🎓 Lesson 9
D5
Data Ontology Mapping: ISO 15926, IFC Mining & Custom Schemas
Data ontology mapping is like creating a universal dictionary that helps different mining software systems understand and share data correctly.
🎯 Learning Objectives
- ✓ Explain the semantic mismatch between ISO 15926 Part 2 classes and IFC Mining entities using concrete examples
- ✓ Design a bidirectional mapping rule between an IFCMining:ExcavationActivity and ISO15926:WorkTask using OWL property chains
- ✓ Analyze interoperability gaps in a real mine planning-to-production data flow using ontology alignment metrics (e.g., coverage ratio, ambiguity score)
- ✓ Apply SPARQL-based validation queries to verify consistency of mapped instances against shared constraints
📖 Why This Matters
In modern mine digital twins, data flows from geological modeling (Leapfrog) → blast design (OASIS, SHOTPlus) → equipment telematics (Caterpillar MineStar, Komatsu Frontline) → ERP (SAP Mine Management). Without precise ontology mapping, 'bench height' may mean elevation above sea level in one system and vertical distance from muck line in another—causing cascading errors in production forecasting, safety compliance, and regulatory reporting. Mapping isn’t optional plumbing—it’s the semantic foundation of trustworthy automation.
📘 Core Principles
Ontology mapping operates at three levels: lexical (term similarity), structural (class/property hierarchy alignment), and logical (axiomatic equivalence or subsumption). ISO 15926-2 provides a reference ontology for process industries with top-level classes like 'Thing', 'Class', and 'Relationship', while IFC Mining (IFC4x3_ADD2) extends IFC4 with mining-specific entities like IfcMiningBench, IfcDrillPattern, and IfcBlastEvent. Custom schemas (e.g., internal ERP asset taxonomies) often introduce pragmatic but non-standardized concepts—requiring 'bridge ontologies' or context-aware alignment rules. Successful mapping requires distinguishing identity (same real-world entity), equivalence (interchangeable meaning), and specialization (e.g., 'drilling activity' ⊑ 'work task').
📐 Mapping Coverage Ratio (MCR)
MCR quantifies how completely a target ontology (e.g., IFC Mining) is aligned to a reference ontology (e.g., ISO 15926). A high MCR indicates robust semantic coverage for integration scenarios—but low values reveal critical gaps requiring custom extension or schema harmonization.
Mapping Coverage Ratio (MCR)
MCR = N_mapped_strong / N_total_targetMeasures the proportion of target ontology classes with verifiable semantic alignment to a reference ontology.
Variables:
| Symbol | Name | Unit | Description |
|---|---|---|---|
| N_mapped_strong | Number of strongly mapped classes | unitless | Classes linked via equivalence, subsumption, or functional property chain (not lexical similarity) |
| N_total_target | Total number of classes in target ontology | unitless | Total count of defined classes in the source schema (e.g., IFC Mining) |
Typical Ranges:
Production-ready digital twin: 0.85 – 1.0
Pilot integration: 0.4 – 0.7
💡 Worked Example
Problem: A mining company maps its IFC Mining schema (127 defined classes) to ISO 15926-2 (1,842 classes). Of the 127 IFC classes, 93 have at least one asserted semantic equivalence or subsumption link to ISO 15926; 14 are linked only via lexical similarity; 20 have no verifiable mapping.
1.
Step 1: Identify mapped classes with strong semantic links (equivalence/subsumption): 93
2.
Step 2: Exclude classes linked only by lexical similarity (14) and unmapped (20) — these do not satisfy interoperability requirements
3.
Step 3: Compute MCR = 93 / 127 = 0.732
Answer:
The result is 0.732 (73.2%), which falls below the industry-recommended minimum of 85% for production-critical digital twin integrations.
🏗️ Real-World Application
At BHP’s South Flank iron ore operation (Western Australia), ontology mapping enabled automated reconciliation of blast design data (SHOTPlus exports in IFC Mining format) with SAP S/4HANA asset maintenance records. By defining OWL equivalences between IfcMiningBlastEvent and ISO15926:WorkTask, and linking IfcDrillPattern to ISO15926:GeometricRepresentationItem, engineers reduced manual data re-entry by 92% and cut blast plan-to-execution cycle time from 4.7 to 1.3 days—validated via ISO/IEC 15926-11 conformance testing.
✏️ Alignment Validation Exercise
Given: In your company’s custom blasting schema, 'BlastHoleDiameter' is defined as a property of class BlastHole with unit 'mm'. In IFC Mining (IFC4x3), IfcDrillHole has IfcPositiveLengthMeasure for 'NominalDiameter', and in ISO 15926-2, 'diameter' is a property of 'ShapeRepresentation' with unit 'metre'. Task: (a) Write an OWL equivalence axiom linking the three properties; (b) Identify the required unit conversion factor; (c) Propose a SPARQL query to detect inconsistencies where BlastHoleDiameter > 400 mm but NominalDiameter < 0.4 m.