Folder-gluer artwork inspection
Check the smallest artwork difference between similar blanks before folding hides it.
Folder gluer inspection is a camera check on the folder gluer that confirms each carton blank has glue on the glue flap, the right print and a readable barcode before it is folded. The glue check works only where the glue fluoresces under UV light or contrasts visibly with the board, and that is proven on your samples first. addaScan builds it on HIKROBOT cameras; an indicative installed budget is R 170 000 – R 970 000 excluding VAT.
Folding cartons and many corrugated boxes in South Africa are folded and glued on a folder gluer machine before they reach the packer. Mpact alone runs nine corrugated converting plants in South Africa, with fresh produce farmers among its customers (Mpact), and Southern African citrus growers packed 203.4 million 15 kg export cartons in 2025 (CGA, 2025). A corrugated box or folding carton that leaves the gluer without glue on the flap opens on the customer’s erector or case packer; one with the wrong print goes onto the wrong product.
For packaging makers that run FSSC 22000, print is part of the food safety system: version 6 asks packaging manufacturers for print control, including approval of each run against the master sample and a process to detect and identify printing errors during the run (FSSC 22000 v6). Barcodes on shipping cartons are scanned all the way down the supply chain, and the GS1 General Specifications set a minimum symbol quality for ITF-14 and GS1-128 in general distribution (GS1, 2026). A camera on the gluer catches carton printing errors and unreadable codes on every blank, not only on the samples pulled for the run approval.
Glue presence has no regulation behind it; it is a customer complaint and a cost. It is also the weakest part of this station, and we say so up front. A camera can only see glue that looks different from the board: glue with a fluorescent tracer that glows under UV light, or glue that is visibly darker, lighter or glossier than the board. HIKROBOT’s light catalogue states that UV fluorescence can be used to detect UV glue (HIKROBOT), but HIKROBOT shows no folder-gluer case, and clear glue on brown board may show almost no contrast at all. Whether your glue can be seen is decided on samples before any glue check is quoted as firm.
| What the line reports | What a station checks | Where it stops |
|---|---|---|
| “The customer says boxes pop open on their erector” | Glue line on the glue flap after the applicator: present, unbroken and in the right place, where the glue fluoresces under UV or contrasts with the board. | Glue that is there but too thin, too cold or not bonding is not seen; nor is glue that looks the same as the board. |
| “A glue gun blocked and we ran a pallet without glue” | Glue line missing on consecutive blanks: each one is ejected and an alarm tells the crew to stop and clear the gun. | If the glue has no tracer and no visible contrast, the camera cannot tell; a fluorescent glue is a question for your glue supplier. |
| “An old print version got mixed into the job” | Print on each blank compared with the approved job: one panel with a vision sensor, or the full width with a line scan camera in the high case. | Colour shade against the colour standard is not measured; that stays with the press and its colour control. |
| “The barcode on the carton would not scan at the customer” | Barcode present, decodes and matches the job’s expected code. | Decoding on this camera is not an ISO/IEC grade; a verifier on samples gives the grade. |
The station sits on the folder gluer between the glue applicator and the folding section, while the glue flap is still open. A blank sensor and the machine encoder track each blank from the cameras to the kicker. In the low case two vision sensors look at the glue flap and one print panel; in the high case a 4K line scan camera images the full width for print and code, and two area cameras watch the glue lines.
Swipe sideways to see the whole drawing →
| Component | HIKROBOT example | Role in this station |
|---|---|---|
| Glue and print (low) | Two HIKROBOT SC3000-series colour vision sensors (MV-SC3016C) with HIKROBOT ring lights (MV-LRSS-H-80) | One over the glue flap, one over the print and code panel; existence, pattern, OCR and code tools in the sensor |
| Print, full width (high) | HIKROBOT CL-series 4K GigE line scan camera (MV-CL042-91GM) with a 25 mm 10 MP lens and an MV-LTDS-H line-scan light | Images the whole blank as it passes, for print and code across the width |
| Glue cameras (high) | 2 × HIKROBOT CS-series 6 MP mono cameras (MV-CS060-10GM) with 12 mm lenses and bar lights (MV-LLDS-H-600) | Watch the glue lines; two cameras where a carton has more than one glue point, such as crash-lock bottoms |
| Processing (high) | HIKROBOT VC3000 vision controller (MV-VC3704H) running HIKROBOT VisionMaster with the deep-learning tier | Runs the line scan and glue cameras and compares each blank with the job |
| UV lighting (only where the glue fluoresces) | HIKROBOT lists a UV365 wavelength for its MV-LLDS-H bar lights in the light catalogue, but no UV bar light appeared in its online product list; orderability is confirmed per project, otherwise a third-party UV light | Makes fluorescent glue glow against the board; not included in the budget below |
| Trigger and encoder | HIKROBOT MV-PE photoelectric sensor for the blank edge; local supply: incremental encoder on the machine drive | Line rate for the line scan camera and blank tracking to the kicker |
| Reject | Not HIKROBOT: the folder gluer’s own kicker or ejector, driven by the station | Removes failed blanks before the delivery section |
| Tool or software | What it does here |
|---|---|
| SCMVS on the vision sensors (low) | Existence and pattern tools find the glue line and check the print panel; code recognition reads the barcode. |
| HIKROBOT VisionMaster blob and pattern tools | Measure the glue line’s presence, position and continuity in the UV or white-light image. |
| VisionMaster deep learning (high) | Compares the print with the approved job and separates real print errors from normal variation in board and ink. |
| VisionMaster code reading and OCR | Reads the carton barcode and any printed text, and compares them with the job data. |
| Job recipe | One recipe per carton job: glue positions, print reference and expected code, selected at make-ready. |
Brackets on the folder gluer frame between the glue section and the folding section, clear of belts and carriers that move at format change; covers against board dust.
UV or angled white light over the glue flap, chosen on samples, with a shroud that keeps plant light out and UV light away from operators’ eyes; the line-scan light spans the working width.
The machine encoder sets the line scan rate and tracks each blank to the kicker, so tracking holds when the machine speeds up or slows down.
Pass/fail per blank to the folder gluer’s control as digital I/O; PROFINET or EtherNet/IP is used where the chosen HIKROBOT device lists it. The job number comes from the machine or the operator screen.
The machine’s kicker ejects failed blanks and a sensor confirms each ejection; if the vision system is not ready, the machine ejects or stops rather than passing blanks unchecked.
Recipe per job chosen at make-ready; the screen shows the last rejected blanks with the reason, so the crew sees a blocked glue gun at once.
Indicative budget for a typical application of this kind. Every project involves custom requirements (your products, line speed, layout, PLC and acceptance criteria), so the final price differs from case to case. A firm price follows the sample test.
The low figure is two vision sensors on the folder gluer, one on the glue flap and one on the print and code, signalling the machine’s kicker. The high figure adds a 4K line scan camera for print across the full width, two glue cameras, an encoder and VisionMaster deep learning on a controller. Line scan sits outside the standard ranges on the cost guide, which is why it is priced here separately. UV lights for fluorescent glue are not included: no price was collected and availability is confirmed per project. Changes to your machine, fluorescent glue, utilities and VAT are excluded.
Hardware is priced from public list prices converted at the ECB reference rate, plus freight, clearance and margin; engineering at R 900 – R 1 300 per hour. How these ranges are built →
Swipe the table sideways to see the high figure and the basis →
| Cost block | Low | High | Basis |
|---|---|---|---|
| Vision hardware (landed, with margin) | R 74 800 | R 296 800 | Low: two colour vision sensors with ring lights. High: 4K line scan camera, lens and line-scan light, two 6 MP cameras with lenses and bar lights, VC3000 controller. UV lights not included |
| Software licence | R 0 | R 120 000 | Low: vision sensor software carries no licence fee. High: HIKROBOT VisionMaster with the deep-learning tier |
| Local materials and reject | R 20 000 | R 80 000 | Low: mounts on the machine frame, cabling. High: encoder and mounts, panel and screen; the machine’s own kicker is used |
| Engineering (recipes, PLC link) | R 45 000 | R 260 000 | 50 h at R 900 to 200 h at R 1 300 |
| Installation, FAT, SAT, training, documents | R 21 600 | R 104 000 | 24 h to 80 h |
| Project management | R 9 000 | R 52 000 | 10 h to 40 h |
| Travel (Gauteng) | R 2 000 | R 6 000 | Outside Gauteng quoted separately |
| Total (rounded) | R 170 000 | R 970 000 | Low: two vision sensors. High: line scan with deep learning |
The technologies behind this station, how much each matters here, and where each is explained in depth.
| Technology | Role | Why it matters in this scenario | Read more |
|---|---|---|---|
| Lighting technique | Core | Glue is only visible with the right light: UV for fluorescent glue, angled white light for glue that shines or darkens the board. | Machine vision lighting |
| Line scan imaging | Core | A blank moving through a folder gluer is imaged best line by line, at machine speed, across the full width. | Web and line scan |
| Vision sensor (HIKROBOT SC1000–SC3000) | Supporting | Two vision sensors cover the glue flap and one print panel on a small budget. | Choosing a smart camera |
| Deep learning (classification, segmentation, DL OCR) | Supporting | Print comparison must ignore normal board and ink variation; deep learning learns it from good blanks. | Rule-based vs deep learning |
| Code content matched to PLC, MES or WMS data | Supporting | The code on each blank is compared with the job’s expected code, not only decoded. | Carton and pallet verification |
| PLC link and reject with confirmation | Supporting | Rejection uses the machine’s own kicker, so encoder tracking and a confirmed ejection matter. | PLC and reject integration |
Every station is accepted on your products, against numbers agreed in writing before the build. How we deliver →
Blanks with and without glue, from your board and glue, imaged under UV and white light, plus blanks with print and code faults; the result states plainly whether your glue can be seen. If it cannot, the glue check is not offered.
Before delivery, marked blanks without glue, with a broken glue line, with wrong print and with an unreadable code must each be rejected, and good blanks must pass.
On your folder gluer at production speed, the challenge blanks are run again, and false ejections and escapes are counted over an agreed run of production.
A converter’s print check and a food factory’s packaging-version check solve different problems. Keep the print-job master, material identity and production-order mapping explicit at the handover.
| Decision or event | Check or action | Acceptance boundary |
|---|---|---|
| Approved carton | Verify the version-specific print/code regions | Artwork approval remains upstream. |
| Glue and folds | Check visible glue/position before hidden surfaces are joined | Hidden bond strength is not established by an external image. |
| Mixed blanks | Challenge another product or revision with similar artwork | A matching general colour is not a version identifier. |
| Handover to packer | Link carton identity to the approved production packaging master | The food packer must still verify the carton against its product order. |
Agree how old revisions, mixed blanks, splices and changed reels are labelled and isolated. Only claim inspection of the areas visible at the defined conversion or packing step.
Downstream packaging-version verification · Changeover and reload challenge plan
Only when the glue looks different from the board. Glue with a fluorescent tracer glows under UV light and gives a clear glue line; glue that is visibly darker, lighter or glossier than the board can be seen under angled white light. Clear glue on brown board may show almost nothing. addaScan tests your glue on your board before quoting, and if the glue cannot be seen, says so and offers only the print and code checks.
UV glue inspection relies on a fluorescent tracer in the glue. Under UV light the glue glows while the board stays dark, so the camera sees a bright glue line and can check that it is there and unbroken. HIKROBOT lists a UV365 wavelength for its bar lights in its light catalogue (HIKROBOT catalogue), but no UV bar light appeared in its online product list, so orderability is confirmed per project. UV lights are not in the budget on this page; they are priced once your glue has been tested.
No. Dedicated glue detection systems from glue equipment suppliers can control the guns and monitor the glue pattern with its own sensors. This station only checks, with a camera, what can be seen after the applicator, and ejects failed blanks through the machine’s kicker. Glue gun control, glue pattern settings and skip-and-reject logic in the glue controller stay with the glue equipment or machine supplier.
Yes, for print the camera sees as the blank passes: a wrong or old print version, missing print, smears and register shifts against the approved master. Colour measurement to a standard is a separate instrument, not this camera.
An indicative installed budget is R 170 000 – R 970 000 excluding VAT, at 2 October 2026 prices. The low figure is two vision sensors on the glue flap and one print panel, signalling the machine’s kicker. The high figure adds a 4K line scan camera across the full width, two glue cameras and deep learning on a controller. UV lights are extra and priced after the glue test. A firm price follows a sample test on your blanks. This is a typical application: every project involves custom requirements, so the final price depends on your products, line and acceptance criteria.
Yes, for what the camera sees. In the low case a vision sensor compares one print panel with the approved job; in the high case a 4K line scan camera images the full width and deep learning compares each blank with the reference, ignoring normal board and ink variation. Wrong artwork, missing print and a wrong or unreadable code are found. Colour shade against the colour standard is not measured.
Not necessarily. The station decodes the code and compares it with the job; that proves the code is there and reads on this camera. GS1 sets a minimum symbol quality for ITF-14 and GS1-128 in general distribution, measured with an ISO/IEC verifier under defined conditions. Check samples from each job on a verifier; the difference is explained in our guide to barcode verification and grading.
Describe the line, the products and the fault. We test on your good and faulty samples and reply with a feasibility note and a budget.