OEE for Egg Grading and Packing: Measuring a Grader and Its Lanes

An egg grader is not one machine with one outfeed. It is one infeed that fans out into many packing lanes, and every lane can hold the rest back.

That is why a plain OEE setup, built for a filler and a packer in a row, gives odd numbers in a packing station.

This page shows where to measure, what rate to use, how to treat downgrades, and what to check in software before you buy. We sell Fabrico, so test our claims like any other vendor's.

The line, and where it stops

Most packing stations follow the same order. Eggs come from a loader that lifts them off trays, or straight from the houses on conveyors.

Then come candling and detection (cracks, dirt, leakers, and on many lines blood), weighing, and the grader that sends each egg to a lane by weight and quality.

The lanes do the packing: denester, fill, closer, printer or labeller. A real grader has many more lanes than the five drawn below, often several per grade.

After the lanes come case packing and palletising.

The losses by station are typical, not measured. Use the table as a first reason list and cut what your line does not have.

Measure at the grader, explain with the lanes

Pick one point for the OEE number. In a packing station that point is the grader, counted in eggs.

The grader sets the pace for the whole station, and its rate is the one on the nameplate: eggs per hour, or cases per hour.

Lanes are different. A lane that is idle because no eggs of that grade are coming is not losing anything.

So lane OEE, averaged, tells you little. What you need from each lane is its state on the same clock as the grader: running, blocked, starved, in changeover.

Then when the grader drops to half speed at 10:40, the report can show that the L lane ran out of packs at 10:39.

This is the linked-line problem in a wide form. Our page on finding the real constraint on linked packaging lines covers the logic for lines in a row.

Set the ideal rate once

Use the grader's rated eggs per hour as the ideal rate, and keep it the same for all flocks and all order mixes.

It is tempting to lower the rate for a hard day: many small eggs, weak shells, a lane short of capacity.

Do not. A lower target hides the loss you want to see.

If you report in cases, convert with the eggs per case you really pack. Store the count in eggs, because packs of 6, 10, 12 and 30 all leave the same grader.

Count eggs, not chain speed. A grader can run at full speed with empty positions on the infeed, and only an egg count shows that gap.

Set the planned time

Planned time, as used on this page, is the time you meant to grade. Breaks and the standard end-of-shift wash-down are planned stops: take them out, at a fixed length you agree once.

If the wash-down runs over that standard, the overrun is downtime inside planned time. A clean in the middle of a run, after a broken egg on the rollers, is an unplanned stop with its own reason.

Waiting for eggs is harder. With house conveyors, the hens set the supply, so agree up front which hours count as planned.

With a loader, "no eggs at the infeed" inside planned time is a real stop. Log it as starved, not as a grader fault.

Size mix against orders: a loss with no broken part

The flock decides the size mix. A young flock lays more small eggs, and the mix moves as the flock ages.

Orders do not follow the flock. If today's orders need large eggs and the houses send mostly medium, the M lanes fill and the L lanes wait.

A lane can only pack so fast. Some graders can hold or divert a short surplus, but when one grade keeps flooding its lanes, the grader has to slow down to what those lanes can take, or stop while another lane is changed over to help.

Nothing is broken, so this loss often gets no reason at all. It shows up as "ran slow" and nobody owns it.

Give it a code of its own, such as "size mix, lane full". It belongs to planning and egg supply, not to the operators or the technicians.

Downgrades: split what arrived from what the line made

This is the point where egg OEE goes wrong most often. A grader exists to find cracks, dirts and other seconds.

An egg that arrived cracked and was sent to the seconds lane is the grader doing its job. It is input quality, and the line did not lose it.

An egg that arrived sound and was broken at a transfer, a drop-off or the closer is a line loss. Only this belongs in the OEE quality factor.

How do you split them in practice? Use a mass balance per batch: eggs counted in at the grader, minus eggs counted out in packs, seconds included.

The gap is what the line broke or lost after the count point. Eggs broken before it, at the loader, never reach the count, so track those apart.

One honest limit: the crack detector sits after the loader, so it also sees cracks the loader made. It cannot tell them from farm cracks.

To catch that, compare the crack rate of the same flock across days and loaders, and check by hand a sample of trays before the loader when the rate jumps.

Worked example: one shift at the grader

These are invented round numbers for one shift. They describe no real plant, and the line loss is set high on purpose so it shows in the figure.

• Ideal rate: 120,000 eggs an hour, which is 2,000 eggs a minute.
• Planned time: 400 minutes (a 480 minute shift, less 30 minutes of breaks and a 50 minute wash-down).
• Stops: 40 minutes. Run time is 360 minutes.
• Eggs handled: 612,000.
• Broken or lost on the line: 12,000. Good count is 600,000.
• Of the good count: 30,000 arrived cracked or dirty and went to seconds. 570,000 were packed as first quality.

Availability is 360 of 400 minutes, or 90%. Performance is 612,000 of the 720,000 eggs the run time allowed, or 85%.

Quality is 600,000 of 612,000, or 98.0%. Multiply the three and you get 75.0%.

Now check it the short way. 600,000 good eggs at 2,000 a minute is 300 minutes of ideal running.

300 of 400 planned minutes is 75.0%. The two agree.

See what happens if the 30,000 farm downgrades are booked as a quality loss. Good count falls to 570,000, quality to 93.1%, and OEE to 71.25%.

That is 3.75 points taken from the line for eggs it never damaged. The full method is in our guide to OEE calculation.

Pack and label changeovers, by customer

A packing station changes packs and labels many times a day: another customer, another pack size, another date code.

Most of these happen on one lane while the grader keeps running. So record two things: the changeover time on the lane, and what it did to the grader.

If eggs of that grade had another lane to go to, the grader lost nothing. If not, the grader slowed or stopped, and that is the real cost of the changeover.

A flock or supplier change is its own event. Many stations run the grader empty between batches to keep them apart.

Log that gap as a changeover with a standard length. It still counts against availability, and anything over the standard should show as an overrun.

Short jams matter here too: a denester that misfeeds a few times an hour never makes the stop sheet. Our page on micro-stops shows why they add up.

Where the counts come from

The grader's own controller knows the egg count per lane and per grade. Ask the grader supplier how that data can be read out, and at what cost.

If it cannot be read, a run signal and a pulse count from the grader still give stops and speed.

At the lanes and the case packer, mixed pack formats make a fixed sensor hard to trust.

This is a place where a camera that counts parts can help when there is no usable PLC count. Test it on your own pack formats before you rely on it.

Whatever counts, store the pack format with it, so packs turn back into eggs for the mass balance.

What to check in OEE software for an egg packing station

Run these in the demo with every vendor, and with us. Bring one real shift with a lane blockage and a flock change in it.

1. One timeline for grader and lanes. Ask to see a grader slowdown next to the lane that caused it. If lanes live on separate pages, you will do that match by hand.
2. Blocked and starved as states. A lane with no eggs coming, and a grader held back by a full lane, should not look like breakdowns.
3. Speed loss with a reason. Ask how "running at 70% of rated speed for 25 minutes" gets a cause such as size mix or no packs.
4. Count in eggs, show in packs and cases. Ask how a lane that packs 10s in the morning and 30 trays in the afternoon is turned back into eggs.
5. Downgrades apart from quality loss. Ask if seconds can be reported by flock or supplier without lowering the OEE quality factor.
6. Wash-down as a planned stop with a standard. Ask what the report shows when a 50 minute clean takes 70.
7. Changeovers per lane. Ask for the count and length of pack and label changes by lane and by customer.
8. Your data out. Ask for an export of one shift, stops and counts, in a file your planner can open.

For a wider comparison of tools, see our review of OEE software.

Where Fabrico fits

Fabrico is a cloud-based OEE and light MES platform. It runs on the web, iOS and Android.

• Stops: they come from the machine signal, through the PLC or a Fabrico IoT device. The operator adds the stop reason on a tablet or phone.
• OEE and downtime: availability, performance and quality, by line and by shift. Micro-stops are detected from the machine signal.
• Losses in money: losses can be shown as a cost, not only as minutes.
• Counting: the AI camera is an add-on that counts parts. Ask us to test it on your packs or cases. Read more on camera-based OEE.
• Your data: REST API and webhooks, and any list exports to Excel or CSV.
• Follow-up: a stop reason can be set up to open a maintenance task.
• Not in Fabrico today: production planning and scheduling. Matching orders to the size mix stays with your planning tool.

How Fabrico would model your grader and its lanes depends on the signals your grader can give. Ask us to show it on your layout, and run the eight checks above on us.

Book a demo or Contact us.

What people ask next

Should we measure OEE per lane or for the grader?

For the grader, in eggs. Lanes wait on the size mix, so lane OEE swings for reasons the lane does not control.

Keep lane stops and changeovers as a log that explains the grader number.

Do cracked and dirty eggs count against OEE quality?

Only the ones the line made. Eggs that arrived cracked or dirty are input quality: report them by flock or supplier.

Eggs broken between the infeed and the pack are the OEE quality loss.

Is wash-down downtime?

The standard daily wash-down is a planned stop at an agreed length, outside the time you meant to grade. The overrun is downtime, and so is a clean in the middle of a run after a broken egg.

What is a good OEE for an egg grader?

We know of no public figure worth trusting for egg graders, so we give none. Compare your own line week to week with fixed rules, and read OEE for food and beverage for the wider picture.

How do we count eggs if the grader controller will not share its data?

Start with a run signal and a pulse count from the grader for stops and speed. Count packs or cases at the outfeeds, with a sensor or a camera, and convert them to eggs by pack format.

Do we need a full MES for a packing station?

Often not for this job. Stops, speed, counts and reasons are the core of a light MES.

Traceability by flock and order planning are separate needs, so check where they live today.

Want to run these checks on your grader and lanes? Book a demo or Contact us.