Research libraryThe Industrial Data Spine
Chapter 02 of 14 · Part I · Draft web edition
A workable governance model fits into existing factory roles and makes every definition, exception and release somebody's explicit decision.
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 mapping of the six functional governance roles and three decision anchors across technical execution closeness (x-axis) and operational decision authority (y-axis), derived from role assignment in S02, human threshold setting in S11, AI risk integration in S14 and enterprise risk governance in S23. Coordinates are illustrative design positions for a small factory, not empirical measurements.
Unit: illustrative governance coordinates (0 to 10 scale). Status: illustrative. Sources: S02, S11, S14, S23. Illustrative figures are teaching examples, not measured factory results.
A small factory does not need a standalone data department to govern its records and models; it needs named decision rights attached to existing operational roles.[S02] When a mill or garment plant records production, quality, maintenance and order events, ambiguity over who may define a metric, alter a schema, grant access, quarantine a broken feed or approve a model release corrupts the evidence base.[S01] In GFT's Cameroon textile programme, whose evidence register prepared 31 January 2024 records about 95 percent of cotton fibre exported raw and about 5 percent processed locally against a 50 percent local-processing target by 2030 (inherited entries not refreshed in 2026 with unrecorded original publication dates, stating strategic ambition rather than current factory output), digitalisation exists to make domestic processing reliable and investable.[R06][R07] Meeting that challenge with a lean team requires one business owner, one domain steward and one technical custodian for every critical dataset, indicator and model, with explicit escalation into existing factory meetings.[S02][S23]
Without explicit decision rights, data work defaults to convenience.[S19] An engineer modifies a spreadsheet column for a new style, an analyst excludes rework from first-pass yield to tidy a slide, or a vendor updates an inspection threshold during maintenance without recording the change.[S18][S38] Such informal edits break provenance between physical work and management action.[S09] International data-quality, cybersecurity and AI risk standards require organisations to establish clear roles and responsibilities before relying on technical controls, keeping technical choices subordinate to operational risk appetite.[S02][S11][S23][S0201]
This chapter translates those standards into an operating charter for small and mid-sized factories.[S02][S14] It assigns one owner to each of six factory domains, separates ownership from stewardship and custody, protects line stewards with a weekly time budget and quarantine authority, subjects definition changes and model releases to versioned approval and independent challenge, and routes cross-domain decisions through WP02's management cadence.[S01][S11][S23][S36][S38]
2.1
Related · chapter 058 paragraphs step it
Read-out
Derived minimum architecture separating capture, immutable raw evidence, contextual canonical records, published data products and operational decisions, based on interface boundaries in S05, provenance modelling in S09 and S0501, smart manufacturing standards in S10 and pipeline lineage in S15.
Unit: . Status: derived. Sources: S05, S09, S10, S15, S0501.
Read this exhibit in its chapter ↗Accountability fails whenever a dataset is owned by "everyone" or left to whichever technician administers the database.[S02] Under ISO 8000-150:2022, establishing appropriate roles and responsibilities is foundational for data quality management.[S02] In a small or mid-sized textile and garment plant, creating new executive posts for data governance is neither affordable nor necessary; instead, the factory assigns data ownership to the existing heads of six bounded operational domains: production, quality, maintenance, orders, people and sustainability.[S02][S05]
Dividing the data spine into six domains aligns governance with the interfaces between enterprise planning and shop-floor execution in IEC 62264-1 (ISA-95).[S05] Each domain covers a distinct family of operational events, master records and management decisions:[S02][S05]
Assigning a domain owner does not mean a department head keys in hourly logs or writes queries.[S02] Transaction capture belongs at the point of work, where operators, inspectors, storekeepers and technicians observe the event.[S05][S08] The domain owner holds decision authority over four structural questions:[S02]
Separating governance decisions from routine entry prevents a recurring failure in turnarounds.[R12][S19] Public reporting on Cameroon's legacy textile manufacturer CICAM in July 2023 noted capacity utilisation of 33.7 percent at Garoua, 48.9 percent at CICAM 1 and 10.83 percent at CICAM 2, with underutilisation costing about CFAF 2 billion in 2020 (an inherited grade-C report not refreshed in 2026).[R12] Under utilisation pressure, managers need clarity on whether lost output stems from material starvation, mechanical fault, power loss or rework; if supervisors define stoppage categories differently on local sheets, leadership cannot direct engineering time to the binding constraint.[R12][S02][S38]
The people domain demands stricter boundaries than machine telemetry because its records describe identifiable workers.[S29][S30] Under the data-minimisation principle in Recital 39 of Regulation (EU) 2016/679 (GDPR), personal data must be adequate, relevant and limited to what is necessary for the stated purposes.[S29] Tracking line-level or bundle-level completion supports scheduling and traceability, whereas continuous individual biometric or video tracking is rarely justified for line balancing.[S08][S29]
Furthermore, Principle 5.5 of the International Labour Office (ILO) code of practice on workers' personal data states that decisions concerning a worker should not be based solely on automated processing of that worker's personal data, which GFT treats as a mandatory engineering constraint.[S30] In Cameroon, Section 19(1) of Law No. 2024/017 of 23 December 2024 relating to personal data protection provides that processing personal data is subject to prior authorisation by the Personal Data Protection Authority.[S31] Because current enforcement requires local legal verification, the people-domain owner must obtain qualified Cameroonian legal advice before authorising collection, transfer or algorithmic analysis of worker records; this paper provides engineering design, not legal advice.[S31]
2.5
A small factory team often collapses all data work into one job description, hiring a single "IT and data officer" expected to define metrics, clean shift sheets, maintain servers, build dashboards and validate models.[S18][S20] That concentration fails because a technical administrator lacks operational authority to fix capture habits on the floor ([S02], [S19]), and allowing the person who administers storage or builds a model to alter KPI formulas and approve releases removes independent quality and risk control.[S11][S14]
To preserve checks and balances without inflating headcount, GFT defines six functional roles across the data spine, combining role principles from ISO 8000-150 with risk-governance structures from NIST CSF 2.0, NIST AI RMF 1.0 and ISO/IEC 23894.[S02][S11][S14][S23] These roles are bundles of decision rights, not separate full-time salaries:[S02][S0201]
As Exhibit 5 illustrates, high operational authority sits in the upper quadrants with the domain owner, model owner and security approver, while technical execution sits in the lower-right quadrant with the custodian and builder.[S02][S11][S23] The chart clarifies which hat a person wears for each decision and which combinations create a conflict of interest.[S02][S14]
In a small factory with an illustrative workforce of 150 to 500 staff and two or three technical specialists, one person will hold more than one role in the charter.[S02][S16] For instance, the quality assurance manager may serve as both quality domain owner and operating model owner for an inspection system, while an industrial engineer acts as both production and quality steward.[S02][S14] Combining compatible roles keeps overhead low, provided the factory enforces three separation rules from NIST AI RMF Playbook guidance and ISO/IEC 23894:[S11][S14][S0201]
Decision rights remain abstract unless each right is tied to a trigger, an approver, a required artefact and an escalation forum.[S02][S0201] Synthesising role assignment from ISO 8000-150 ([S02]), provenance modelling from W3C PROV-O ([S09]), pipeline lineage from MLOps architecture ([S15]) and incident-response verification from NIST SP 800-61 ([S27]), Exhibit 6 specifies the seven lifecycle decisions governing the data spine.
Read-out
| 1. Define KPI or schema | Domain steward | Domain owner | Versioned KPI or contract entry with impact note | Weekly operations review |
|---|---|---|---|---|
| 2. Grant data access | Data custodian | Domain owner and security approver | Scoped access ticket with purpose and expiry | Monthly governance review |
| 3. Quarantine feed | Domain steward or automated gate | Domain steward (immediate hold) | Quarantine log with failed rule ID and batch scope | Daily shift tier-two meeting |
| 4. Correct historical record | Source role and steward | Domain owner | Appended superseding event linked to original ID | Weekly operations review |
| 5. Release model version | Model builder | Model owner and security approver | Candidate record, local slice metrics and fallback test | Monthly risk and S&OP gate |
| 6. Roll back active model | Model owner, steward or OT lead | Model owner (immediate revert) | Incident ticket, alert trigger and restored baseline pointer | Daily shift meeting and post-mortem |
| 7. Retire dataset or model | Domain or model owner | Domain owner and custodian | Archival manifest, lineage freeze and consumer notice | Monthly governance review |
Derived governance specification linking seven lifecycle decisions across the industrial data spine to their proposing or executing role, accountable approver, mandatory evidence artefact and escalation forum, synthesised from S02, S09, S15 and S27.
Unit: governed lifecycle decision classes. Status: derived. Sources: S02, S09, S15, S27.
Two asymmetric rules in Exhibit 6 deserve emphasis.[S02][S27] Opening a new exposure, such as changing a definition, granting access or releasing a model, requires prior approval backed by documented evidence.[S02][S15] Conversely, containing a failure, whether quarantining a corrupted feed or rolling back a misbehaving model to manual fallback, is a pre-authorised protective action that the steward, model owner or OT lead executes immediately on the shift, with formal review at the next scheduled meeting.[S15][S22][S27]
2.9
Production steward (line logs and states)
Production steward (line logs and states): 4.5 illustrative protected hours per week · illustrative
Data stewardship fails when management treats it as an unbudgeted chore added to a full workload.
Quality steward (defects and quarantine)
Quality steward (defects and quarantine): 4 illustrative protected hours per week · illustrative
Internal adoption champion (coordination)
Internal adoption champion (coordination): 4 illustrative protected hours per week · illustrative
Orders steward (styles, lots and dispatch)
Orders steward (styles, lots and dispatch): 2.5 illustrative protected hours per week · illustrative
Maintenance steward (assets and fault codes)
Maintenance steward (assets and fault codes): 2.5 illustrative protected hours per week · illustrative
Sustainability steward (meters and waste)
Sustainability steward (meters and waste): 1.5 illustrative protected hours per week · illustrative
People steward (rosters and training logs)
People steward (rosters and training logs): 1 illustrative protected hours per week · illustrative
The method behind it: Worked example.
Illustrative weekly time budget allocated to domain stewardship and adoption coordination within Cameroon's 40-hour standard working week (R16), applying role assignment from S02 and internal champion coordination from S37. Values are illustrative planning allowances to be resized after a 30-day factory workload baseline, not universal staffing benchmarks.
Unit: illustrative protected hours per week. Status: illustrative. Sources: R16, S02, S37. Illustrative figures are teaching examples, not measured factory results.
2.9 · the bodyExhibit 07 above; its figures are illustrative where marked.
Data stewardship fails when management treats it as an unbudgeted chore added to a full workload.[S19][S35] Sambasivan et al. found in high-stakes AI systems that upstream data cascades were pervasive (92 percent of studied cases), invisible and delayed because organisations undervalued data work.[S19] Although that study examined high-stakes AI broadly rather than Cameroonian manufacturing, the shop-floor mechanism is identical: without protected weekly hours to inspect rejected records, verify master-data codes and coach operators, defects accumulate until reports and models fail.[S04][S19]
…upstream data cascades were pervasive (92 percent of studied cases)
S19
Rather than hiring full-time data bureaucrats, a small factory protects weekly working time for existing supervisors, planners and engineers to perform stewardship.[S02][S35] In Cameroon, labour reference data record a standard 40-hour week, a 48-hour maximum and overtime at 1.2 times base pay, while the statutory minimum wage (SMIG) rose 5 percent from XAF 41,875 in 2023 to XAF 43,969 a month from 23 February 2024 (about US$76), and multi-sector average wages sit near US$359 a month in inherited grade-C sources.[R14][R15][R16] Pushing stewardship beyond the 40-hour week without workload relief adds overtime cost or leaves checks undone when orders run late.[R16][S35]
To make the commitment explicit before a site baselines its own defect rates, Exhibit 7 shows an illustrative weekly allocation across the six stewards and internal champion inside a 40-hour week.[R16][S02][S37]
As shown in Exhibit 7, the illustrative allowance totals 20.0 hours per week across seven staff members, equivalent to 0.50 of a full-time equivalent (FTE) spread across the plant rather than a new hire.[R16][S02] Production and quality stewards receive the largest allocations (4.5 and 4.0 illustrative hours per week) because hourly line logs and defect sheets carry the highest transaction volume.[S04][S38] A plant migrating legacy spreadsheets needs more stewardship time during its first 90 days than one with stabilised barcode capture.[S35][S41]
Protected time is wasted if a steward lacks authority to stop a broken feed from polluting downstream reports and training sets.[S04][S19] Under ISO/IEC 5259-3:2024, managing data quality for analytics and machine learning requires systematic lifecycle controls.[S04] GFT grants every domain steward three operational rights:[S02][S04]
Interlude · the cloth half-woven
Woven: sections 2.0 to 2.9Still to weave: 2.12 to 2.21
Read-out
Inherited GFT programme evidence base prepared 31 January 2024. R06 and R07 were not refreshed in 2026 and their original publication dates are unrecorded. Figures describe sector-level baseline and policy ambition, not a current GFT factory measurement or causal productivity claim.
Unit: percent of cotton fibre. Status: sourced. Sources: R06, R07.
Read this exhibit in its chapter ↗2.12
Nothing erodes trust faster than two managers arriving at a meeting with different numbers for the same indicator because a formula changed without notice.[S01][S38] ISO 22400-2:2014 specifies that manufacturing KPIs must be defined by formula, inputs, time behaviour, unit and characteristics.[S38] ISO 8000-1:2022 frames quality around explicit requirements, while W3C's Data on the Web Best Practices requires a version history explaining each change.[S01][S0202]
ISO 22400-2:2014 specifies that manufacturing KPIs must be defined by formula
S38
Read-out
Derived eight-step workflow for modifying a KPI formula, event schema or taxonomy mapping, combining quality requirements from S01, role accountability from S02, manufacturing KPI structure from S38 and version-history documentation from S0202.
Unit: governed change-control sequence. Status: derived. Sources: S01, S02, S38, S0202.
first_pass_yield moving from v1.2 to v2.0) and effective shift timestamp.[S09][S0202]provisional-waiver or estimated rather than observed.[S04][S09]2.15
When a factory moves from descriptive reporting to predictive or vision models, the consequences of a flawed decision right multiply.[S11][S18] A model compresses historical records into recommendations or actions.[S13][S15] Left unchecked, the builder who trained a model has a bias to emphasise aggregate accuracy and downplay local failures on specific styles, night shifts or lighting conditions.[S04][S19]
Standards place human accountability at the centre of AI risk control.[S11][S12][S14] NIST AI RMF 1.0 states that human judgement should decide specific metrics and threshold values for AI trustworthiness.[S11] ISO/IEC 23894:2023 integrates risk management into AI activities ([S14]), while ISO/IEC 42001:2023 and ISO/IEC 5338:2023 govern management systems and lifecycle processes ([S12], [S13]).
Applying identical bureaucracy to a monthly yarn forecast and a real-time machine-control loop wastes management attention.[S11][S14] GFT scales challenge across four consequence tiers, combining NIST AI RMF Playbook lines-of-defence concepts ([S0201]), operational-technology safety guidance in NIST SP 800-82 Revision 3 ([S22]) and the ILO worker-protection principle ([S30]).
2.18
Role charters fail if nobody drives cross-functional follow-through under daily production pressure.[S35][S37] World Bank case-study evidence on SME digital adoption in Viet Nam found that firms successfully taking up the platform assigned an internal point person who championed adoption.[S37] OECD SME research similarly notes that converting traditional processes into digital ones requires sizeable effort and a longer learning process rather than an immediate payoff.[S35]
In GFT's governance architecture, the internal adoption champion coordinates execution across the six domains, reporting to the factory director without replacing domain owners or becoming a solitary "chief data officer".[S02][S23][S37] Instead, the champion holds three cross-cutting responsibilities:[S36][S37]
If the digitalisation programme lives inside one champion's laptop, the system collapses when that person falls ill or leaves.[S18][S37] NIST AI RMF Playbook guidance underscores that roles, responsibilities and communication lines must be documented across the organisation, with ownership metadata attached to work products.[S0201]
2.21
Model rollback
Model rollback: Decision horizon in WP02 cadence (0 to 10) 1.2, Cross-domain and risk consequence (0 to 10) 8.4. Quadrant: Emergency containment (Shift / OT) · derived
Scheduling a second tier of "data committee" meetings that duplicate the daily shift handover, weekly operations review and monthly sales and operations planning (S&OP) cycle breeds resentment.
OT safety halt
OT safety halt: Decision horizon in WP02 cadence (0 to 10) 0.8, Cross-domain and risk consequence (0 to 10) 9.4. Quadrant: Emergency containment (Shift / OT) · derived
Quarantine hold
Quarantine hold: Decision horizon in WP02 cadence (0 to 10) 1.8, Cross-domain and risk consequence (0 to 10) 3.5. Quadrant: Daily shift tier-two board · derived
Cross-domain KPI change
Cross-domain KPI change: Decision horizon in WP02 cadence (0 to 10) 5.5, Cross-domain and risk consequence (0 to 10) 6.8. Quadrant: Monthly risk and S&OP gate · derived
Waiver expiry audit
Waiver expiry audit: Decision horizon in WP02 cadence (0 to 10) 6.2, Cross-domain and risk consequence (0 to 10) 3.8. Quadrant: Weekly operations review · derived
Model release sign-off
Model release sign-off: Decision horizon in WP02 cadence (0 to 10) 8.6, Cross-domain and risk consequence (0 to 10) 8.8. Quadrant: Monthly risk and S&OP gate · derived
Access exception review
Access exception review: Decision horizon in WP02 cadence (0 to 10) 8.2, Cross-domain and risk consequence (0 to 10) 6.2. Quadrant: Monthly risk and S&OP gate · derived
The method behind it: Warning.
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.
2.21 · the bodyExhibit 09 above; its figures are illustrative where marked.
Scheduling a second tier of "data committee" meetings that duplicate the daily shift handover, weekly operations review and monthly sales and operations planning (S&OP) cycle breeds resentment.[S18][S35] In the GFT library, WP02 owns the operating cadence, visual boards and meeting method; WP14 refuses to redesign those forums or invent a parallel hierarchy.[S23][S36]
WP14 injects decision-ready escalations into the WP02 cadence only when an issue crosses domains or exceeds a steward's local authority.[S02][S23][S0201] Drawing on ISO 8000-150 role escalation ([S02]), NIST AI RMF threshold authority ([S11]) and NIST CSF 2.0 risk governance ([S23]), Exhibit 9 maps eight recurring escalations across decision horizon (x-axis) and cross-domain consequence (y-axis).
Every decision in the daily, weekly or monthly forum is recorded in the single WP02 action log, never in a separate data-team sheet.[S36][S37] Each entry carries four fields: artefact or defect ID, named owner, due date and verification condition (such as seven consecutive shifts with zero unmapped downtime codes on weaving line two, verified by the production steward).[S02][S36][S38]
kpi-cascade tool: Uses the six-domain ownership assignments and ISO 22400-2 contracts ([S38]) to trace how capture defects or unapproved formula changes propagate into plant OEE, first-pass yield and margin.[S02][S38]maturity tool: Evaluates named domain owners, protected weekly stewardship hours, versioned change control and practised rollback authority before scoring readiness under staged MLOps maturity ([S16]).[S02][S16]risk-matrix tool: Operationalises the four-tier consequence classification and challenge rules ([S11], [S14], [S22], [S30]) to weigh operational, worker-privacy and OT-safety exposure before model admission or release.[S14][S23]Anchoring every dataset, formula, access grant and model release to a named role inside the factory's management rhythm gives a small team institutional control without bureaucracy.[S02][S23] Once decision rights are clear, those owners need one shared language across machines, lines and ledgers, which the next chapter establishes.[S05][S38]
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.
38 sources · 406 citation marks on this pageMost cited: S02, 57 times
Grade AGrade BGrade CKnot size: citations on this page
All 38 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.
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.
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.
c02-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 ↗c02-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 ↗c02-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 ↗c02-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 ↗c02-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 ↗c02-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 ↗c02-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 ↗c02-S12 · Grade A
ISO/IEC 42001:2023: Information technology, Artificial intelligence, Management system
ISO/IEC 42001 is an international standard that specifies requirements for establishing, implementing, maintaining, and continually improving an Artificial Intelligence Management System (AIMS) within organizations.
2023. Register access date: 2026-10-30. Source provenance is supplied by the paper evidence register.
Open the recorded source ↗c02-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 ↗c02-S14 · Grade A
ISO/IEC 23894:2023: Information technology, Artificial intelligence, Guidance on risk management
The guidance also aims to assist organizations to integrate risk management into their AI-related activities and functions.
2023. Register access date: 2026-10-30. Source provenance is supplied by the paper evidence register.
Open the recorded source ↗c02-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 ↗c02-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 ↗c02-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 ↗c02-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 ↗c02-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 ↗c02-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 ↗c02-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 ↗c02-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 ↗c02-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 ↗c02-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 ↗c02-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 ↗c02-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 ↗c02-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 ↗c02-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 ↗c02-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 ↗c02-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 ↗c02-S20 · Grade B
Machine Learning Operations: Overview, Definition, and Architecture
MLOps includes several aspects, such as best practices, sets of concepts, and development culture.
2022-05-14. Register access date: 2026-10-30. Source provenance is supplied by the paper evidence register.
Open the recorded source ↗c02-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 ↗c02-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 ↗c02-S0201 · Grade A
NIST AI RMF Playbook: Govern
Roles and responsibilities and lines of communication related to mapping, measuring, and managing AI risks are documented and are clear to individuals and teams throughout the organization.
2023-01-26. Register access date: 2026-10-30. Source provenance is supplied by the paper evidence register.
Open the recorded source ↗c02-S0202 · Grade A
Data on the Web Best Practices
Provide a complete version history that explains the changes made in each version.
2017-01-31. Register access date: 2026-10-30. Source provenance is supplied by the paper evidence register.
Open the recorded source ↗S02 International Organization for Standardization. 2022.
S01 International Organization for Standardization. 2022.
R06 Cameroon Trade Hub. Publication date unrecorded; inherited source-text check.
R07 Ecofin Agency. Publication date unrecorded; inherited source-text check.
S23 National Institute of Standards and Technology. 2024-02-26.
S19 Association for Computing Machinery, CHI 2021. 2021.
S18 Advances in Neural Information Processing Systems. 2015.
S38 International Organization for Standardization. 2014.
S09 World Wide Web Consortium. 2013-04-30.
S11 National Institute of Standards and Technology. 2023-01-26.
S0201 National Institute of Standards and Technology. 2023-01-26.
S14 ISO/IEC. 2023.
S36 World Bank. 2023.
S05 International Electrotechnical Commission. 2013.
S04 ISO/IEC. 2024.
S08 GS1 AISBL. 2022.
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.
S25 National Institute of Standards and Technology. 2020-12-10.
R12 Business in Cameroon. Publication date unrecorded; inherited source-text check.
S20 arXiv, author version of an IEEE Access article. 2022-05-14.
S13 ISO/IEC. 2023.
S15 Google Cloud. 2024-08-28.
S16 Microsoft Learn, Azure Architecture Center. 2026-06-03.
S27 National Institute of Standards and Technology. 2025.
S35 OECD. 2022-11-21.
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.
S37 World Bank. 2025-05-20.
S41 GHG Protocol, World Resources Institute and World Business Council for Sustainable Development. 2013.
S0202 World Wide Web Consortium. 2017-01-31.
S12 ISO/IEC. 2023.
S44 The Journal of Technology Transfer, Springer Nature. 2024-07-27.
S43 Frontiers in Artificial Intelligence. 2021-06-14.
R29 MDPI; industry inspection literature. Publication date unrecorded; inherited source-text check.
Keep the thread · chapter 02 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.