Batch Traceability Without Entering the Same Information Twice

Follow one batch through the records

A receiver copies a supplier lot number into the ERP. An operator writes it on a production sheet. Quality types it into an inspection record, and shipping enters it again for the customer. Each person needs the information, but four transcriptions create four opportunities for the records to diverge. Asking people to be more careful leaves the underlying handoff unchanged.

Choose one recent batch and follow it from the supplier document to the shipment. Compare the purchase order, receipt, location, material issue, production record, inspection and dispatch. Watch how people identify the physical material between those steps. A consistent database cannot compensate for two unlabelled containers beside a workstation.

The goal is to capture a fact at its authoritative point, then reuse it with its source intact. That requires more than connecting software. The team must agree which fact has been recorded, who can correct it and whether later users are seeing an observation, a plan or a release decision. Start with a bounded product flow that exposes those distinctions.

Excellence industrielle’s 2024 report describes traceability-related technology in its demonstration environment. A demonstration can make a connection easier to understand; your acceptance test still needs the awkward events in your own operation. Ask the team to bring one split, one rework record and one disputed receipt to the first mapping session. Those examples often expose the missing relationships more clearly than a diagram of the normal flow. [1]

Separate identity from status

Keep the supplier’s lot reference alongside your internal lot identifier. A lot number alone may not be unique across suppliers or products. Define the combination that identifies the received material, and retain the supplier and item references rather than guessing from a familiar-looking code. Include the unit of measure and any conversion that affects quantities.

A purchase order says what was ordered. A receipt says what arrived. A quality release says what may be used. Those records should be linked, but they should not be collapsed into one editable status field with no history. The planner needs to distinguish stock physically present from stock available for the intended order.

Assign authoritative ownership field by field. Receiving records the observed supplier lot and quantity; quality owns the disposition; production records what was consumed; shipping records what left and where it went. Existing systems can retain these responsibilities. A connected view can reference them without allowing every department to overwrite every other department’s evidence. Document the permitted direction of each update and the response to conflicting values before enabling automatic synchronization between systems.

Record what changes the genealogy

Imagine, as an illustrative example, a received material lot divided between two production orders. Each order produces a separate finished lot. Some output from the first lot is held, then reworked with additional material from another receipt. A simple column called “original batch” cannot describe all of those relationships.

Record the inputs and outputs of each transformation, with quantities, units, time, equipment or location, and the production order. A split retains the relationship between the parent lot and each child. A merge records all contributing lots. Rework links the original output, added inputs and resulting output rather than replacing the original history with a cleaner-looking record.

Packaging relationships need their own care. Putting identified cases onto a pallet groups them for handling; it does not necessarily create a new material transformation. When cases are removed or reassigned, update the grouping history. Otherwise a search for the pallet may return items that were no longer on it when shipped.

GS1’s traceability and EPCIS standards offer common concepts for identity and events when interoperability matters. Adopting a standard can help partners exchange meaning, but it does not decide your lot boundaries or establish regulatory compliance. Those choices need review against the product, customer requirements and applicable obligations. [2]

Make capture shorter without making it blind

At a workstation, scan the material identifier and select or scan the active order. Populate known references from the source system, then ask for the new fact: quantity consumed, location changed or inspection result. Show enough context for the operator to catch a wrong selection. Automatically filling the wrong order faster only makes the error travel further.

Use a stable event identifier so retrying a transmission does not consume the material twice. The receiving system should recognize the same event and return its existing result. Keep event time separate from the time the server received it. Offline work can then arrive late without being misrepresented as work that happened after the next shift began.

Define an explicit offline queue and reconciliation process. An operator needs to know whether a scan is recorded locally, synchronized or rejected. If a lot is on hold while a disconnected device still shows an earlier release, the site needs an approved operating response. Do not silently treat a lost connection as permission to continue.

Prove the history by using it

Run a trace exercise in both directions. From a finished shipment, identify the contributing receipts and relevant production and quality events. From a suspect input lot, identify affected output, remaining stock, rework and shipments. Reconcile quantities within the agreed process rules, recording scrap, samples and normal losses rather than forcing an artificial balance.

Test an incorrect scan and a correction. Preserve the original event with a linked correction, reason, author and timestamp. Limit who may make consequential changes, and record approvals where required. A tidy current screen is useful; an explainable history is what lets the team investigate how it became current.

Use the accompanying event-mapping template with receiving, production, quality and shipping. Measure missing links, duplicate events and the time needed to resolve a trace question before expanding. Quindart’s role is to connect the relevant records and decisions, not to promise that a new label or database removes every uncertainty. The practical result is less repeated transcription and a history that people can actually interrogate.