Batch records and traceability in growing

In insect, mushroom and plant production the batch is the unit of work and the unit of accountability. To trace a batch, hand it over and later explain what happened, a record needs five things: identity, shared time, context, actions and deviations. What follows is a record-design recommendation, not a compliance determination; which rules apply to your operation depends on your jurisdiction and activity.

Identity comes first

A record has to say which batch it belongs to, which room it was in, which stage it was at and when it was placed there. Those operation-specific fields are design suggestions, not a universal mandatory list — every operation defines its own identifiers. Food-chain guidance defines traceability as the ability to follow a product through specified stages of production, processing and distribution — one step back, one step forward at any chosen stage (Codex CAC/GL 60-2006). A dedicated international standard for designing such a system in the feed and food chain is deliberately flexible about each organisation's objectives (ISO 22005:2007).

Time has to be shared

Every entry needs a timestamp from the same clock: measurements, stage and setpoint changes, operator notes, device events. A record with dates but no times cannot reconstruct what happened overnight, and one device that disagrees with the rest corrupts every later comparison.

Context makes a number usable

Measurements belong with the equipment state behind them and the limits that applied at the time — which device was running, which stage, what the agreed range was. In its data-integrity guidance for drug CGMP records, the US FDA puts it plainly: "the number “23” is meaningless without metadata, such as an indication of the unit “mg.”", and it counts date/time stamps, user IDs and instrument IDs among the metadata that gives a value meaning (FDA, December 2018). That guidance targets drug CGMP records, not growing operations; the design point is general.

Actions are the part people forget

A batch record is also a record of what people did: feeding, weighing, cleaning, moving, treating, photographing, and every setpoint change with the name of whoever authorised it. Free text is acceptable; an entry without a time is not. Photographs deserve a place in growing records — design advice, not a rule — because much of what matters in a batch is visible rather than numeric.

Deviations belong in the record

A record that contains only good data is not a record. Record what went out of range, who reviewed it and when, what action was authorised, by whom and when, and what remained open at the end of the stage. Keep deviations visible so later readers can carry the lesson into the next batch.

What makes a record untrustworthy

Entries without a time. Numbers without a room or a batch. Gaps that were dropped instead of shown. Corrections that overwrite the original entry instead of adding to it. The FDA guidance describes an audit trail as a secure, computer-generated, time-stamped record allowing reconstruction of the course of events — a drug-CGMP scope, but a useful design aim. Agree these record-keeping habits in advance, alongside the technical requirements.

Write it for the next reader

Write the record for a later reader who needs to answer a question. A simple editorial test: if the record were not good enough to hand over to the next shift, it is not good enough to support an investigation either.

What the cited guidance covers

Each source above has its own scope. The Codex text serves competent authorities within food inspection and certification systems. EU Regulation (EC) No 178/2002, Article 18, sets identification and labelling obligations for food and feed business operators (EUR-Lex). The US records rule in 21 CFR Part 1 Subpart J specifies, for food businesses under that rule, example record content — source and recipient name and address, product description, dates, lot or code, quantity and packaging — and defines retention by product class (eCFR); cited only as an example that retention must be settled, not a recommended duration. This article is a record-design checklist: the method is general, the specifics stay with your operation.

Where MIND fits

MIND's process-record layer — operator logging with photos, batch codes and traceability links, a governed batch lifecycle, and climate grow schedules with an audit trail of every setpoint change — keeps measurements, equipment state and operator actions in one place for a batch. The batch lifecycle was built for Manna's own black soldier fly (BSF) production, and a new species or crop is configured through customisation work, never ready-made support. The intended value is better visibility and useful records. If you want to discuss a project, contact Manna.

Prepared with AI assistance for Manna Insect Oy.