Six Big Losses in Manufacturing: Where Your Line Loses Performance and How to Check It

Explaining the six big losses in manufacturing gives you a clear way to talk about why a line misses the production plan: it stops, slows down, produces defects, or takes too long to return to stable operation. This article will help you see which losses to measure first, how to connect them with OEE, and when machine data becomes more useful than another spreadsheet.

On the shop floor, you usually see the symptoms: the plan was not completed, operators are trying to catch up, maintenance is handling another issue, and quality comes back with comments on the batch. The cause is often less obvious.

Time leaks away in pieces. A few minutes of adjustment after a changeover, a short stop at the feeder, a slower cycle because the machine is running more cautiously after a failure, or a series of defects after startup.

One event may not look serious. A series of those events, repeated every day, starts to affect unit cost, deadlines, and available capacity.

The six-loss model, rooted in TPM and Lean, helps you start a more focused conversation. Instead of a broad statement that we have a performance problem, you get a specific category: availability, speed, or quality. That shortens the path from reporting to decision-making.

OEE is based on three questions that do not require theory:

Those questions create a practical map of losses.

This breakdown is useful because it narrows the analysis right away. With low availability, the team looks for sources of downtime. With lower performance, it checks cycle time and short stops. With quality issues, it looks at the conditions in which defects occur.

A breakdown is the easiest loss to notice. The machine stops, the plan shifts, and someone calls maintenance. A report is usually created, but the quality of that report determines whether anything can be improved.

A note that says “machine failure” does not help much. The team should know whether the issue was mechanical, electrical, material-related, connected with the setup, or related to operation. Without that, every analysis starts with guessing.

Two short breakdowns may cost less than one stop in the middle of an important batch. So the number of events matters less than their impact on production flow.

Good questions for the report:

When these answers are available right away, maintenance can move from reacting to failures toward preventing repeat breakdowns.

A changeover is not a problem by itself. The real issue is the gap between the standard and the actual return to production.

That is why it is better to measure more than the tool change time. Measure the whole period from the last good part of the previous batch to the first good part of the next batch. This window includes adjustments, trials, waiting for quality approval, and startup rejects.

Simple math shows the scale quickly. Three changeovers per day, each 10 minutes longer than the standard, create 30 minutes of loss on product changes. Over 22 working days, that is 11 hours per month on one line. In a plant with several lines, that gap can affect additional batches, overtime, or delays.

A good changeover standard has to be measured in real production time and compared across shifts.

A minor stop is short. The operator adjusts the material, resets a sensor, clears a jam, restarts the machine, and gets back to work. It often leaves no trace in the report, but the loss remains in the result.

If a line stops for 90 seconds 20 times during a shift, it loses 30 minutes. With two shifts per day, that becomes one hour of lost production time. Over a month, it adds up to more than 20 hours that do not show up in reports focused only on longer downtime.

This is one reason manual reporting reaches its limits quickly. An operator will not record every short stop, especially while trying to keep production running. Machine data shows the frequency of events, and the operator’s comment adds the cause. Together, they give a fuller picture than a list of breakdowns alone.

Slow running can be deceptive. The machine is working, people are busy, and product is coming off the line. At first glance, the process looks normal. Only a comparison between actual cycle time and target cycle time shows how much performance is being lost through speed.

The reasons may be technical, material-related, or organizational: worn components, more cautious settings after a failure, variable raw material quality, lack of a work standard, or intentionally slowing the machine during a more difficult batch.

The most useful comparison looks at speed by product, machine, and shift. If the same SKU on the same line gives different results depending on the shift, check settings and ways of working. If the slowdown is tied to a specific material, the discussion should move toward process parameters and supplier quality.

The biggest risk is getting used to it. Slower running can start to look like the new standard, even though it continues to reduce margin.

A defect is material, but not only material. It also includes machine time, operator time, energy, inspection, rework, replanning, and the risk of a customer complaint.

The number of defects shows the scale, but it does not show the source. You need to connect the reject with the product, batch, machine, parameters, and moment in the process. Without that, the team sees the result but does not know what to change.

The most useful question is: under what conditions do most defects occur?

If the answer requires searching through spreadsheets, paper reports, and messages from quality, the reaction will come too late. Data should lead to the batch, setting, shift, material, or time of the event.

After the start of a shift, a stop, or a changeover, the line needs time to reach stable parameters. During this period, the risk of trial parts, corrections, and rejects increases.

This loss often gets mixed with quality or changeover losses, so it is worth separating. It shows whether the process is ready to start: whether initial settings are known, material is ready, and data from previous runs is available to the operator.

Start with a simple measurement: how many minutes and how many parts are needed to reach stable production after startup. When the result differs across shifts, the issue probably lies in the startup standard or in dependence on the experience of specific people.

One line, one process, or one loss group is a good starting point. A broad initiative may sound ambitious, but it often ends in a table that no one uses for decisions.

A practical sequence:

In many plants, the hardest part is data. ERP stores the plan, product, and batch. The production system shows the process flow. Spreadsheets collect comments that are hard to compare later. When this information does not connect into one view, loss analysis takes too long and the team returns to intuition.

The clearest signal is when the result is known, but the cause is still debated. The plan was missed, OEE is down, and the discussion moves between breakdowns, material, operators, changeovers, and quality.

Take a closer look if, in your plant:

In this situation, production monitoring and OEE analysis based on machine data shorten the path from event to decision. We help you see where the line is losing time, which type of loss keeps coming back most often, and whether corrective actions are actually improving performance.

The same missed plan can have different sources. A breakdown requires a different response than a minor stop, and slower cycle time leads to different actions than defects after startup.

Start by identifying where you lose the most: availability, performance, or quality. Then choose one loss, assign an owner, and check the change week by week.

Data helps when it leads to decisions. Everything else is only reporting.

They are a breakdown of the most common sources of lost production efficiency. They include breakdowns, changeovers and adjustments, minor stops, reduced speed, quality defects, and startup losses.

The most common types are availability losses, performance losses, and quality losses. Availability is reduced by downtime and changeovers, performance drops because of minor stops and slower cycles, and quality is reduced by defects, rework, and scrap.

Choose one line, measure OEE, break the result into loss types, and identify the three largest sources of the problem. Plan corrective actions after that.

Downtime lasts longer and usually appears in reports. A minor stop is short and often handled immediately by the operator. Over a month, these short interruptions can take many hours of line time.

They are enough to start. Later, their limits become clear: short events disappear, causes are entered from memory, and comparing data takes too long. Machine data gives a more accurate picture of line performance.

When you know the overall result but cannot quickly point to the causes of losses. A system helps connect downtime, speed, quality, and operator comments into one analysis.