Food manufacturing traceability software links every batch of ingredients and packaging to the finished products made from them, and to the customers those products were sent to. In the UK, being able to show that link is a legal duty. When a withdrawal or recall happens, how quickly you can answer “where did this go?” depends on what was captured at goods in, during production and at despatch. This guide covers what the software needs to record, what UK law asks for, how recall tracing works in practice, and where a warehouse management system (WMS) fits alongside production software.
What food traceability software does
Food traceability software records batch, lot and expiry data for each movement of food, and keeps the link between what came in and what went out. Good systems capture that data at the point of scan, so the record is made when the product moves rather than rebuilt from paperwork afterwards.
For a food manufacturer the job breaks into four parts. The system captures the identity and date of every incoming batch, keeps that identity attached through storage and processing, records which batches went into which finished lots, and records which customer received each finished batch. Miss one of those steps and the trail stops at that point.
What UK food businesses are required to do
UK food businesses must be able to identify the businesses that supplied them and the businesses they supplied, and make that information available to the authorities on request. That is the core of retained Regulation (EC) 178/2002, Article 18, which requires traceability to be established at all stages of production, processing and distribution. The Food Standards Agency (FSA) traceability guidance notes that references to EC legislation should now be read as assimilated law where they apply to Great Britain.
Den FSA’s Managing food safety guidance adds practical duties. You must keep records of the suppliers that provide your food and ingredients and of the businesses you supply, often held on the invoice. Those records must be kept up to date and be available for inspection at all times. The table below maps each duty to what the system has to capture.
| Requirement | Where it comes from | What the system needs to capture |
|---|---|---|
| Identify who supplied each ingredient or packaging batch | Article 18(2) | Supplier, batch or lot number, receipt date and delivery document |
| Identify which businesses received your products | Article 18(3) | Customer, despatch date, batch or lot shipped and quantity |
| Keep traceability through production and distribution | Article 18(1) | Links between incoming batches, production runs and finished lots |
| Label or identify products so they can be traced | Article 18(4) | Batch and lot identifiers on pallets, cases and units |
| Records available at inspection | FSA, Managing food safety | Searchable records you can export without rebuilding them by hand |
How fast should a food recall trace be?
The FSA does not set a minute count for a trace. Its withdrawals and recalls guidance says that if you believe food you have supplied is unsafe, you should immediately withdraw or recall it from the market and tell your local authority.
One published standard does set a number. The BRCGS Storage and Distribution standard, as reproduced in its site self-assessment tool (SD401), includes a traceability test in clause 12.3.2 and says traceability shall be achieved within 4 hours of notification. That clause belongs to the storage and distribution standard, so check which BRCGS standard applies to your site before you quote it to a customer. Manufacturers working to the BRCGS Food Safety standard have their own traceability requirements to check against.
The number that matters most in practice is how long your own recall takes to run. Campeys fra Selby, a food logistics business, reports sub-5 minute recall tests on its Clarus case study. Treat that as Campeys’s own measure of its own process, not a benchmark for every site.
Capturing batch and lot data at goods in
Most of the traceability problem is solved, or lost, at the goods-in dock. If the batch and expiry date are not captured when the pallet is received, they are rarely recovered later. GS1 UK’s case study on IPC and Subway describes GS1-128 barcodes that encode the GTIN, batch or lot number and expiration date, and scanning them to capture that data is what lets the business trace and recall product quickly.
Clarus reads the identifiers already printed on a supplier’s GS1 label. A scan can capture the batch and the quantity without anyone re-keying them. You choose which identifiers the handheld reads, and the rest switch off. Not every supplier uses the same label. Two Benelux customers had different GS1 layouts, with the article number and sell-by date in one barcode set and the batch and production date in the other. Both needed a configuration that matched their label, not a generic one.
Where a supplier sends an advance shipping notice with an SSCC pallet barcode, a single scan can accept the whole pallet and all the data sent with it. Serial numbers can be captured on arrival, on dispatch, or both.
Batch control is set per account and can be switched off for individual products. Where batch control is on, the handheld asks for the batch as part of the task and warns the operator if the batch scanned does not match the one expected. Packaging that does not need batch control skips that screen entirely. The batch and lot control feature covers how that configuration works in practice.
Expiry, rotation and quarantine
Rotation rules decide which stock gets picked first, and getting them wrong is one of the most common causes of short-dated stock going out. In Clarus, FIFO uses the date the stock record was created, down to the millisecond. Expiry-based rotation is handled separately as a sell-by rule. Keeping those two ideas apart matters, because a system that treats FIFO and first-expiry-first-out as the same thing will pick the wrong pallet when dates and receipt order disagree.
Physical layout changes the answer. FIFO works well in racking, where the oldest pallet is reachable. In bulk stacks the oldest pallet can sit at the back of the stack, and moving fifty pallets to reach it is not practical. Clarus’s switch function lets an operator scan a pallet at the front instead, provided it has the same product, date and batch. The sporing av utløpsdato feature covers the date rules in more detail.
Quarantine works best when it is automatic. Stock can be set to go on hold the moment it is put away into a quarantine location, so nobody has to remember to block it. One manufacturer made an animal feed product that needed veterinary sign-off before release. It went straight into a 3PL, and an automation put it on hold on receipt. The hold, and who released it and when, sit in the transaction history.
Recall readiness and mock recalls
When a recall starts, the question is not only where the affected batch is, but where it went. The sequence works like this. If you captured what was sent, to which customer, on which order and on which vehicle, you can trace the batch to those drops quickly. Then you put the remaining stock of that batch on hold from the stock list in seconds, so it cannot be allocated to another order. You can also see stock that is on an order but has not yet left, stop those picks, reallocate that stock to a batch that is not affected, and then hold whatever remains.
The trace is only as good as the dispatch data. Not every operation records the vehicle and order at despatch, and a trace that stops at the dock door cannot answer the customer question. That is a process choice, and the software can only enforce it if you configure it.
Test this before you need it. Run a mock recall on a real batch, time how long it takes to produce the customer list, and write down every point where someone had to search a spreadsheet or ask a colleague. The gaps you find are the ones a real recall would expose.
Audit evidence for BRCGS and customer audits
Auditors ask for evidence that checks were done, by whom, and when. Clarus stamps each captured check against the transaction line with the date, time and user, so the evidence sits in the same record as the movement it relates to. Checks are captured as fields you set up. Clarus supports renamed text, integer and yes-or-no fields on the handheld, so a vehicle arrival check, a pest check or a damage check can each become a step in the goods-in or pick flow.
There is no built-in HACCP module. HACCP is normally a local process for managing raw materials and goods, and Clarus supports it through generic fields added to the workflow, not through a pre-built compliance screen. Temperature control works the same way. Clarus does not map temperatures or control the chilled or frozen environment, so a temperature check is a recorded manual step, not an automated reading.
Campeys reports a full BRC audit pass at the first attempt, and says the auditor described its traceability as exceptional for a first-time audit. Campeys is also using the system to segregate allergen stock and enforce FIFO, which is the kind of control a food auditor looks for.
Where a WMS fits, and where production software fits
A WMS manages stock once it is in the warehouse: goods in, putaway, holds, picking, despatch and the movement history behind them. It does not plan production runs, hold recipes or run material requirements planning (MRP). If you make food as well as store it, you need production software for those jobs, and the two should exchange batch and quantity data.
Clarus replenishment and pick faces run as standalone functions, so they do not depend on an MRP link. Mathew Buttar makes the broader point that a WMS “doesn’t necessarily have to do the function” of every business process. It should “complement the function and process of the business that needs it.” For a food manufacturer, that means the WMS should own the warehouse record, and production software should own the recipe and the run.
If you are choosing a warehouse system for a food business, our guide to the best WMS for food and beverage is a useful next read, and Clarus also has a food and beverage solutions page that sets out how the system is used in that sector.
What software won’t fix
Traceability software amplifies the process underneath it. Several limits matter before you buy.
- Poor data at source. If suppliers send inconsistent batch codes or labels, the software records the inconsistency faithfully. Fix supplier data requirements before go-live.
- Gaps in scanning. The trail only exists where something was scanned. A pallet that leaves without a scan breaks the chain at that point.
- The truck. The system cannot control what happens on the truck: “What it doesn’t control is what happens on the truck.” Once a load leaves the back door, the system has the dispatch record, not the physical condition of the load.
- Implementation time. A standard project with the right support on both sides is usually live within around 12 weeks. Complex, high-volume projects take longer, and skipping user acceptance testing is the usual reason a go-live goes wrong.
Snakk med en lageransvarlig
Hvis du vurderer ulike alternativer og ønsker å se hvordan et spesialutviklet lagerstyringssystem (WMS) fungerer i praksis, er Clarus verdt å ta en prat med. Vi samarbeider med 3PL-leverandører og distributører over hele Storbritannia for å implementere lagerstyringsprogramvare som tilpasses din driftsmodell – ikke omvendt.
Ta kontakt med teamet vårt å snakke gjennom dine krav.