Twelve changeovers a shift.
Every one of them a risk.

AI inspection for consumer goods production — where the line runs fast, the variants change constantly, and a labelling mistake becomes a return.

Test it on your product

The Hard Case

The defect is rarely the product.
It is what is printed on it.

Consumer goods lines mostly do not fail on the product itself. They fail on the pack: the wrong label for the variant, a date code that smeared, a seal that caught product, a case short by one.

These are verification problems rather than judgement problems, which makes them well suited to a camera and badly suited to a person watching several hundred packs a minute.

The difficulty is changeover. A line running six variants in a shift needs the inspection to change with it — and a system that has to be reconfigured at every switch will quietly be turned off by the second week.

  • A single xis.rt handles several variants at once.

    Different products running on the same line, each with its own checks, without reconfiguring the system between runs.

  • The cost of a labelling error is not the pack.

    It is the recall, the retailer relationship and the retrospective check on everything shipped since the error started — which is why the per-unit record matters as much as the detection.

What it checks

The checks a packaging line
actually needs.

  • Right label, right variant?

    Label verification

    Confirms the correct label for the product running, applied straight and in the right position, undamaged and legible.

  • Is the code there and readable?

    Date & lot code

    Reads printed date and lot codes and checks each against what should be on that pack, catching smears, drop-outs and misprints.

  • Is the pack filled correctly?

    Fill level

    Checks fill height or level against the specification for that variant, recorded as a value for every unit.

  • Will the pack hold?

    Seal integrity

    Seal condition, product trapped in the seal, closure seating and cap presence before the pack leaves the line.

  • Is the print legible and complete?

    Print quality

    Detects missing, faint or misaligned print on packaging and sleeves, including partial printhead failures.

  • Is the case right?

    Pack & case count

    Counts units per case or tray and checks orientation and position before closing.

  • Is it on straight and tight?

    Cap & closure

    Cap presence, cross-threading, tilt and colour against the variant specification.

  • Does the pack look sellable?

    Surface & cosmetic

    Dents, scuffs, creases and contamination on the pack face that a retailer would reject.

  • Is this the product we think it is?

    Variant confirmation

    Confirms the product running matches the job, which is where changeover errors are caught before a pallet is built.

Built for the environment

Fast lines, frequent changes,
mixed materials.

Consumer goods lines are less hostile than a bakery but less forgiving on speed. These are the conditions the hardware is specified against.

ENVIRONMENT

Built for stability. Designed for real-world conditions.

  • Line speedSpecified against your actual rate, not a demonstration rate
  • ChangeoverVariant switching without reconfiguring the inspection
  • VibrationMounting isolated from conveyor and filler vibration
  • Dust and splashIP-rated housings appropriate to the product and cleaning regime
  • SpaceMounting designed around existing guarding and access
Industrial vision camera

IMAGING

Clearer vision. Smarter inspection.

  • Print and labelRGB industrial cameras at resolution matched to the smallest character
  • CodesOCR and OCV against the expected string for that run
  • Reflective film and foilControlled lighting geometry to manage specular packaging
  • Continuous webLine-scan for sleeving and wrapping
  • Fill and levelBacklight or structured lighting depending on container
  • Multiple facesSeveral inspection points where more than one face matters

On lighting. Printed and filmic packaging is reflective, and reflective surfaces are where lighting geometry decides whether a system works at all. Dome, backlight or structured illumination is specified per application and validated on your material, not assumed from the camera.

Where the record matters

Most of these checks are
already obligations.

Labelling, date coding and net quantity are not quality preferences in the EU — they are legal requirements, and your certification scheme adds verification duties on top.

  • Labelling and traceability

    Label legibility, date and lot code presence and one-up one-down traceability are required by EU food information and food law. Inspection is how you meet them at line speed.

  • Net quantity

    Average quantity rules under the EU prepackaging directive, enforced nationally, make fill control a legal as well as a commercial matter.

  • Scheme verification

    GFSI-recognised schemes such as BRCGS require start-of-run, interval and end-of-run verification with equipment self-test. A system that logs every unit produces that evidence automatically.

The platform

Same three components,
tuned to your product.

Quality

xis.rt

Runs at the line on an industrial PC, at 150 fps, signalling your PLC over OPC-UA or MQTT.


xis. Dashboard

Defect and form trends across lines, shifts and sites — on your own servers.

Develop

xis. Safety

Your team trains the model on your own bakes, and retrains it when the recipe, flour or season changes.

Partner tracks

The obligations
behind the inspection.

The published criteria your customers, auditors and quality team already reference. We train models against your interpretation of these standards — we do not certify conformance to them, and an inspection result is not a conformity assessment.

  • Reg (EU) 1169/2011

    What it covers

    Food information to consumers. Requires the label to be easily visible, clearly legible and indelible, with allergen and ingredient declaration and date marking.

    Where it applies

    Label verification, print legibility and date coding

  • Reg (EU) 178/2002

    What it covers

    General food law. Establishes one-up, one-down traceability for food business operators.

    Where it applies

    Lot code verification and the per-unit record

  • Directive 76/211/EEC

    What it covers

    Making-up by weight or volume of certain prepackaged products. Sets the average quantity rules enforced nationally.

    Where it applies

    Fill level and net quantity

  • GFSI schemes (BRCGS, IFS)

    What it covers

    Certification scheme requirements including start-of-run, interval and end-of-run verification with equipment self-test.

    Where it applies

    Scheme evidence produced automatically from inspection logs

  • ISO/IEC 15416 & 15415

    What it covers

    Print quality grading for 1D and 2D codes, A to F. Applies wherever a code has to be read reliably by someone downstream.

    Where it applies

    Barcode and data matrix verification

Consumer goods questions

What packaging teams ask first.

  • No. A single xis.rt instance handles several product variants at once, each with its own checks, so the inspection follows the job rather than needing to be switched with it. Adding a new variant is a task for your own team in xis.dev.

  • Can it read our date and lot codes reliably?
  • Our packaging is shiny film. Is that a problem?
  • How fast can it run?
  • What happens when we launch a new SKU?

Still deciding

The one question this page can't answer:
does it work on your defect?

  • Start here

    Send us your images

    Share examples of the defect you care about. We train a model on your product and come back with a straight answer — including if it's no.

    Start a feasibility check
  • Or ask directly

    Talk to an engineer

    Thirty minutes with someone who builds this. Bring your line speed, your defects and your integration constraints.

    Get a demo