How to Evaluate Pouch Filling Accuracy

An accuracy number means little without its conditions. A figure with no product, pouch, speed or sample size can't be compared with another. This guide shows how to define accuracy, run a test and agree acceptance criteria before a trial. We don't publish accuracy figures for any machine, and you shouldn't carry a figure over from a different product or machine type.

Define Weight or Volume

First decide what you control: the weight in each pouch or the volume. Don't mix them. A liquid dosed by volume changes weight with density and temperature. A powder dosed by auger can fill a set volume and still weigh differently as it aerates or settles.

Write the target in the unit you'll check, such as net weight in grams. Say whether the target includes the pouch, and which scale checks it. If your label claims are by weight, test by weight. If you have a volume spec, say how you'll verify it. Measuring volume in a flexible pouch is harder than weighing it.

Product and Operating Conditions

A result belongs to its conditions. Record these with it:

• product name, batch, temperature, density or bulk density, viscosity or particle size;
• pouch type, size and material;
• machine model, dosing method and setup;
• running speed in pouches per minute;
• feed condition, such as hopper level or tank pressure;
• operator, date and scale.

Speed needs care. Speed with an empty machine isn't output with your product. Accuracy usually changes with speed, so a slow-speed result doesn't carry over to a fast run. Test at the speed you plan to run. Our liquid and paste and powder pages cover how product behavior affects dosing.

Sampling and Measurement

Plan the sample before the test. Decide how many pouches to check and when in the run to take them. Consecutive pouches show short-term variation. Pouches spread across the run show drift. A common mistake is weighing only the first ten pouches, which misses drift entirely.

Use a scale suited to the target weight, and record its resolution. A coarse scale can't show small variation. Weigh the empty pouch or use a tare method. Let the scale settle, and keep drafts and vibration away.

A simple record sheet looks like this:

Add a header block for the conditions above and keep the sheet with the trial report.

Bias and Variation

Two different things hide inside "accuracy". Bias is how far the average net weight sits from the target. Variation, or spread, is how much single pouches differ from each other.

For example, with made-up numbers that are not a LINEVO result: say the target is 100 g. Line A fills 98, 102, 99, 101 and 100 g. The mean is exactly 100 g, but pouches swing by 2 g either way. Line B fills 104, 104, 105, 104 and 103 g. The mean is 104 g, but every pouch is within 1 g of the next. Line A has no bias and wide spread. Line B has 4 g of bias and tight spread. A single "accuracy" figure would hide that difference.

The fixes differ too. Bias often goes away by changing the set point. Spread usually comes from the product, the feeding and the dosing hardware, so moving the target won't cure it.

Calculate both from your records: the mean against the target, and the range or standard deviation for spread. Report them separately.

Underfill and Overfill

Decide before the trial what happens to pouches outside the limit. Underfill can breach label claims. Overfill gives product away. The two limits aren't always symmetric.

Agree on:

• the lower and upper limits, and whether they apply per pouch or to the average;
• whether out-of-limit pouches are rejected, by a checkweigher or by hand;
• what happens to rejects, and whether product is recovered;
• how many out-of-limit pouches a sample may contain, if any.

Automatic rejection is an inspection item and belongs in the quote scope. See what affects price.

Agree on Acceptance Criteria

Write the criteria before the trial and sign them off with the supplier. A usable set names:

A result for one material and configuration doesn't carry to another, and a short trial doesn't show long-term performance. Our sample testing page works the same way: agree the method first, and the record states product, packaging, configuration, settings and observations.

If you have a trial in mind, send us your target and your own method and we'll go through them with you. The selection guide shows where accuracy fits in the larger brief.

Frequently Asked Questions

Can I compare accuracy claims from two suppliers?

Only if product, pouch, speed, sample size and method match. Otherwise the numbers describe different tests. Ask each supplier for the conditions, and test both under yours if they differ.

How many pouches should I weigh?

There's no single right number. More pouches give a better estimate of bias and spread, and a longer run shows drift. Agree the number in advance and weigh enough that one odd pouch doesn't decide the outcome.

Do I need a checkweigher?

If out-of-limit pouches must not reach customers, yes, an inspection step helps. It catches errors but doesn't prevent them. Your limits and what you do with rejects decide whether you need one.

Common Pitfalls in Accuracy Testing

Some habits make results look better or worse than they are. Weighing only the first pouches misses drift. A coarse scale hides variation. Changing the set point mid-test without a note ruins the record. Testing a clean, free-flowing sample when production material is sticky or lumpy gives a result that won't hold.

Don't mix weight and volume checks either. A pouch that looks full isn't evidence of correct net weight.

Keep records: conditions, raw weights, calculated bias and spread, and any adjustments. If you repeat the test later, the old record is your reference. Share the template with the supplier before the trial so both sides know what's measured.