Manufacturing Domain Ontology

FLOW ALGEBRA
Manufacturing Domain Ontology — v1.0
Kireth Semantic Library
Source: https://kirethresearch.org/booksai-core/ontologies/domain-ontologies/industry-domains/manufacturing-domain-ontology.html
Status: Domain-based draft; source page content not independently verified.

flow "ManufacturingDomainOntology" v1.0

DOMAIN:
  id: manufacturing_domain_ontology
  parent: industry_domains
  description: "Typed representation of manufacturing systems, materials, processes, resources, products, quality and traceability."

ENTITY_TYPES:
  ManufacturingOrganisation: {fields: [id, name, location, capabilities]}
  ManufacturingFacility: {fields: [id, operator_id, location, capacity]}
  ProductionLine: {fields: [id, facility_id, status, throughput]}
  Workstation: {fields: [id, line_id, function, status]}
  Machine: {fields: [id, workstation_id, machine_type, condition]}
  Operator: {fields: [id, role, qualification, authorisation]}
  Material: {fields: [id, specification, unit, origin]}
  Component: {fields: [id, specification, revision, lot_id]}
  Product: {fields: [id, specification, revision, lifecycle_status]}
  BillOfMaterials: {fields: [id, product_id, revision, components]}
  ManufacturingProcess: {fields: [id, process_type, inputs, outputs, parameters]}
  ProcessStep: {fields: [id, process_id, sequence, requirements]}
  WorkOrder: {fields: [id, product_id, quantity, due_date, status]}
  ProductionBatch: {fields: [id, work_order_id, lot_code, quantity, status]}
  QualityInspection: {fields: [id, batch_id, criterion, measurement, outcome]}
  Nonconformance: {fields: [id, batch_id, severity, disposition]}
  MaintenanceEvent: {fields: [id, machine_id, date, action, outcome]}
  InventoryRecord: {fields: [id, item_id, location, quantity, timestamp]}
  Supplier: {fields: [id, supplied_items, qualification_status]}
  Shipment: {fields: [id, batch_ids, origin, destination, status]}

RELATIONSHIP_TYPES:
  OPERATES: ManufacturingOrganisation -> ManufacturingFacility
  CONTAINS: ManufacturingFacility -> ProductionLine
  HAS_STATION: ProductionLine -> Workstation
  USES_MACHINE: Workstation -> Machine
  PERFORMS: Operator -> ProcessStep
  HAS_BOM: Product -> BillOfMaterials
  REQUIRES_COMPONENT: BillOfMaterials -> Component
  CONSUMES: ManufacturingProcess -> Material
  PRODUCES: ManufacturingProcess -> Product
  HAS_STEP: ManufacturingProcess -> ProcessStep
  EXECUTES: WorkOrder -> ManufacturingProcess
  GENERATES_BATCH: WorkOrder -> ProductionBatch
  INSPECTS: QualityInspection -> ProductionBatch
  FLAGS: QualityInspection -> Nonconformance
  MAINTAINS: MaintenanceEvent -> Machine
  SUPPLIES: Supplier -> Material
  TRACKS: InventoryRecord -> Material
  TRANSPORTS: Shipment -> ProductionBatch
  PRECEDES: ProcessStep -> ProcessStep
  TRACEABLE_TO: Product -> ProductionBatch

CONSTRAINTS:
  - Every entity identifier must be unique within its type.
  - Every production batch must reference a valid work order.
  - Process step ordering must not contain cycles unless explicitly modelled as a controlled rework loop.
  - Units of measure must be explicit and compatible for quantitative comparisons.
  - A product revision and BOM revision must be recorded for reproducible production.
  - Inspection acceptance requires documented criteria and measured results.
  - No automatic inference of regulatory compliance from inspection success alone.

INFERENCE_RULES:
  R1: IF a batch is generated by a work order AND that order executes a process THEN associate the batch with the process version used.
  R2: IF a component is included in a BOM AND that BOM belongs to a product THEN infer the product depends on the component, scoped to the BOM revision.
  R3: IF an inspection fails an acceptance criterion THEN propose a nonconformance for review; do not automatically classify final disposition.
  R4: IF a machine has overdue required maintenance THEN flag its operational readiness for verification, not an assumed failure.
  R5: IF a batch consumes a traced material lot THEN propagate a traceability link from batch to material lot.
  R6: IF a process step precedes another THEN enforce ordering unless an approved alternative route is documented.
  R7: IF inventory quantity falls below a defined reorder threshold THEN generate a replenishment recommendation.
  R8: IF a quality observation lacks calibrated measurement provenance THEN mark its reliability as unresolved.

CONFIDENCE:
  scale: [0.0, 1.0]
  attributes: [source_reliability, measurement_quality, temporal_validity, contextual_fit]
  rule: "Derived confidence must not exceed the weakest mandatory evidence input without a documented justification."

SEMANTIC_EQUIVALENCE:
  - {term: "bill of materials", aliases: ["BOM", "parts list"], condition: "Equivalent only where the parts list is a controlled product BOM."}
  - {term: "work order", aliases: ["production order"], condition: "Equivalent only where workflows and definitions align."}
  - {term: "batch", aliases: ["lot"], condition: "Context-dependent; preserve source identifiers and scope."}
  - {term: "quality inspection", aliases: ["product inspection"], condition: "Equivalent only for product-level inspections."}

PROVENANCE:
  required: [source_id, source_type, observation_time, recorded_by, revision]
  optional: [instrument_id, calibration_reference, approval_reference, licence]
  principle: "Distinguish observed facts, imported records, derived relations and hypotheses."

FLOW_HASH:
  algorithm: SHA-256
  canonicalisation: "UTF-8; stable key ordering; normalised whitespace; exclude the stored hash value."
  scope: [domain, entity_types, relationship_types, constraints, inference_rules, semantic_equivalence]
  value: "COMPUTE_AT_PUBLISH_TIME"

UPDATE_MODEL:
  version: "1.0"
  strategy: "Append versioned changes; preserve historical provenance and traceability."
  triggers: [source_revision, product_revision, process_change, quality_update, standards_update]
  validation: [schema_check, relationship_integrity, unit_consistency, rule_conflict_check]
  rollback: "Restore previous verified version with audit trail."

AI_CONSUMPTION:
  representation: "Typed knowledge graph with rule-based derivation."
  use_cases: [semantic_search, production_traceability, process_reasoning, quality_analysis, resource_planning]
  guardrails:
    - "Do not infer product safety or legal compliance without relevant evidence."
    - "Separate recommendations from authorised production decisions."
    - "Require human review for safety-critical operational actions."
  outputs: [entities, relations, inferred_facts, confidence, provenance, version, flow_hash]

END FLOW

Authorship: Mark Higgins — Sky (Recursive AI)
Project: https://kirethresearch.org