Feeder verification should establish a traceable relationship between the actual material, the feeder carrying it, and the active assembly job. Matching a package outline or reading a familiar value is insufficient when two components look alike but differ in rating, tolerance, temperature range or approved manufacturer.
A practical control plan covers initial setup, every replenishment and every changeover. It also defines what happens when identity cannot be proved. This article focuses on preventing similar-part substitutions after material reaches the line; the broader component kitting guide covers build readiness and packaging.
Define the identity chain before choosing a scanner
Start with the released product revision and assembly variant. Associate each populated BOM line with its approved manufacturer part number and any formally accepted alternatives. Then associate the physical reel or other carrier with that material identity, its relevant lot information and its current handling status.
The feeder record connects the material carrier to the machine’s supply position or logical allocation. The active job connects that supply to the reference designators being placed. These links need not all appear on one label, but an investigation must be able to follow them without relying on an operator’s memory.
A barcode is an identifier, not an independent analysis of the parts inside a reel. If the receiving label was associated with the wrong material, scanning it perfectly can reproduce the original error. Define checks for label creation, relabeling and cut-tape repackaging as carefully as checks at the machine.
Separate approved alternatives from visual similarity
Consider an illustrative build using two capacitors with the same nominal capacitance and package. Their voltage ratings differ. A feeder sheet that lists only value and package gives an operator too little information to distinguish them. A correct identity comparison must use the released material definition and permitted alternatives.
Do not turn an unavailable part into an approved substitute by editing a feeder description during setup. Route the proposed replacement through the agreed engineering disposition. Once approved, check whether the replacement requires changes to package recognition, orientation or supply data, as well as the BOM record.
FUJI’s description of Nexim option functions illustrates this distinction: its Dynamic AVL function can associate registered shape, package and supply-direction data with alternative manufacturers. The same page describes traceability links between panel and part identities. These are examples of available functions, not a claim that every assembly line has them enabled.
Verify the installed configuration, not just a printed plan
At initial setup, compare the material identity, feeder identity and active job association. Confirm orientation using the exact component packaging information. If the equipment permits flexible feeder allocation, verify the allocation recognized by the system rather than insisting that a remembered slot number is authoritative.
FUJI documents free feeder allocation as a specific option. On a line without an equivalent validated function, moving a feeder may require a different controlled workflow. The work instruction should reflect the installed equipment, software and configuration instead of describing a generic “smart feeder” capability.
For manual verification, define the fields that a second check must compare and retain the result. Two signatures do not establish independent verification if both people simply copy the same incorrect setup sheet. Choose a source of identity evidence that is independent of the action being checked wherever practicable.
Treat replenishment as a new verification event
A correct start-of-shift setup does not prove that the next reel is correct. Require a new identity check when material is replaced, spliced or transferred to another carrier. Retain the transition point so affected boards can be bounded if an error is discovered later.
The relevant boundary may be a panel serial range, a placement record or another validated production identifier. A timestamp alone can be ambiguous when boards are buffered or reworked. Define how the material transition relates to actual placements and how any uncertainty expands the containment range.
For partial reels and cut tape, preserve the link back to the original identity and handling records. Hold material whose label or chain of custody is unclear. Relabeling an uncertain reel with the expected job number would make the paperwork agree while leaving the underlying identity unresolved.
Control the job change as well as the material change
Common setups allow several products to share installed material, but the correct reels alone do not prove that the correct placement program is active. Confirm the product identifier, board side, variant and job revision at the start of the next run.
Yamaha’s automatic setup examples describe using program or board codes to coordinate downstream changeovers. That illustrates why program selection and material verification are related controls. It does not establish compatibility with an unspecified mixed-equipment line.
Test the changeover workflow with deliberately mismatched identifiers under controlled, nonproduction conditions. Verify that the expected warning or stop occurs and that restarting requires the intended correction. Record the result after relevant software or configuration changes; a documented feature is not evidence that the installed workflow was validated.
Respond to a mismatch without losing the evidence
Stop the affected operation according to the line procedure, identify the suspect material and preserve its label and feeder association. Determine the last verified correct state and the first uncertain placement. Segregate potentially affected assemblies until engineering defines inspection, electrical test, rework or other disposition.
Do not assume optical inspection can distinguish unmarked passive values or every component rating. The assembly process guide places inspection and test within the wider build. Use the specific defect and available evidence to choose containment checks rather than relying on a generic final-test pass.
Walk through a replenishment error before it happens
Consider a hypothetical line that is running a board with two visually similar resistor reels. One carrier reaches its replacement point, and an operator brings a reel from a nearby cart. The label is readable and the package fits the feeder, but the material belongs to another approved product. A check against the active job should reject that association even though the reel itself is legitimate inventory.
Define the expected response before production starts. The operator should know how to hold the incorrect material, obtain the required reel and re-establish the identity chain. The procedure should also state whether an interrupted splice, a removed feeder or a partially completed transaction requires a fresh verification. Ambiguous restart instructions create opportunities to bypass a control that worked correctly at the first warning.
If the mismatch is discovered after placement, use the available production records to bound the suspect assemblies. Do not assume the last scanned board is the last affected board when panels can wait between stations. Establish how machine events relate to board identifiers and consider buffers, rework loops and interrupted runs. Where the mapping is uncertain, the containment boundary must include that uncertainty rather than present an unjustifiably precise serial range.
Engineering can then choose checks that distinguish the actual wrong part from the intended one. Two resistors with different values may be distinguishable with an appropriate electrical measurement, but two capacitors with the same value and different ratings present a different problem. The recovery plan must address the material difference, not merely demonstrate that the assembly powers on.
Validate exceptions and offline operation
Normal scanning is only part of the workflow. Evaluate damaged labels, unreadable codes, duplicate identifiers, an unavailable verification system and material returned from a previous job. For each exception, define who may authorize an alternative check and what evidence must be retained. An exception should be traceable to its reason and affected build rather than disappear into a generic operator note.
A fallback paper process needs a controlled source for the released BOM and approved alternatives. Record how the operator confirms that the document is current and how verification results are reconciled when the normal system returns. If those conditions cannot be established, holding the job may be more appropriate than improvising an untraceable setup. The decision belongs in the agreed production procedure.
Challenge the workflow using designated test material and a nonproduction validation plan. Include a correct part in the wrong association and an incorrect part with a similar description, not only an obviously missing barcode. Record the observed response, the expected response and the recovery sequence. These tests validate the configured process without introducing deliberate defects into saleable assemblies.
Review the records as evidence of a controlled process
A traceability report should connect a meaningful event to the assembly it affected. Review a small set of representative initial setups, replenishments and changeovers with the supplier. Follow each record back to the material identity, active job and applicable approval. This reveals whether the exported report contains enough context for an actual investigation.
Distinguish missing information from a confirmed absence of an event. A blank field may mean that no replenishment occurred, that the system did not capture it, or that a report filter excluded it. Agree the interpretation and the retention period before relying on the report for product containment. Avoid promising component-level genealogy unless the installed workflow can actually produce and reconcile it.
Use discrepancies to improve the control where the identity link was lost. Repeated relabeling problems may point to incoming-material procedures, while correct reels associated with obsolete jobs may point to revision management. Asking operators to be more careful does not resolve either record-system problem by itself. Assign corrective actions to the specific failure mechanism and verify them on a subsequent controlled setup.
Records to request before release
- Released BOM, approved alternatives, assembly variant and active job revision.
- Material-carrier identity and feeder association at initial setup.
- Verification results and the authority for resolving exceptions.
- Replenishment, splice and changeover events linked to affected assemblies.
- Orientation and recognition-data changes for approved alternatives.
- Containment boundaries and disposition records for any mismatch.
For turnkey assembly planning, provide the released BOM and identify parts whose alternatives require particular scrutiny. Discuss the material-verification and traceability requirements before the build is released.