Skip to content
Quality, standards & complianceFood & Beverage

Listeria environmental monitoring by zone

AI tracks every swab with zone, spot and result, spots clusters and recurring positive sites and triggers follow-up swabs.

Today

In ready-to-eat food production, a Listeria finding close to the product is a recall risk. Every week dozens of swabs from zones ranging from contact surfaces to floor drains go to the lab, the results come back as PDFs and end up in a spreadsheet. Whether a drain is positive for the third time, whether findings are moving closer to the line, and whether they relate to a cracked joint or a change in cleaning is often seen only by someone who lays all the reports side by side, and by then it is late.

An example

At Tarvina, floor drain B-17 next to the salad line in hall 2 tests positive for Listeria three times in nine weeks, each time in the Monday swab after the weekend. The agent lays the results side by side and finds a maintenance ticket from August: tile joint at the drain cracked, still open. Since the cleaning contractor changed, the cleaning records no longer show the drain basket being removed. It creates eight vector swabs around the drain, including the frame of the vegetable cutter in zone 2, and product samples for the salads of the next production day. The lab confirms Listeria monocytogenes in the drain; frame and product are negative. After the joint is repaired and the cleaning changed, three follow-up swabs come back negative and the quality manager closes the case; at the IFS audit in November the trend analysis with every action is ready.
Fictional company, typical workflow.

Your data, wherever it lives

Your environmental monitoring data is scattered today: sampling plan and result list in spreadsheets, lab reports as PDFs in the mailbox or on the lab portal, floor plans as PDFs, cleaning records on paper, building defects in maintenance and photos on phones. demi reads, understands and links them per sampling point and zone, with your rights. AI spots a cluster at the second finding instead of the fifth and proposes actions; what happens to product and line is decided by your quality assurance.

Where the data lives today

  • PDFs
  • E-mails
  • Office files
  • Databases
  • QM system
  • Scans & paper
  • Photos, video & audio
Data LayerOne view of everything, with your rights

How demi solves it

Three layers in this order: connect the data first, then the rights, then the AI.

  1. 1 · ConnectData Layer

    Data sources in this case: lab reports, the sampling plan, floor plans, cleaning records, maintenance and production planning. demi replaces nothing and links every swab to its zone, sampling point, timing before or after cleaning, result and lab report. Cleaning runs, open building defects such as cracked joints or dripping condensation, alterations and the products that ran at that spot come on top.

  2. 2 · PermitPermission Layer

    Results, floor plans and trends are visible to quality assurance, production and plant management; the external cleaning contractor sees only the actions for its areas. Whether lots are held, tested or released is decided by quality assurance alone, with the reason logged. What goes to customers or the authority is released by plant management.

  3. 3 · ActIntelligence Layer

    The agent reads every lab report, assigns the result to its sampling point and detects clusters in place and time, spots that keep testing positive and findings that move from zone 3 towards contact surfaces. It links them to cleaning, maintenance and production and, following your own plan, proposes vector swabs around the site, intensified cleaning and product samples. Clusters and trends are computed from the recorded swabs; the language model only explains.

How it runs

  1. 1

    The sampling plan sets which spot in which zone is swabbed how often, and demi generates the week's sampling list from it.

  2. 2

    Lab reports arrive as PDFs or from the lab portal; demi assigns each result to its sampling point on the floor plan.

  3. 3

    The agent detects clusters, recurring positive sites and findings that move closer to the line, and links them to cleaning, building defects and production.

  4. 4

    After a positive result it creates vector swabs, cleaning and repair orders and, where needed, product samples according to your plan.

  5. 5

    Quality assurance decides on actions and lots and closes the case only after negative follow-up swabs.

What changes

  • Recurring positive sites flagged at the second finding, not the fifth

  • Every finding documented with action, follow-up swab and closure

  • Trend analysis for audits and authorities at the push of a button

What does it bring you?

Check it with your own numbers

Every starting value is an assumption you can change. The calculation sits right below, step by step.

Assumption, editable

Your numbers

Result

+72 hfreed per year

Assumptions and calculation

demi tracks every swab with zone and result, spots clusters and sets up follow-up swabs; you decide.

6 min → 4.2 min per case

How does it run today?

What is already digital saves less: the factor lowers the share demi brings.

How we calculate

200 environmental swabs × 6 min × 30% × 1.0 (manual) × 12 months = 4,320 min = 72 h

Value per year, before demi costs

Freed per year+72 h

Assumption: labour cost of €49.50 per hour in German manufacturing in 2025 (Destatis), plus workplace, IT and overheads.

= Value per year, before demi costs

+€4,680

gained per year

An estimate, not a promise. The starting values are assumptions for an industrial company with 150 to 500 people.

Combine with more cases and get the package

Questions about this use case

How does demi know the zones and sampling points?

From your sampling plan and floor plans. Each sampling point gets its zone and its place on the plan once, for example zone 1 for contact surfaces such as conveyor belts and cutting tools, zone 3 for floors and drains. If the plans exist only as PDFs or photos, demi reads them in and quality assurance confirms the assignment.

Does demi recognise whether it is the same Listeria strain?

demi does no lab work itself, but it keeps the typing results when the lab supplies them, for example from whole genome sequencing. If they show the same strain at the same spot over months, the agent flags the spot as a possible harbourage site. Without typing, it shows the pattern of place, time and frequency and recommends typing where it pays off.

Who decides whether product is put on hold?

Quality assurance. The agent shows which lots ran on that line at the time of the finding, which product samples exist and what your plan says for this case. Hold, testing or release is decided by an authorised person, and the decision is logged with its reason.

Does this help with IFS or BRCGS audits?

Yes. Auditors want to see that results are analysed, trends recognised and actions closed effectively. demi compiles the trend analysis per zone and sampling point, every action with its follow-up swab and closure, and links the lab reports.

See this case with your data

In the demo we show you the workflow on your own systems.

Your configurator

No use cases yet

Your configurator is still empty

Add a use case from any card or case page and see what it brings per year, with the calculation. Free, no sign-up, no commitment: try it out and change any number.

Browse use cases