

The OPC UA MCP Server
Oct 1, 2025
In this engaging discussion, Armin Hadzalic, a Senior Software Developer at Siemens, shares his insights on the transformative Model Context Protocol (MCP) for industrial AI. He explains how MCP connects large language models with real-world data and tackles the challenges of integrating multiple systems. Armin delves into the future of seamless production systems powered by agents and AI, while also addressing criticisms of MCP. With practical advice for engineers, he advocates for an early embrace of these technologies to stay ahead in an evolving industry.
AI Snips
Chapters
Transcript
Episode notes
MCP Connects LLMs To The Real World
- MCP (Model Context Protocol) gives LLMs standardized access to external data and tools so they can act on current information.
- It provides primitives: tools, resources, and prompt templates to make AI connectable and consistent.
Solves The N-by-M Integration Problem
- MCP solves the N-by-M integration problem by decoupling LLMs from each tool implementation.
- This lets you swap LLMs or tools without re-implementing every connector.
Anomaly Workflow Example
- Anomaly handling used to force engineers to manually query OPC UA, check spare parts, and create tickets.
- With MCP and agents the LLM plans, invokes connectors, and shows a proposed plan to the engineer for approval.