Start with the pack failures that must be controlled
List the defects that could make the pack unacceptable: wrong or missing dose, product in the seal, weak or open seal, leak, incorrect dimensions, registration drift, wrong or unreadable code, damaged film, missing component or incorrect lane/strip separation. Rank them by product, customer and process risk.
Each control should have a defined purpose. Adding a camera, checkweigher or sensor without a detection requirement, test method and response can create data without protecting released product.
Control incoming product, film and code data
Verify the product batch and condition, film reel and artwork revision, reel winding, code data and machine recipe before production. Check that the declared product and pack match the approved machine/format family.
Incoming film checks may include identification, dimensions, print repeat, eye mark and visible damage. Product checks depend on the process and may include temperature, density, viscosity or other characteristics used to define the approved operating case.
Design dose control around the commercial requirement
Define whether the result is mass, volume or count and specify the measurement method, instrument, tare, units, sample frequency and acceptance rule. A process average alone can hide individual packs or lanes outside the permitted range.
For multi-lane equipment, identify samples by lane and across time. Include start-up, steady running, refill, low-level, stop and restart where these conditions can affect product feed or dosing.
Separate visual seal checks from integrity evidence
Visual inspection can find wrinkles, product contamination, channel defects, misalignment or damaged film. It does not by itself prove the barrier or integrity required for the product. Select an appropriate integrity or leak-test method from the pack, product and quality requirement.
Document sample conditioning, test equipment, method, acceptance and action on failure. Retain failed and accepted examples during trials so the test and defect language are understood consistently.
Verify registration, variable code and artwork separately
Registration checks the position of artwork relative to seals and cuts. Code verification checks variable information such as batch or date. Artwork control confirms the correct printed reel. Treat them as separate checks with defined fault responses.
For automatic inspection, define the readable area, expected content, contrast, pack-tracking and reject confirmation. Challenge the system with controlled defects during acceptance.
Engineer reject handling and containment
The line must know which pack failed, move it to a controlled reject path and confirm that the rejection occurred where required. Define full-bin, jam, failed-reject and access responses. Prevent rejected packs from being returned to saleable output without authorised review.
Where inspection is sampled rather than automatic, define how production is held, traced and assessed when a sample fails. The quality plan—not the machine alone—sets the disposition rules.
Make the sampling plan represent the process
Samples concentrated at the start of a run do not show refill, temperature drift, reel changes, lane variation or stop/restart behaviour. Spread checks across the declared production states and retain enough context to link a result to time, lane, reel and product batch.
The final frequency and limits belong to the manufacturer or packer’s quality system and applicable requirements. Machinery trials can demonstrate a proposed method under controlled conditions but should not invent release criteria.
Quality-control plan for a sachet line
| Input identity | Product batch/condition, film reel and artwork revision, code data, recipe and approved format/change parts. |
|---|---|
| Dose | Mass/volume/count basis, instrument and tare, sample frequency, lane identity, limits and action on failure. |
| Pack dimensions | Width, length, seal bands, notch, perforation, strip/single format and relationship to approved drawing. |
| Seal/integrity | Visual defects plus the agreed integrity or leak method, conditioning, sample plan, acceptance and retained examples. |
| Print and code | Artwork/reel identity, registration, cut position, variable-code content, readability and fault response. |
| Automatic inspection | Detection challenge, pack tracking, reject, reject confirmation, containment and failed-reject response. |
| Process states | Start-up, normal run, refill, low level, reel change, stop/restart, speed change and end-of-batch as relevant. |
| Records | Time/lane/reel identity, results, rejects, deviations, action, retained samples, authorisation and traceable revision. |
Use the same definitions at FAT, SAT and routine production
Factory and site tests should refer to the approved pack drawing, product/film conditions, measurement methods and accepted-output definition. Differences between factory and site equipment or methods should be recorded and resolved.
Routine production may use a different sampling frequency from acceptance, but defect names and measurement principles should remain traceable. Keep retained samples and test evidence with the relevant machine recipe and material revisions.