Inspection and rejection
How should a capping reject system be specified?
Specify a capping reject system by defining the finished-pack faults that matter, deciding which faults can be detected reliably inline, proving the inspection with real acceptable and representative bad packs, selecting a controlled reject method, confirming reject success and defining what the line does when faults repeat or the reject route is unavailable.
Specification starting point
Define the pack failure before choosing inspection hardware
Use the sequence below to connect each required pack decision with a detectable condition, a verified reject action and an escalation rule. Inline inspection and offline sampling should support the same accepted-pack standard.
Start with approved packs and representative failures rather than a sensor catalogue. The inspection method, tracking distance, reject device and line response can then be challenged against the real bottle, closure, decoration and normal production variation.
Start with failure modes, not a sensor catalogue
List the pack failures that create commercial, quality or safety risk: missing cap, grossly high or tilted cap, damaged closure, incorrect orientation, tamper-feature fault, cross-thread, unacceptable torque or failed seal. Then separate properties that are visible or measurable inline from those that require an offline test.
A sensor can only control a fault that produces a stable, detectable difference. The same cap may present differently because of bottle position, colour, reflection or decoration, so detection must be proven across the approved format and normal variation.
Match inspection method to what it actually proves
| Inspection route | May help check | Does not automatically prove |
|---|---|---|
| Cap-presence sensor | Missing closure or gross position fault | Correct thread engagement, torque or sealing |
| Height or distance sensor | Gross cap-height variation against a stable reference | Subtle tilt, liner seating or leakage |
| Vision inspection | Presence, orientation, visible tilt, height or cosmetic features when trained and validated | Internal thread condition or opening torque unless another measurement is integrated |
| Reject confirmation | That the intended bottle left the main product path | That the reject reason was diagnosed correctly |
| Offline torque or integrity test | Defined sample properties under a controlled method | Every bottle unless the process and sampling plan support that conclusion |
Define the reject and escalation sequence
- Identify the bottle from detection to reject point so the correct pack is removed.
- Confirm that the reject action occurred.
- Prevent rejected bottles rejoining accepted product.
- Define a full-bin, failed-reject or missing-air response.
- Set warning and stop rules for repeated or clustered faults.
- Record reject reason, time, format and batch for investigation.
The exact sequence must be included in the line risk assessment and tested during acceptance. A reject mechanism should reduce escape risk without allowing the underlying feeder, bottle or capping fault to continue unseen.
Prove performance with representative samples
Create approved packs and controlled examples of the faults the system is intended to detect. Test the expected position, colour, reflection, label and bottle variation, then challenge the reject sequence at normal line conditions. Retain the validation record and define how it will be checked after changeover, maintenance or inspection-system adjustment.
Related guidance
Continue the investigation or specification
Capping quality control plan
Define the finished-pack failures and the checks used to accept them.
Cap height and tilt inspection
Understand what presence, height and vision systems may detect.
Capping line integration
Connect inspection, tracking, reject confirmation and line response.
Discuss an inspection requirement
Send good packs, representative faults and the required line outcome.