Buyer questions answered
Practical questions for specification and approval.
Each answer starts with the decision, then identifies the evidence or limitation that changes it.
What run-identification data should be recorded?
Record the product and batch, approved recipe, target weight, pack format, line or machine identity, operator or authorised login where the quality system requires it, start and stop times, and the verification equipment used. The record should show which physical configuration was installed.
For multi-product machines, include change-part or outlet identity when it affects the process. If settings are changed during the run, retain the previous and new values with the reason and authority for the change.
Are individual fill weights needed, or are summaries enough?
Individual consecutive weights provide the strongest evidence for drift, cycles and refill effects. Statistical summaries may be sufficient for routine monitoring when the sampling method, calculation and retention rules are controlled, but they should not erase the data needed to investigate an out-of-control event.
Agree the sampling and retention approach with the buyer’s quality and legal-metrology advisers. A machine supplier can support data capture, but cannot decide the organisation’s regulatory record-keeping obligations without the application context.
How should filler and checkweigher data be separated?
Label the measurement source, units, timestamp and pack identity clearly. Filler data describes the measured dose or weighing cycle; checkweigher data describes the completed pack passing a downstream dynamic scale. Differences can arise from tare, product loss, closure addition and measurement uncertainty.
Use a pack-tracking or time-alignment method appropriate to the line before feeding checkweigher results back to the filler. A delayed or mismatched result can cause the controller to correct the wrong process condition.
Which events and alarms should be logged?
Log events that can change weight quality or explain lost output: recipe changes, manual target adjustments, hopper low or high level, refill faults, scale zero or calibration checks, overload, discharge timeout, pack-not-ready, reject events and safety or downstream stops.
The log should distinguish an informational event from a condition that invalidates production. Clear timestamps and acknowledgement records help maintenance and quality teams reconstruct what happened without relying on memory.
How should automatic feedback adjustments be controlled?
Automatic feedback should use a defined measurement source, averaging method, deadband, step size, limits and timeout, with a safe response when corrections reach the permitted boundary. It should not chase individual noisy readings or continue adjusting when pack tracking is uncertain.
Record each correction and the values that caused it. The upstream filler still needs a stable mechanical and feeding process; feedback can trim a controlled process but should not hide product-flow faults, spills or an unstable weighing system.
How long should weigh filling data be retained?
Retention time depends on the product, customer contracts, quality system, legal obligations and the purpose of the record. The website cannot prescribe one universal period. Define the owner, retention rule, backup method and access control before deciding the data architecture.
Keep records in a format that remains interpretable after software or personnel changes. Document units, field meanings and calculation methods, especially when summaries or standard-deviation values are exported to another system.
