OT to MES Integration with Industrial Edge and Opcenter

Download detailed architecture PDF

Download architecture details

Field level: Industrial Edge as the OT foundation

Industrial Edge Devices sit directly on the shop floor and connect to PLCs, drives, robots, and any other automation equipment using pre-configured connectors for PROFINET, S7, OPC UA, EtherNet/IP, Modbus TCP, and others. The connector library removes most of the integration work that traditionally slows down MES rollouts, and because it covers equipment from any vendor, it fits brownfield environments without requiring hardware replacement.

A set of local apps runs alongside the connectors on the same device:

• WinCC Unified for operator visualization and HMI
• Virtual PLC and LiveTwin for control and simulation
• AI Inference Server for on-device machine learning (defect detection, anomaly detection, predictive maintenance)
• Energy Manager and Performance Insight for operational KPIs such as energy consumption and OEE
• Industrial Information Hub, which provides a unified semantic data model so that raw PLC tags are mapped to meaningful, consistently named data objects before leaving the edge
• OPC Router for dedicated MES-bound routing when needed
• Mendix on Edge for role-based operator UIs that span both the edge and the MES

The Databus, based on MQTT, connects these apps to each other on the device and provides the publish-subscribe backbone for event-driven communication up to the MES integration layer.

Factory level: the edge-to-MES integration layer

This is the layer where most integration projects succeed or fail. Two Opcenter components handle the boundary between Industrial Edge and the MES, and they are designed to work together rather than as alternatives. Each one handles a fundamentally different class of data.

Opcenter connect MOM: event and transaction routing

Connect MOM is a general-purpose integration framework tuned for MOM use cases. Its job is to move discrete messages between systems, transform them along the way, and guarantee delivery even when endpoints are temporarily unavailable. Use Connect MOM for:

• Shopfloor events and transactions: station completions, scrap events, batch releases, changeovers, genealogy records
• Edge app results: AI inference outputs (defect classifications, anomaly events, predicted failures), energy consumption readings from Energy Manager, OEE data from Performance Insight
• Bidirectional ERP integration: receiving work orders and BOMs from SAP, sending confirmations and goods movements back
• Any data that needs to reach multiple destinations: Opcenter Execution, Opcenter Quality, Opcenter Intelligence, a data lake, or an external system
• Store-and-forward buffering: Connect MOM queues messages and replays them if Opcenter or the network is temporarily unavailable

Connect MOM supports REST, MQTT, OPC UA, and file-based adapters, and is deployed with high-availability support. It is the integration hub for everything that is event-shaped or transaction-shaped rather than tag-shaped.

Opcenter execution foundation Automation Gateway: in-process parameter binding

The Automation Gateway lives inside Opcenter Execution rather than alongside it. Its job is narrow but high-value: it binds specific OPC UA nodes (PLC tags) to data collection fields on active work order operations, so that in-process measurements are captured automatically at runtime instead of being typed in by operators.

Use the Automation Gateway (available for Opcenter Execution Foundation) for:

• Capturing in-process measurements directly from the automation layer: dimensions, torques, temperatures, cure times, pressures, serial numbers read from vision systems
• Eliminating manual operator data entry into MES data collection forms
• Linking machine parameters to specific work order operations, so values are stamped against the right production context automatically

The Gateway is not a general integration tool. It supports OPC UA for live tag binding, direct REST API calls, and the Remote Interop Connector (RIC) for additional connectivity patterns. It operates on live values during active operations and writes into Opcenter Execution's data collection model.

Why both are deployed together

A realistic deployment uses both in parallel because they handle different problems. Consider a pump assembly line:

• A work order arrives from SAP via Connect MOM (REST or IDoc adapter)
• Opcenter dispatches the order to a torque station. As the operator runs the step, the Automation Gateway reads the torque value and the serial number from the PLC via OPC UA and writes them into the data collection form on that operation automatically
• When the unit completes the station, a completion event with genealogy and measurement data flows back through Connect MOM to SAP and to Opcenter Intelligence
• Meanwhile, the AI Inference Server flags a vibration anomaly. That result is published to the Databus, picked up by Connect MOM via MQTT, and delivered to Opcenter Quality as a non-conformance transaction

Without the Gateway, operators are typing measurements. Without Connect MOM, there is no clean path for events, AI results, ERP flows, or multi-destination routing.

IT and enterprise level: the Opcenter MOM portfolio

Once data reaches Opcenter through either integration path, the modular MOM portfolio orchestrates production across the enterprise:

• Opcenter Execution (Discrete, Process, Pharma, Electronics, Semiconductor): Manufacturing execution

Work orders, BOMs, recipes, routing, confirmations, and traceability across the production lifecycle

• Opcenter Quality: Quality management

Inspection plans, SPC, non-conformance handling, and CAPA workflows. AI inference results from the edge land here as non-conformances or quality holds

• Opcenter APS: Production planning and scheduling

Finite-capacity scheduling across resources, materials, and time. Predicted failures from edge AI can trigger schedule re-optimization here

• Opcenter Intelligence: Enterprise manufacturing intelligence

Production dashboards, KPI reporting, and root-cause analysis fed by contextualized data from the edge and the MES

• Mendix: Low-code UI layer

Role-based operator interfaces that can span both the MES and the edge without requiring custom development for each touchpoint

Lifecycle management across the stack

Industrial Edge Management (Virtual, Pro, or Cloud) and the Industrial Edge Hub provide a single place to deploy apps, push updates, monitor device health, and manage AI models across all sites. This includes deploying updated inference models to the AI Inference Server on devices at remote locations, without requiring on-site intervention. On the MES side, Opcenter administration handles configuration, user roles, adapter lifecycle, and Opcenter Connect MOM adapter management. Combined, this gives OT and IT teams a shared operating model without forcing either side to give up control of their own domain.