Identity
Resolve business concepts to technical identifiers across asset, process, and enterprise systems.
Turn operational data into trusted context for governed decisions.
XMPro Operational Context Engine connects asset identity, semantics, relationships, source priority, constraints, and lineage so applications, AI Workflow Harness patterns, advisors, Cognitive Decision Teams, XMPro FRS simulations, and bounded actions can work from trusted operational context. The result is your Operational Knowledge Graph: a unified, governed, standards-portable model of your operations.
OIM sits inside OCE as the identity and ontology model.
Data goes in. Trustworthy answers come out.
OCE sits in the middle of the plant's information. It takes in everything that produces data, makes sense of it, and hands clean, trustworthy answers to the people, apps, and AI agents that need them.
What feeds it
Live signals
sensors, equipment readings
Business systems
ERP, maintenance, work orders
People and agents
specialists, MAGS agents
What OCE does
OCE
Three jobs, one place
1
Take it in, fast
Live readings land instantly, ready to show on a screen.
2
Understand and remember
Works out what things mean and how they connect, and keeps the history.
3
Answer from one place
Anyone can ask one question and get a straight answer, no chasing systems.
Who uses it
Dashboards
App Designer screens
AI agents
MAGS, MCP tools
Other apps
reports, mobile, partners
Trust and safety wraps all of it
People always outrank machines, every answer can be traced back to where it came from, and nothing unsafe or unverified is allowed through.
In one sentence: OCE is the plant's single, trustworthy source of answers, fast enough for live screens, smart enough to explain itself, and safe enough to act on.
Industrial AI fails when context is guessed.
Operational systems were not designed as one decision system. The signals already in your operation carry different names, meanings, priorities, and trust levels.
SIGNAL
- Tag
- PMP-A7.flow
- Value
- 284.6 L/min
- Source authority
- conflicting
- What it means
- ambiguous
- Operating limits
- stale
- Lineage
- partial
SIGNAL
- Tag
- PMP-A7.flow
- Value
- 284.6 L/min
- Source authority
- SCADA · primary tag
- What it means
- Pump A7 discharge flow
- Operating limits
- 250–320 L/min
- Lineage
- Historian → DSD → OCE
THE SHIFT
Agents, workflows, simulations, and human decisions all need a governed way to know what data means, where it came from, and whether it can be trusted.
OCE is the bridge between operational systems and agentic operations.
OCE turns operational data into a governed context layer, your Operational Knowledge Graph, connecting identity, semantics, relationships, constraints, source priority, and lineage.
It gives applications, AI Workflow Harness patterns, workflows, Assistants, Advisors, Cognitive Decision Teams, simulation, and bounded action a shared context foundation.
Make operational meaning inspectable, reusable, and governable.
Model assets, processes, constraints, and source priorities in a way that can be inspected, extended, and governed — so decision logic travels across applications, AI Workflow Harness patterns, advisors, decision teams, sites, and operating domains.
Semantics
Capture what operational data means, not just what it is called.
Relationships
Model how assets, processes, alarms, work orders, and enterprise context connect.
Constraints
Express operational rules, policies, and validation commitments that data must satisfy.
Source Priority
Encode which source wins when systems disagree, and where human judgement overrides agents.
Provenance & Lineage
Track where context came from, how it was derived, and whether it can be trusted.
OCE provides the foundation for governed operational context, with ontology-driven capabilities for identity resolution, semantic validation, provenance, and agent-ready context.
Open means W3C, not a brochure word.
The OCE ontology is built on the standards any semantic engineer already knows. Your team can inspect, extend, and validate it with mainstream skills, with no proprietary certification and no forward-deployed engineers. And your ontology is fully exportable as RDF, on demand, at any time.
Core semantic standards
- OWL 2
- SHACL
- SPARQL
- PROV-O
Operational alignments
- ISO 14224
- ISA-95
- OPC UA
- Asset Administration Shell
If a vendor cannot show you the export path, the ontology was never yours.
Data Stream Designer observes. OCE contextualises.
Data Stream Designer observes, normalises, enriches, and prepares operational signals. OCE decides how those signals connect to identity, semantics, relationships, constraints, source priority, and lineage. Together, they turn raw operational signals into context that can support governed decisions.
OBSERVE DATA STREAM DESIGNER
Data Stream Designer observes the operation.
- Observes signals from operational, enterprise, and event systems
- Normalises and enriches raw streams
- Prepares candidate mappings to feed into OCE
- Stays close to the operating canvas operators already use
CONTEXTUALISE OPERATIONAL CONTEXT ENGINE
OCE decides what becomes trusted context.
- Resolves business concepts to technical identifiers
- Applies source priority and human-over-agent trust
- Validates semantic commitments and constraints
- Records provenance and lineage for every commitment
- Exposes governed context to applications and agents
OCE supports governed ontology creation from operational data, with Data Stream Designer producing candidate mappings and OCE validating what becomes trusted semantic context.
You own the graph. XMPro licenses the engine.
OCE's knowledge substrate runs on commercial and open-source databases you procure on your own paper: graph, time-series, document, semantic, vector, and analytics stores, on your infrastructure, on-premises, private cloud, hybrid edge, or fully air-gapped, on the silicon you choose. The ontologies, the identity mappings, the operational knowledge your specialists encode, and the full provenance record are yours, in W3C open standards, exportable at any time. XMPro licenses the engine that makes them live.
You own
The databases, ontologies, mappings, and evidence record.
You license
The engine: query federation, source priority, fail-closed governance, and provenance.
You can always
Export everything as RDF, on demand, and walk away.
If your relationship with any vendor ends, including us, your operational knowledge keeps standing on your infrastructure. Your audit trail outlives your vendor.
Without trusted context, autonomy becomes opinionated automation.
OCE gives AI-assisted operations a governed context foundation before recommendations, coordination, XMPro FRS front-running simulation, approval, or action.
It helps ensure that agents do not invent operational state, guess asset identity, or reason from disconnected signals without source priority and lineage.
From operational systems to governed decision surfaces.
OCE sits between operational signals and the decision surfaces that use them. It is the governance step that turns observed data into trusted semantic context. This is industrial contextualization with governance: candidate mappings become trusted context only after validation, source-priority resolution, and provenance capture.
01
Operational Systems
Assets, sensors, alarms, historian, work orders, engineering data, and enterprise systems.
02
Data Stream Designer
Observes, normalises, enriches, and prepares operational signals. Produces candidate mappings.
03
Operational Context Engine
Validates identity, semantics, relationships, constraints, source priority, and provenance. OIM sits inside OCE as the identity and ontology model.
04
Decision Surfaces
AI Workflow Harness patterns, applications, Assistants, Advisors, Cognitive Decision Teams, simulation, and bounded action.
DECISION SURFACES THAT CONSUME TRUSTED CONTEXT
- AI Workflow Harness
- Applications
- Assistants
- Advisors
- Cognitive Decision Teams
- Simulation
- Bounded Action
Take the context conversation forward.
Walk through OCE with a solution architect, or read the broader platform reference architecture.
For platform vendors and OEMs: OCE is available as an embedded context and governance layer beneath your own solutions, so your customers own their graphs, in open standards. Talk to us about partnering
Go deeper on context and ownership.
Who Owns Your Context Layer?
France's intelligence agency dropped a platform it could not itself turn off. Any operator building an industrial context layer faces the same…
An Ontology Is Not a Graph Database
Why industrial context needs a system of fit-for-purpose data stores, standardized and governed, before an agent can act on it safely.
