Research libraryEdge to Dashboard
Chapter 09 of 14 · Part III · Draft web edition
The factory's data store is an evidence system, not a growing pile of values, so raw observations, contextual events, KPI results and corrections remain distinguishable and replayable.
The cloth is woven from this chapter's 5 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.
Read-out
Illustrative GFT historian store architecture informed by ISO 22400 KPI structure (S08), EPCIS visibility event sharing (S15), ISO/IEC 27001 change control (S24) and edge database persistence (S29).
Unit: . Status: illustrative. Sources: S08, S15, S24, S29. Illustrative figures are teaching examples, not measured factory results.
A factory historian earns its authority as a system of record only when it stores raw machine observations, contextualised operational events, versioned key performance indicator calculations and audited human corrections in distinct, traceable layers rather than overwriting values inside one mutable table.[S08][S15][S24][S0902] In a connected textile plant operating under intermittent grid power and variable wide-area network links, telemetry rarely arrives in an uninterrupted chronological stream.[S32][S40] Gateways buffer readings during blackouts, wireless sensors experience clock drift or radio contention, and supervisors reclassify defective garment bundles after physical re-inspection.[S26][S29][S31] If the central store treats arrival time as occurrence time, silently interpolates across communication gaps, or lets a manual adjustment erase the original sensor reading, management cannot reconstruct why a production decision was taken.[S0901][S0903] Green Future Tech Limited (GFT), Hong Kong SAR, therefore specifies the historian as an engineered chain of evidence that preserves context, quality state and provenance from the machine edge to the management review.[S08][S0904]
This discipline directly supports GFT's three-part Cameroon textile programme (feasibility study, performance-management system design and digital strategy, prepared 31 January 2024 with source-refresh notes retained in 2026). Cameroon's inherited sector baseline records roughly 95 percent of cotton fibre exported raw and about 5 percent processed locally, while national policy targets 50 percent local processing by 2030.[R06][R07] Closing that gap requires plants that prove control over utilisation, first-pass quality and resource intensity from their first shifts, avoiding the unmeasured underutilisation that historically cost domestic mills billions of CFA francs (XAF).[R12] Where a plant starts with a manual or spreadsheet system of record for daily shift huddles, transitioning to automated capture must not sever historical continuity or alter indicator formulas.[S08][S30] Preceding chapters defined how operational meaning travels across protocols and semantic models; this chapter specifies how the storage spine persists, reconciles, retains and recovers that meaning.[S04][S07][S15][S40]
9.1
Related · chapter 117 paragraphs step it
Read-out
Illustrative GFT architecture, not a normative schema. ISO 23247-4 informs information exchange. ISO 8000-120 is limited to characteristic master-data provenance and excludes change control. CSF informs risk governance, not the exact release fields.
Unit: proposed evidence responsibilities. Status: illustrative. Sources: S06, S41, S18. Illustrative figures are teaching examples, not measured factory results.
Read this exhibit in its chapter ↗An industrial data spine fails its first audit test when it conflates what a physical instrument measured with what execution logic concluded.[S01][S15] A current transformer on a ring-spinning frame, an optical counter on a sewing station, or a level sensor on a dye vat produces a raw observation: a value bound to a unit, a source timestamp, a receipt timestamp, a quality code and a point identifier.[S27][S29][S0903] That observation does not know which customer order was running, which crew was assigned, or whether the machine was in maintenance warm-up rather than commercial production.[S01][S02] Those attributes belong to the interpreted event layer, where raw signals join manufacturing operations management context from Level 3 schedules, work-in-process (WIP) scans and quality specifications.[S02][S25][S28] Storing both layers separately ensures that if a work-order mapping table was misconfigured at shift handover, engineers can re-run the contextualisation rule against untouched raw observations without losing physical telemetry.[S24][S0902]
At the lowest persistence tier, the raw observation store acts as an append-only ledger of instrument readings as they crossed the acquisition boundary.[S29][S0903] In the OPC Unified Architecture (OPC UA) Historical Access specification (OPC 10000-11), RawData and HistoricalDataNodes are queried through their SourceTimestamp, while allowing clients to inspect both SourceTimestamp and ServerTimestamp (BOTH) so transport delay never masks the instant of physical measurement.[S04][S0903] Each raw record retains the asset-hierarchy point identifier, engineering value, unit symbol, source clock reading, receiving timestamp and acquisition status code, with no downstream filter or script permitted to overwrite it.[S01][S24][S0902]
Interpreted production events belong in a separate contextual event store.[S15][S16] GS1's Electronic Product Code Information Services (EPCIS Release 2.0) standard enables disparate applications to create and share visibility event data within and across enterprises, capturing the what, when, where and why of object movements.[S15] In a denim or garment plant using radio-frequency identification (RFID) at cutting and sewing checkpoints, an antenna read is a raw observation that a tag passed within radio range.[S25][S31] The interpreted event asserts that bundle B-1042 completed sewing operation OP-20 on Line 2.[S25][S28] Because proximity reads can bounce at workstation boundaries, transforming raw RFID reads into an EPCIS visibility event relies on a versioned state-machine rule that can be replayed cleanly if the raw log remains intact.[S15][S31][S0902]
The separation extends upward to calculated key performance indicators (KPIs) and outward to media and audit records, forming four distinct persistence stores across the factory spine, as shown in Exhibit 39.[S08][S15][S24][S29]
As Exhibit 39 illustrates, each store enforces its own mutability contract.[S08][S15][S24][S29] The raw observation store records machine counters, metre totals and sensor readings at the edge and central archive.[S14][S29] The contextual event store binds observations to ISA-95 and IEC 62264 Level 3 entities: work centres, orders, lots, shifts and EPCIS milestones.[S01][S02][S15] Above it, the versioned KPI store holds reproducible ISO 22400 results, while the audit and media store preserves configuration releases, override approvals and selected inspection images linked by reference identifiers.[S08][S24][S26] Because Exhibit 39 is an illustrative architectural model, it defines logical boundaries rather than prescribing a specific database product or claiming measured plant throughput.[S08][S29]
Engineers must also distinguish which standards govern provenance.[S41][S0904] The World Wide Web Consortium (W3C) PROV overview defines provenance as information about entities, activities and people involved in producing data, used to assess quality, reliability or trustworthiness.[S0904] Within industrial standards, however, ISO 8000-120:2016 specifies provenance requirements strictly for master data consisting of characteristic data, supplementing ISO 8000-110.[S41] Characteristic master data covers durable asset properties, material grade specifications and nameplate parameters such as those inside an Asset Administration Shell (AAS) submodel, not telemetry streams or Sparkplug states.[S07][S40][S41]
GFT therefore applies ISO 8000-120 where the factory exchanges characteristic master data, while specifying telemetry transformation and KPI lineage as GFT's own architecture rule governed under ISO/IEC 27001 change control and OPC 10000-11 modification tracking.[S07][S24][S41][S0902] Furthermore, OPC 10000-11 Section 4.4 notes that multiple historical events can share a timestamp; a historian storing events must preserve the EventId generated during event processing and must never allow a modification to alter that EventId.[S0902]
9.5
A numeric value stripped of its acquisition health misleads automated dashboards and supervisors alike.[S31][S40] If a steam-flow metre freezes because a serial converter lost power, or if a warehouse humidity node transmits with an unsynchronised clock, storing the bare number as an ordinary observation corrupts downstream averages and alerts.[S31][S40][S0901] Quality state cannot be an optional comment added during monthly data cleaning.[S08][S24] It must travel as a mandatory, machine-readable attribute inside every observation and event envelope, governing whether the historian may feed that value into a live tile, alert or KPI calculation.[S08][S40][S0901]
In MQTT publish-subscribe architectures, report-by-exception publishes updates only when values change.[S05][S40] Without session-state management, a subscribing historian cannot tell whether a quiet metric is constant or offline.[S40] The Eclipse Sparkplug Specification (Version 3.0.0) resolves this through birth and death certificates (NBIRTH, DBIRTH, NDEATH, DDEATH).[S40] When an Edge Node disconnects and its NDEATH message is published, Sparkplug specifies that all metrics in the associated NBIRTH and DBIRTH messages under that Edge Node are immediately implied to be STALE.[S40]
The historian must persist this transition as an explicit time-stamped state change.[S40][S0901] In OPC 10000-11 Section 4.3, when data collection breaks and recovery reveals a discontinuity between the last stored SourceTimestamp and the newly arrived initial SourceTimestamp, the historian must record a bad status at the break timestamp and report gaps to querying clients with Bad_DataLost.[S0901] Similarly, if an event subscription buffer overflows during an interruption, the historian must record EventQueueOverflowEventType or Bad_DataLost so queries never assume an incomplete event stream is whole.[S0901]
Wireless sensor networks and shop-floor RFID deployments depend on node time synchronisation and spectrum-sharing discipline to protect data integrity and avoid uncertainty in sensed data.[S31] If an edge gateway loses lock with its Network Time Protocol (NTPv4, capable of tens-of-microseconds accuracy on fast LANs with modern workstations) or Precision Time Protocol (IEEE 1588-2019, capable of sub-microsecond accuracy with minimal resources) reference, its local clock drifts relative to neighbouring machines.[S12][S13][S31] Unsynchronised readings may still count shift totals, but they cannot correlate sub-second stoppages across linked machines.[S12][S31]
Computer-vision inspection cells carry a similar physical dependency: textile defect-detection research shows that lighting variations and background noise significantly degrade machine-learning detection accuracy.[S26] When illumination dims or a lens fouls, the historian must capture the resulting degraded-acquisition flag alongside the inspection event so a drop in flagged defects is recognised as a sensor fault rather than improved cloth quality.[S26][S40]
To standardise how services and dashboards treat degraded readings, GFT derives a six-state quality dictionary from the underlying protocols and security standards, presented in Exhibit 40.[S24][S31][S40][S0901]
Read-out
| GOOD_LIVE | Active session and synchronised clock | Persist raw value and source time | Eligible for live display and KPI calculation |
|---|---|---|---|
| STALE_OFFLINE | Sparkplug NDEATH or DDEATH received | Retain last value with offline marker | Dim tile and suppress threshold alerts |
| UNSYNCHRONISED | Clock stratum lost or radio contention | Store with receipt time and skew warning | Exclude from sub-second correlation |
| BAD_DATA_LOST | Gap detected after collection break | Insert explicit gap boundary record | Block linear interpolation across gap |
| ESTIMATED | Approved fallback calculation applied | Store separately from raw observation | Display with provisional badge only |
| CORRECTED | Authorised update with reason and approver | Append new version and retain original | Use latest approved version with audit link |
Derived quality-state dictionary combining Sparkplug session and stale-metric semantics (S40), wireless clock and spectrum risks (S31), OPC UA historical gap reporting (S0901) and controlled correction audit rules (S24).
Unit: . Status: derived. Sources: S24, S31, S40, S0901.
The derived rules in Exhibit 40 translate protocol diagnostics into deterministic storage and display rules.[S24][S31][S40][S0901] Only GOOD_LIVE observations enter real-time threshold alerts and unflagged KPI aggregations.[S08][S40] When STALE_OFFLINE or BAD_DATA_LOST is active, the historian blocks linear interpolation across the gap, preventing queries from fabricating synthetic production across a power outage.[S32][S40][S0901] Where ESTIMATED or CORRECTED values enter under ISO/IEC 27001 change control, they are stored as versioned records so auditors can compare the raw stream against the adjusted operational view.[S24][S0902] Because Exhibit 40 is a derived governance dictionary, edge gateways must map device-specific fault codes (such as Modbus exception responses) into these six canonical states.[S09][S40]
9.9
Persist in edge buffer
Persist in edge buffer. Starting point. Leads to Detect host STATE online (WAN outage). · illustrative
An architecture that relies on a single central or cloud database to ingest machine signals directly loses evidence whenever site power or external links fail.
Detect host STATE online
Detect host STATE online. From Persist in edge buffer (WAN outage). Leads to Flush historical batch (Link restored). · illustrative
Flush historical batch
Flush historical batch. From Detect host STATE online (Link restored). Leads to Check EventId and source time (Replay stream). · illustrative
Check EventId and source time
Check EventId and source time. From Flush historical batch (Replay stream). Leads to Insert into time domain (Unique record). · illustrative
Insert into time domain
Insert into time domain. From Check EventId and source time (Unique record). Leads to Publish new KPI version (Close gap). · illustrative
Publish new KPI version
Publish new KPI version. From Insert into time domain (Close gap). End point. · illustrative
Illustrative store-and-forward replay sequence based on edge persistence (S29), Sparkplug host STATE and historical flush rules (S40), immutable OPC UA EventId tracking (S0902) and controlled recalculation (S24).
Unit: . Status: illustrative. Sources: S24, S29, S40, S0902. Illustrative figures are teaching examples, not measured factory results.
9.9 · the bodyExhibit 41 above; its figures are illustrative where marked.
An architecture that relies on a single central or cloud database to ingest machine signals directly loses evidence whenever site power or external links fail.[S14][S16][S32] In the World Bank's Cameroon 2024 Enterprise Survey (fieldwork June 2024 to February 2025 across formal private non-agricultural firms), 93.3 percent of surveyed firms experienced electrical outages, averaging 10.4 outages in a typical month.[S32] Although this covers all surveyed firms rather than textile mills alone, it confirms that power and communication interruptions are routine operating conditions.[S32] GFT therefore divides persistence duties between a local edge buffer in the production security zone and a central historian in the operations zone.[S16][S19][S29]
The edge buffer's duty is uninterrupted local capture and immediate line-level calculation.[S16][S29] ETSI's Multi-access Edge Computing (ETSI GS MEC 003) runs software applications on virtualised infrastructure close to the network edge, while W3C's Web of Things (WoT Architecture 1.1) describes an industrial edge device collecting selected controller data for local analysis and downstream monitoring.[S14][S16] A peer-reviewed industrial energy-monitoring system demonstrates this pattern: when the edge receives metre data, it stores the reading in a local database server and computes nine additional analytical parameters locally before transfer.[S29]
Writing raw observations and local parameters to non-volatile storage at the edge node (backed by uninterruptible power) keeps line huddles, local human-machine interfaces (HMIs) and inspection cells running when the plant network or WAN is down.[S17][S26][S29] An illustrative edge gateway polling 50 registers once per second (1 Hz) with an illustrative 120-byte compressed envelope writes 6,000 bytes per second, or 518.4 megabytes per day (50 * 120 * 86,400 = 518,400,000 bytes). A 32-gigabyte industrial solid-state partition holds weeks of illustrative telemetry and cropped defect images without dropping primary evidence.[S26][S29][S40]
While the edge store optimises for write survivability, the central historian optimises for cross-line reconciliation, long-term query, enterprise-control integration under ISA-95 and IEC 62264, and governed reporting.[S01][S02][S29] Under Sparkplug 3.0, when a Primary Host Application publishes an online STATE message, the Edge Node resumes live publishing and flushes historical data stored while offline.[S40]
If the central historian indexes messages by broker arrival time (ServerTimestamp), a three-hour flush from Line 1 will be interleaved at the reconnection instant rather than placed in the afternoon window when the garments were sewn.[S40][S0903] The central ingestion service must therefore route replayed batches through the identity and source-time reconciliation workflow shown in Exhibit 41.[S24][S29][S40][S0902]
As Exhibit 41 shows, the edge node writes every observation and event locally with its SourceTimestamp, sequence counter and EventId.[S29][S40][S0902] Once the primary host publishes its online STATE message, the edge node flushes its queue in bounded batches so replay traffic does not starve live operational telemetry.[S17][S40] The central service checks the composite key (asset_id, point_id, SourceTimestamp, EventId), inserts unique records into the OPC 10000-11 time domain, clears BAD_DATA_LOST markers where continuity is proven, and triggers a versioned KPI recalculation.[S08][S40][S0901][S0903] Because Exhibit 41 is an illustrative workflow, it defines the required logical steps without promising a fixed replay duration, which depends on link bandwidth and outage length.[S14][S40]
If an acknowledgement packet drops during a historical flush, the edge node re-transmits the unacknowledged batch.[S05][S40] To prevent double-counting output, the historian enforces two idempotency rules.[S0902][S0903] For scalar observations (HistoricalDataNodes), an identical (asset_id, point_id, SourceTimestamp, source_version) tuple is treated as an idempotent retry that updates only the replication log.[S0901][S0903] For discrete production and visibility events (HistoricalEventNodes and EPCIS events), uniqueness is enforced on the immutable EventId generated at the edge (OPC 10000-11 Section 4.4).[S15][S0902] Machine cycle counters are also published as monotonic cumulative totals alongside delta increments so any unrecovered gap still reveals the exact number of elapsed cycles, flagged as ESTIMATED until verified.[S08][S24][S0901]
Interlude · the cloth half-woven
Woven: sections 9.0 to 9.9Still to weave: 9.13 to 9.21
Read-out
GFT arithmetic from World Bank Enterprise Surveys Cameroon 2024 Country Profile (S32). Outage-incidence gap against Sub-Saharan Africa: 93.3 percent minus 69.7 percent equals 23.6 percentage points. Outage-incidence gap against lower-middle-income economies: 93.3 percent minus 55.0 percent equals 38.3 percentage points. Monthly frequency excess: 10.4 minus 6.9 equals 3.5 outages per typical month. Compares all surveyed firms, not textile factories.
Unit: percentage points and monthly events. Status: derived. Sources: S32.
Read this exhibit in its chapter ↗9.13
Provisional central
Provisional central: total 820. Running total 820 illustrative garment count · illustrative
When a plant manager reviews weekly Overall Equipment Effectiveness ( OEE ), first-pass yield, or kilowatt-hours per kilogram of fabric, that number is the output of a formula evaluated over selected observations, events…
Edge buffer flush
Edge buffer flush: adds 340. Running total 1,160 illustrative garment count · illustrative
Duplicate filter
Duplicate filter: removes 18. Running total 1,142 illustrative garment count · illustrative
Quality-hold reject
Quality-hold reject: removes 22. Running total 1,120 illustrative garment count · illustrative
Verified version 2
Verified version 2: total 1,120. Running total 1,120 illustrative garment count · illustrative
The method behind it: Method.
Illustrative shift reconciliation case showing how a provisional central count of 820 garments is updated by +340 buffered RFID reads (S29, S40), -18 duplicate scans, and -22 quality-hold rejects (S08, S24) to reach 1,120 verified good garments.
Unit: illustrative garment count. Status: illustrative. Sources: S08, S24, S29, S40. Illustrative figures are teaching examples, not measured factory results.
9.13 · the bodyExhibit 42 above; its figures are illustrative where marked.
When a plant manager reviews weekly Overall Equipment Effectiveness (OEE), first-pass yield, or kilowatt-hours per kilogram of fabric, that number is the output of a formula evaluated over selected observations, events, time boundaries and exclusion rules.[S08][S29] If dashboards and spreadsheets compute KPIs independently with drifting queries, review meetings stall over conflicting numbers.[S08][S30] The historian must store KPI results as versioned calculation artefacts locked to a governed dictionary.[S08][S24]
ISO 22400-2:2014 presents manufacturing KPIs through their formula, source elements, time behaviour, unit or dimension and related characteristics.[S08] In GFT's architecture, every stored KPI record must carry five provenance attributes: 1. The indicator identifier and version (KPI-OEE-LINE-V1.2, specifying numerator and denominator elements under ISO 22400-2).[S08] 2. The target entity in the ISA-95 hierarchy and the operational scope (order, lot or shift).[S01][S02] 3. The evaluation window (window_start_utc and window_end_utc on SourceTimestamp) and execution time (computed_at_utc).[S08][S0903] 4. The input coverage ratio (valid GOOD_LIVE and approved CORRECTED events divided by expected events) and worst-case quality flag.[S08][S40][S0901] 5. A hash or snapshot pointer to the input event set and the calculation version (calc_version = 1, 2, ...).[S24][S0904]
As OPC 10000-11 Section 4.4 notes, historical data may require updating because of a failed sensor, communication drop, calibration error, or offline inspection and laboratory findings arriving after the shift.[S0902] A quality supervisor may re-inspect 30 fabric rolls flagged by a line-scan camera under variable lighting ([^S26], [^R29]), confirm 22 defective rolls, and release 8 rolls as first-quality goods.[S26][S0902]
When such an update enters, OPC 10000-11 requires recording the original value and who made the change (HistoryUpdateDetails) and generating an audit event (AuditHistoryValueUpdateEventType or AuditHistoryEventUpdateEventType).[S0902] Under ISO/IEC 27001, GFT requires an authenticated user identity, reason code, evidence link and role approval.[S24] Rather than overwriting Version 1 of a shift KPI, the historian appends Version 2 and marks Version 1 SUPERSEDED, preserving what was visible at the morning huddle alongside the verified weekly record.[S08][S24][S0902]
The signed contributions from this illustrative shift case appear in Exhibit 42, reconciling the provisional central count to the verified closing balance.[S08][S24][S29][S40]
As Exhibit 42 shows, each step between Version 1 (820 illustrative garments) and Version 2 (1,120 illustrative garments) maps to a distinct mechanism: edge persistence (+340), EventId deduplication (-18) and audited quality disposition (-22).[S08][S24][S29][S40][S0902] Because Exhibit 42 is an illustrative method case, it demonstrates reconciliation logic rather than reporting live factory output.[S08]
9.17
A historian cannot keep every raw camera frame, millisecond sample, visibility event and KPI record for the same duration on the same storage tier.[S26][S29][S36] Retaining uncompressed vision streams indefinitely exhausts storage budgets, whereas purging production and resource events after thirty days destroys the evidence required for quality claims, life-cycle carbon accounting and digital product passports.[S15][S26][S36][S39] Retention policy must classify records by evidentiary obligation, reconstruction cost and byte volume rather than applying one uniform retention period.[S24][S36][S39]
Sustainability and traceability standards show why monthly plant rollups cannot replace batch and order histories.[S36][S38][S39] The GHG Protocol Product Life Cycle Accounting and Reporting Standard quantifies greenhouse-gas emissions and removals associated with a specific product across its life cycle.[S36] Similarly, ISO 14046:2014 specifies life-cycle water-footprint assessment for products, processes and organisations, while ISO 50001:2018 governs energy management systems and baselines.[S37][S38] Allocating electricity ([^S29]) or water ([^S27]) to a garment lot requires preserving the production denominators (good units, rework and run windows) at the lot and order boundary ([^S08], [^S15]).[S36][S38]
In the European Union, Regulation (EU) 2024/1781 (ESPR) defines a digital product passport as data specific to a product, specified in the applicable delegated act under Article 4 and accessible electronically through a data carrier.[S39] Preserving export-ready evidence requires keeping EPCIS visibility events, lot genealogy and batch resource allocations across commercial horizons rather than discarding them while garments remain in the supply chain.[S15][S36][S39]
At the high-volume extreme, industry inspection literature reports line-scan fabric inspection running at up to 60 metres a minute across more than 40 defect types.[R29] Archiving every raw frame from defect-free cloth wastes capacity without adding analytical value.[S26][R29] GFT therefore positions each data class by storage cost per record versus required evidence horizon, as shown in Exhibit 43.[S08][S15][S24][S26][S36][S39]
Read-out
Illustrative positioning of six factory data classes by storage cost per record (x-axis) and long-term audit or life-cycle evidence horizon (y-axis), informed by S08, S15, S24, S26, S36 and S39. Actual retention durations require factory policy approval.
Unit: illustrative index (0 to 10). Status: illustrative. Sources: S08, S15, S24, S26, S36, S39. Illustrative figures are teaching examples, not measured factory results.
As Exhibit 43 illustrates, raw defect-free camera frames (x = 9, y = 2 illustrative index) stay in an edge circular buffer for short-term diagnostics before pruning, while cropped defect images (x = 7, y = 6) are indexed by EventId in selective archives for claim review and model retraining.[S16][S26][S0902] High-rate telemetry (x = 5, y = 3) remains at full resolution in warm storage before compaction into validated rollups and state boundaries.[S29][S0901] In the upper-left quadrant (Durable compact ledgers), EPCIS visibility events (x = 2, y = 9), versioned ISO 22400 KPI results (x = 1, y = 8) and ISO/IEC 27001 / OPC 10000-11 audit logs (x = 2, y = 10) combine small byte footprints with multi-year life-cycle and compliance value.[S08][S15][S24][S36][S39][S0902] Because Exhibit 43 uses an illustrative zero-to-ten scale, exact retention durations remain a site governance decision set with legal and commercial owners.[S24][S39]
Long-term archives fail if they depend on a proprietary historian licence that the plant may later replace.[S04][S24] Cold-tier archives and compliance exports must be written in open, self-describing formats bundled with their ISA-95 hierarchy snapshot ([^S01]), AAS submodels ([^S07]), ISO 22400 KPI dictionary version ([^S08]), EPCIS vocabulary context ([^S15]), ISO 8000-120 master-data provenance records ([^S41]) and a SHA-256 manifest logged under ISO/IEC 27001 ([^S21], [^S24]).
9.21
The final test of a historian is how it recovers from server failure, storage corruption or cybersecurity isolation.[S18][S21][S22] Checking only that a nightly backup script completed is insufficient: a network-accessible backup share can be encrypted by the same malware that hit the primary server ([^S21]), and restoring a noon snapshot at 18:00 leaves a six-hour gap between the database and the shop floor ([^S22], [^S40], [^S0901]). True recovery requires isolated restoration followed by evidence reconciliation.[S21][S22]
CISA (Advisory AA22-110A) advises operators to test backup procedures regularly and isolate backups from network connections that could spread malware.[S21] Within an IEC 62443-3-2 zone-and-conduit design ([^S19]) aligned with NIST SP 800-82 Rev. 3 ([^S17]) and NIST CSF 2.0 ([^S18]), the backup store must sit behind a controlled conduit with separate credentials and immutable retention locks ([^S19], [^S21], [^S24]). To rebuild the system of record cleanly, the isolated backup package must include six artefacts: 1. Raw observation and contextual event partitions up to the checkpoint ([^S15], [^S29], [^S0903]). 2. The versioned ISO 22400 KPI store and formula dictionary history ([^S08]). 3. The ISA-95 hierarchy, AAS submodels and ISO 8000-120 master-data provenance tables ([^S01], [^S07], [^S41]). 4. The OPC 10000-11 modification records, quality-flag rules and ISO/IEC 27001 audit logs ([^S24], [^S0902]). 5. Edge-gateway configuration manifests, Sparkplug topic contracts and broker access lists ([^S19], [^S21], [^S40]). 6. The restoration runbook and integrity-verification scripts ([^S21], [^S22]).
When a historian is restored from a backup taken at $t_{\text{backup}}$ and brought back online at $t_{\text{restore}}$, it initially lacks all data from $[t_{\text{backup}}, t_{\text{restore}}]$.[S21][S22][S0901] Because edge buffers continued recording machine cycles, energy metres and RFID scans during the central outage ([^S25], [^S29], [^S40]), recovery personnel execute a four-step catch-up sequence before clearing degraded-mode status:
SourceTimestamp and sequence number as of $t_{\text{backup}}$ ([^S01], [^S40], [^S0903]).[^S40]) and replay edge-buffered observations and events ([^S29]) across $[t_{\text{backup}}, t_{\text{restore}}]$ using the EventId and composite-key deduplication rules in Exhibit 41 ([^S0902]).[^S22], [^S30]), inserting validated entries as ESTIMATED or CORRECTED with approver audit events ([^S24], [^S0902]) and marking any unrecoverable gap BAD_DATA_LOST ([^S0901]).ISO 22400 KPIs ([^S08]) as new calculation versions linked to the recovery ticket ([^S22], [^S24]).NIST SP 800-34 Rev. 1 distinguishes three activities: testing validates recovery capabilities, training prepares personnel for plan activation, and exercising identifies planning gaps to improve preparedness.[S22] GFT requires a live restoration and reconciliation exercise on a staging instance before operational acceptance ([^S21], [^S22], [^S30]), measuring actual Recovery Time Objective (RTO) and Recovery Point Objective (RPO) after edge replay and manual log reconciliation and logging any variance as a corrective action under ISO/IEC 27001 ([^S08], [^S22], [^S24]).
By separating raw observations from interpreted events, persisting quality state as operational data, pairing edge buffers with identity-based central reconciliation, locking KPIs to ISO 22400 formula versions, tiering retention by evidence need, and proving recovery through exercised reconciliation, the factory turns its historian into a trustworthy spine for the decision dashboards examined next.[S02][S08][S15][S22][S40]
Check the record
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.
40 sources · 329 citation marks on this pageMost cited: S40, 34 times
Grade AGrade BGrade CKnot size: citations on this page
All 40 records shown
R06 · Grade C · inherited record
Inherited evidence R06
About 95 percent of cotton fibre exported raw against 5 percent processed locally; industrialisation target of 50 percent local processing.
Publication date unrecorded. Inherited source-text record. The original publication has not been independently reopened for this web edition.
R07 · Grade C · inherited record
Inherited evidence R07
SND30 restates the 50 percent local-processing target by 2030.
Publication date unrecorded. Inherited source-text record. The original publication has not been independently reopened for this web edition.
R12 · Grade C · inherited record
Inherited evidence R12
CICAM capacity utilisation, July 2023 report: Garoua 33.7 percent, CICAM 1 48.9 percent, CICAM 2 10.83 percent; underutilisation cost about CFAF 2 billion in 2020.
Publication date unrecorded. Inherited source-text record. The original publication has not been independently reopened for this web edition.
R29 · Grade C · inherited record
Inherited evidence R29
Manual defect inspection 60 to 70 percent accurate, missing 20 to 30 percent of defects; computer vision above 90 percent across more than 40 defect types; line-scan inspection up to 60 metres a minute.
Publication date unrecorded. Inherited source-text record. The original publication has not been independently reopened for this web edition.
c09-S01 · Grade A
Update to ISA-95 Standard Addresses Integration of Enterprise and Manufacturing Control Systems
ANSI/ISA-95.00.01-2025, also known as ISA-95 Part 1, summarizes the scope of the manufacturing operations and control domain, discusses how physical assets of a manufacturing enterprise are organized, lists the functions associated with the interface between control functions and enterprise functions and describes the information shared among these functions.
2025-04-10. Register access date: 2026-10-30. Source provenance is supplied by the paper evidence register.
Open the recorded source ↗c09-S02 · Grade A
IEC 62264-1:2013: Enterprise-control system integration, Part 1: Models and terminology
IEC 62264-1:2013 describes the manufacturing operations management domain (Level 3) and its activities, and the interface content and associated transactions within Level 3 and between Level 3 and Level 4.
2013-05-22. Register access date: 2026-10-30. Source provenance is supplied by the paper evidence register.
Open the recorded source ↗c09-S04 · Grade A
OPC-10000-1: OPC Unified Architecture, Part 1: Overview and Concepts
The OPC UA is layered to isolate the core design from the underlying computing technology and network transport.
2025-10-22. Register access date: 2026-10-30. Source provenance is supplied by the paper evidence register.
Open the recorded source ↗c09-S05 · Grade A
MQTT V5.0 OASIS Standard published
These characteristics make it ideal for use in many situations, including constrained environments such as for communication in Machine to Machine (M2M) and Internet of Things (IoT) contexts where a small code footprint is required and/or network bandwidth is at a premium.
2019-03-07. Register access date: 2026-10-30. Source provenance is supplied by the paper evidence register.
Open the recorded source ↗c09-S07 · Grade A
Specification of the Asset Administration Shell Part 1: Metamodel, IDTA 01001-3-0-1
The aim of this document is to define selected specifications of the structure of the Administration Shell to enable the meaningful exchange of information about assets and I4.0 components between partners in a value creation network.
2024-06. Register access date: 2026-10-30. Source provenance is supplied by the paper evidence register.
Open the recorded source ↗c09-S08 · Grade A
ISO 22400-2:2014: Automation systems and integration, Key performance indicators for manufacturing operations management, Part 2: Definitions and descriptions
The KPIs are presented by means of their formula and corresponding elements, their time behaviour, their unit/dimension and other characteristics.
2014-01. Register access date: 2026-10-30. Source provenance is supplied by the paper evidence register.
Open the recorded source ↗c09-S09 · Grade A
MODBUS Application Protocol Specification V1.1b3
MODBUS is a request/reply protocol and offers services specified by function codes.
2012-04-26. Register access date: 2026-10-30. Source provenance is supplied by the paper evidence register.
Open the recorded source ↗c09-S12 · Grade A
IEEE 1588-2019: IEEE Standard for a Precision Clock Synchronization Protocol for Networked Measurement and Control Systems
The protocol supports synchronization accuracy and precision in the sub-microsecond range with minimal network and local computing resources.
2020-06-16. Register access date: 2026-10-30. Source provenance is supplied by the paper evidence register.
Open the recorded source ↗c09-S13 · Grade A
Network Time Protocol Version 4: Protocol and Algorithms Specification
NTPv4 includes fundamental improvements in the mitigation and discipline algorithms that extend the potential accuracy to the tens of microseconds with modern workstations and fast LANs.
2010-06. Register access date: 2026-10-30. Source provenance is supplied by the paper evidence register.
Open the recorded source ↗c09-S14 · Grade A
Web of Things Architecture 1.1
An industrial edge device collects selected data from various controllers and makes it available to a cloud backend service, e.g., for remote monitoring via a dashboard or analyzes it for preventive maintenance.
2023-12-05. Register access date: 2026-10-30. Source provenance is supplied by the paper evidence register.
Open the recorded source ↗c09-S15 · Grade A
EPCIS Standard, Release 2.0
The goal of EPCIS is to enable disparate applications to create and share visibility event data, both within and across enterprises.
2022-06. Register access date: 2026-10-30. Source provenance is supplied by the paper evidence register.
Open the recorded source ↗c09-S16 · Grade A
ETSI GS MEC 003 V3.2.1: Multi-access Edge Computing, Framework and Reference Architecture
Multi-access Edge Computing enables the implementation of MEC applications as software-only entities that run on top of a Virtualisation infrastructure, which is located in or close to the network edge.
2024-04. Register access date: 2026-10-30. Source provenance is supplied by the paper evidence register.
Open the recorded source ↗c09-S17 · Grade A
NIST SP 800-82 Rev. 3: Guide to Operational Technology Security
This document provides guidance on how to secure operational technology (OT) while addressing their unique performance, reliability, and safety requirements.
2023-09. Register access date: 2026-10-30. Source provenance is supplied by the paper evidence register.
Open the recorded source ↗c09-S18 · Grade A
The NIST Cybersecurity Framework 2.0
The NIST Cybersecurity Framework (CSF) 2.0 provides guidance to industry, government agencies, and other organizations to manage cybersecurity risks.
2024-02-26. Register access date: 2026-10-30. Source provenance is supplied by the paper evidence register.
Open the recorded source ↗c09-S19 · Grade A
IEC 62443-3-2:2020: Security for industrial automation and control systems, Part 3-2: Security risk assessment for system design
A key concept in this document is the application of IACS security zones and conduits.
2020-06. Register access date: 2026-10-30. Source provenance is supplied by the paper evidence register.
Open the recorded source ↗c09-S21 · Grade A
Russian State-Sponsored and Criminal Cyber Threats to Critical Infrastructure
Regularly test backup procedures and ensure that backups are isolated from network connections that could enable the spread of malware.
2022-05-09. Register access date: 2026-10-30. Source provenance is supplied by the paper evidence register.
Open the recorded source ↗c09-S22 · Grade A
NIST SP 800-34 Rev. 1: Contingency Planning Guide for Federal Information Systems
Testing validates recovery capabilities, whereas training prepares recovery personnel for plan activation and exercising the plan identifies planning gaps; combined, the activities improve plan effectiveness and overall organization preparedness.
2010-11-11. Register access date: 2026-10-30. Source provenance is supplied by the paper evidence register.
Open the recorded source ↗c09-S24 · Grade A
ISO/IEC 27001:2022: Information security, cybersecurity and privacy protection, Information security management systems, Requirements
The ISO/IEC 27001 standard provides companies of any size and from all sectors of activity with guidance for establishing, implementing, maintaining and continually improving an information security management system.
2022-10. Register access date: 2026-10-30. Source provenance is supplied by the paper evidence register.
Open the recorded source ↗c09-S25 · Grade B
An RFID-based tracking system for denim production processes
The most striking result from the case study is that a radio frequency identification application could be successfully implemented to support the tracking and tracing of work in process in denim production processes, an area in which studies about radio frequency identification applications have not been done.
2016-09-05. Register access date: 2026-10-30. Source provenance is supplied by the paper evidence register.
Open the recorded source ↗c09-S26 · Grade B
Textile Defect Detection Using Artificial Intelligence and Computer Vision: A Preliminary Deep Learning Approach
Data acquired in complex industrial settings are highly susceptible to external variables such as lighting variations and background noise, which can significantly degrade the detection accuracy of machine learning (ML) models [14].
2025-09-18. Register access date: 2026-10-30. Source provenance is supplied by the paper evidence register.
Open the recorded source ↗c09-S27 · Grade B
Digital and Sustainable Transition in Textile Industry through Internet of Things Technologies: A Pakistani Case Study
Figure 8 illustrates changes in water level, depicting fluctuations that enable the industry to optimize water usage.
2024-06-21. Register access date: 2026-10-30. Source provenance is supplied by the paper evidence register.
Open the recorded source ↗c09-S28 · Grade B
RFID-enabled real-time manufacturing execution system for mass-customization production
RFID devices are deployed systematically on the shop-floor to track and trace manufacturing objects and collect real-time production data.
2013-04. Register access date: 2026-10-30. Source provenance is supplied by the paper evidence register.
Open the recorded source ↗c09-S29 · Grade B
Industrial IoT-Based Energy Monitoring System: Using Data Processing at Edge
Once the edge receives the meter’s data, it stores the information in the database server, followed by the data processing method to find nine additional analytical parameters.
2024-09-28. Register access date: 2026-10-30. Source provenance is supplied by the paper evidence register.
Open the recorded source ↗c09-S30 · Grade B
Six-Sigma Reference Model for Industry 4.0 Implementations in Textile SMEs
The implementation of small pilot projects is recommended to test and validate the effectiveness of selected I4.0 technologies.
2023-08-19. Register access date: 2026-10-30. Source provenance is supplied by the paper evidence register.
Open the recorded source ↗c09-S31 · Grade B
A Review of IoT Sensing Applications and Challenges Using RFID and Wireless Sensor Networks
Also, time synchronization between the nodes and spectrum sharing techniques are of the utmost importance to ensure the integrity of the data or avoid uncertainty in the sensed data, without consuming all the radio resources [37].
2020-04-28. Register access date: 2026-10-30. Source provenance is supplied by the paper evidence register.
Open the recorded source ↗c09-S32 · Grade A
Enterprise Surveys: What Businesses Experience, Cameroon 2024 Country Profile
Figure 8 shows the extent to which firms face failures in the provision of electricity as measured by the percent of firms experiencing electrical outages and by the number of power outages.
2025. Register access date: 2026-10-30. Source provenance is supplied by the paper evidence register.
Open the recorded source ↗c09-S36 · Grade B
Product Life Cycle Accounting and Reporting Standard
The GHG Protocol Product Life Cycle Accounting and Reporting Standard (referred to as the Product Standard) provides requirements and guidance for companies and other organizations to quantify and publicly report an inventory of GHG emissions and removals associated with a specific product.
2011-09. Register access date: 2026-10-30. Source provenance is supplied by the paper evidence register.
Open the recorded source ↗c09-S37 · Grade A
ISO 50001:2018: Energy management systems, Requirements with guidance for use
ISO 50001:2018 is an international standard that provides a framework for establishing, implementing, maintaining, and improving an Energy Management System (EnMS).
2018-08. Register access date: 2026-10-30. Source provenance is supplied by the paper evidence register.
Open the recorded source ↗c09-S38 · Grade A
ISO 14046:2014: Environmental management, Water footprint, Principles, requirements and guidelines
ISO 14046:2014 specifies principles, requirements and guidelines related to water footprint assessment of products, processes and organizations based on life cycle assessment (LCA).
2014-08. Register access date: 2026-10-30. Source provenance is supplied by the paper evidence register.
Open the recorded source ↗c09-S39 · Grade A
Regulation (EU) 2024/1781 establishing a framework for the setting of ecodesign requirements for sustainable products
‘digital product passport’ means a set of data specific to a product that includes the information specified in the applicable delegated act adopted pursuant to Article 4 and that is accessible via electronic means through a data carrier in accordance with Chapter III;
2024-06-28. Register access date: 2026-10-30. Source provenance is supplied by the paper evidence register.
Open the recorded source ↗c09-S40 · Grade A
Sparkplug Specification 3.0.0
The intent and purpose of the Sparkplug Specification is to define an MQTT Topic Namespace, payload, and session state management that can be applied generically to the overall IIoT market sector, but specifically meets the requirements of real-time SCADA/Control HMI solutions.
2022-11-16. Register access date: 2026-10-30. Source provenance is supplied by the paper evidence register.
Open the recorded source ↗c09-S41 · Grade A
ISO 8000-120:2016: Data quality, Part 120: Master data: Exchange of characteristic data: Provenance
ISO 8000-120:2016 specifies requirements for the representation and exchange of information about the provenance of master data that consists of characteristic data, and supplements the requirements of ISO 8000‑110.
2016-10. Register access date: 2026-10-30. Source provenance is supplied by the paper evidence register.
Open the recorded source ↗c09-S0901 · Grade A
OPC 10000-11: OPC Unified Architecture, Part 11: Historical Access, Release 1.05.04, Section 4.3: Historians and interruption of data collection
The Historian shall report any gaps when a client is accessing the stored historical data with an error code of Bad _ DataLost.
2024-10-15. Register access date: 2026-10-30. Source provenance is supplied by the paper evidence register.
Open the recorded source ↗c09-S0902 · Grade A
OPC 10000-11: OPC Unified Architecture, Part 11: Historical Access, Release 1.05.04, Section 4.4: Modification of Historical Data/Events
A Historian that historizes Events shall store the EventId that it receives; Modification of Events shall not allow modification to EventIds.
2024-10-15. Register access date: 2026-10-30. Source provenance is supplied by the paper evidence register.
Open the recorded source ↗c09-S0903 · Grade A
OPC 10000-11: OPC Unified Architecture, Part 11: Historical Access, Release 1.05.04, Section 4.5: Timestamps
For HistoricalDataNodes , the SourceTimestamp is used to determine which historical data values are to be returned.
2024-10-15. Register access date: 2026-10-30. Source provenance is supplied by the paper evidence register.
Open the recorded source ↗c09-S0904 · Grade A
PROV-Overview: An Overview of the PROV Family of Documents
Provenance is information about entities, activities, and people involved in producing a piece of data or thing, which can be used to form assessments about its quality, reliability or trustworthiness.
2013-04-30. Register access date: 2026-10-30. Source provenance is supplied by the paper evidence register.
Open the recorded source ↗S08 International Organization for Standardization. 2014-01.
S15 GS1 AISBL. 2022-06.
S24 International Organization for Standardization and International Electrotechnical Commission. 2022-10.
S0902 OPC Foundation. 2024-10-15.
S32 World Bank Enterprise Surveys. 2025.
S40 Eclipse Foundation. 2022-11-16.
S26 MDPI, Electronics. 2025-09-18.
S29 MDPI, IoT. 2024-09-28.
S31 MDPI, Sensors. 2020-04-28.
S0901 OPC Foundation. 2024-10-15.
S0903 OPC Foundation. 2024-10-15.
S0904 World Wide Web Consortium. 2013-04-30.
R06 Cameroon Trade Hub. Publication date unrecorded; inherited source-text check.
R07 Ecofin Agency. Publication date unrecorded; inherited source-text check.
R12 Business in Cameroon. Publication date unrecorded; inherited source-text check.
S30 MDPI, Sustainability. 2023-08-19.
S04 OPC Foundation. 2025-10-22.
S07 Industrial Digital Twin Association. 2024-06.
S01 International Society of Automation. 2025-04-10.
S27 MDPI, Applied Sciences. 2024-06-21.
S02 International Electrotechnical Commission. 2013-05-22.
S25 Springer Nature, International Journal of Advanced Manufacturing Technology. 2016-09-05.
S28 Elsevier, Robotics and Computer-Integrated Manufacturing. 2013-04.
S16 European Telecommunications Standards Institute. 2024-04.
S14 World Wide Web Consortium. 2023-12-05.
S41 International Organization for Standardization. 2016-10.
S05 OASIS Open. 2019-03-07.
S12 IEEE Standards Association. 2020-06-16.
S13 Internet Engineering Task Force and RFC Editor. 2010-06.
S09 Modbus Organization. 2012-04-26.
S19 International Electrotechnical Commission. 2020-06.
S17 National Institute of Standards and Technology. 2023-09.
R29 MDPI; industry inspection literature. Publication date unrecorded; inherited source-text check.
S36 GHG Protocol, World Resources Institute and World Business Council for Sustainable Development. 2011-09.
S39 European Parliament and Council of the European Union, Official Journal of the European Union. 2024-06-28.
S38 International Organization for Standardization. 2014-08.
S37 International Organization for Standardization. 2018-08.
S21 CISA and joint international cybersecurity authorities. 2022-05-09.
S18 National Institute of Standards and Technology. 2024-02-26.
S22 National Institute of Standards and Technology. 2010-11-11.
Keep the thread · chapter 09 of 14
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
01Part I: Decisions before devices
GFT starts with the production decisions and infrastructure constraints that a connected architecture must serve, because visualisation alone cannot create trustworthy operating evidence.
02Part I: Decisions before devices
Each connected-factory use case needs an event, an owner, a permitted delay, a fallback and a measurable consequence before any sensor or platform is selected.
03Part I: Decisions before devices
A useful reference architecture assigns each function to a responsibility boundary and information contract before it assigns that function to a product.
04Part II: Capture at the production edge
A value becomes production evidence only when its asset, occurrence time, uncertainty and transformation history remain inspectable.
05Part II: Capture at the production edge
The safest digital retrofit begins by observing equipment through its documented interfaces and proving that acquisition does not alter control timing, protective functions or maintainability.
06Part II: Capture at the production edge
Local inspection earns its place when a valid image becomes a traceable quality decision before the fabric leaves the decision point. Acquisition, persistence and an approved fallback matter as much as inference speed.
07Part II: Capture at the production edge
The connected factory must lose remote services before it loses local evidence, and lose local evidence before it ever loses safe machine operation.
08Part III: One governed data spine
Two systems can exchange packets and still misunderstand production, so the architecture must define information, state, events and asset meaning separately.
09Part III: One governed data spine
The factory's data store is an evidence system, not a growing pile of values, so raw observations, contextual events, KPI results and corrections remain distinguishable and replayable.
10Part III: One governed data spine
The dashboard is a role-specific control surface whose freshness, definitions and next action are visible, not a gallery of independently calculated charts.
11Part III: One governed data spine
Applications should arrive as bounded, observable and reversible changes to a working production architecture, with ownership and rollback proven before scale.
12Part IV: Prove value and scale safely
This chapter specifies evidence capture rather than accounting or compliance methods: metres, denominators, allocations and life-cycle boundaries must make every displayed result reconstructable.
13Part IV: Prove value and scale safely
A connected factory is not accepted when its screens work on a normal day, but when its controls, fallbacks and recovery perform under adversarial and degraded conditions.
14Part IV: Prove value and scale safely
GFT recommends a four-phase build that earns each expansion through evidence quality, operating value, safe failure and recovery rather than device count.