Overall equipment effectiveness tells you how much of an asset's potential output you are capturing, but a single percentage does not tell you why you are losing the rest. The Six Big Losses are the diagnostic framework that answers that question by naming the six distinct ways equipment falls short of ideal, and by mapping each one to the availability, performance, or quality factor it degrades. This guide walks through all six loss categories, shows which OEE factor each attacks, and explains how runtime and reject data expose them so a team can act on the right problem.
The Six Big Losses (OEE) in one line: The Six Big Losses are the standard categorization of the ways equipment loses productivity, used to diagnose what is dragging down OEE. They are breakdowns and setup/adjustments (which reduce availability), small stops and reduced speed (which reduce performance), and startup rejects and production rejects (which reduce quality). Each loss maps to one of OEE's three factors, so decomposing a low OEE into these six categories points directly at whether the problem is downtime, slow running, or defects, and where to focus improvement.
The first two losses are availability losses, meaning they stop the machine from running when it was scheduled to. Breakdowns are unplanned stops from equipment failure, from a tripped motor to a jammed line, and they are the losses people notice most because the machine is visibly down. Setup and adjustment losses are the time spent changing over from one product or run to the next and dialing the process back in, including calibration and first-piece checks. Both eat into the time the equipment could have been producing, which is why they show up as lost availability.
The next two are performance losses, meaning the machine is running but not at full speed. Small stops, sometimes called minor stops or idling, are brief interruptions that clear quickly, such as a misfeed, a sensor false-trip, or a momentary blockage, individually trivial but collectively significant. Reduced speed loss is the machine running below its ideal rate for any reason, from a worn component to a deliberately conservative setpoint. These do not register as downtime because the equipment never fully stops, which is precisely what makes them easy to overlook.
The final two are quality losses, meaning the machine produced output that cannot be sold as first-quality. Startup rejects are the defective units made while a process stabilizes after a start or changeover, before it reaches steady conditions. Production rejects are defects generated during otherwise stable running. Both represent capacity that was consumed to make something that must be scrapped or reworked, so they subtract from the good output the equipment is credited with.
OEE is the product of three factors, and the Six Big Losses partition cleanly into pairs against them. Availability is reduced by breakdowns and by setup and adjustment, so those two losses together explain the gap between the time the equipment was scheduled to run and the time it actually ran. When availability is low, the diagnosis lives in these two buckets, and separating them matters because chronic breakdowns and slow changeovers call for very different fixes.
Performance is reduced by small stops and by reduced speed, which together explain why a machine that was running still did not produce as many units as its ideal cycle time would predict. A low performance factor with few recorded stops usually means speed loss or a swarm of micro-stops too brief to have been logged, while frequent short stoppages point at the small-stops category. Because neither shows up as classic downtime, they are the losses most often hidden inside a disappointing OEE.
Quality is reduced by startup rejects and production rejects, which together explain the gap between total units produced and good units. A quality factor that sags right after every changeover implicates startup rejects and points at process stabilization, while a steady trickle of defects during normal running implicates production rejects and points at the process or materials. Because each of the six losses lands on exactly one factor, computing the three factors first and then attributing each one's shortfall to its two candidate losses turns a vague low score into a specific, actionable list.
Diagnosing the Six Big Losses requires data that most facilities already generate but rarely organize this way. A run/stop state tag tells you when the machine was down; pairing each stop with a reason code separates breakdowns from setups, giving you the two availability losses directly. Without reason codes, all downtime collapses into one undifferentiated block and the diagnosis stalls, so the discipline of coding stops is what makes availability losses actionable rather than merely visible.
Performance losses need rate data rather than state data. Comparing a live production count or flow rate against the ideal rate exposes reduced speed as a running-but-slow condition, and detecting the brief drops and recoveries in that count reveals small stops that were too short to trip a downtime record. This is where a historian earns its keep: micro-stops and creeping speed loss only become visible when you can look at the count rate at fine time resolution over a long window, not at a shift summary.
A cloud SCADA such as Merobix is well suited to feeding this analysis because it historizes the same tags used to run the equipment: state, reason codes, production counters, and reject or off-spec counts read from the field over Modbus, DNP3, OPC UA, or MQTT. With good and total counts logged, the two quality losses fall out of the difference between them, and startup rejects can be isolated by tying reject counts to the periods just after a start. Because the platform already collects and time-stamps these signals continuously, the Six Big Losses can be computed from live operations rather than reconstructed by hand, letting a team see which of the six is actually costing them the most.
They are the six standard ways equipment loses productivity: breakdowns and setup/adjustment losses, which reduce availability; small stops and reduced speed, which reduce performance; and startup rejects and production rejects, which reduce quality. Each loss maps to one of OEE's three factors, so categorizing losses this way shows whether an OEE shortfall comes from downtime, slow running, or defects.
The six losses pair up against the three OEE factors. Breakdowns and setups are availability losses, small stops and reduced speed are performance losses, and startup and production rejects are quality losses. Because each loss belongs to exactly one factor, you can compute the three factors and then attribute each factor's shortfall to its two candidate losses, turning a low OEE number into a specific diagnosis.
Both are performance losses in which the machine is running rather than fully stopped, so they never register as classic downtime. Small stops are brief enough to clear before a downtime record is created, and reduced speed is simply the machine running below its ideal rate without any stop at all. Detecting them requires comparing the actual production count or rate against the ideal rate at fine time resolution, which is why they usually stay hidden until a historian is used to look for them.
This page references the protocol specifications published by the organizations below. Editions, product capabilities, and documentation change over time - confirm current requirements and specifications directly with the source.
Last reviewed: July 27, 2026. Merobix is not affiliated with, endorsed by, or sponsored by these organizations; their names are used only to identify the standards and products discussed.
Merobix reads your field devices into a cloud SCADA - the real thing behind these terms, live in days from any browser.