Skip to main content
In most factories, quality control happens at the end. A batch comes off the line, an inspector checks a sample, and if something is wrong the entire run has to be reviewed, reworked, or scrapped. By then, the machine has been making the same mistake for hours and the cost is already locked in. Bold Factory’s quality controls work differently — they are built into the manufacturing process itself, so defects are caught at the step where they are created, not after the damage is done.

​
Quality as Part of the Process, Not Bolted On After

When you define a recipe in Bold, you can attach quality checks to any process step. The check becomes a required gate — the operator cannot mark the step complete until the check has been recorded. There is no separate quality module to open, no paper form to fill in later, and no risk of the check being skipped because it felt like optional overhead. This approach has three important consequences:
  1. Early detection — a defect found at step 3 costs a fraction of one found after step 10.
  2. Precise traceability — every quality record is stamped with the step, work order, machine, operator, and timestamp, so you always know exactly where and when an issue occurred.
  3. Operator accountability — operators are part of the quality process, not observers of it. When they record a measurement, they are the first line of defence.

​
Types of Quality Checks

Bold supports three check types that cover the vast majority of in-process quality requirements:
A binary confirmation — the condition is either met or it is not. Use this for checks like “Is the surface free of visible scratches?”, “Has the part been deburred?”, or “Is the label applied correctly?”. If the operator answers No, Bold flags the non-conformance and can trigger a supervisor alert.
The operator enters a numeric measurement — a dimension, a temperature, a torque value, a weight — and Bold compares it against the minimum and maximum you have defined in the recipe. If the value falls outside the range, the check fails automatically and the operator is prompted to escalate before continuing. Use this for any check that involves a measurement instrument.
The operator takes a photo using the tablet camera and attaches it to the work order. Use this for visual checks where a human judgement call is required — weld bead appearance, cosmetic finish, packaging integrity — and where having a photographic record for the customer or auditor is valuable.
You can combine check types freely within a single step. A surface treatment step, for example, might require a thickness measurement (value range), a visual inspection (yes/no), and a photo of the finished part.

​
When Checks Trigger

Quality checks trigger inline, exactly when the operator reaches the step they are attached to. The sequence is:
1

Operator reaches the step

Operator Work Mode displays the step instructions and then presents the quality checks in order before the completion button becomes active.
2

Operator records the check

The operator enters the measurement, answers the yes/no question, or attaches the photo. Bold validates the input immediately.
3

Pass or fail decision

If the check passes, the operator proceeds to the next step. If it fails, Bold shows a clear failure alert and — depending on your configuration — either blocks progress until a supervisor releases it or allows the operator to record the non-conformance and continue.
4

Data is recorded

Every check result — pass or fail, value entered, photo attached — is stored against the work order with a full timestamp and operator identity.
Do not configure every step to have multiple mandatory checks from day one. Start with the steps in your process where defects are most frequent or most costly to fix downstream. Add coverage progressively as you see where it adds the most value.

​
Finding Root Causes in Your Processes and Machines

Individual quality records become genuinely powerful when you look at them in aggregate. Bold’s quality analytics let you slice the data in the ways that matter:
  • By product — which products generate the most non-conformances, and at which steps?
  • By machine — is a particular press or lathe consistently producing out-of-tolerance parts compared to the others doing the same work?
  • By shift — are defect rates higher on the night shift? That might indicate a training gap, a fatigue pattern, or a temperature-related process variation.
  • By operator — is a specific operator’s check results diverging from the rest of the team? That is a coaching conversation waiting to happen, not a blame exercise.
  • By supplier — if you record the material lot against each step, you can identify whether a defect cluster traces back to a specific incoming batch.
When you find a root cause, update the recipe. Add a tighter quality check at the step where the issue originates, or add a clarifying instruction. The recipe becomes your best-practice record, not just a checklist.
Bold generates quality reports automatically from the data your operators are already recording. Key reports include:
ReportWhat it shows
Non-conformance logAll failed checks in a date range, with step, order, machine, and operator
First-pass yield by productThe percentage of units that pass all checks first time, by product and date range
Defect ParetoThe most frequent failure types ranked by occurrence, highlighting where to focus improvement effort
Machine quality comparisonSide-by-side comparison of defect rates across machines performing the same operation
These reports are available to production managers and quality teams in real time — no waiting for a monthly review meeting to find out what went wrong three weeks ago.
Quality data from the MES module feeds directly into Bold Factory’s Analytics module, where you can build custom dashboards and track quality KPIs alongside your other operational metrics.