Custom app development for batch linking on the production line
A traceability test can only pass if someone on the line has recorded which batch ended up in which batch. That is not a reporting question but a three-second action, a hundred times a day, performed by people wearing gloves. If that action doesn't happen, there is nothing left to salvage afterwards.
Why it goes wrong on the line
The chain is only as strong as its weakest link, and that is almost always the same link. Receiving goes well, because there is a weighbridge and a system for it. Dispatch goes well, because there is an order attached. In between sits the moment a bag is cut open and its contents disappear into a mixer, and that moment is rarely recorded.
Reprocessing is the sharpest example. A residual batch from yesterday goes back onto the line this morning. That happens in practice, it is permitted, and it is exactly the step nobody records. The result: the mass balance does not close, and in the event of a recall it cannot be said which end products contain that residual batch. That is a bigger problem than a balance that doesn't add up.
The reason it doesn't happen is rarely unwillingness. It is an environment of gloves, wet hands, a splash guard and a pace that doesn't pause for an input screen. Any action that takes more than a few seconds or needs two hands gets skipped as soon as things get busy. And that is precisely when the most material passes through.
How we build this
The design starts from the action, not from the data model. What doesn't work on the line never ends up in any report.
Where does a batch open, where does something come together, where does something split apart? Those moments shape the design, not the structure the office system happens to have.
A code on the pallet, the tote or the label. Choosing from a list of two hundred batches works at a desk, not at a mixer.
Residual batches get their own record with their origin. As long as that is an exception someone has to report later, it won't happen.
A production hall full of steel can have spots with no coverage. The app records locally and synchronises later, keeping the original timestamp.
What the app does in practice
The partner integration carries the whole. What you add on top depends on how many steps your process has and how often work needs to be redone.
Connecting by scanning twice
The incoming batch and the outgoing order, nothing more. That is the action that carries the entire trace, and it has to be quick enough to happen every time.
Reprocessing with its origin
A residual batch gets its own code and keeps hold of the batches that went into it. Without that, the mass balance does not close and a recall is incomplete.
Works without coverage
In a hall full of steel, the signal drops out. Everything is recorded locally and synchronised later, with the time of the action rather than the time of synchronisation.
Weights and quantities at the point of action
What went in and what came out, recorded where it happens. Those are the figures the mass balance later relies on.
Blocking what must not be released
A batch awaiting a test result must not simply go onto the line. The app can enforce that block at the moment of scanning.
Evidence of an incident
Photograph any damaged packaging or illegible label at the point it is found. It saves an argument afterwards about what actually happened.
Who we build for
Where the chain breaks depends on the process. Four situations.
Production and processing
Mixing, cutting and packing. This is where batches come together and split apart, often within the same order, and that is where the registration needs to sit. See also software for the food industry.
Repacking and distribution
A batch splits into smaller units for many customers. The challenge is keeping the outbound side complete without slowing the pace.
Sites with multiple lines
A batch that moves between lines or sites is usually where the chain gets interrupted. Transfer moments therefore deserve an action of their own.
Fresh products with a short lead time
With a short shelf life, traceability is mainly about speed. A reconstruction taking three days is too late in a recall, even if it is accurate.
Test your idea first: a working prototype in 1 day
With OneDayBuild, we turn your idea into something tangible in one day for €1,150, so you can see whether further development is worth the investment. Decide to go ahead with the full build? Then we credit the full cost.
Explore OneDayBuild →Technology and integrations
Processes change more often than software. Anything to do with steps, code types and hold rules should be configurable and not hard-wired into the app.
Why Appfront
Three seconds or it doesn't happen
We design the action first and the screen second. A data entry form on a production line gets skipped the moment things get busy.
Reprocessing is the breaking point
We make it a normal action rather than an exception someone has to report. That is where most balances go wrong.
A hall full of steel has no signal
We build offline as the starting point, preserving the original timestamp. The latter determines whether the record counts as evidence.
We walk the line first
Where a batch opens cannot be deduced from a data model. You need to see it in order to build it properly.
Security and privacy
Batch data reveals your recipes and your suppliers. We set access by role and by line, so an operator sees what they need and not the whole picture, and we log every view. For agency and temporary staff, access lapses automatically.
For traceability, the record itself is the evidence. A batch integration that can be altered afterwards undermines the entire chain; in a recall, it is the first thing a regulator will examine. We record integrations as unalterable entries with time and person, even when synchronisation happens later, and any correction appears as a visible amendment alongside the original value. Our own approach to security is set out in our information security policy; reports from outside go through our vulnerability disclosure policy.
Frequently asked questions about the batch integration app
The production system remains the source of truth; we do not replace it. The difference lies in where the integration happens: at the line, not behind a desk. We retrieve orders and batches through an integration and write the registration back, so there remains a single source of truth.
That is the core of the design: large touch targets, few steps, and a handheld scanner in place of the camera where needed. We test this in your own hall, because a demo at a clean desk says nothing about a wet cutting department.
The app keeps working. Everything is recorded locally and synchronised as soon as coverage returns, with the original timestamp preserved. That last point is essential: an integration that receives the time of synchronisation no longer matches the production order.
As a standard action with its own code. The residual batch keeps track of which batches went into it, so the origin travels with it when it re-enters the line. As long as that remains an exception someone has to report afterwards, it won't happen and the balance won't reconcile.
Yes. A batch awaiting a laboratory result, or one that has been blocked, can be refused at scan with the reason shown. That is stronger than a list on a noticeboard, because it works at the moment someone is about to open the bag.
The BRCGS software handles the trace, the mass balance and follow-up after an audit. This app supplies the data that those rely on. Without the recording on the line, there is nothing to trace, whichever software you use.
The IFS app covers control rounds: temperatures, hygiene and visual checks. This app covers batches and their relationships to one another. These are two different tasks, often carried out by the same people, and they can sit within a single app.
That depends on the number of process steps, whether codes are already in use, and whether there is a production system to connect to. The scanning action with the batch integration is usually quick to put into use; reprocessing and blocking take more effort. We give a reasoned estimate after the discovery phase.
Finding the missing link?
Walk through your process and identify the point where a batch is opened without anyone recording it. That single point determines whether your traceability test passes. We build this as a standalone app and as part of a broader custom app development project.