Research libraryThe Industrial Data Spine
Chapter 01 of 14 · Part I · Draft web edition
The binding constraint on industrial AI is often the broken chain between work, record and decision, not the absence of an algorithm.
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.
Read-out
Illustrative sewing-line event traced from shop-floor capture through context, KPI calculation, training snapshot, model output and management decision. Failure mechanics are derived from technical-debt entanglement in S18 and upstream data cascades in S19.
Unit: . Status: illustrative. Sources: S18, S19. Illustrative figures are teaching examples, not measured factory results.
A factory becomes reliably data-driven only when a supervisor, an engineer and a machine-learning pipeline can read the same operational record and know what physical event occurred, who owns its definition, whether the value passed quality checks and how to recover the history if a system fails.[S01][S02][S09][S26] In small and mid-sized manufacturing plants, the binding constraint on industrial analytics is rarely the absence of a statistical model.[S19][S0101] The constraint is the broken chain between shop-floor work, recorded evidence and management action.[S05][S08][S43] Green Future Tech Limited (GFT), Hong Kong SAR, anchors its work in the principle of Warp and Signal: physical production is the warp, and operational measurement is the weft. In GFT's three-part Cameroon textile programme, prepared on 31 January 2024 with source-refresh annotations retained in 2026, the sector baseline records that about 95 percent of Cameroon's cotton fibre is exported raw while about 5 percent is processed locally, against a national industrialisation target of 50 percent local processing by 2030.[R06][R07] Those baseline and target entries come from the inherited programme register, were not independently refreshed in 2026 and carry unrecorded original publication dates, so they frame the scale of the industrial ambition rather than certifying current mill output.[R06][R07] Moving from raw-fibre export to competitive yarn, fabric and garment production requires factories that can defend their daily numbers on yield, quality, downtime, order trace and resource use without hiring an enterprise data department.[R06][S02][S38] This paper specifies that alternative: an industrial data spine built on named decision rights, explicit operational meaning, enforced quality gates, reversible model releases and measured operational value.[S01][S04][S15][S36]
The urgency of governed records rises sharply when a sector seeks to rebuild domestic processing capacity under tight capital and margin constraints.[R06][R12] Inherited sector evidence illustrates the cost of operating without reliable utilisation and market alignment: CICAM's reported domestic market share fell from about 85 percent three decades ago to about 5 percent in the inherited register, while a July 2023 report recorded capacity utilisation at 33.7 percent in Garoua, 48.9 percent at CICAM 1 and 10.83 percent at CICAM 2, with underutilisation costing about CFAF 2 billion (XAF 2 billion) in 2020.[R11][R12] Meanwhile, announced third-party industrial expansion, such as the Camtext SA integrated cotton-to-textile project reported on 12 November 2025 at US$300 million (180 billion XAF) with a planned processing volume of 12,000 tonnes a year, remains an announced third-party investment target reaching optimal capacity around 2033 rather than realised factory performance.[R09][R10] Whether an operating team is rehabilitating an existing mill or commissioning a new garment line, capital equipment alone does not produce control.[S05][S36] As Exhibit 1 summarises, the gap between the inherited 5 percent local-processing baseline and the 50 percent national target by 2030 places a high penalty on untrusted operational evidence.[R06][R07]
Exhibit 1 establishes the industrial context and its evidentiary limits.[R06][R07] It shows why local value retention is a strategic priority in Cameroon, but a macroeconomic target cannot tell a spinning supervisor why roving breaks rose on the afternoon shift or tell a garment director whether an automated fabric-inspection camera is reducing customer claims.[R06][R07][S38] Answering those operational questions requires a governed chain from the shop floor to the management review table.[S05][S38]
1.1
Read-out
Derived routing matrix mapping eight recurring data and model escalations across decision horizon (x-axis: 0 = immediate shift action, 10 = monthly portfolio review) and cross-domain risk consequence (y-axis: 0 = single-line capture fix, 10 = plant-wide risk or policy gate) into WP02's existing management forums, based on S02, S11 and S23.
Unit: horizon (x) and consequence (y) index (0 to 10). Status: derived. Sources: S02, S11, S23.
Read this exhibit in its chapter ↗Read-out
Illustrative scoring model (0 = untested, 1 = manual spot check, 2 = entry syntax check, 3 = automated gate with quarantine, 4 = versioned gate with full lineage) across four factory record sets and five quality dimensions, derived from ISO/IEC 25012 (S03), ISO/IEC 5259-3 (S04) and W3C DQV (S0401). Values are illustrative and not a measured GFT factory result.
Unit: illustrative coverage score (0 to 4). Status: illustrative. Sources: S03, S04, S0401. Illustrative figures are teaching examples, not measured factory results.
Read this exhibit in its chapter ↗1.1 · Introduction
Data debt is the future correction, reconciliation and redesign cost that a factory accumulates when it captures operational numbers with undefined fields, inconsistent identifiers and no named owner.
Exhibit 09, chapter 02: Model rollback
1.2 · Upstream capture defects compound silently across the lifecycle
When initial capture lacks explicit rules for syntax, meaning and provenance , defects multiply across reporting and analytical layers.
Exhibit 09, chapter 02: Model rollback
1.3 · Informal spreadsheet workarounds mask structural gaps
In the early phases of GFT’s Cameroon textile programme, manual spreadsheets and shift logs serve as the starting system of record .
Exhibit 15, chapter 04: Production logs
1.1 · the bodyThe film above follows these sub-steps.
Data debt is the future correction, reconciliation and redesign cost that a factory accumulates when it captures operational numbers with undefined fields, inconsistent identifiers and no named owner.[S01][S02][S18] Sculley and colleagues showed that real-world machine-learning systems incur heavy ongoing maintenance costs because models depend on unstable upstream data feeds, undeclared consumers and improvised correction scripts.[S18] In a factory, that obligation starts on the first shift when an operator writes a bundle count on a clipboard, a technician logs a fault in a chat message or a supervisor edits a spreadsheet formula to force weekly output to match the packing list.[S01][S18][S38]
When initial capture lacks explicit rules for syntax, meaning and provenance, defects multiply across reporting and analytical layers.[S01][S09][S19] Sambasivan and colleagues studied high-stakes artificial-intelligence deployments and defined data cascades as compounding events caused by upstream data problems that produce negative downstream consequences.[S19] Among the high-stakes AI practitioners they interviewed, 92 percent reported experiencing at least one data cascade, noting that such defects were typically invisible at capture, delayed in impact and avoidable through deliberate data practices.[S19] That 92 percent figure describes the interviewed sample of high-stakes AI practitioners rather than a prevalence estimate for manufacturing firms or small and medium-sized enterprises (SMEs), yet the mechanism it identifies operates directly on factory floors.[S19] Treating data entry as clerical overhead rather than engineered measurement hides upstream omissions until a management decision or model fails in production.[S04][S19]
Consider how an unowned capture defect propagates through a garment plant, as traced in the illustrative sequence of Exhibit 2.[S18][S19] Suppose a sewing line loses forty minutes of output because a batch of dyed fabric arrives from finishing with uneven shade across the roll width, forcing operators to re-sort cut panels at the line feeder (an illustrative shop-floor scenario). If the line tally sheet offers only two downtime checkboxes, mechanical breakdown and operator absence, the clerk ticks mechanical breakdown to avoid leaving a mandatory field blank (illustrative). That forced misclassification corrupts every downstream stage.[S18][S19]
As Exhibit 2 illustrates, the error does not correct itself.[S18][S19] At the second node, the link between the sewing stoppage and the upstream fabric lot disappears because the tally sheet records only the sewing line number.[S08][S09] At the third node, the weekly key performance indicator (KPI) calculation charges the lost forty minutes against equipment availability under ISO 22400-2 time elements rather than material starvation or quality hold.[S38] At the fourth and fifth nodes, an analyst training a predictive-maintenance classifier on twelve months of logs ingests those mislabelled stoppages as mechanical failures.[S04][S18][S19] At the sixth node, management approves spare-parts expenditure for sewing heads while the dye-lot shade problem remains untouched.[S18][S19] What Exhibit 2 cannot establish is the exact frequency of such errors in a given mill before an audit; rather, it shows why ISO/IEC 5259-3 requires data-quality management across the full analytics and machine-learning lifecycle rather than as a late cleaning step.[S04][S19]
In the early phases of GFT's Cameroon textile programme, manual spreadsheets and shift logs serve as the starting system of record. Structured manual logs are an appropriate choice when capital and connectivity are constrained, provided those sheets enforce field definitions, identifier stability and version control.[S01][S04][S16] Data debt arises when workbooks mutate privately across shifts: a morning supervisor inserts a column, an afternoon supervisor merges cells for a split order and a clerk pastes values over formulas.[S01][S09][S18] Reconciling those files at month-end consumes engineering hours needed for process improvement, and any dashboard built on top inherits their contradictions.[S18][S0101]
1.4
A manufacturing plant is a physical system in which material, machines, people and orders interact simultaneously, yet most small factories record those interactions in separate silos.[S05][S08] When a disruption occurs on the floor, it leaves partial footprints in three or four departmental logs that cannot be joined without manual investigation.[S05][S08][S0101]
Consider a single operational event reconstructed across four typical factory logs: the sewing-line hourly board, the maintenance technician's note, the quality inspector's defect sheet and the commercial order tracker (an illustrative reconstruction grounded in IEC 62264-1 integration boundaries and GS1 EPCIS 2.0 visibility events).[S05][S08] Suppose an overlock machine on Line B begins skipping stitches at 10:12 on a Tuesday while running a cotton polo-shirt order.[S05][S08]
OVL-14 from 10:25 to 10:50 for a worn looper, without recording which line OVL-14 was mounted on or which order was interrupted.[S05][S08]ORD-2026-114 short by 18 pieces at packing two days later, logging "sewing short-ship" with no link to OVL-14 or the morning quality hold.[S05][S08]Read-out
| Line hourly sheet | Clock hour (10:00 to 11:00) | Line B only | 35-unit shortfall | No asset ID, lot ID or exact stop duration |
|---|---|---|---|---|
| Maintenance note | Repair window (10:25 to 10:50) | Machine OVL-14 | Looper replaced | No line assignment, wait time or order ID |
| Quality tally sheet | Morning shift aggregate | Style description text | 18 seam-skip defects | No bundle ID, machine ID or rework vs scrap state |
| Order progress tracker | Packing date (two days later) | Order ORD-2026-114 | 18-unit short-ship | No link to quality hold or machine fault event |
| Governed spine event | Event + ingestion timestamps | Line B + OVL-14 + Lot + Order | State transitions + 18 rework/scrap | Reconciles time, object, location and business step |
Derived comparison applying IEC 62264-1 enterprise-control interface domains (S05), GS1 EPCIS 2.0 visibility event dimensions (S08) and ISO 22400-2 KPI time and quantity elements (S38) to a single shop-floor stoppage.
Unit: record attributes captured. Status: derived. Sources: S05, S08, S38.
As Exhibit 3 shows, none of the four logs is falsified, yet the factory cannot reconstruct the event from them.[S05][S08] The maintenance note shows 25 minutes of active repair (10:25 to 10:50), whereas the line lost 38 minutes between the stop at 10:12 and restart clearance at 10:50 (illustrative timestamps).[S38] Because the 13-minute wait from 10:12 to 10:25 is unrecorded in maintenance and merged into a clock hour on the line sheet, management cannot separate technician response delay from active wrench time, nor can quality prove whether the 18 defective pieces came from OVL-14 just before 10:12.[S08][S38] What Exhibit 3 does not prescribe is the transport protocol or database engine; Chapter 3 specifies the vocabulary contract and Chapter 5 the minimum architecture.[S05][S07][S08]
The maintenance note shows 25 minutes of active repair (10:25 to 10:50)
S38 · illustrative
Two international standards explain how a small team can repair these broken links without buying a monolithic software suite.[S05][S08] First, IEC 62264-1 (the ISA-95 standard) defines models and terminology for the interfaces between enterprise business systems and manufacturing operations and control systems.[S05] It organises work schedules, production performance, equipment, material lots and personnel into explicit interface domains so that an order identifier and a machine identifier meet at a defined boundary.[S05]
Second, the GS1 EPCIS 2.0 standard enables disparate applications to create and share visibility event data within and across enterprises.[S08] Rather than overwriting a status cell, an EPCIS-aligned event records what object or lot was involved, when the event occurred and was captured, where it happened down to the line or station, and why it occurred under a defined business step and disposition.[S08] Adopting that four-part event grammar even in initial structured sheets links the stoppage, repair, quality hold and order variance into one inspectable chain.[S05][S08][S09]
1.7
When production, quality and finance numbers conflict in a weekly meeting, management is often tempted to buy a dashboard tool or an enterprise resource planning (ERP) module.[S02][S34][S0101] Software procurement cannot fix an accountability vacuum.[S02][S23] If nobody holds the right to decide whether reworked garments count inside hourly output, a new platform merely displays the dispute faster and at higher cost.[S02][S38]
ISO 8000-150:2022 specifies the key considerations for organisations establishing roles and responsibilities for data-quality management.[S02] Its core principle is organisational: data quality improves only when specific roles are accountable for defining data requirements, monitoring conformance and executing corrections.[S02] Similarly, the NIST Cybersecurity Framework 2.0 states that governance activities are critical for incorporating risk management into broader enterprise strategy.[S23] A factory must therefore assign who owns domain meaning, who stewards daily capture quality and who holds technical custody before comparing software vendors.[S02][S23]
In a small factory, assigning decision rights does not create a separate bureaucracy.[S02][S0101] NIST's smart-manufacturing programme observes that a typical SME cannot afford a dedicated data-analytics specialist on staff and that commercial tools remain difficult to integrate with shop-floor routines.[S0101] Governance must therefore sit inside existing operating roles: the production manager owns output and time definitions, the quality lead owns defect taxonomies and dispositions, the maintenance lead owns asset hierarchies and fault codes, and the commercial lead owns order and lot identifiers.[S02][S05][S38] Chapter 2 formalises this six-role charter with conflict-of-interest rules where one person holds multiple roles.[S02][S11][S23]
Paper role descriptions fail if daily adoption is left to external consultants.[S37] In a May 2025 World Bank case study of SME digital technology adoption in Viet Nam, firms that successfully adopted the platform assigned an internal point person who championed adoption within the firm.[S37] That champion coordinates training, follows up on late logs, escalates cross-domain definition disputes and protects line stewards' time.[S02][S37]
Designers must also avoid assuming that integrated enterprise software already exists on the ground.[S34] In the OECD's June 2026 Western Balkans Enterprise Survey, tools supporting integrated operations, such as cloud computing, ERP and customer relationship management (CRM) systems, were used by only 10 to 20 percent of surveyed SMEs, with supplier-customer systems even rarer.[S34] That 10 to 20 percent finding describes the surveyed Western Balkans SME sample rather than a Cameroon estimate, but it warns against assuming that an SME possesses a clean ERP backbone ready for machine learning.[S34][S42] As SCITEPRESS Industry 4.0 research notes, the starting technical infrastructure and target degree of automation strongly govern MLOps complexity.[S42]
Interlude · the cloth half-woven
Woven: sections 1.0 to 1.7Still to weave: 1.10 to 1.19
Read-out
Derived monthly measurement pack synthesised from S02, S04, S15, S17, S26, S36 and S1402. Tile 2 pairs semantic contract coverage (Measure 2) and quality-gate resolution (Measure 3) so all seven capabilities are surfaced without universal numeric targets before site baseline.
Unit: derived monthly governance and MLOps capability pack. Status: derived. Sources: S02, S04, S15, S17, S26, S36, S1402.
Read this exhibit in its chapter ↗1.10
Related · chapter 095 paragraphs step it
Read-out
Illustrative reconciliation from gross annual operational benefit (100 index units) to net lifecycle contribution (-12 index units) for a custom history-dependent model with bespoke pipeline preparation. Cost categories derive from S18 and the upper-bound data-processing warning in S39; values are illustrative arithmetic, not a universal factory ratio.
Unit: illustrative index (gross benefit = 100). Status: illustrative. Sources: S18, S39. Illustrative figures are teaching examples, not measured factory results.
Read this exhibit in its chapter ↗When a factory treats data as a by-product of isolated projects, every new question triggers another custom extraction, cleaning script and identifier mapping.[S10][S39][S0101] One effort builds a sheet for overall equipment effectiveness (OEE), a second pulls defect tallies for a camera trial and a third extracts utility bills for carbon reporting, each redefining shift boundaries, rework and good output.[S10][S38][S39]
In a 2025 NIST workshop report on AI and scaled data sharing in semiconductor manufacturing, manufacturers reported that data processing for a single application can reach up to 70 percent of total cost.[S39] That 70 percent figure is a manufacturer-reported upper-bound warning from semiconductor fabrication rather than a universal SME average, yet it highlights the penalty of building separate pipelines for every application.[S39] NIST's smart-manufacturing research confirms that substantial time goes into data acquisition and formatting because analytics tools use proprietary interfaces and specialised data formats.[S0101]
The alternative is to capture each operational event once with governed identity, timestamps, units and quality states, catalogued under a standard metadata model such as the W3C Data Catalog Vocabulary (DCAT) Version 3, and publish controlled views to multiple consumers.[S08][S09][S10][S0102] Across the GFT white-paper library, one governed record spine feeds four downstream uses without re-cleaning the floor logs:
1.12
A factory record is not trustworthy simply because its schema validates.[S22][S29][S30] An architecture that disrupts machine reliability, exposes control networks, collects excessive personal telemetry on workers or breaches national data-protection law creates severe operational exposure.[S22][S29][S31]
Factory systems interact with rotating machinery, dyeing vessels, boilers and switchgear.[S22] NIST Special Publication 800-82 Revision 3 guides operational technology (OT) security around performance, reliability and safety requirements.[S22] Data collection and model inference must remain subordinate to physical process safety: no gateway, broker or model may write unvalidated commands into safety-critical control loops, and any data-layer failure must leave machines operating safely under local control.[S22][S33] During network outages, edge devices use local store-and-forward buffering to retain event records without disturbing shop-floor control.[S22][S33]
Shop-floor logs often touch operator sign-ins, bundle scans, skills matrices and camera feeds.[S29][S30] Two rules govern these records.[S29][S30] First, Regulation (EU) 2016/679 (GDPR, Recital 39) states the data-minimisation principle: personal data should be adequate, relevant and limited to what is necessary for the stated purpose.[S29] Line balance, availability, yield and energy metrics require station, line, shift or role identifiers rather than personal identity; where personal records are required (such as high-voltage maintenance certification), access follows least privilege under NIST SP 800-53 Revision 5.[S25][S29][S38]
Second, Principle 5.5 of the International Labour Office (ILO) 1997 code of practice states that decisions concerning a worker should not be based solely on automated processing of that worker's personal data.[S30] While the ILO code is non-binding guidance, GFT enforces it as an operating rule: no productivity score, defect tally or camera output may automatically trigger discipline, pay changes or dismissal without informed human review of material, machine and line context.[S11][S30]
Factories in Cameroon must also address Law No. 2024/017 of 23 December 2024 relating to personal data protection in Cameroon.[S31] Section 19(1) of Law No. 2024/017 provides that the processing of personal data shall be subject to prior authorisation by the Personal Data Protection Authority.[S31]
…must also address Law No. 2024/017 of 23 December 2024 relating to personal data…
S31
1.16
Order ledger (baseline)
Order ledger (baseline): Ownership and meaning (Tests 1-2) 6.5, Quality and recovery (Tests 3-4) 4.5. Quadrant: Unverified contract (enforce quality and backup) · derived
Across GFT’s Cameroon textile programme, the sequencing principle is strict: capture cleanly from the first shift, establish one governed vocabulary and admit machine learning only after records pass explicit readiness gates.
Production logs (baseline)
Production logs (baseline): Ownership and meaning (Tests 1-2) 4, Quality and recovery (Tests 3-4) 3. Quadrant: Data pile (fix ownership and capture) · derived
Quality sheets (baseline)
Quality sheets (baseline): Ownership and meaning (Tests 1-2) 3.5, Quality and recovery (Tests 3-4) 2.5. Quadrant: Data pile (fix ownership and capture) · derived
Maintenance notes (baseline)
Maintenance notes (baseline): Ownership and meaning (Tests 1-2) 2, Quality and recovery (Tests 3-4) 2. Quadrant: Data pile (fix ownership and capture) · derived
Utility/carbon logs (baseline)
Utility/carbon logs (baseline): Ownership and meaning (Tests 1-2) 3, Quality and recovery (Tests 3-4) 4. Quadrant: Data pile (fix ownership and capture) · derived
Spine admission gate (target)
Spine admission gate (target): Ownership and meaning (Tests 1-2) 8, Quality and recovery (Tests 3-4) 8. Quadrant: Governed data spine (admit to analytics) · derived
Derived readiness framework scoring five factory record domains across two composite axes: Ownership and Meaning (Tests 1 and 2, from S02 and S05; 0 = unowned free text, 10 = contracted owner, IDs and KPI formula) and Quality, Lineage and Recovery (Tests 3 and 4, from S01, S09 and S26; 0 = unchecked/unbacked, 10 = gated, provenance-linked and restoration-tested). Point coordinates represent an illustrative pre-governance factory baseline versus the minimum admission threshold (score >= 7 on both axes) required to enter the governed spine.
Unit: derived governance readiness score (0 to 10 scale). Status: derived. Sources: S01, S02, S05, S09, S26.
1.16 · the bodyExhibit 04 above; its figures are illustrative where marked.
In Exhibit 4, a typical pre-governance baseline leaves production, quality, maintenance and utility logs in the lower-left "Data pile" quadrant while order ledgers sit near the middle (illustrative baseline coordinates on the derived 0 to 10 rubric).[S01][S02][S05] Training models on lower-left records imports unowned definitions and unflagged missingness directly into predictions.[S04][S18][S19] Reaching the upper-right "Governed data spine" gate (8.0, 8.0) requires passing all four admission tests during an on-site audit using the checklists in Chapters 2 through 8.[S01][S02][S09][S26]
1.19
World Bank guidance on MSME digitalisation stresses embedding monitoring and evaluation throughout design and implementation to track impact and improve learning.[S36] On the shop floor, the primary operational measure is the cycle time from an abnormal physical event to a verified corrective action across five steps:[S10][S36][S38]
In manufacturing data mining, the CRISP-DM lifecycle runs through six phases: business understanding, data understanding, data preparation, modelling, evaluation and deployment.[S43] A model with high offline accuracy in phase four still fails in evaluation or deployment if alerts arrive after fabric is cut, if operators cannot act on the output or if data-preparation and maintenance costs exceed gross savings.[S18][S39][S43][S0101] When resources, skills, costs or expected benefits are not met, postponing or stopping an AI project is a valid governance decision.[S44]
The fourteen chapters of this paper build that governed chain across four parts:
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.
52 sources · 361 citation marks on this pageMost cited: S05, 25 times
Grade AGrade BGrade CKnot size: citations on this page
All 52 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.
R09 · Grade C · inherited record
Inherited evidence R09
Camtext SA integrated cotton-to-textile project, US$300 million (180 billion XAF); Panafritex and Arise with SODECOTON, CNPS and Marlo; 12 November 2025.
Publication date unrecorded. Inherited source-text record. The original publication has not been independently reopened for this web edition.
R10 · Grade C · inherited record
Inherited evidence R10
Camtext SA: 12,000 t a year processed; about 12,000 jobs at the Dibamba hub and 3,000 at Garoua; return on investment near 11 years; optimal capacity around 2033.
Publication date unrecorded. Inherited source-text record. The original publication has not been independently reopened for this web edition.
R11 · Grade C · inherited record
Inherited evidence R11
CICAM market share from about 85 percent three decades ago to about 5 percent today.
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.
R14 · Grade C · inherited record
Inherited evidence R14
Cameroon minimum wage (SMIG) XAF 41,875 a month in 2023, raised to XAF 43,969 a month from 23 February 2024, about US$76, a rise of 5 percent.
Publication date unrecorded. Inherited source-text record. The original publication has not been independently reopened for this web edition.
R15 · Grade C · inherited record
Inherited evidence R15
Average salary across sectors in Cameroon about US$359 a month.
Publication date unrecorded. Inherited source-text record. The original publication has not been independently reopened for this web edition.
R16 · Grade C · inherited record
Inherited evidence R16
Informal share of employment about 90 percent; standard 40-hour week, 48-hour maximum, overtime at 1.2 times.
Publication date unrecorded. Inherited source-text record. The original publication has not been independently reopened for this web edition.
R27 · Grade B · inherited record
Inherited evidence R27
AI at scale in fashion: inventory cost down 35 to 40 percent; forecast accuracy up 20 to 30 percent; revenue per customer up 15 to 25 percent; forecast errors cut 20 to 50 percent; lost sales down up to 65 percent; 75 percent of executives prioritising; sector operating-profit potential US$150 to 275 billion over three to five years.
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.
R32 · Grade C · inherited record
Inherited evidence R32
CFA franc to US dollar reference rates; franc pegged to the euro at 655.957. The GFT study's method note places the franc near 590 to 610 per US dollar across the 2023 study window and near 563 to 577 more recently, and treats currency conversions as estimates.
Publication date unrecorded. Inherited source-text record. The original publication has not been independently reopened for this web edition.
c01-S01 · Grade A
ISO 8000-1:2022: Data quality, Part 1: Overview
The series defines which characteristics of data are relevant to data quality, specifies requirements applicable to those characteristics, and provides guidelines for improving data quality.
2022. Register access date: 2026-10-30. Source provenance is supplied by the paper evidence register.
Open the recorded source ↗c01-S02 · Grade A
ISO 8000-150:2022: Data quality, Part 150: Data quality management: Roles and responsibilities
This document specifies the key considerations for organizations that are establishing appropriate roles and responsibilities for data quality management.
2022. Register access date: 2026-10-30. Source provenance is supplied by the paper evidence register.
Open the recorded source ↗c01-S03 · Grade A
ISO/IEC 25012:2008: Software engineering, Software product Quality Requirements and Evaluation, Data quality model
Data quality requirements and data quality measures can be categorized with the data quality characteristics in 5.2 and used by an evaluation process to analyse data independently from other computer system's components.
2008. Register access date: 2026-10-30. Source provenance is supplied by the paper evidence register.
Open the recorded source ↗c01-S04 · Grade A
ISO/IEC 5259-3:2024: Artificial intelligence, Data quality for analytics and machine learning, Part 3: Data quality management requirements and guidelines
ISO/IEC 5259-3 defines the requirements and guidance for managing the quality of data used in analytics and machine learning (ML).
2024. Register access date: 2026-10-30. Source provenance is supplied by the paper evidence register.
Open the recorded source ↗c01-S05 · Grade A
IEC 62264-1:2013, Edition 2.0: Enterprise-control system integration, Part 1: Models and terminology
This part of IEC 62264 provides standard models and terminology for describing the interfaces between the business systems of an enterprise and its manufacturing operations and control systems.
2013. Register access date: 2026-10-30. Source provenance is supplied by the paper evidence register.
Open the recorded source ↗c01-S06 · Grade A
OPC Unified Architecture, Part 1: Overview and Concepts, OPC 10000-1, version 1.05.06
It provides a comprehensive framework for modelling information, secure communication, and ensuring interoperability between diverse systems and devices.
2025-10-22. Register access date: 2026-10-30. Source provenance is supplied by the paper evidence register.
Open the recorded source ↗c01-S07 · Grade A
MQTT Version 5.0
MQTT is a Client Server publish/subscribe messaging transport protocol.
2019-03-07. Register access date: 2026-10-30. Source provenance is supplied by the paper evidence register.
Open the recorded source ↗c01-S08 · 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. Register access date: 2026-10-30. Source provenance is supplied by the paper evidence register.
Open the recorded source ↗c01-S09 · Grade A
PROV-O: The PROV Ontology
It provides a set of classes, properties, and restrictions that can be used to represent and interchange provenance information generated in different systems and under different contexts.
2013-04-30. Register access date: 2026-10-30. Source provenance is supplied by the paper evidence register.
Open the recorded source ↗c01-S10 · Grade A
Current Standards Landscape for Smart Manufacturing Systems, NISTIR 8107
Standards are fundamental 1) to facilitate the delivery of the right information at the right time, 2) to enable actions based on that information and 3) to reduce risk of technology adoption and development.
2016-02-23. Register access date: 2026-10-30. Source provenance is supplied by the paper evidence register.
Open the recorded source ↗c01-S11 · Grade A
Artificial Intelligence Risk Management Framework, AI RMF 1.0, NIST AI 100-1
Human judgment should be employed when deciding on the specific metrics related to AI trustworthiness characteristics and the precise threshold values for those metrics.
2023-01-26. Register access date: 2026-10-30. Source provenance is supplied by the paper evidence register.
Open the recorded source ↗c01-S13 · Grade A
ISO/IEC 5338:2023: Information technology, Artificial intelligence, AI system life cycle processes
This document defines a set of processes and associated concepts for describing the life cycle of AI systems based on machine learning and heuristic systems.
2023. Register access date: 2026-10-30. Source provenance is supplied by the paper evidence register.
Open the recorded source ↗c01-S15 · Grade B
MLOps: Continuous delivery and automation pipelines in machine learning
Information about each execution of the ML pipeline is recorded in order to help with data and artifacts lineage, reproducibility, and comparisons.
2024-08-28. Register access date: 2026-10-30. Source provenance is supplied by the paper evidence register.
Open the recorded source ↗c01-S16 · Grade B
MLOps Maturity Model
Use this model to develop your MLOps capability gradually instead of facing the full complexity of mature implementation upfront.
2026-06-03. Register access date: 2026-10-30. Source provenance is supplied by the paper evidence register.
Open the recorded source ↗c01-S17 · Grade B
Model monitoring in production
You can set thresholds for these metrics to trigger alerts about model or data anomalies via Azure Machine Learning or Azure Event Grid.
2026-01-27. Register access date: 2026-10-30. Source provenance is supplied by the paper evidence register.
Open the recorded source ↗c01-S18 · Grade B
Hidden Technical Debt in Machine Learning Systems
Using the software engineering framework of technical debt, we find it is common to incur massive ongoing maintenance costs in real-world ML systems.
2015. Register access date: 2026-10-30. Source provenance is supplied by the paper evidence register.
Open the recorded source ↗c01-S19 · Grade B
Everyone wants to do the model work, not the data work: Data Cascades in High-Stakes AI
Data cascades are pervasive (92% prevalence), invisible, delayed, but often avoidable.
2021. Register access date: 2026-10-30. Source provenance is supplied by the paper evidence register.
Open the recorded source ↗c01-S21 · Grade B
Learning under Concept Drift: A Review
Concept drift describes unforeseeable changes in the underlying distribution of streaming data over time.
2020-04-13. Register access date: 2026-10-30. Source provenance is supplied by the paper evidence register.
Open the recorded source ↗c01-S22 · Grade A
Guide to Operational Technology Security, NIST SP 800-82 Revision 3
This document provides guidance on how to secure operational technology (OT) while addressing their unique performance, reliability, and safety requirements.
2023. Register access date: 2026-10-30. Source provenance is supplied by the paper evidence register.
Open the recorded source ↗c01-S23 · Grade A
The NIST Cybersecurity Framework 2.0
Governance activities are critical for incorporating cybersecurity into an organization's broader enterprise risk management (ERM) strategy.
2024-02-26. Register access date: 2026-10-30. Source provenance is supplied by the paper evidence register.
Open the recorded source ↗c01-S25 · Grade A
Security and Privacy Controls for Information Systems and Organizations, NIST SP 800-53 Revision 5
Employ the principle of least privilege, allowing only authorized accesses for users (or processes acting on behalf of users) that are necessary to accomplish assigned organizational tasks.
2020-12-10. Register access date: 2026-10-30. Source provenance is supplied by the paper evidence register.
Open the recorded source ↗c01-S26 · Grade A
#StopRansomware Guide
Maintain offline, encrypted backups of critical data, and regularly test the availability and integrity of backups in a disaster recovery scenario [CPG 2.R].
2023. Register access date: 2026-10-30. Source provenance is supplied by the paper evidence register.
Open the recorded source ↗c01-S27 · Grade A
Incident Response Recommendations and Considerations for Cybersecurity Risk Management: A CSF 2.0 Community Profile, NIST SP 800-61 Revision 3
Procedures can be tested or exercised periodically to verify their accuracy and can be used to help train new personnel.
2025. Register access date: 2026-10-30. Source provenance is supplied by the paper evidence register.
Open the recorded source ↗c01-S29 · Grade A
Regulation (EU) 2016/679, General Data Protection Regulation
The personal data should be adequate, relevant and limited to what is necessary for the purposes for which they are processed.
2016-04-27. Register access date: 2026-10-30. Source provenance is supplied by the paper evidence register.
Open the recorded source ↗c01-S30 · Grade A
Protection of workers' personal data: An ILO code of practice
Decisions concerning a worker should not be based solely on the automated processing of that worker's personal data.
1997. Register access date: 2026-10-30. Source provenance is supplied by the paper evidence register.
Open the recorded source ↗c01-S31 · Grade A
Law No. 2024/017 of 23 December 2024 relating to personal data protection in Cameroon
The processing of personal data shall be subject to prior authorization by the Personal Data Protection Authority.
2024-12-23. Register access date: 2026-10-30. Source provenance is supplied by the paper evidence register.
Open the recorded source ↗c01-S32 · Grade A
Regulation (EU) 2023/2854 on harmonised rules on fair access to and use of data, Data Act
This Regulation ensures that users of a connected product or related service in the Union can access, in a timely manner, the data generated by the use of that connected product or related service and that those users can use the data, including by sharing them with third parties of their choice.
2023-12-13. Register access date: 2026-10-30. Source provenance is supplied by the paper evidence register.
Open the recorded source ↗c01-S33 · Grade B
Understand extended offline capabilities for IoT Edge devices, modules, and child devices
If your devices go offline, the IoT Edge parent device stores all device-to-cloud messages until the connection is reestablished.
2024-06-10. Register access date: 2026-10-30. Source provenance is supplied by the paper evidence register.
Open the recorded source ↗c01-S34 · Grade B
Uptake of digital tools among SMEs: Western Balkans Enterprise Survey
Tools that support more integrated business operations, such as cloud computing, enterprise resource planning (ERP) and customer relationship management (CRM) systems, are used by between 10-20% of surveyed firms, while e-commerce and supplier-customer management systems are less common.
2026-06-26. Register access date: 2026-10-30. Source provenance is supplied by the paper evidence register.
Open the recorded source ↗c01-S35 · Grade B
Employee empowerment through digital tools: The case of a German bakery SME
However, following managing director Holste the change of traditional processes into digital ones requires sizeable efforts that do not directly pay off but involve a longer learning process.
2022-11-21. Register access date: 2026-10-30. Source provenance is supplied by the paper evidence register.
Open the recorded source ↗c01-S36 · Grade B
The Missing Element of Firm Digitalization: Lessons for the EU Member States: Best practices for enhancing MSMEs Digitalization and Managerial Capacity Building
Embedding monitoring and evaluation throughout each step of the program design and implementation process is crucial for generating knowledge, tracking impact, and improving learning for the further improvement of interventions pursuing the digitalization of MSMEs.
2023. Register access date: 2026-10-30. Source provenance is supplied by the paper evidence register.
Open the recorded source ↗c01-S37 · Grade B
Challenges with digital technology adoption by SMEs: A case study from Viet Nam
Firms that successfully took up the platform assigned a point person who championed adoption within the firm.
2025-05-20. Register access date: 2026-10-30. Source provenance is supplied by the paper evidence register.
Open the recorded source ↗c01-S38 · 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. Register access date: 2026-10-30. Source provenance is supplied by the paper evidence register.
Open the recorded source ↗c01-S39 · Grade A
Artificial Intelligence with Open and Scaled Data Sharing in Semiconductor Manufacturing: Workshop Report
Also, manufacturers have reported that the cost of data processing for a single application can be as high as 70% of the total cost.
2025. Register access date: 2026-10-30. Source provenance is supplied by the paper evidence register.
Open the recorded source ↗c01-S41 · Grade B
Technical Guidance for Calculating Scope 3 Emissions, version 1.0
Collecting data, assessing data quality, and improving data quality is an iterative process.
2013. Register access date: 2026-10-30. Source provenance is supplied by the paper evidence register.
Open the recorded source ↗c01-S42 · Grade B
Is MLOps different in Industry 4.0? General and Specific Challenges
Also the starting-level of the technical infrastructure as well as the aimed at degree of automation influence very significantly the overall complexity of implementing MLOps in Industry 4.0.
2022. Register access date: 2026-10-30. Source provenance is supplied by the paper evidence register.
Open the recorded source ↗c01-S43 · Grade B
Ensuring the Robustness and Reliability of Data-Driven Knowledge Discovery Models in Production and Manufacturing
Briefly, CRISP-DM divides the data mining-related knowledge discovery process into six phases: 1) business understanding, 2) data understanding, 3) data preparation, 4) modeling, 5) evaluation and 6) deployment.
2021-06-14. Register access date: 2026-10-30. Source provenance is supplied by the paper evidence register.
Open the recorded source ↗c01-S44 · Grade B
Supporting SME companies in mapping out AI potential: a Finnish AI development case
One must remember that in IT investments, the decision to postpone, e.g. an AI project, is also valid if resources, skills, costs, or expected benefits are not met.
2024-07-27. Register access date: 2026-10-30. Source provenance is supplied by the paper evidence register.
Open the recorded source ↗c01-S0101 · Grade A
Data Analytics for Smart Manufacturing Systems
Much time is spent on identifying the target performance objectives, selecting or developing the appropriate DA tool, and managing input data (including data acquisition and data formatting.)
2025-03-26. Register access date: 2026-10-30. Source provenance is supplied by the paper evidence register.
Open the recorded source ↗c01-S0102 · Grade A
Data Catalog Vocabulary (DCAT) - Version 3
DCAT enables a publisher to describe datasets and data services in a catalog using a standard model and vocabulary that facilitates the consumption and aggregation of metadata from multiple catalogs.
2024-08-22. Register access date: 2026-10-30. Source provenance is supplied by the paper evidence register.
Open the recorded source ↗S01 International Organization for Standardization. 2022.
S02 International Organization for Standardization. 2022.
S09 World Wide Web Consortium. 2013-04-30.
S26 CISA, MS-ISAC, NSA and FBI. 2023.
S19 Association for Computing Machinery, CHI 2021. 2021.
S0101 National Institute of Standards and Technology. 2025-03-26.
S05 International Electrotechnical Commission. 2013.
S08 GS1 AISBL. 2022.
S43 Frontiers in Artificial Intelligence. 2021-06-14.
R06 Cameroon Trade Hub. Publication date unrecorded; inherited source-text check.
R07 Ecofin Agency. Publication date unrecorded; inherited source-text check.
S38 International Organization for Standardization. 2014.
S04 ISO/IEC. 2024.
S15 Google Cloud. 2024-08-28.
S36 World Bank. 2023.
R12 Business in Cameroon. Publication date unrecorded; inherited source-text check.
R11 Business in Cameroon; MMI News. Publication date unrecorded; inherited source-text check.
R09 Ecofin Agency. Publication date unrecorded; inherited source-text check.
R10 Ecofin Agency. Publication date unrecorded; inherited source-text check.
S18 Advances in Neural Information Processing Systems. 2015.
S16 Microsoft Learn, Azure Architecture Center. 2026-06-03.
S07 OASIS Open. 2019-03-07.
S34 OECD. 2026-06-26.
S23 National Institute of Standards and Technology. 2024-02-26.
S11 National Institute of Standards and Technology. 2023-01-26.
S37 World Bank. 2025-05-20.
S42 SCITEPRESS, IN4PL 2022. 2022.
S10 National Institute of Standards and Technology. 2016-02-23.
S39 National Institute of Standards and Technology. 2025.
S0102 World Wide Web Consortium. 2024-08-22.
R29 MDPI; industry inspection literature. Publication date unrecorded; inherited source-text check.
R27 McKinsey, The State of Fashion 2025. Publication date unrecorded; inherited source-text check.
S41 GHG Protocol, World Resources Institute and World Business Council for Sustainable Development. 2013.
R14 WageIndicator. Publication date unrecorded; inherited source-text check.
R15 wage.is. Publication date unrecorded; inherited source-text check.
R16 WageIndicator. Publication date unrecorded; inherited source-text check.
R32 Wise; XE. Publication date unrecorded; inherited source-text check.
S22 National Institute of Standards and Technology. 2023.
S29 European Parliament and Council of the European Union. 2016-04-27.
S30 International Labour Office. 1997.
S31 Presidency of the Republic of Cameroon. 2024-12-23.
S33 Microsoft. 2024-06-10.
S25 National Institute of Standards and Technology. 2020-12-10.
S03 ISO/IEC. 2008.
S44 The Journal of Technology Transfer, Springer Nature. 2024-07-27.
S17 Microsoft Learn, Azure Machine Learning. 2026-01-27.
S21 arXiv, author manuscript of an IEEE Transactions on Knowledge and Data Engineering article. 2020-04-13.
S06 OPC Foundation. 2025-10-22.
S27 National Institute of Standards and Technology. 2025.
S32 European Parliament and Council of the European Union. 2023-12-13.
S13 ISO/IEC. 2023.
S35 OECD. 2022-11-21.
Keep the thread · chapter 01 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: Govern meaning before scale
The binding constraint on industrial AI is often the broken chain between work, record and decision, not the absence of an algorithm.
02Part I: Govern meaning before scale
A workable governance model fits into existing factory roles and makes every definition, exception and release somebody's explicit decision.
03Part I: Govern meaning before scale
Shared identifiers, event semantics and KPI formulas let unlike systems describe the same work without hiding their differences.
04Part I: Govern meaning before scale
A small set of enforced checks, visible exceptions and complete lineage is more useful than an unmeasured claim that the data are clean.
05Part II: Build the minimum durable spine
A durable architecture preserves source evidence, adds governed context once and lets each decision consume a controlled product.
06Part II: Build the minimum durable spine
A value becomes useful evidence only when the factory can establish what it describes, when it happened and which state or transformation it represents.
07Part II: Build the minimum durable spine
The data spine must protect production, people and legitimate access at the same time, with controls proportionate to a small team.
08Part II: Build the minimum durable spine
A factory proves resilience by operating through disconnection, restoring its records and leaving a vendor without losing its history.
09Part III: Control models as production change
The first model gate is operational and economic: nobody should train before the action, baseline, data rights and stop conditions are explicit.
10Part III: Control models as production change
A model becomes a release candidate only when another competent person can reconstruct what data, code, configuration and evaluation produced it.
11Part III: Control models as production change
Release authority follows demonstrated local fitness, safe fallback and accountable acceptance, not a strong aggregate validation score.
12Part III: Control models as production change
A useful monitor detects change, establishes whether it matters and routes a named person through a rehearsed response.
13Part IV: Prove value and scale deliberately
The spine earns further investment when people use it, decisions improve and net operational benefit survives a comparison with the baseline.
14Part IV: Prove value and scale deliberately
GFT should scale the spine only as ownership, meaning, recovery, reproducibility and net value are demonstrated in production.