Cement bag counting on a dispatch conveyor
A cement-bag counter must count physical bags crossing the dispatch point once each, including stop-start motion and closely spaced bags. A hidden boundary requires an exception response.
Line solution
Test this on your samplesCement bag inspection: codes and bag counting
// Short answer
Cement bag inspection is a camera check after the cement packing machine that confirms each bag carries a legible packing date or code and the right pre-printed information, and counts the bags going to each order. addaScan builds it on HIKROBOT cameras in sealed housings with purge air, with a bag reject flap in the full version. An indicative installed budget for a typical application is R 140 000 – R 840 000 excluding VAT.
// At a glance
The station in three lines
- Faults caught
- Missing or illegible packing code, wrong pre-printed bag for the cement being packed, bag count per order
- Configuration
- HIKROBOT vision sensor with OCR in a purged IP67 housing (low), or two 20 MP cameras with deep-learning OCR and a reject flap (high)
- Line link
- PROFINET or EtherNet/IP from the vision sensor (where the chosen HIKROBOT device lists it), or digital I/O from the VC3000; count per order to the loading PLC
// Fit & limits
What this station will not do, and what it needs
What it will not do
- Bag mass: weight is the checkweigher’s job; a camera does not see mass.
- Print grading: it checks that the code reads on this camera, not character height, indelibility or an ISO grade.
- The cement inside: the bag print is checked against the recipe, not the cement in the bag.
- Hidden faces: a code on a face the cameras do not see is not read.
What makes it work
- Spaced bags: bags lying flat and separated at the camera.
- Clean purge air: dry, filtered air for the housings, and a cleaning routine.
- A fixed code position: the packing code printed in the same place on every bag.
- Recipe data: the cement type and expected codes from the packer PLC.
// On South African lines
What a cement bag must show in South Africa
Cement sold in bags in South Africa falls under the compulsory specification for cement, VC 9085. According to a published copy of the compulsory specification, the minimum information must be printed legibly and indelibly on each bag in English, in print at least 6 mm high, and includes the factory identity (which may be coded), a date or code that provides traceability to the week of packing, and the NRCS “LOA No” (VC 9085, published copy). Producers make the same point: Sephaku Cement notes that proper coding assists traceability, identifying the product’s origin and production date (Sephaku Cement). Check the current NRCS text for the full list; the copy cited here is a secondary transcription.
Most of the bag’s information is pre-printed on the valve bag by the bag maker. The packing date or week code is printed at the packer, and that is the part that fails on the line: a blocked print head, a dusty bag face, a code printed off the bag. The specification sets the outcome, a legible and traceable code; it does not prescribe how it is checked. A camera is one way to check it on every bag rather than on a sample.
Fertiliser packing has a similar need. According to a published copy of the 2007 fertiliser regulations under Act 36 of 1947, each fertiliser bag label shows the trade and registered names, the nutrients, the mass, the registration holder and the registration number in the form “Reg Nr. … Act No. 36 of 1947” (fertiliser regulations, published copy); revised regulations may apply, so check the current text. The same station reads the code and checks the bag print on fertiliser lines.
What the packing plant reports, and what cement bag inspection checks
| What the line reports | What a station checks | Where it stops |
|---|---|---|
| “The date code is missing on half a truckload” | Packing code present and readable on every bag: OCR of the printed code, compared with the expected week and factory code. | A code on a face the camera does not see is not read; the high case adds the second face. |
| “The code is there but you can’t read it through the dust” | Legibility on this camera: deep-learning OCR in the high case reads faint and smudged print, and codes it cannot read are rejected, not guessed. | Character height and indelibility are not tested; the check is whether the code reads. |
| “Bags for the other cement type went onto the packer” | Pre-printed bag identity: the product panel matched to the cement the packer is running, from the PLC recipe. | The cement inside is not seen; the check is that the bag print matches what the packer was told to fill. |
| “The bag counter and the loader don’t agree” | Bags counted at the camera per order and sent to the loading PLC, with failed bags counted separately. | Touching or overlapping bags must be separated for a reliable count; bags taken off by hand after the camera are not known to the station. |
| “Underweight bags at the hardware store” | Not checked by the camera; the reject flap can also take the checkweigher’s rejects and keep them out of the count. | Mass is weighed, not seen: that stays with the checkweigher. |
// The complete solution
Cement bag inspection station: hardware, software and installation
The station sits on the bag conveyor between the packer and the truck loader or palletiser, where bags lie flat and spaced. A trigger sees each bag; the camera reads the packing code and checks the bag print; the count goes to the loading PLC. Cameras sit in sealed housings with filtered purge air, because cement dust settles on everything.
[Schematic]Plan view, not to scale
Swipe sideways to see the whole drawing →
[1]: A photoelectric sensor sees each bag and drives the count.
[2]: Camera in a purged, sealed housing reads the code and print on the top face.
[3]: A second camera reads the other face or the gusset, wherever your code is printed.
[4]: A flap or diverter takes failed bags off before the truck loader.
[5]: Confirms each rejected bag left the line.
Hardware
| Component | HIKROBOT example | Role in this station |
|---|---|---|
| Code and print (low) | HIKROBOT SC3000-series mono vision sensor (MV-SC3050M) with OCR and code tools in the sensor | Reads the packing code and checks the bag print on one face; counts bags from the trigger |
| Cameras (high) | 2 × HIKROBOT CS-series 20 MP mono GigE cameras (MV-CS200-10GM) with 25 mm 10 MP lenses, or CT-series PRO cameras rated IP67 with lens and lens cap fitted | Read the code and print on two faces |
| Processing (high) | HIKROBOT VC3000 vision controller (MV-VC3302P) running HIKROBOT VisionMaster with deep-learning OCR | Reads dusty and smudged codes and compares the bag print with the packer recipe |
| Lighting | HIKROBOT bar lights: MV-LLDS-H-600 (low), two MV-LLDS-H-1400 (high) | Low-angle light across the woven or paper bag face |
| Housings and purge | Not HIKROBOT: IP67 camera housings with filtered purge air and dust shrouds (HIKROBOT lists no purge housing) | Keep cement dust off windows and lenses |
| Trigger | HIKROBOT MV-PE photoelectric sensor or the existing bag sensor | One pulse per bag for the camera and the count |
| Reject (high) | Not HIKROBOT: bag reject flap or diverter with confirmation sensor | Takes failed bags off before the loader |
Software and vision tools
| Tool or software | What it does here |
|---|---|
| SCMVS OCR and code recognition (low) | Reads the packing code in the vision sensor and compares it with the expected week and factory code. |
| HIKROBOT VisionMaster deep-learning OCR (high) | Reads low-contrast, smudged and deformed characters that rule-based OCR struggles with on dusty bags. |
| VisionMaster OCV and pattern tools | Check the code against the expected string, and the bag’s pre-printed product panel against the recipe. |
| Count and order logic | Counts good and failed bags per order in the PLC from the trigger and the vision result. |
| Week code recipe | The expected week and factory code follow the plant calendar automatically, so nobody types them in at the start of each week. |
Installation and integration
01
Mounting and enclosure
Cameras in sealed housings on a frame over the bag conveyor, clear of the packer’s spill zone; filtered purge air keeps the windows clean, and housings open for cleaning without moving the camera.
02
Lighting and shrouding
Bar lights at a low angle across the bag face and a shroud, so the woven or paper surface and the dust film look the same on every shift.
03
Trigger and tracking
One trigger per bag; bags must arrive spaced, not touching. The PLC tracks each bag from the camera to the flap so the right bag is removed.
04
PLC link
Result, read code and count to the packing or loading PLC; the vision sensor lists PROFINET and EtherNet/IP, and the VC3000 controller has opto-isolated digital I/O.
05
Reject and confirmation
Reject flap with a confirmation sensor; if the station is not ready, bags are flapped off or the conveyor stops, so no bag is loaded unchecked. Rejected bags come off the order count.
06
Recipes and operator screen
Cement type and bag recipe selected by the packer PLC at changeover; the screen shows the last unreadable codes with images, so the printer is cleaned before a whole shift is affected.
// Indicative cost in South Africa
What this station costs, excluding VAT
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.
- Low
- R 140 000
- one face, count from trigger
- High
- R 840 000
- two faces, deep learning, reject flap
- Basis
- Typical application of this kind, indicative
- South African rand, excluding VAT
- Prices as at 2 October 2026
- Travel (Gauteng): outside Gauteng quoted separately
// Included and excluded
The low figure is one vision sensor with OCR in a purged IP67 housing, reading one bag face and counting bags from the trigger, with the result sent to an existing reject or alarm. The high figure reads two faces with deep-learning OCR on a VisionMaster controller and adds a bag reject flap with confirmation. Changes to your conveyor, the compressed air supply for purging, utilities and VAT are excluded. Travel is budgeted for Gauteng; most cement plants are outside Gauteng, so travel for your site is quoted separately.
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 55 400 | R 183 700 | Low: one vision sensor with OCR, bar light, IP67 housing. High: two 20 MP cameras with lenses, two long bar lights, VC3000 controller, two IP67 housings |
| 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 25 000 | R 150 000 | Low: dust shroud, purge air, mounts, cabling. High: dust-tight mounts and purge, bag reject flap with confirmation, panel |
| Engineering (recipes, PLC link) | R 32 400 | R 195 000 | 36 h at R 900 to 150 h at R 1 300 |
| 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 10 000 | Outside Gauteng quoted separately |
| Total (rounded) | R 140 000 | R 840 000 | Low: one face, count from trigger. High: two faces, deep learning, reject flap |
// Related technology
Which technologies this scenario uses, and why
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 |
|---|---|---|---|
| OCR/OCV of printed codes | Core | The packing code must be read and compared with the expected week and factory code on every bag. | OCR/OCV inspection |
| Presence detection and counting | Core | Bags are counted at the same point the code is read, so the loading count and the checked bags agree. | Presence and counting |
| Vision sensor (HIKROBOT SC1000–SC3000) | Supporting | One vision sensor with OCR in a purged housing covers one bag face on a small budget. | Choosing a smart camera |
| Deep learning (classification, segmentation, DL OCR) | Supporting | Faint, smudged print on dusty woven or paper bags is where deep-learning OCR earns its place. | Rule-based vs deep learning |
| Lighting technique | Supporting | Low-angle light across the bag face stops texture and dust from hiding the characters. | Machine vision lighting |
| PLC link and reject with confirmation | Supporting | A failed bag must leave before the loader, with confirmation, and come off the order count. | PLC and reject integration |
// Acceptance
How the station is proven before handover
Every station is accepted on your products, against numbers agreed in writing before the build. How we deliver →
01
Sample test
Bags from your packer with good, faint, smudged and missing codes, clean and dusty, in each bag type; the result states which codes are reliably read and which are rejected as unreadable.
02
FAT with challenge bags
Before delivery, marked bags (no code, wrong week code, smudged code, wrong pre-printed bag) must each be rejected, and good bags must pass and be counted.
03
SAT on the line
Between the packer and the loader at production rate, the challenge bags are run again; false rejects, escapes and the count against the loader are compared over an agreed run of production.
// FAQ
Questions about cement bag inspection
What must be printed on a cement bag in South Africa?
According to a published copy of the compulsory specification for cement (VC 9085), the minimum information must be printed legibly and indelibly on each bag in English, at least 6 mm high, and includes the factory identity, a date or code traceable to the week of packing and the LOA number. The specification sets the outcome; it does not require a camera. Check the current NRCS text for the full list before relying on it.
Can a camera read date codes through cement dust?
Usually, if the dust is a film rather than a crust and the camera window is kept clean. Deep-learning OCR reads low-contrast and smudged characters better than rule-based OCR, and purged housings keep the optics clear. Codes covered by caked dust or spilled cement are not read, and the station rejects them as unreadable rather than guessing. The sample test on your bags shows where your line sits.
Can the same station count bags for loading?
Yes. Each bag seen at the camera is counted, and the PLC adds it to the order being loaded, with failed bags counted separately. In that sense it is a cement bag counting system and a code check in one place: a camera-based bag counter on the conveyor belt, instead of a separate bag counter machine built on photo-eyes. Bags must arrive spaced: touching or overlapping bags have to be separated mechanically for a reliable count, and bags removed by hand after the camera are not known to the station.
How much does cement bag inspection cost in South Africa?
An indicative installed budget is R 140 000 – R 840 000 excluding VAT, at 2 October 2026 prices. The low figure is one vision sensor with OCR in a purged housing, reading one face and counting bags. The high figure reads two faces with deep-learning OCR on a controller and adds a bag reject flap with confirmation. Both include hardware, software, engineering, installation, FAT and SAT in Gauteng; plants elsewhere add travel. A firm price follows a sample test. This is a typical application: every project involves custom requirements, so the final price depends on your products, line and acceptance criteria.
Does it work on fertiliser bags too?
Yes. On fertiliser lines the station reads the batch or date code and checks that the pre-printed bag matches the product being packed. According to a published copy of the 2007 fertiliser regulations, the label shows the registration number with “Act No. 36 of 1947”, the nutrients and the mass; the camera checks that the right printed bag is used, not the nutrient content or the mass in it.
Where on the line should the camera go?
After the cement packing machine and before the truck loader or palletiser, where bags lie flat and spaced and the packing code has been printed. Right at the packer the air is dustiest; a little further down the line, after bag cleaning if you have it, codes read better and windows stay cleaner. The reject flap must sit before the loader, so a failed bag never reaches the truck.
Sources
- Compulsory specification for cement (VC 9085, GN R544 of 2007), Annex B bag marking: published copy on tradebarriers.org (NTM 2550) (secondary transcription, NRCS original not retrieved, checked 2026-10-02)
- Sephaku Cement: understanding the legalities of supplying cement in South Africa (producer guide, checked 2026-10-02)
- Regulations regarding fertilizers (Act 36 of 1947, 2007), regulation 17 labelling: published copy on tradebarriers.org (NTM 4635) (secondary transcription, revised regulations may apply, checked 2026-10-02)
- HIKROBOT VisionMaster: deep-learning OCR for complex-background, low-contrast and deformed characters (vendor page, checked 2026-10-02)
- HIKROBOT CT-series PRO camera (MV-CT1MG0016): IP67 with lens and lens cap installed (vendor product page, checked 2026-10-02)
Related:
More line solutions in building materials and metals:
// Next step
Send us bags with good, faint and missing codes
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.