← Back to catalogue
Published

Digital Twin

vr.wm-vrt-001 · wm-vrt-001-digital-twin

Virtual representations kept in sync with a specific physical asset, process or system.

World Models Information and virtual systems INF.VRT.TWN

Bundle → Layer → Finding → Questions Filled

3 bundles · 3 layers · 5 findings · 10 questions

Twin and counterpart What is twinned.

Binding

The physical counterpart and the scope of the twin.

Counterpart

The real asset, process or system the twin represents.

  1. Which real asset, process or system does the twin represent, and by which identifier?
  2. Which parts or aspects of the counterpart are in scope?

Purpose

What the twin is used for.

  1. Is the twin used for monitoring, prediction, control or design?
  2. Who relies on its outputs?
Synchronization and fidelity How closely it mirrors reality.

Data link

Data sources, frequency and latency.

Synchronization

How and how often the twin is updated.

  1. Which data sources feed the twin, and how often?
  2. When was the twin last synchronized?

Fidelity and validation

How accurate the twin is and how that was checked.

  1. What fidelity does the twin claim, and how was it validated against the counterpart?
  2. Which models and their versions does the twin use?
Actuation and control Whether the twin can act on reality.

Write-back

Commands from the twin to the counterpart.

Control path

Whether and how the twin sends commands back.

  1. Can the twin send commands to the counterpart, and through which interface?
  2. Which safety interlocks and approvals apply to those commands?

Classifiers Filled

Family
World Models
Category
Information and virtual systems
Entry kind
relationship
Navigation path
NAV.INF.VRT.TWN
Domain
INF.VRT.TWN
Industry
Cross-industry
Tags
digitaltwininf.vrt.twn

What it is Filled

A digital twin is a virtual representation of a specific physical asset, process or system that is kept in sync with it through data at a stated frequency and fidelity, and used to monitor, analyse or predict its behaviour. It differs from a generic model or simulation by being bound to one real counterpart, and from the digital thread, which links lifecycle data rather than mirroring current state.

In scope

  • Endpoint identity and scope, intended use, representation revisions and semantic mappings
  • Observation bindings, timing policy, synchronization evidence and qualified state views
  • Fitness evidence, isolated scenario references, capability declarations, controlled relationship revision and federation links

Out of scope

  • Physical asset, device, process, dataset and software master lifecycles
  • Simulation engine internals, telemetry transport, estimation algorithms, automatic model training and physical control execution
  • Generic product templates or unbound 3D models as active twins; identity cloning of persons
  • Safety certification, legal permission and domain-specific operational procedures; dangerous applications remain policy-level only

Why it exists Filled

Virtual representations kept in sync with a specific physical asset, process or system.

Distinguishing features Filled

  • It is bound to one real counterpart and kept in sync with it, unlike a generic model or simulation.
  • Its value depends on stated synchronization frequency and validated fidelity.
  • It may write back to the counterpart, which raises safety duties absent from passive models.
  • Distinct from the digital thread, which connects lifecycle records rather than mirroring current state.

What robots and AI may and may not do Filled

Must not

  • Send commands to the physical counterpart without the configured approvals and interlocks.
  • Present twin predictions as measured reality.
  • Use a twin whose synchronization is stale for safety decisions.
  • Expose the twin of critical infrastructure to unauthorized parties.

Only with a human decision

  • Enabling write-back control from the twin.
  • Acting on twin predictions in safety-critical operations.
  • Accepting a twin as validated for a new purpose.

May

  • Query the twin for current and predicted state.
  • Run what-if analyses on a copy of the twin.
  • Flag divergence between twin and counterpart.
  • Report the twin's last synchronization time and fidelity.

Moral aspects Filled

  • Twins of buildings, cities or people can enable surveillance if they include personal data.
  • Over-trust in a twin can lead to unsafe decisions when it diverges from reality.
  • Twins of critical infrastructure are security-sensitive.

Who is affected

  • Operators of the physical counterpart
  • People in or near the twinned asset
  • Owners and maintainers

Owners Filled

Steward

The owner or operator of the physical counterpart, who answers for the twin's fidelity and use.

Roles

Relationship steward
Maintain scope, endpoint binding and revision accountability
Data custodian
Qualify source mappings, access and retention
Fitness reviewer
Assess evidence for the stated baseline, purpose and envelope
Authorized consumer
Use qualified views within purpose and freshness limits
External control authority
Own any operational action authorization and safety boundary outside this model

Master systems

  • Asset management system
  • Industrial data platform
  • Building or city information model repository

Links to other meta-models Filled

references

  • WM-OBJ-001 - Candidate conditional item endpoint binding; use the appropriate physical system or process master instead where relevant. Version pin deferred.
  • WM-OBJ-008 - Candidate sensor or computing endpoint reference; calibration and device operations remain external. Version pin deferred.
  • WM-VRT-003 - Candidate optional simulation model and run binding, without solver or experiment ownership. Version pin deferred.
  • WM-VRT-013 - Candidate optional lineage service reference; lifecycle traceability is not a local twin execution function. Version pin deferred.
  • wm-obj-001-physical-item-instance - A twin is bound to a specific physical instance.

aligned

  • OGC SensorThings API 1.1 - Conceptual observation mapping only; no service implementation or conformance claim.
  • SOSA/SSN 2017 - Conceptual feature, sensor, procedure and observation mapping; simulation output needs explicit classification.
  • W3C PROV-O 2013 - Conceptual derivation and revision mapping for evidence, not proof of truth.
  • W3C WoT Thing Description 1.1 - Conceptual interface and security declaration binding, not authority to invoke.
  • ISO/IEC 30173:2023 - Terminology alignment lead only; catalogue scope reviewed and full clauses unverified.

neighbor

  • Registry record plane - standalone-mm is not a semantic kind. The registry flag calls for a synchronization relationship, adopted here as relationship. This is a repository modeling decision, not a claim that every external standard uses this classification.
  • WM-OBJ-001 Physical Item Instance - Conditional endpoint reference for item-level twins. A bounded physical process or system can instead use its own authoritative master; no universal item-only or one-twin-per-item constraint.
  • WM-VRT-003 Simulation - Optional model and run references. Solver operation and scenario execution remain external. Simulated output must not replace physical observations or the authoritative connected state.
  • WM-VRT-013 Digital Thread - Optional lifecycle lineage binding. Cross-system lifecycle traceability remains external; this model records only the relationship-specific evidence and synchronization assertions.
  • WM-OBJ-008 Device / Sensor / Compute HW - Sensor and host references do not transfer calibration, device configuration, firmware or actuation ownership into the twin relationship.
  • Definition profiles - The selected UK definition is stricter about prediction and two-way communication than this repository minimum. Record the selected profile and unmet conditions. Monitoring-only records cannot claim UK-profile conformance.
  • wm-vrt-013-digital-thread - The digital thread links lifecycle data that a twin may draw on.

related

  • wm-vrt-003-simulation - Twins run simulations to predict behaviour.

What else AI and robots need to interact with it Filled

Identity and identifiers required Filled

  • A twin is identified by its own identifier together with the identifier of the counterpart it represents.
  • In industry the Asset Administration Shell gives globally unique identifiers to the shell and the asset.

Direct properties not applicable Not applicable

Not applicable

A digital twin is an information or software construct; its measurable attributes are configuration and operational data, not physical properties.

Recognition optional Filled

  • A twin names its counterpart, its data sources and its synchronization frequency.
  • Often confused with a 3D model, a dashboard or a one-off simulation.

Capabilities and actions required Filled

  • A twin can be synchronized, queried, used for prediction and, where enabled, send commands.
  • It can be versioned as its models and data sources change.

Hazards and failure modes required Filled

  • Unsafe physical actions triggered from a wrong or stale twin.
  • Cyber attacks reaching physical assets through the twin's control path.
  • Decisions based on unvalidated predictions.

Standards and interfaces required Filled

  • Asset Administration Shell (IEC 63278) for industrial assets.
  • ISO 23247 digital twin framework for manufacturing.
  • OPC UA and OGC SensorThings for data links.

Context of use required Filled

  • Used in manufacturing, energy, buildings, infrastructure, cities and health care.
  • Safety and cybersecurity rules of the counterpart's sector apply to any control path.

Sources Filled

  1. Use Case Scenarios for Digital Twin Implementation Based on ISO 23247 - National Institute of Standards and Technology
  2. Digital Twin (official) - UK Government
  3. OGC SensorThings API Part 1: Sensing Version 1.1 - Open Geospatial Consortium
  4. Semantic Sensor Network Ontology - World Wide Web Consortium
  5. PROV-O: The PROV Ontology - World Wide Web Consortium
  6. Web of Things (WoT) Thing Description 1.1 - World Wide Web Consortium
  7. Security and Trust Considerations for Digital Twin Technology - National Institute of Standards and Technology
  8. Credibility consideration for digital twins in manufacturing - National Institute of Standards and Technology
  9. ISO/IEC 30173:2023 - Digital twin - Concepts and terminology - International Organization for Standardization
  10. ISO 23247 Digital twin framework for manufacturing (ISO)
  11. IEC 63278 Asset Administration Shell for industrial applications (IEC)
  12. ISO/IEC 30173 Digital twin - Concepts and terminology (ISO/IEC)

Open questions

  • Run the supplied source checker outside the sandbox, verify selected editions and full-text claim support, and assess source licensing before closing the source gate.
  • Develop adopted profiles and nested schemas with fixtures for endpoint replacement, server-assigned timestamps, stale and partial data, simulated sensors, scenario isolation, federation restrictions and retirement holds.
  • Pin neighbor model versions and test semantic mappings and purpose-specific fitness evidence before operational use.
  • Restore independent external review before any canonical or publishable-draft promotion.
  • Independent external provider review is absent; the separate local self-audit cannot replace it.
  • Direct source HTTP checks are unattempted under the owner-reported sandbox block. Full ISO clauses and the credibility paper body were not reviewed.
  • Nested instance schemas, algorithm implementations, conformance fixtures, neighbor version pins and sector adoption profiles remain incomplete.
  • Nonphysical-only referents, personal twins and domain-specific control procedures are outside this physical-context boundary.

Machine files

Provenance

world-models research · reviewable-draft

Built from: models/wm-vrt-001-digital-twin/spec.yaml, ver-cy/world-models/card-supplements/wm-vrt-001-digital-twin.json