SPONSORED BY: Integration Objects
Accelerate industrial projects and build domain-intelligent agents with SIOTH®
Industrial data becomes more valuable when applications can understand what it represents, how equipment relates to a process, and which operating conditions matter. OPC UA Companion Specifications provide a common framework for that understanding. Implementing them, however, can demand substantial engineering effort.
The object-oriented SIOTH® Data Model simplifies this work by collecting data from diverse OT sources, organizing it into meaningful industrial objects, and mapping it to applicable OPC UA Companion Specifications. Reusable models help reduce project time, effort, and cost. Combined with the SIOTH® AI reasoning engine, the same foundation can be extended into domain-intelligent agents that apply industrial knowledge to operational data.
1. Why create an OPC UA Companion model?
Industrial systems often describe the same equipment differently. A pump may appear as hundreds of tags in a control system, a separate asset in a maintenance application, and another structure in a historian. Applications must reconstruct the relationships between those records before they can use the data effectively.
An OPC UA information model based on a Companion Specification gives these systems a common industrial language. It defines recognizable equipment and process types, their properties, measurements, states, and relationships. A pressure reading can be identified as the discharge pressure of a particular pump, with its engineering unit and equipment context, instead of being presented as an isolated value.
This shared meaning delivers value across the industrial lifecycle. Integration teams spend less time interpreting proprietary tag conventions. Software developers can build applications around consistent structures. Operations teams gain a clearer view of equipment and processes. Analytics and AI teams receive contextualized inputs, reducing the effort required to prepare data for each use case.
The benefits grow through reuse. When multiple assets and sites implement the same model, dashboards, calculations, and analytics can be adapted more efficiently. Organizations can connect additional applications without repeatedly reconstructing the meaning of their source data. Standardized models also reduce dependence on individual vendors’ naming conventions and make industrial knowledge easier to share.
Creating an OPC UA Companion model therefore establishes a reusable foundation for interoperability, enterprise integration, and industrial AI. Its business value comes from making data easier to understand and less expensive to use across successive projects.
2. Why implementation takes time and effort
The standardized blueprint does not automatically organize a plant’s existing data. Industrial information remains fragmented across PLCs, DCSs, SCADA systems, historians, condition-monitoring systems, and other OT sources, each with its own protocols, formats, and conventions.
Engineers must identify the relevant signals, understand their meaning, reconcile names and units, define equipment relationships, create object instances, and bind their variables to live sources. They must then validate data quality, model structure, and application behavior. Missing context or inconsistent source data can create rework and delay commissioning.
Without reusable modeling tools, much of this work is repeated for every asset, site, and project. The result is a lengthy, complex, and labor-intensive implementation process that increases engineering cost and postpones the benefits of analytics and AI. Developing a new industry Companion Specification adds a separate effort in domain agreement and standardization; SIOTH® addresses the practical modeling and implementation work.
3. SIOTH® Data Model: faster deployment at lower cost
SIOTH® Data Model reduces this implementation burden through an agile, object-oriented approach. It brings multi-source data collection, industrial object modeling, and mapping to applicable OPC UA Companion Specifications into a reusable workflow.
- Collect and unify data
Using the connectors configured for the deployment, SIOTH® gathers data from diverse OT sources. Measurements belonging to one asset can originate in different systems and still be represented together. A pump object, for example, can combine control-system measurements with vibration indicators and historical operating information.
- Structure data into meaningful objects
Engineers organize information into reusable equipment and process templates. Each template defines the relevant properties, measurements, states, and relationships. The object-oriented approach allows a common structure to be defined once and instantiated across multiple assets, with source bindings configured for each instance.
- Map and expose through OPC UA
The modeled objects are mapped to applicable Companion Specifications and exposed through the SIOTH® OPC UA server. Consuming applications access a consistent information model, reducing the need to maintain separate interpretations of each underlying source. Project-specific extensions can add required information while preserving standardized meanings.
- Reuse the work across projects
Templates, model structures, and established mapping patterns become reusable engineering assets. New projects can start from existing work, accelerating configuration and deployment. This reduces repetitive engineering, simplifies application onboarding, and lowers implementation effort and cost. The savings depend on source complexity and the extent of reuse, but the value mechanism is direct: work completed for one asset or project contributes to the next.
4. Extend the model into a domain-intelligent agent
Once industrial data is structured and contextualized, the next step is to apply knowledge to it. SIOTH® Data Model provides the equipment and process context; the SIOTH® AI reasoning engine evaluates that context using domain rules, calculations, operating states, and event logic. Together, they enable the creation of an AI-ready, domain-intelligent agent.
This extension builds on the same objects and data bindings already created for the OPC UA model. Engineers enrich those objects with operational knowledge rather than rebuilding the data foundation for a separate intelligence application. The model supplies the meaning, while the reasoning engine determines which conditions warrant attention and which configured conclusions or recommendations should be produced.
A practical example: an intelligent pump agent
A pump model can bring together suction and discharge pressure, flow, motor current, vibration, bearing temperature, and running state. These measurements become more useful when assessed together and in the context of how the pump is operating.
For example, the SIOTH® AI reasoning engine can evaluate elevated vibration alongside rising bearing temperature while the pump is running. An approved domain rule can identify a potential bearing anomaly and publish a finding linked to the pump, the contributing measurements, and a recommended inspection. Additional rules can detect excessive cycling, inconsistent equipment states, or deviations from an expected operating envelope.
Applications and operators then receive an interpreted finding with supporting context. The same rule structure can be reused across suitable pumps, with asset-specific limits and operating conditions. Domain experts validate the logic so that findings remain appropriate to the equipment and process.
The operational and strategic value
Accelerating the AI strategy. Contextualized objects provide a reusable data foundation for analytics and AI applications. Teams can spend more time developing useful capabilities and less time reconstructing asset context for each initiative.
Improving uptime and maintenance decisions. Domain reasoning can identify abnormal conditions earlier and help teams prioritize investigation. Failure prediction can build on this foundation when suitable historical data and validated predictive rules or models are available.
Supporting process optimization. Calculations and domain rules can assess performance against expected operating conditions, helping identify efficiency losses and improvement opportunities. Their value depends on the process knowledge encoded and the actions taken in response.
Capturing and retaining knowledge. Engineering expertise becomes explicit, reusable logic associated with industrial objects. Organizations can preserve approved practices, share them across teams, and apply them consistently as equipment, personnel, and projects evolve.
SIOTH® connects project delivery with operational intelligence: create contextualized OPC UA Companion models faster, reuse them to reduce engineering time and cost, and extend them with domain knowledge and AI reasoning to support better industrial decisions.
About Integration Objects
Integration Objects is a global provider of Industry 4.0 and industrial digital transformation solutions. The company combines deep domain expertise with advanced technologies such as OPC, artificial intelligence, cybersecurity, Industrial IoT (IIoT), big data analytics, C4ISR, and process automation to deliver scalable and reliable systems.
Through its solutions, Integration Objects enables secure and seamless integration between IT and OT environments, supporting organizations across industrial, defense, and government sectors in achieving operational excellence and accelerating their digital transformation initiatives.