Generation of the MCP from the Asset Administration Shell

  • Subject:Industry 4.0, Digital Twins, LLMS, AI
  • Type:Bachelor’s Thesis (Master’s Thesis possible with extended scope)
  • Date:25.09.2025
  • Supervisor:

    Moritz Dorn

  • Links:Tender

  • Are you interested in the future of industrial digitalization? In this thesis, you will dive deep into the world of the Asset Administration Shell (AAS) – the digital twin for industrial components. The goal is to enrich the AAS with semantic annotations to make complex relationships in smart factories more understandable. You will develop a so-called Model Context Protocol (MCP), which serves as a semantic bridge between LLMs and components within the plant. Sounds exciting? Then this thesis is for you!

MOTIVATION:

Industry 4.0 is no longer a vision of the future – it is reality. However, as industrial systems become more complex, managing them efficiently becomes increasingly challenging. The Asset Administration Shell offers a standardized solution, but semantic interpretation remains difficult. This is where your thesis comes in: you will combine modern AI technologies with industrial standards and develop an innovative concept that helps engineers better understand and manage complex systems. You will work at a cutting-edge intersection between research and practice – and actively shape the future.

OBJECTIVES:

  • Familiarize yourself with the creation and testing of MCP servers
  • Understand the technology and structure of the Asset Administration Shell
  • Develop an MCP server for managing and semantically linking Asset Administration Shells
  • Evaluate the concept in terms of its suitability for working with complex digital twins

RECOMMENDED PRIOR KNOWLEDGE:

  • No fear of trying out some programming
  • Interest in the lectures of Prof. Barth
  • Enjoyment in exploring new and largely unexplored technologies