percent across 40 production lines in 12 countries. Nutriset achieved plus 14 productivity points with payback under one month. Typical payback: 8 to 14 months.
CMMS Integration: Closing the Loop Between Loss Detection and Maintenance
Detecting all six big losses generates full value when the loss data connects to the maintenance management system. TEEPTRAK integrates with major CMMS platforms through open REST APIs. Unplanned downtime events classified in real time trigger automatic CMMS work orders with cause context. Historical stop data populates MTBF calculations for preventive maintenance scheduling. JEMBA root cause findings connect directly to maintenance action — identifying not just that a machine stopped, but the specific condition that caused it and should be corrected.
0 Comments