Application guide
What does a date-code inspection system check?
An inline date-code inspection system uses controlled lighting and an industrial camera to assess the printed area on every presented pack. Depending on the risk and print format, it can confirm code presence, read characters, compare them with the active recipe and track a failed result to automatic rejection or a controlled line stop.
The inspection must be defined around the actual printer, substrate, code position and normal production variation. A high-contrast laser mark on a flat carton behaves very differently from an inkjet code on a curved, glossy bottle or a flexible pouch.
OCR, OCV and code-presence checks are not the same
Optical character recognition (OCR) turns an image into characters when the system needs to determine what was printed. Optical character verification (OCV) compares the observed characters with an expected string or permitted format. A simpler presence or quality tool may only confirm that an adequately sized printed pattern exists in the defined region.
The right method depends on the consequence of an error. If the quality plan requires the exact use-by date to match the current job, a presence check alone is not enough. If only a clearly visible batch mark is required, full OCR may add unnecessary complexity and false rejects.
Date and batch-code failures that can be targeted
- Missing date, batch, lot or expiry code
- Wrong characters or a code that does not match the active recipe
- Incomplete, broken, doubled or poorly contrasted print
- Code printed outside the permitted area
- Unexpected date format, separator or character count
- Incorrect relationship between the pack, label, barcode and printed code
- No-read or uncertain results sent to a controlled fail condition
Build the check around normal production variation
Representative packs should include the normal movement, colour, gloss, print density and material variation expected on the line. Trials establish the practical exposure, light angle, field of view and character size. They also show whether the pack needs to be stabilised or spaced before imaging.
Curved containers may require a narrow inspection window or more than one view. Reflective films can need diffuse or polarised lighting. Flexible packaging may move through the focal plane, so guide design and trigger timing can be as important as the recognition software.
Recipe, data and reject integration
The expected code may be entered locally, selected from a protected recipe or supplied by a printer, PLC or production database. The interface and authority for changing that value must be agreed during specification. Systems can also record counts, no-reads, failed strings and selected images when the chosen controls and storage design support it.
A failed pack can be rejected with a pusher, air blast, drop gate or other suitable device. Where a single rejected pack cannot be positively separated, a controlled stop may be more appropriate. Reject confirmation, secure collection and consecutive-failure logic can be added to suit the quality risk.
Information needed for a useful feasibility review
- Good, poor and failed print samples from normal production
- Printer type, font, code height and permitted formats
- Pack materials, colours, curvature and code locations
- Line speed, product pitch and conveyor movement
- Expected string source and recipe-change process
- Required response to a no-read, mismatch or poor-quality code
- Data retention, audit and user-access requirements
Related inspection stages
Combine code inspection with barcode and 2D-code checks, label inspection, carton inspection and confirmed rejection.