• 1. WHAT is IT/OT Integration
  
  

  IT / OT integration is the convergence of Information Technology (IT) with Operational Technology (OT). IT technologies include cloud and web platforms and IT systems, e.g., for ERP and PLM. OT technologies comprise machines and devices, e.g., IPCs, controllers, sensors, and actuators, as well as control systems like MOM, DCS, SCADA and HMI. While IT technologies enable openness, flexibility, and scalability, OT technologies address qualities like real-time, determinism, and reliability. IT / OT integration aims at utilizing and retaining the strengths of both worlds while allowing their seamless integration in application workflows that crosscut IT and OT domain boundaries. Hence, IT / OT integration enables the digital transformation of enterprises, supports an increase of their operational efficiency and transparency, and contributes to optimizing their production and business processes.

• 2. WHY is IT/OT Integration important for Siemens?
  
  

  IT / OT integration is central to Siemens' digitalization strategy because the company has domain expertise and products in both fields. Combining these capabilities lets Siemens deliver unified systems instead of asking customers to connect separate automation and IT products. This can distinguish its portfolio from those of traditional automation vendors and IT-only suppliers, particularly for customers that want one accountable provider across the system lifecycle.

• 3. AI/IT/OT Integration
  
  

  Digital integration is not only for IT and OT software and hardware, nor only for human engineers to understand; it is also for AI agents, enabling AI to comprehend the model, structure and behavior of industrial systems.

This research view studies how industrial engineering artifacts can move from static diagrams and disconnected configuration files into executable, versioned, and agent-readable runtime patterns. The central question is how a workflow can be described once and then be deployed across web apps, edge runtimes, industrial controllers, simulation environments, and AI tools without losing semantics.

• Patent mapping: Workflow Canvas patent families, Industrial Logic Emitter, Robotic WASM Aggregated Model, and AGV orchestration show how engineering models can become executable runtimes.
• Publication mapping: technical manuals and papers document the model structure, engineering rules, and reusable patterns behind these runtimes.
• Research output: a pattern language for describing workflow generation, runtime orchestration, tool integration, and deployment across IT/OT boundaries.

Behavior Modeling is the foundation for making industrial systems understandable to both humans and AI agents. Instead of treating assets only as static objects, this research describes what assets can do, which states they can enter, which events they react to, and how their behavior can be composed into workflows, procedures, and autonomous operations.

• Patent mapping: Industrial Logic Emitter fuses FSM, behavior tree, and rule-engine logic; CREEM-related invention families define behavior-driven workflow ecosystems.
• Publication mapping: CREEM and Dynamic Behavior Modeling materials provide the common information model, semantic definitions, and engineering rules.
• Research output: a reusable behavior layer for machines, robots, controllers, human work steps, software services, and AI agent tools.

AI-native industrial software is not simply traditional engineering software with a chatbot attached. It requires structured semantic inputs, verifiable generated outputs, tool contracts, model lineage, and runtime environments where AI agents can reason over assets, states, workflows, and constraints before taking action.

• Patent mapping: Workflow Canvas, Robotic WASM, and AGV orchestration demonstrate agent-ready abstractions for workflow generation, robotic behavior execution, and heterogeneous fleet coordination.
• Publication mapping: white papers and technical manuals explain how semantic models, behavior models, and runtime interfaces make AI-assisted engineering auditable and reusable.
• Research output: AI-native patterns for intent-to-workflow generation, model-to-runtime transformation, safe tool invocation, simulation-based validation, and human-in-the-loop governance.

Data Mesh Architecture provides the organizational and technical pattern for cross-domain industrial semantics. Each domain owns its data products, while shared semantic contracts, mapping products, governance services, and resolution engines allow the products to be discovered, composed, and consumed by applications and AI agents.

• Patent mapping: Workflow Canvas and SiteScope MultiModal-related concepts connect data products, operational graphs, workflows, and agent-facing APIs.
• Publication mapping: cross-domain semantic mapping papers and architecture notes describe how Data Mesh and Data Fabric ideas can be applied to industrial ontology interoperability.
• Research output: a federated semantic access layer with lineage, confidence, mapping versions, quality policies, and domain-owned semantic data products.

Cross-domain semantic ontology research focuses on translating between the vocabularies of automation, manufacturing, process industry, robotics, buildings, energy, carbon accounting, and IT systems. The work creates governed mappings and reusable semantic anchors while allowing each domain to retain its own ontology.

• Patent mapping: behavior-model and workflow patents provide executable semantics that complement structural ontology mappings.
• Publication mapping: ontology mapping papers connect DEXPI, AAS, OPC UA, ISA-95/ISA-88, MTP, Brick, SAREF, ECLASS, DPP, and other standards through explicit mapping products.
• Research output: a semantic interoperability method that links assets, documents, events, behavior models, and runtime services across domains.

Vertical research validates the general AI/IT/OT integration method in concrete domains. Each vertical provides different constraints: real-time behavior in automation, safety and procedure logic in process industries, navigation and fleet coordination in logistics, physical behavior in robotics, and lifecycle data in carbon or energy systems.

• Patent mapping: Robotic WASM represents robotics simulation; AGV map fusion represents logistics orchestration; Industrial Logic Emitter represents control logic; Workflow Canvas links them through engineering and runtime patterns.
• Publication mapping: vertical papers, manuals, and presentations organize these cases into reusable methods rather than isolated demos.
• Research output: domain-specific validation scenarios for smart factories, logistics systems, process automation, industrial metaverse, energy systems, and AI-enabled engineering tools.