Egg-tray pocket occupancy before pack closure
Egg-tray checking before pack closure needs a pocket-by-pocket occupancy map. Flag the two positions separately.
Egg pack inspection is a camera check of every egg carton or tray after the grader and packer: an egg in every position, the lid closed, the grade, size and best-before marks present and readable, and, in open trays only, visibly cracked or dirty eggs. addaScan builds it on HIKROBOT vision sensors or cameras with a reject signal; candling and crack detection stay in the grading machine. An indicative installed budget for a typical application is R 180 000 – R 850 000 excluding VAT.
Eggs are packed on farms and packing stations across South Africa. SAPA’s avian influenza database for the second quarter of 2026 lists 212 egg-producing layer farms holding 21 484 881 birds, notes that laying-hen numbers on egg farms are considerably under-reported, and shows Gauteng with the highest share of egg-sector farms and birds (SAPA, 2Q2026). On a packing station the egg packaging, cartons and trays, is filled by the packer behind the grader.
The grading, packing and marking of eggs for sale is set by R345 of 20 March 2020 under the Agricultural Product Standards Act (R345, 2020). Containers must be closed properly, packed with eggs to their full capacity, and not mix grades. The size and grade go on the front or top panel; the best-before date goes on any panel, preceded by “best before” or “BB”, in day-month-year order, and may not be removed or altered; markings must be clearly legible on a contrasting background. The grade standards allow no cracks, and an inspector checks external defects by eye and internal defects with a candling device. The rule sets the outcome, not the method; egg pack inspection is one way to check the count, the lid and the marks on every pack.
It does not replace the grader. The egg grading machine South African packing stations run candles and weighs the eggs and, where fitted, detects cracks: that is an egg inspection machine built as turnkey equipment by the grader maker, and addaScan does not build graders. Pack inspection works after it, on what the packer puts in the box.
| What the line reports | What a station checks | Where it stops |
|---|---|---|
| “The shop found a dozen with an egg missing” | Every position in the open carton or tray before the lid closes: egg present or empty. | An egg broken in its position still counts as present unless the crack check (high case) flags it. |
| “Cartons are opening on the pallet” | The closed pack after the lid closer: lid edge and lock tab in the position taught on good packs. | A lock that looks closed but has not latched is confirmed on samples only. |
| “The date printer ran dry and nobody saw it” | Best-before date present and readable on every pack, compared with the date the PLC or operator set for the run. | The camera does not know the lay date; it checks the printed date against the expected one. |
| “Large eggs went out in jumbo boxes” | Grade and size mark on the top panel read and matched to the recipe of that packer lane. | It checks the mark against the lane, not the eggs: weight and grade come from the grader. |
| “Cracked and dirty eggs are reaching the trays” | In open trays and cartons (high case): eggs that are visibly cracked, broken or soiled, by deep-learning classification. | Hairline cracks, internal defects and the underside of the egg are not seen from above. |
One view looks down into the open cartons or trays before the lid closes, and a second looks at the closed and marked pack. The result per pack goes to the packer PLC, which rejects or stops; in the high case a pusher moves failed packs off the line and a sensor confirms they left.
Swipe sideways to see the whole drawing →
| Component | HIKROBOT example | Role in this station |
|---|---|---|
| Open-pack view (low) | HIKROBOT SC3000X colour vision sensor with learning-based count (budget priced on MV-SC3050XC), with a HIKROBOT MV-LRDS-H ring light | Counts every egg position before the lid closes |
| Closed-pack view (low) | HIKROBOT SC3000-series 5 MP vision sensor with OCR (budget priced on MV-SC3050M), with an MV-LLDS-H bar light | Lid closed, grade and size mark, best-before date |
| Cameras (high) | HIKROBOT MV-CS200-10GC 20 MP colour GigE camera over the open pack and MV-CS050-10GM 5 MP mono camera over the closed pack, each with a 25 mm lens (MVL-HF2524M-10MP) | Colour shows dirt and stains on brown and white shells; mono reads the print |
| Processing (high) | HIKROBOT VC3000 vision controller (MV-VC3704H) running HIKROBOT VisionMaster with deep learning | Runs count, crack and dirt classification and print checks for both views |
| Lighting | HIKROBOT MV-LBES-H flat dome light over the open pack (high); MV-LLDS-H bar light over the closed pack | Diffuse light on curved shells so cracks and dirt show without glare; even light on the printed lid |
| Trigger | HIKROBOT MV-PE photoelectric sensor or the packer’s cycle signal | Fires the cameras once per pack or row of packs |
| Reject | Not HIKROBOT: the packer’s own reject or stop (low); a pneumatic pusher with confirmation sensor and reject table (high) | Takes failed packs off the line and proves they left |
| Mounting and panel | Local supply: brackets over the packer outfeed, shrouds, cabling, panel and operator screen | Holds the views fixed and shows rejects to the packing team |
| Tool or software | What it does here |
|---|---|
| Learning-based count (SC3000X) or VisionMaster presence tools | Each egg position in the carton or tray: egg present or empty, one recipe per pack format. |
| VisionMaster deep-learning classification (high case) | Separates visibly cracked, broken and soiled eggs from good eggs; trained on your own eggs and shell colours. |
| OCR and character comparison | Reads the best-before date and the grade and size mark and compares them with the run set by the PLC or operator. |
| HIKROBOT VisionMaster OCV (high case) | Verifies each printed character against the expected string, so a broken digit fails. |
| Recipe per pack and lane | One recipe per carton or tray type, egg count and grade lane, selected at changeover. |
| Result and image log | Reject reason per pack and saved images of rejects for the packing supervisor. |
The open-pack view sits over the packer outfeed between egg placement and the lid closer, where the packer leaves room, or over open trays; the closed-pack view after marking. Covers keep dust and feathers off the optics.
A flat dome or ring light over the open pack and a shroud against the plant lights; white and brown shells and each carton colour get their own exposure settings.
One trigger per pack or per row across the lanes; the PLC tracks each pack per lane from the first camera through the lid closer to the reject, so the right pack is removed.
Pass/fail and reason per pack by digital I/O to the packer PLC or reject (the SC3000X lists opto-isolated inputs and outputs); network protocols only where the chosen device lists them and the packer supports them.
The packer’s reject or a pusher with a confirmation sensor and a full-table signal; if a camera is not ready, packs are rejected or the packer stops, never passed unchecked.
Recipe per pack format and grade lane chosen at changeover; the expected best-before date comes from the PLC or is entered at the start of the run; a screen shows the last rejects with the reason.
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.
This is a typical application with custom requirements: your pack formats, lanes, packer and acceptance criteria set the final price. The low figure is two vision sensors, one counting eggs in the open pack and one checking the lid and reading the marks, signalling the packer’s reject or stop. The high figure adds a colour camera with deep-learning classification of visibly cracked and dirty eggs in open packs, print verification on a controller, and a pusher reject with confirmation. Graders, candling and crack detection, changes to your packer or conveyor, 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 103 600 | R 206 500 | Low: SC3000X colour vision sensor with ring light for the count, SC3000 vision sensor with bar light for lid and marks. High: 20 MP colour camera with flat dome light, 5 MP camera with bar light, lenses, VC3000 controller |
| Software licence | R 0 | R 120 000 | Vision sensor software is free; high case with the VisionMaster deep-learning tier |
| Local materials and reject | R 20 000 | R 125 000 | Low: two brackets, shrouds, cabling, signal to the packer’s reject or stop. High: stainless mounts and shrouds, pusher reject with confirmation and reject table, panel, UPS, screen |
| Engineering (recipes, PLC link) | R 36 000 | R 208 000 | 40 h at R 900 to 160 h at R 1 300, including deep-learning training on your eggs |
| Installation, FAT, SAT, training, documents | R 18 000 | R 93 600 | 20 h to 72 h |
| Project management | R 7 200 | R 41 600 | 8 h to 32 h |
| Travel (Gauteng) | R 3 000 | R 8 000 | Outside Gauteng quoted separately |
| Total (rounded) | R 180 000 | R 850 000 | Low: two vision sensors. High: two cameras with deep learning and own reject |
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 |
|---|---|---|---|
| Presence detection and counting | Core | Every egg position in a carton or tray is a presence check; a short pack breaks the R345 rule that containers are packed to full capacity. | Presence and counting |
| OCR/OCV of printed codes | Core | The best-before date and the grade and size mark are printed on the pack and must be present and legible on every one. | OCR/OCV inspection |
| Vision sensor (HIKROBOT SC1000–SC3000) | Supporting | Two vision sensors cover the count and the marks on a small packer without a controller. | Choosing a smart camera |
| Deep learning (classification, segmentation, DL OCR) | Supporting | Cracks and dirt vary egg by egg and between shell colours; classification trained on your eggs handles that better than fixed rules. | Rule-based vs deep learning |
| Lighting technique | Supporting | Curved shells reflect point light; a flat dome gives the even light that shows cracks and stains. | Machine vision lighting |
| PLC link and reject with confirmation | Supporting | Several packer lanes run side by side, so each result must be tracked to the right pack and the reject confirmed. | PLC and reject integration |
Every station is accepted on your products, against numbers agreed in writing before the build. How we deliver →
Good packs and real faults (missing egg, open lid, missing or smeared date, wrong grade mark, cracked and dirty eggs) in each carton type and shell colour, imaged with the proposed cameras and lights; the result states which faults are reliably found and which are not.
Before delivery, marked challenge packs of each fault type must each be rejected, and good packs of every format must pass.
On your packer, at production rate, the challenge packs are run again in every lane, and false rejects and escapes are counted over an agreed run of production.
Egg pack inspection checks every carton or tray after the packer: an egg in every position, the lid closed, the grade and size mark and the best-before date present and readable, and, in the high case, eggs in open trays or cartons that are visibly cracked, broken or dirty. Failed packs are rejected or the packer is stopped. It does not grade, weigh or candle eggs.
No. The grader candles and weighs eggs one by one before they are packed and, where fitted, checks them for cracks; that is turnkey equipment from the grader maker, and addaScan does not build it. Egg pack inspection works after the grader, on the packs: count, lid, marks, and visible cracks and dirt where the pack is open. The two do different jobs and work together.
No. A camera over the pack sees cracks that are open and visible from above, broken shells and dirt. Hairline cracks are found by the crack detection in the grader, and internal defects by candling; R345 also has the inspector check internal defects with a candling device (R345, 2020). The camera check catches visible damage that happens between the grader and the pack, in open trays and cartons only.
An indicative installed budget is R 180 000 – R 850 000 excluding VAT, at 2 October 2026 prices. The low figure is two vision sensors: one counts eggs in the open pack, one checks the lid and reads the marks, signalling your packer’s reject. The high figure adds deep-learning classification of visibly cracked and dirty eggs, print verification and a pusher reject with confirmation. Both include engineering, installation, FAT and SAT in Gauteng. This is a typical application: every project involves custom requirements, so the final price depends on your packs, lanes and acceptance criteria.
No. R345 sets outcomes: containers closed properly and packed to full capacity, the size and grade on the front or top panel, and a best-before date preceded by “best before” or “BB” that is clearly legible (R345, 2020). How a packer checks that is its own choice. A camera is one way to check the count, the lid and the marks on every pack instead of on samples.
It can tell whether the printed date matches the date expected for the run, which the PLC sends or the operator enters at the start. R345 limits the best-before date of unpasteurised eggs to 40 days from the date of lay (R345, 2020), but the camera does not know when the eggs were laid; setting the right date stays with the packing station. The camera finds missing, smeared and wrong dates.
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.