Back to Blog
Corporate Solutions

Quality Control Software: Inspections and Sampling

What should quality control software do? Incoming, in process and final inspection plans, tolerances and critical characteristics, sampling tables, measurement records, accept or reject decisions and supplier quality reports.

Kalite KontrolKalite YönetimiÜretimKurumsal Yazılım

Quality control software records whether incoming material, work in progress and finished goods meet defined specifications, following an inspection plan, and makes an accept or reject decision from that record. The short answer: the plan defines in advance what is measured, with which instrument, on how many samples and to what tolerance; the inspector only enters the measurement and the decision follows the rule. That way the same lot does not get a different result depending on who inspected it.

Why can quality control not run on paper forms?

In many factories inspection still runs on printed forms: the inspector measures, writes the value, signs, and the form goes into a binder. The trouble starts with an audit or a customer complaint. Answers to "how many lots from this supplier were rejected in March?" or "is this dimension drifting toward its upper limit over the last three months?" are scattered across binders. A paper form also does not stop a bad lot from entering production; hours can pass between the reject decision and physically separating the lot in the warehouse.

The value of quality control is not in individual measurements but in linking them to a decision and to stock movements. If a rejected lot is not locked in the system, the inspection has not really happened.

Core modules of quality control software

  • Inspection plan: per product or material, the characteristics to measure, lower and upper tolerance, instrument and whether the characteristic is critical.
  • Incoming inspection: at goods receipt the lot is automatically set to "awaiting inspection" and cannot be used until a result exists.
  • In process inspection: measurement reminders at given operations and intervals (for example 5 parts every 2 hours).
  • Final inspection: a check before shipping; a lot that fails cannot be added to a shipment.
  • Sampling table: sample size and acceptance number by lot size.
  • Decision and quarantine: a reject locks the lot, and using or scrapping it requires an authorised approval.
  • Reporting: reject rate per supplier, trends per characteristic, inspection lead time.

How are sampling and the accept decision handled?

Measuring every part is usually too expensive, so samples are taken. The sample size and the number of defective parts a lot may contain and still pass come from a sampling table; the tables widely used in industry give values by lot size and chosen inspection level. In the software that table is a setting, and the inspector does not pick the sample size by hand. If a critical characteristic (for example a safety related dimension) is out of tolerance on even one sample, the lot is rejected outright. The example below shows the logic; the numbers in the table are illustrative only, real values must come from your own standard or customer specification.

type Characteristic = { id: string; min: number; max: number; critical: boolean };
type Measurement = { characteristicId: string; value: number };

// Sample size comes from a plan table chosen by lot size, never typed by hand.
const PLAN = [
  { maxLot: 50, sample: 5, accept: 0 },
  { maxLot: 500, sample: 20, accept: 1 },
  { maxLot: 3200, sample: 50, accept: 2 },
];

function planFor(lotSize: number) {
  return PLAN.find((p) => lotSize <= p.maxLot) ?? PLAN[PLAN.length - 1];
}

function decide(lotSize: number, chars: Characteristic[], samples: Measurement[][]) {
  const plan = planFor(lotSize);
  if (samples.length < plan.sample) return { result: 'incomplete', plan };
  let defective = 0;
  for (const piece of samples) {
    const bad = piece.filter((m) => {
      const c = chars.find((x) => x.id === m.characteristicId)!;
      return m.value < c.min || m.value > c.max;
    });
    if (bad.some((m) => chars.find((c) => c.id === m.characteristicId)!.critical)) {
      return { result: 'rejected', reason: 'critical characteristic out of tolerance', plan };
    }
    if (bad.length > 0) defective++;
  }
  return { result: defective <= plan.accept ? 'accepted' : 'rejected', defective, plan };
}

Two details matter. First, no decision is made with an incomplete sample; the result stays "incomplete". Second, the plan row used for the decision is stored with the record. Even if the table changes later, you can see which rule each past decision followed.

Where does measurement data come from?

The simplest route is manual entry on a tablet or PC; the screen shows only the fields from that product's plan and turns an out of tolerance value red immediately. Reading data straight from digital calipers, micrometers or scales removes typing errors but needs integration per instrument model. The instrument itself must be trustworthy too: a measurement taken with an instrument whose calibration has expired should not be accepted. Linking this check to calibration tracking software answers the classic audit question "which instrument took this measurement, and was it valid that day?" in one click.

What happens after a reject?

A rejected lot has three paths: return to the supplier, rework or scrap. Sometimes the customer or quality manager grants a conditional acceptance (a deviation approval); that decision must always be recorded with a reason and the approver. Repeated rejects call for root cause analysis. At that point the quality control software opens a record in the nonconformance and corrective action process, which we described in detail in our nonconformance and CAPA software article. Even if the two are separate products, they should share the same lot and supplier identifiers.

Off the shelf quality software or custom development?

Packaged quality management suites usually handle inspection plans, sampling and reporting well. The hard part is integration: inspection must talk to goods receipt, production work orders and shipping in real time. If your current ERP or warehouse system cannot enforce those locks, or you want data straight from instruments, a custom layer means fewer compromises. A good start is to digitise the inspection plans of your three most rejected products and link the result to stock movements; it shows the value of the project quickly.

Conclusion

Good quality control software defines in advance what is measured and how, takes the sample size from a table, rejects a lot on a single critical failure, turns the reject into a stock lock and shows trends per supplier. To build inspection into your production and warehouse processes, explore our corporate solutions page or describe your needs and request a quote.

Let's Build Your Project

Get a free consultation for your website, mobile app, or corporate software project.

Get a Free QuoteExplore our Corporate Solutions service