Digital Twin Standards Landscape

Industry standards, conventions, and frameworks from major players relevant to MetaForge’s Digital Twin architecture

Table of Contents

1. Purpose
2. Standards Already Adopted
   1. ISO/IEC 30173:2025 — Digital Twin Framework
   2. glTF/GLB 2.0 — 3D Interchange Format
   3. STEP AP242 — Managed Model-Based 3D Engineering
3. Standards Evaluated — High Relevance
   1. Asset Administration Shell (AAS) — Industry 4.0 Digital Twin Standard
   2. Universal Scene Description (USD) — Large-Scale 3D Collaboration
   3. SysML v2 — Model-Based Systems Engineering
4. Standards Evaluated — Lower Relevance
   1. DTDL — Digital Twins Definition Language (Microsoft)
   2. W3C Web of Things (WoT)
   3. OPC UA (IEC 62541)
   4. NGSI-LD (ETSI)
   5. IFC (Industry Foundation Classes)
5. What the Big Players Are Doing
   1. Siemens (Xcelerator / Teamcenter)
   2. NVIDIA (Omniverse)
   3. Microsoft (Azure Digital Twins)
   4. Dassault Systemes (3DEXPERIENCE)
   5. PTC (ThingWorx + Creo)
   6. AWS (IoT TwinMaker)
6. Gap Analysis: MetaForge vs. AAS
   1. Covered
   2. Gaps (Small)
   3. Gaps (Larger — Phase 3+)
7. Recommendations
8. Related Documents

Purpose

This research document surveys the digital twin standards ecosystem as of early 2026, evaluates relevance to MetaForge’s hardware design platform, and identifies specific alignment actions. It informs ADR-004: Digital Twin Standards Alignment.

Standards Already Adopted

ISO/IEC 30173:2025 — Digital Twin Framework

• Owner: ISO/IEC JTC 1/SC 41
• Status: Published 2025
• MetaForge alignment: The L1-L4 maturity model (Digital Thread, Operational Twin, Live Twin, Simulation Twin) in the Digital Twin Evolution doc is directly based on this standard.
• Action: None — already aligned.

glTF/GLB 2.0 — 3D Interchange Format

• Owner: Khronos Group
• Status: Stable, widely adopted
• MetaForge alignment: GLB is the viewer format per ADR-002. Server-side OCCT converts STEP to GLB.
• Action: None — already adopted.

STEP AP242 — Managed Model-Based 3D Engineering

• Owner: ISO 10303
• Status: Edition 3 (2024)
• MetaForge alignment: Primary CAD input format. Handled by OpenCascade in the conversion pipeline.
• Relevance: AP242 extends AP203/AP214 with PMI (Product Manufacturing Information), tessellated geometry, and composite materials. The OCCT conversion pipeline should eventually extract PMI data (Phase 3+).
• Action: Ensure MET-151 (OCCT container) preserves AP242 PMI when available.

Standards Evaluated — High Relevance

Asset Administration Shell (AAS) — Industry 4.0 Digital Twin Standard

• Owner: Plattform Industrie 4.0 (Siemens, Bosch, SAP, ZVEI), standardized as IEC 63278
• Status: V3.0 specification, IDTA maintains submodel templates
• Adoption: Mandatory in German automotive supply chains (VDA), spreading across European manufacturing. Catena-X (automotive) and Manufacturing-X (cross-industry) both built on AAS.

Architecture:

An AAS wraps a physical or digital asset with:

Key Submodel Templates (IDTA):

AAS REST API (IDTA-02001):

Standardized REST API for querying AAS registries and retrieving submodels. Defines endpoints like:

• GET /shells — list all administration shells
• GET /shells/{aasId}/submodels — list submodels for an asset
• GET /submodels/{submodelId}/submodel-elements — browse submodel content

Relevance to MetaForge:

• European manufacturing customers (automotive, industrial) will increasingly require AAS-compliant digital twins
• The Catena-X dataspace (automotive supply chain) mandates AAS for data exchange
• MetaForge’s graph schema already captures most of the data — the gap is export format, not data model
• Adding a globalAssetId (URN/IRI) to key nodes enables AAS export without schema changes

Assessment: High relevance. Align graph schema properties now, build export adapter in Phase 3.

Universal Scene Description (USD) — Large-Scale 3D Collaboration

• Owner: Pixar, Alliance for OpenUSD (AOUSD) — Apple, NVIDIA, Autodesk, Adobe, Pixar
• Status: Rapidly evolving, AOUSD formed 2023
• Adoption: NVIDIA Omniverse (BMW, Siemens, Foxconn, Mercedes-Benz for factory digital twins), Apple Vision Pro (spatial computing)

Architecture:

Key Capabilities:

• Composition arcs (references, payloads, inherits) enable massive scene assembly
• Physics schema (rigid body, collision, joints) maps to simulation twin
• Industrial extensions via NVIDIA Omniverse (IoT connectors, simulation)
• Binary format (USDC) and ASCII format (USDA) with packaging (USDZ)

Relevance to MetaForge:

• Primary relevance is for L4 (Simulation Twin) — factory-floor or production-line digital twins
• Not needed for product-level 3D viewing (GLB is sufficient and simpler)
• If MetaForge integrates with NVIDIA Omniverse for physics simulation, USD becomes the interchange format
• Apple Vision Pro adoption could make USDZ relevant for AR visualization of hardware prototypes

Assessment: Medium relevance. Track but don’t adopt. Relevant only if Omniverse or spatial computing integration is planned.

SysML v2 — Model-Based Systems Engineering

• Owner: Object Management Group (OMG)
• Status: Finalized 2024, tooling still maturing
• Adoption: Aerospace, defense, automotive (Lockheed Martin, Airbus, Boeing, GM). MBSE tools: Cameo (Dassault), Rhapsody (IBM), Capella (Eclipse)

Architecture:

SysML v2 is a complete redesign of SysML v1:

Key Innovation: SysML v2 has a standardized REST API (SysML v2 API and Services) for model exchange — tools can interoperate without file-based exchange.

Relevance to MetaForge:

• Enterprise customers using MBSE tools may want to import/export SysML v2 models
• MetaForge’s graph schema has significant overlap with SysML v2’s structural model
• The constraint engine maps to SysML v2’s constraint definitions
• Cross-domain ontology relationships (THERMALLY_COUPLED_TO, ELECTRICALLY_CONNECTED_TO, etc.) mirror SysML v2 connection/interface modeling

Assessment: Medium relevance. The graph schema is already structurally similar. A SysML v2 import/export adapter would be a Phase 3+ differentiator for enterprise customers.

Standards Evaluated — Lower Relevance

DTDL — Digital Twins Definition Language (Microsoft)

• Owner: Microsoft (Azure Digital Twins)
• Status: V3, open specification but Azure-centric
• Assessment: JSON-LD schema language for IoT twins. Strong for building/infrastructure twins, weak for hardware manufacturing. MetaForge’s graph model is more expressive. Adopting DTDL would constrain the schema unnecessarily.
• Action: Do not adopt. Study telemetry modeling patterns only.

W3C Web of Things (WoT)

• Owner: W3C
• Status: Thing Description 1.1 (Recommendation 2023)
• Assessment: Protocol-agnostic device capability descriptions. Useful for L3 (Live Twin) device telemetry — WoT Thing Descriptions could standardize how TelemetrySource nodes describe sensor capabilities.
• Action: Low priority. Consider for Phase 3 telemetry integration.

OPC UA (IEC 62541)

• Owner: OPC Foundation
• Status: Mature, widely deployed in industrial automation
• Assessment: Already referenced in MetaForge’s L3 Live Twin spec as a telemetry transport alongside MQTT. OPC UA Companion Specifications exist for many device types.
• Action: Already planned. No additional alignment needed.

NGSI-LD (ETSI)

• Owner: ETSI ISG CIM / FIWARE Foundation
• Status: V1.8 (2024)
• Assessment: Context information management for smart cities/IoT. Property graph model similar to MetaForge’s but oriented toward geospatial/smart-city use cases. Not relevant for hardware design.
• Action: Do not adopt.

IFC (Industry Foundation Classes)

• Owner: buildingSMART International
• Status: IFC4x3 (2024)
• Assessment: Architecture, Engineering, and Construction (AEC) standard. Not relevant for hardware product design.
• Action: Do not adopt.

What the Big Players Are Doing

Siemens (Xcelerator / Teamcenter)

• Full lifecycle digital twin platform
• AAS-compliant export from Teamcenter
• JT format for lightweight 3D visualization
• Parasolid geometry kernel (not open source)
• Catena-X integration for automotive supply chain
• NVIDIA Omniverse connector for factory simulation

NVIDIA (Omniverse)

• Physics-accurate simulation twins at factory scale
• USD as the universal scene description format
• OpenUSD connectors for CAD tools (SolidWorks, Creo, NX, Revit)
• Isaac Sim for robotics, Drive Sim for autonomous vehicles
• Real-time ray tracing for visual fidelity

Microsoft (Azure Digital Twins)

• Cloud-native IoT twin platform
• DTDL as schema language
• Integration with Azure IoT Hub, Time Series Insights
• Primarily building/infrastructure/energy use cases
• Less relevant for hardware product design

Dassault Systemes (3DEXPERIENCE)

• Product lifecycle management with virtual twin
• 3DXML and CATIA kernel proprietary formats
• ENOVIA for governance and configuration management
• Strong in aerospace and automotive

PTC (ThingWorx + Creo)

• Connected product twins bridging CAD design to IoT
• Creo for CAD, ThingWorx for IoT platform
• Vuforia for AR-based service and maintenance
• Strong in industrial equipment and medical devices

AWS (IoT TwinMaker)

• Entity-component model for composing twins from multiple data sources
• Grafana-based visualization
• Less opinionated about standards
• Primarily infrastructure monitoring use cases

Gap Analysis: MetaForge vs. AAS

The table below maps MetaForge’s current graph schema to AAS submodel requirements, identifying gaps.

Covered

Gaps (Small)

Gaps (Larger — Phase 3+)

Recommendations

See ADR-004: Digital Twin Standards Alignment for the formal decision record.

Related Documents