The cloth is woven from this chapter's 4 exhibits: each weft is one exhibit's own data, normalised; each warp group is a section. The probe reads the exhibit's own value where it crosses. Procedural visualisation, not a measurement.
EX 16
EX 17
EX 18
The chapter as a draft: 6 sections, 3 exhibit wefts, one shuttle.
Lead exhibit · 16FlowillustrativeMatching dispatch and receipt closes a handover
01 / 07
Read-out
Method and source
Illustrative GFT handover procedure using distinct shipping and receiving observations, scheme-specific counterparty acknowledgement as an example, and a retained mismatch branch. It is a proposed control design, not a record of transit time or an implemented integration.
Unit: . Status: illustrative. Sources: S09, S10, S11. Illustrative figures are teaching examples, not measured factory results.
The chapter's warpBar: share of the reading · square: an exhibit
A custody transfer closes only when the sender's dispatch observation and the receiver's independent receipt observation match the same material, quantity basis, parties, sites and transport reference. An invoice, a dispatched message or a carrier scan cannot close that gap on its own. Green Future Tech Limited (GFT), Hong Kong SAR, therefore proposes a paired handover record: preserve both observations, compare them under explicit rules, and hold the affected claim whenever identity, evidence or quantity remains unresolved. The physical goods may move while their stronger provenance claim stays restricted.[S01][S03][S10][S11]
This chapter supplies the dispatch-receipt matching procedure and discrepancy decision tree for GFT's Cameroon textile programme. The authorised programme record covers feasibility study, performance-management system design and digital strategy. The controls below are proposed extensions for a pilot, not evidence of a running factory, certified supplier, completed integration or accepted customs outcome. All shipment identifiers, quantities and operating scenarios are illustrative. The chapter is not investment, legal or tax advice.
¶ 02
4.14 min · 2 sub-steps · 1 exhibit
Dispatch and receipt describe different observations
4.2 An invoice cannot stand in for a physical handover
A seller can issue an invoice before loading, a carrier can depart before the buyer inspects the goods, and a warehouse can receive material while its commercial owner remains unchanged. These are related but distinct occurrences. GS1's Electronic Product Code Information Services (EPCIS) Release 2.0 records completed visibility events, while its Core Business Vocabulary (CBV) gives shared meanings to business steps such as shipping and receiving. A purchase-order reference can connect those observations to a commercial agreement, but it does not turn either observation into the other.[S10][S11]
The minimum handover therefore contains two signed observations linked by one proposed transfer identifier. The dispatch side states what the sender released. The receipt side states what the receiver actually observed. Neither side should be copied silently into the other. Where a value was not observed, the record says UNKNOWN rather than inheriting a convenient value from the dispatch advice. EPCIS distinguishes event time from the business context in which an identifier was read, so the proposed record also separates the physical event time from later data entry and transmission.[S10][S11]
The Aid by Trade Foundation's June 2025 Transparency Standard provides a scheme-specific example of the documentary boundary. Its purchase and sales provisions call for identified parties, product descriptions, unique product and transaction references, quantities, dates and transport details. Its receiving provisions call for inspection of accompanying documents against the material received. Those requirements should be applied only where that scheme and edition govern the claim, not treated as a universal template for every cotton movement.[S06]
The Sustainable Cotton Tracker (SCOT) FAQ supplies another bounded example. It distinguishes tracking from commercial ordering and describes counterparty acknowledgement of transactions entered by a seller. This supports the design principle that a commercial order and a custody confirmation are different records. It does not establish that SCOT is required for this programme, that any proposed supplier participates, or that its workflow applies under other schemes.[S09]
GFT's proposed matching procedure has seven operational steps:
¶ 05
Create the expectation. The releasing custodian records the transfer identifier, intended receiver, material identifiers, quantity basis, transport references and evidence pointers before or at dispatch. The shipment enters the local state EXPECTED; this is a proposed workflow state, not a claim that goods have arrived.[S10][S11]
Preserve the dispatch observation. The sender signs the shipping event with event time, recording time and responsible role. Any later change references the original record rather than replacing it.[S10]
Capture an independent receipt. The receiving custodian identifies what arrived, repeats the relevant measurements and records missing, additional or substituted items. The receiver does not merely accept the sender's electronic form.[S03][S06]
Normalise only for comparison. A calculation service converts units or derives net mass where the declared method permits. It retains the original units, formula, rounding and measurement basis beside the comparison.[S03]
Compare claim-critical fields. The matcher checks transfer reference, counterparties, sites, material IDs, claim class, packages, quantity and evidence scope. A transport delay alone does not prove a custody failure, just as an on-time arrival does not prove identity.[S01][S10]
Route the outcome. A complete match enters RECEIVED, pending any unrelated quality or assurance holds. A mismatch enters DISPUTED; uncertain conformity or unresolved identity enters QUARANTINED. These local states require a documented crosswalk before anyone maps them to a partner vocabulary.[S06][S11]
Authorise release separately. A reviewer releases the affected provenance or scheme claim only after admissible evidence resolves the mismatch. Physical acceptance, commercial acceptance and claim release remain different decisions.[S01][S03][S06]
Exhibit 16 shows the proposed pairing and the branch that preserves a mismatch instead of forcing closure.[S09][S10][S11]
The important output of Exhibit 16 is not a green shipment icon. It is a reviewable pair of observations and an outcome with an owner. A RECEIVED state confirms only the defined receipt checks. It does not prove farm origin, product quality, legal conformity or the truth of every document attached to the shipment. Those propositions retain their own evidence requirements.[S01][S03]
Illustrative role map. Horizontal position represents the role's typical direct responsibility for physical observation in the hypothetical arrangements; vertical position represents its typical authority over the commercial ownership decision. Coordinates are qualitative teaching positions, not measured scores, legal determinations or universal role definitions.
Unit: illustrative qualitative index (0 to 10). Status: illustrative. Sources: S01, S09, S10. Illustrative figures are teaching examples, not measured factory results.
4.4 · the bodyExhibit 17 above; its figures are illustrative where marked.
4.5 The entity making the observation is not always the invoice issuer
Chain-of-custody design must identify who controlled the material and who observed the transfer, not merely who bought or sold it. ISO 22095's public abstract presents a generic framework for custody across materials and products, while warning that a custody system alone does not establish production characteristics or verify a product claim. GFT applies that distinction by recording commercial owner, physical custodian, transport provider and evidence issuer as separate roles. The paid ISO text was not opened, so this chapter does not claim clause-level ISO conformity.[S01]
Consider an illustrative trader selling lint to a spinner while the bales remain in a third-party warehouse. Ownership may change on the contract date, but no bale moves. Later, the warehouse releases the bales to a carrier for delivery. The trader can issue the invoice, yet only the warehouse can attest to the physical removal it observed, and only the carrier can attest to what it accepted for transport. The receiving spinner then observes arrival. Combining all four assertions into one seller event would erase the people and sites whose evidence a reviewer needs.[S01][S10][S11]
These role assignments are design choices, not legal conclusions about title, risk or liability. Applicable contracts and law require separate review. The custody record must identify the physical observer even where that party does not issue the invoice.[S01][S06]
4.6 Role separation makes the evidence pack auditable
4.6
Exhibit 17 places illustrative roles on two qualitative dimensions: authority over the commercial ownership decision and responsibility for the physical observation. The coordinates make the role distinction visible; they are not performance scores or findings about actual organisations.[S01][S09][S10]
A reviewer should use Exhibit 17 to ask whether each required assertion has a competent issuer. A trader's signed statement may support its own commercial transaction, but not a warehouse movement it did not observe. A warehouse record may establish a location and lot list, but not the legal validity of the sale. A scheme acknowledgement can confirm that counterparties used a specified workflow, but not replace all underlying dispatch and receipt evidence.[S01][S06][S09]
Storage is not a pause in traceability. It is a period during which labels can be damaged, lots can move, eligible and ineligible materials can share equipment, and records can drift away from physical stock. The June 2025 Aid by Trade Foundation standard requires identification to be maintained during storage and handling under its HIP provisions. It describes physical and temporal segregation, clear labelling or marking, and accompanying storage documents that identify relevant bales or lots. These are scheme- and edition-specific controls to map where applicable, not a declaration that every warehouse follows them.[S06]
GFT's proposed warehouse record therefore links each material identifier to a site, storage location, claim class and custody state. A move creates a new observation rather than an edit that leaves only the latest aisle. The record retains the prior location, destination, event time, responsible custodian, authorising instruction and any associated cleaning or release evidence. EPCIS location fields and business-step vocabulary can carry parts of that history, while the material's evidence state remains a separate application decision.[S10][S11]
A useful location register answers four questions without consulting a spreadsheet maintained elsewhere:
¶ 03
What is present? The material or lot identifiers, package count and quantity basis observed at that location.
Under which class? The claimed material class, custody model, scheme and edition where relevant, plus any hold or restriction.[S02][S06]
How did it get there? The preceding receipt or internal move, the responsible custodian and the source evidence.
What may happen next? Whether the lot may enter a campaign, move to another custodian, be dispatched, or must remain quarantined pending review.[S03][S06]
The physical arrangement must make those answers credible. Dedicated areas can separate eligible from non-eligible stock. Temporal separation can use a defined campaign, a line-clearance check and a documented release before another material class enters. Under the opened CmiA edition, machinery must be free of non-claimed material before claimed material is processed, and precautions are needed against unintentional mixing. GFT would translate that principle into site instructions only after the applicable scheme, equipment and operating conditions have been confirmed.[S06]
A labelled aisle is a useful signal, but it is not a complete control. If bales can be moved without an identifier scan, written transfer or accountable sign-off, the location register can remain reassuring while becoming false. Likewise, replacing a damaged tag without preserving the original identifier can break the chain even where the replacement is physically correct. The proposed exception record links the old and new identifiers, the reason for replacement, the evidence inspected and the reviewer who authorised it.[S03][S06][S10]
Campaign cleaning needs the same discipline. The proposed record identifies the equipment, prior and next material classes, work performed, responsible role and release decision. Actual cleaning methods and acceptance criteria must be established for the site.[S03][S06]
Transport extends the same boundary beyond the warehouse. The opened CmiA standard calls for suitable packaging, containers or vehicles so that substitution or mixing cannot occur without tampering or damage. It also calls for sales and transport documents that identify corresponding products or materials. GFT therefore proposes recording package condition, vehicle or container reference, seal reference where used, exceptions observed at loading and the receiver's corresponding inspection. A seal number supports a physical check; it does not establish cotton origin or prove that every listed bale is present.[S06]
Quantities must reconcile on compatible measurement bases
Related · chapter 069 paragraphs step it
Related exhibit 27 · chapter 06Decision mapillustrativeLedger defects produce distinct physical and claim effects
01 / 06
Read-out
Method and source
Illustrative arithmetic from the base closing balance of 80 kg. Missing receipt 200: model -120, residual +200, no claim allocation. Duplicate receipt 200: model 280, residual -200, duplicate eligibility 200. Duplicate dispatch 150: model -70, residual +150, duplicate claim 150. Late return 60: model 20, residual +60, no claim allocation. Wrong conversion on 900 eligible kg: 900 x (0.80 - 0.70) = 90 kg unsupported, physical residual 0; both factors are hypothetical. Recovery misclassification: 100 kg incorrectly admitted, physical residual 0. No factor or effect is a scheme benchmark.
Unit: kg, illustrative residual and unsupported claim quantity. Status: illustrative. Sources: S03, S06, S10. Illustrative figures are teaching examples, not measured factory results.
Two scales can produce different observations without either being dishonest, and two documents can report the same nominal mass while using different tare or moisture conventions. The comparison must therefore record unit, gross or net status, packaging treatment, moisture basis, scale identity, observation time and any conversion formula. Textile Exchange's Content Claim Standard overview calls for documentation of receiving, processing and transfer, together with volume reconciliation. The CmiA edition likewise requires receiving documents to be checked against description, quantity, quality, source and blend or mix percentage.[S03][S06]
GFT's comparison routine begins with the original observations. In the illustrative worked case, 1,000 kg net at dispatch and 990 kg net at receipt are comparable only if both net values use compatible packaging and moisture bases. If one document reports gross mass, the routine cannot silently subtract an assumed bale or wrapping weight. It requests the actual tare observation or marks the comparison BASIS_UNKNOWN. Any unit conversion states its input, operation, output unit and rounding. These are proposed controls, not an assumed industry tolerance.[S03][S06]
Measurement uncertainty and administrative discrepancy must also remain separate. Measurement uncertainty concerns the observations and instruments. Administrative discrepancy includes a missing delivery line, a transposed identifier, an unrecorded return or an incompatible unit. A numerical match can hide an administrative error, while a numerical mismatch can have a documented physical explanation. Neither condition should be resolved by changing the receiver's mass to equal the sender's.[S03][S10]
The sign convention must be declared before interpretation. This chapter uses receipt minus dispatch for the transfer difference. A negative result means less mass was observed at receipt on the stated basis. It does not, by itself, establish physical loss, theft, moisture change or scale error. Investigation uses contemporaneous evidence such as weigh slips, packaging observations, vehicle inspection, returns and instrument records. The original pair remains intact throughout that review.[S03][S06][S10]
Exhibit 18 displays the worked net-mass pair without hiding the difference inside a single reconciled value.[S03][S06]
The single comparison in Exhibit 18 is enough to trigger review under this illustrative procedure, but not enough to determine its cause or materiality. The organisation must define investigation and escalation thresholds from the governing scheme, instrument performance, material behaviour and observed operating variation. The opened CmiA edition contains its own volume-deviation investigation rule, but that scheme-specific rule must not be relabelled as a general receiving tolerance or applied automatically to a different claim.[S06]
4.12 Resolution must preserve the evidence that created the exception
A discrepancy log should carry the original dispatch and receipt event IDs, difference and basis, evidence requested, responsible investigator, due review point, decision and claim effect. Where a correction is appropriate, EPCIS 2.0 provides error-declaration concepts for declaring an earlier event erroneous and linking corrective information. GFT would use that structure to retain history, while treating the business acceptance of the correction as a separate review.[S10]
A signed reconciliation adds evidence rather than replacing the underlying records. It states what was agreed, what remains uncertain and who authorised release. The quantity ledger records each physical implication once. The quantity-reconciliation chapter develops the full stock and eligible-claim accounts.[S03][S06]
4.134 min · 2 sub-steps · 1 exhibit
Unresolved handovers restrict downstream claims
Step 1 of 3Intro
Related exhibit 62 · chapter 14FlowillustrativeThe programme expands only after demonstrated control
01 / 08
Read-out
Method and source
Illustrative GFT decision rule. A No at any gate holds the affected claim and creates a remediation record. Independent acceptance covers only the reviewed scope and evidence cut-off. This is not a completed implementation or certification decision.
A handover needs states that describe evidence, not just the location of a truck. GFT proposes EXPECTED for a dispatch-linked transfer awaiting receipt, RECEIVED for a completed observation with no open handover mismatch, DISPUTED for a counterparty or documentary disagreement, QUARANTINED for material whose affected claim must not be released, and RELEASED for an authorised claim decision. These are local workflow labels. CBV supplies shared business steps and dispositions, but a partner crosswalk must verify the exact meaning before treating any local label as a standard equivalent.[S11]
The states are not necessarily sequential for every attribute. A lot may be physically received and immediately quarantined for a provenance question. Its quality review may continue while its scheme claim remains blocked. A disputed commercial document may coexist with a complete physical receipt. Recording one generic accepted flag would conceal those distinctions and could make an unresolved handover appear to support a public origin statement.[S01][S03][S06]
The discrepancy decision tree begins with material identity, because quantity arithmetic cannot repair an unknown lot:
¶ 03
Can the receiver resolve every claim-critical material identifier? If not, isolate the affected items where practicable, mark their identity UNKNOWN, and quarantine the dependent claim. Do not assign the expected lot merely because the package count matches.[S03][S06]
Do the parties, sites and custodian match the authorised route? If not, establish who actually observed and accepted the material. An unexpected subcontractor or storage site requires scope and custody review, not an address edit.[S01][S06]
Are quantities compared on a compatible basis? If not, obtain the missing tare, unit or moisture information. Keep BASIS_UNKNOWN visible rather than importing a standard deduction.[S03]
Does documentary evidence support the claimed class and transaction? If not, request the relevant original records and record their issuer and scope. A certificate or invoice is reviewed for the proposition it can support, not accepted by title alone.[S03][S06]
Can the parties sign one resolution without erasing either observation? If yes, retain the resolution and corrective event, then repeat the affected checks. If no, escalate with both positions and keep the claim restricted.[S10]
Which downstream records inherit the restriction? Trace dependent transformations and dispatches, but limit the hold to the affected attribute and material set. Unknown is not evidence of wrongdoing, and it is not zero.[S01][S03][S10]
Exhibit 19 applies that decision tree to illustrative exceptions. Each row carries a status, accountable role, evidence requirement and release consequence so that closure means a reviewable decision rather than a disappeared warning.[S06][S10][S11]
The cases in Exhibit 19 show why deletion is not correction. A duplicate message should not double the dispatched quantity, but removing all copies from the audit trail would also destroy evidence of the ingestion defect. EPCIS error declarations allow an erroneous event to be identified with corrective references. The proposed operational log additionally records who reviewed the business effect and whether any claim or stock allocation changed.[S03][S10]
4.15 Claim release needs a measurable queue, not an improvised chase
4.15
The release reviewer should work from an exception queue with explicit denominators and observation sources. GFT proposes four pilot measures, each owned by the traceability lead and reviewed weekly during the bounded pilot. These are measurement specifications, not reported performance. An empty queue caused by missing data is reported as UNKNOWN, never as a perfect result.
¶ 06
Proposed measure
Numerator and denominator
Observation source
Decision use
Unmatched handover share
Open or unresolved matched pairs divided by all handovers due for review
Dispatch-receipt register and exception log
Directs investigation capacity; unknown required records stay in the denominator with their status visible.
Receipt observation completeness
Required receipt fields populated from actual observation divided by required fields across due receipts
Signed receipt forms and field audit
Distinguishes a working capture procedure from copied dispatch data.
Exception closure evidence share
Closed exceptions with retained evidence and authorised decision divided by all closed exceptions
Exception log and linked evidence index
Prevents closure by deletion or unsupported supervisor override.[S10]
Restricted-claim propagation completeness
Affected downstream records correctly restricted divided by affected records identified through genealogy
Custody graph and claim-release register
Tests whether a handover hold reaches every dependent claim and no unrelated attribute.[S01][S03]
The denominator, owner and cadence should be frozen before the pilot. Results must include the review period, participating sites and unknown-status count. No threshold is asserted here; buyer or scheme requirements and pilot acceptance criteria need separate agreement.[S03][S06]
An unresolved handover ultimately narrows what the buyer can receive. The evidence pack should show the matched pair, any open exception and the exact claim restriction. If the transfer cannot establish that the claimed material crossed the boundary, the public origin statement is withheld or narrowed to the remaining supported boundary. This does not prove that the original declaration was false. It records that the chain has not preserved enough evidence to repeat it.[S01][S03]
4.163 min · 2 sub-steps · read with a related exhibit
Controls must work across a border and a weak connection
Related · chapter 0110 paragraphs step it
Related exhibit 02 · chapter 01Decision mapillustrativeDifferent labels require different evidence
01 / 05
Read-out
Method and source
Illustrative conceptual placement designed by GFT from S01, S02, S18 and S34. Coordinates place five product propositions by required upstream physical custody depth (x) and rule or verification specificity (y); positions are analytical categories, not empirical scores.
Unit: illustrative position (0 to 10 conceptual scale). Status: illustrative. Sources: S01, S02, S18, S34. Illustrative figures are teaching examples, not measured factory results.
4.17 Transport evidence and customs origin answer different questions
A border crossing adds documents, parties and possible changes of custodian, but it does not change the matching principle. Material identifiers must remain linked to transport and receipt records, while each organisation records what it actually observed. The World Trade Organization defines rules of origin as criteria for determining a product's national source. That legal or customs determination is distinct from evidence that a particular bale or lot moved between two sites. The handover pack must not infer tariff preference or customs eligibility from a delivery note.[S18][S10]
For an illustrative cross-border shipment, the seller's dispatch, carrier acceptance, border-related references, any warehouse transfer and final receipt should share a transfer crosswalk. Each document keeps its issuer and purpose. A customs reference can identify a declaration; a packing list can identify packages; a weigh slip can support a quantity observation. None should be treated as universal proof of every other field. Applicable origin analysis belongs with the relevant legal and trade review, not with an automatic passport rule.[S18][S03]
Weak connectivity creates a different risk: the physical event may be valid but its digital message may arrive late, twice or out of order. EPCIS event data can preserve event time and business context, while application records can separately preserve recording time. GFT's proposed offline procedure uses pre-issued transfer and event identifiers, protected local storage and a signed paper or local electronic fallback. A later upload carries the original identifiers rather than creating a new business event at the time of reconnection.[S10][S11]
Before departure, the sending custodian downloads the minimum transfer record, authorised material list and evidence references needed to work locally. The receiver obtains the expected transfer identifier where the communication channel permits. If no expectation arrives, the receiver still captures an independent observation and marks the transfer UNMATCHED, rather than reconstructing the expected record from memory. These are proposed pilot controls whose usability and security still require testing.[S03][S10]
At reconnection, four reconciliations are required:
Message reconciliation. Compare local queues with received partner messages by event identifier. Repeated uploads must not create repeated business events or quantity allocations.[S10]
Chronology reconciliation. Preserve event time, local time offset and recording time. Investigate impossible sequences without replacing uncertain timestamps with the server arrival time.[S10][S11]
Physical reconciliation. Compare actual stock, packages, vehicle or return observations with the reconstructed transfers. Message delivery alone is not custody confirmation.[S03][S06]
Claim reconciliation. Re-run restrictions and releases against the complete evidence set. A claim held while offline must not become releasable merely because the connection has returned.[S01][S03]
Conflict handling should prefer evidence over recency. If two records share an event identifier but carry different material IDs or quantities, the system quarantines the conflict and retains both payloads, their senders and their recording times. If two different event identifiers describe what may be one physical dispatch, a reviewer compares source documents before merging any operational view. The merge creates a resolution record; it does not delete either original assertion.[S10]
4.18 Recovery tests must include the physical ledger
The pilot should rehearse a delayed dispatch message, a duplicated receipt upload, an out-of-order return and an offline identifier replacement. Each case needs expected physical stock, expected claim status and independently prepared source records for comparison. Success means that another reviewer can reconstruct the handover, the quantities and the claim decision after reconnection. It does not mean that every message reached a server.[S03][S10]
The rehearsal should also test controlled file exchange. A signed export carries transfer identifiers, event times, material references, evidence pointers and correction links; the importer records conversions and repeats matching. Companion papers wp07 and wp18 own infrastructure design, which must preserve custody meaning and restricted release during recovery.[S10][S11]
Check the record
The chapter keeps its sources within reach.
Every knot is a source this chapter cites; select one to read its record, or filter the register by grade. Publication dates and inherited-source limits stay explicit. Read the paper's full evidence summary.
8 sources · 152 citation marks on this pageMost cited: S10, 36 times
Grade AGrade BGrade CKnot size: citations on this page
All 8 records shown
c04-S01 · Grade A
International Organization for Standardization
ISO 22095:2020: Chain of custody, General terminology and models
This document is applicable to all materials and products.
2020. Register access date: 2026-10-30. Source provenance is supplied by the paper evidence register.
Move the shuttle along the rack to preview any chapter: its sections, exhibits and reading time. Draft web edition. This research is not investment, legal or tax advice; illustrative models are not measured factory results.
The paper's rack · 14 boltsBolt length: reading time
1.1
1.3
1.5
1.7
1.10
1.12
01Part I: Claims must match the chain
Cotton value needs evidence that survives the factory
GFT argues that provenance becomes useful when custody controls preserve the evidence behind a saleable claim.
1.1The local-processing gap makes provenance commercially relevant
1.3A provenance claim is a proposition that needs a boundary
1.5A passport distributes evidence without creating it
1.7Greenfield design creates a chance to capture evidence early
The passport should disclose a bounded product record
A useful passport identifies its product, states who makes each claim and reveals the limits of the supporting evidence. Proposed fields must remain distinct from future legal requirements.
9.1Current policy does not settle the final textile schema
9.2A proposed passport needs a field-level evidence map
9.4A data carrier points to information and can be copied
9.5Public and restricted views need different disclosures
Digital signatures preserve assertions while trust needs governance
Cryptography can preserve an assertion and identify its signing authority, but accepting a textile claim still requires physical evidence and an explicit trust policy.
10.1An issuer's identity does not prove a physical claim
10.2Credential verification needs more than a signature check
10.3Identifiers need a recovery and control model
10.4Revocation changes what a reader should accept