1. Define the defect
Use retained good, marginal and failed packs to state what must be detected. Avoid broad wording such as “check the seal” when the proposed device can only see cap presence or position.
Choose the inspection characteristic, process point, reject method and acceptance challenge around the real jar, lid and closure risk.
Use production evidence. Final selection and acceptance depend on representative jars, lids, process conditions and agreed checks.

A jar capping line should inspect the specific closure characteristic that distinguishes an accepted jar from a failed one, track each failed jar to a suitable reject device, confirm that it was physically removed and enter a controlled state if inspection or rejection is unavailable. Presence, cap height, alignment, button state, vacuum, torque and leakage are different checks and should not be treated as interchangeable.
Use retained good, marginal and failed packs to state what must be detected. Avoid broad wording such as “check the seal” when the proposed device can only see cap presence or position.
Inspect application close to the capper. Inspect process-dependent characteristics such as cooled button state or vacuum only when the pack has reached the defined condition.
Challenge detection, tracking, rejection, reject confirmation, full-bin response, reset and reconciliation with deliberate failures across every approved format.
| Inspection target | Possible method | What it can show | What it does not prove by itself |
|---|---|---|---|
| Missing lid | Photoelectric, proximity or vision presence check. | A closure or expected lid feature is present at the inspection point. | Correct engagement, torque, vacuum, leakage or liner condition. |
| High, cocked or misaligned lid | Height sensor, profile measurement or vision. | The lid profile or position differs from the validated accepted range. | The cause of the defect or every hidden engagement failure. |
| Twist-off button/profile | Validated profile, displacement or vision check at a defined process time. | The visible/measurable button condition matches the approved samples. | A universal vacuum value, shelf life or complete seal integrity. |
| Vacuum | Approved non-destructive or destructive measurement, inline or sampled as appropriate. | The measured pack meets the defined method, timing, units and limits. | Why the result changed without reviewing product, headspace, cooling, jar and lid. |
| Application/removal torque | Controlled torque measurement, commonly on a sample or dedicated test station. | The measured torque result under the stated method and conditioning. | Leak integrity, liner compatibility or correct thread start in every case. |
| Leak or closure integrity | Application-specific pressure, vacuum, dye, immersion or other approved method. | The pack passes the defined integrity test under its stated conditions. | Suitability of an arbitrary online sensor unless the method has been validated. |
| Cosmetic damage | Vision or controlled human inspection using approved examples. | Marks, dents, scratches or deformation visible under the validated presentation and lighting. | Internal damage or defects outside the camera/inspection field. |
The inspection supplier and pack owner should agree samples, lighting, jar/lid presentation, recipe controls and challenge procedure. Transparent jars, reflective metal lids, condensation, product residue and decorative finishes can alter sensor performance.
Define the distance and transport from inspection to reject, how jar identity is maintained through variable spacing or stops, and what happens when a jar is removed manually for a sample.
The reject method must suit jar mass, base stability, product condition, line speed and neighbouring jars. A push or air blast suitable for a light bottle may be unsuitable for a heavy glass food jar.
Use a separate confirmation method where required. If rejection is not confirmed, the line should enter the agreed containment state rather than continuing to accept unverified packs.
Define guarding, capacity, segregation, full-bin detection, safe access, reconciliation and disposal or investigation of failed packs.
| Challenge | Expected response to agree | Evidence to retain |
|---|---|---|
| Known failed jar | Detect, track, reject and confirm without removing the adjacent good jar. | Inspection result, reject signal, confirmation and count reconciliation. |
| Reject confirmation blocked | Stop or contain output in the specified fail-safe state. | Fault code, affected time window and recovery check. |
| Reject bin full/unavailable | Prevent uncontained production and require authorised recovery. | Alarm, stop condition and bin-change procedure. |
| Inspection bypass | Visible authorised state with defined disposition of production. | User, time, reason, affected quantity and release decision. |
| Power or control reset | Preserve or safely clear tracked jars and revalidate the system before release. | Restart challenge and reconciliation result. |
Download the inspection and reject checklist CSV
The fastest route to a useful inspection trial is a labelled set of good, marginal and failed packs, plus the approved method used to decide which is which.
Discuss inspection and rejectionNo. Different defects require different evidence. Presence, height, alignment, torque, vacuum, button state, leakage and cosmetic damage are separate characteristics and may need different process points or test methods.
Use 100% inline inspection where the defined risk and method justify it, while retaining sampled or offline checks for characteristics the inline system cannot prove. The pack owner should define the complete control plan.
Cap height can identify gross seating or alignment differences when validated, but it does not directly prove application or removal torque. Use the approved torque method where torque is the acceptance characteristic.
The line should enter the agreed fail-safe or containment state so unverified production is not treated as accepted. The affected jars and time window should be reconciled before restart.
Provide every jar and lid format, known good and failed samples, required defect list, line speed and spacing, process environment, available footprint, reject destination, controls interface and acceptance challenge plan.
Send your jar, lid, product and target output. Lancing will help you identify a practical machine route and the samples needed for confirmation.