A smart camera with built-in lighting sits right at the labeler exit and checks every container. It confirms a label landed and is lying flat, so a flap from a label that lost its glue is rejected. It reads the printed code on the label character by character and compares it with the order the line is running, which the plant's MES sends over EtherNet/IP or TCP/IP. Anything that doesn't match is rejected before it reaches the case packer.
Why check labels at the labeler exit?
Because that is the last point where a labeling mistake is cheap. Two failures matter most:
- A label that didn't stay on. If the glue misses, a label can land loose and lift away as a flap, or come off later in the case.
- The wrong label. After a changeover, a roll from the last order can stay on the labeler, or the wrong roll gets loaded. The container is fine. Its label is wrong.
A wrong label is mislabeled product. If it hides an allergen the contents carry, it becomes a recall. FDA has reported that undeclared allergens were the leading cause of Class I food recalls for at least three years running[1]. A joint industry survey by GMA, FMI and GS1 US put the average direct cost of a food recall at about $10 million[2][3], before lost sales and brand damage.
What does the camera check on every label?
- Presence and position. Did a label land, and is it flat where it should be? A loose flap from lost glue fails.
- The printed code. OCR reads the code on the label character by character.
- The match. The code read is compared with the product in the live production order. A mismatch fails.
Reading character by character matters because the dangerous mistakes are small. Two products from the same plant can share a label design and differ by one letter or digit in the code. A check that only looks for "a code is there" passes both. A check that reads every character and compares it with the running order catches the one-character difference.
The camera is a smart camera with integrated lighting, mounted right at the labeler exit. Camera, light and processing share one housing, so the station fits right where the label has just gone on. A typical strobed exposure is about 60 microseconds[4].
What triggers the camera?
An inductive proximity sensor triggers the camera off the can, so a loose label flap can't false-trigger.
The reason matters. A photoelectric sensor or laser looks for an edge breaking a beam. A label flapping loose can break that beam on its own and fire an extra trigger, which throws off the count and the reject timing. An inductive sensor responds to the metal of the can, not to paper, so each can produces exactly one trigger.
How does the system know which label is right?
It gets the answer from the plant, live. When the line changes over, the customer's MES or production system pushes the new product to the camera over EtherNet/IP[5] or TCP/IP. No one has to type a code into the camera, so a changeover can't leave last order's code in place by mistake.
Labels that fail go to an encoder-tracked pneumatic reject, driven by a PLC output, which removes the exact container that failed. Lines like this one run from 300 to 1,200 cans per minute; 1,200 cans per minute is the ceiling for code-reading tasks like this one.
What do the reports show?
The station runs on the same JM Automated Solutions HMI as our other systems: live image, pass/fail, false-reject counts and the current recipe. For labels, the useful addition is reporting:
- Daily and monthly reject reports, broken down by reason, so quality can see whether loose labels or wrong codes are the bigger problem.
- Studies. Saved images and results let the plant study a problem after the fact, for example whether loose labels cluster after a glue change or on one shift.
- Timestamped logs for traceability.
The hardware is built for washdown areas.
How does it work with the variety code check?
The two checks answer different questions. The variety code check reads the code on the can end before labeling and confirms what is inside. Label verification confirms the label that went on matches the order. Together, they confirm the right product went into the right label.
Want to know if it works on your labels?
Send us a few labeled containers, including a loose label or two and a wrong-code sample if you have one. We'll tell you in writing what a camera can check at your labeler exit, before any proposal or price. Book a call or send JM Automated Solutions a message.
See the label verification station page for the station setup at a glance.