The opened denim case study reports that RFID could support tracking and tracing of work in process in denim production. Its abstract describes design, configuration and deployment as part of the implementation approach. That is useful textile-specific evidence for considering RFID. It does not establish a read-success rate, cost saving or preferred tag technology for the factory being designed here.[S25]
GFT recommends naming the tracked object before specifying the reader. The object might be a bundle carrier, fabric roll or individual item, but the contract must explain what happens when material is split, combined, reworked or transferred to another carrier. A carrier identifier should not become permanent proof of product identity after the carrier is emptied. The owner must define the association, the event that changes it and the evidence required to make that change.
The proposed contract should distinguish a physical read from an accepted execution event. A tag being visible near a station may demonstrate presence within the reader's effective area. GFT would require an additional, explicitly defined rule before interpreting that observation as operation completion. That rule could involve an operator confirmation or an independent process event, according to the approved design. GS1's EPCIS standard concerns sharing visibility events in a business context; this reinforces the distinction between raw observations and contextual records.[S15]
For an illustrative bundle route, GFT would specify arrival, start, completion, transfer and exception as separate event meanings. These names are proposed process states, not a prescribed external vocabulary. The route owner should determine which transitions are valid and what evidence permits each one. An unexpected transition should create a reconciliation task rather than silently move the bundle to the location most convenient for the schedule. A dispatch decision is only as defensible as the interpretation of those transitions.
Duplicate and missing observations require different responses. GFT recommends retaining enough observation identity to recognise repeated delivery of the same read, while allowing legitimate later visits by the same object. A broad instruction to discard repeated tag identifiers would conceal rework and return movement. Conversely, assuming that the next read proves every intermediate operation occurred would invent process history. The requirement should state whether a missed station can be reconciled, who may approve the correction and how the original uncertainty remains visible.
An illustrative reconciliation starts with 120 bundles released to a route. Assume that 112 have valid completion records, 5 have confirmed work still in process and 3 have unresolved location. The assumed balance is 120 = 112 + 5 + 3. The release count therefore reconciles, but the route is not fully explained operationally. Presenting all 120 as completed because they were once observed would erase the unresolved balance. Presenting the unresolved bundles as lost would also exceed the evidence. The appropriate action is a search and reconciliation task under the proposed contract.
GFT would test the tracking requirement with ordinary movement, return movement, tag replacement and a bundle that passes the wrong read point. The acceptance record should connect each scripted physical movement to its expected interpreted event and final location. The production owner should witness representative cases and confirm that the resulting dispatch instruction is usable. This is the proposed test method; the chapter claims no completed factory trial.
Exhibit 7 compares the narrow findings of opened sources. It deliberately records the limit next to each capability so that a purchase specification cannot turn a published example into an unqualified local performance promise.
For handover, GFT recommends an object-association exercise witnessed by production and quality together. The team should follow a chosen physical bundle through a planned split, confirm the resulting identifiers and inspect how its parent relationship appears in the retained record. It should then replace a carrier and demonstrate that the old identifier no longer follows the material. These proposed checks expose ambiguity that a reader-connectivity test cannot resolve. Acceptance should preserve both the expected relationship and the observed result, with unresolved discrepancies assigned to an operating owner before the tracking result is used for dispatch.
The comparison supports selecting a capture approach for investigation. It cannot establish radio performance, camera fitness, meter accuracy or process compliance at an untested installation. Those remain acceptance questions for the relevant use-case owner.