Fraunhofer Institute for Mechanics of Materials IWM
Fraunhofer Institute for Mechanics of Materials IWM

AI-supported knowledge management and decision support

Your most valuable knowledge is often found in test reports, data sheets, and the minds of individual experts—and is lost as soon as they leave the company. We preserve this knowledge and make it usable: We build AI assistants that make proprietary data, technical literature, and process information searchable and provide answers backed by source citations—rather than plausible-sounding but unverifiable chatbot responses. To achieve this, we model your data sources, build knowledge graphs, and automatically align customer requirements with manufacturing data and standards. The result: faster, better-informed decisions in development, quality, and manufacturing; less duplication of effort; reduced risk of overlooking relevant internal data—and a company that preserves its experiential knowledge across generations instead of quietly losing it.

AI-powered access to integrated material and process data (chatbot)

We develop a chatbot interface that allows your teams to access material and process data across silos using natural language. Responses are based on stored company data and are backed by source references. This accelerates R&D decisions, makes better use of existing knowledge, avoids duplication of effort, and reduces the risk of overlooking relevant internal data.

R&D services:

  • Integration of material and process data sources
  • Development of a natural-language chatbot interface
  • Storing source references to ensure the verifiability of every response
  • Definition of access rights and approval levels by user role
  • Integration into existing R&D and documentation systems

Research project: 

AI-powered matching of customer and process data

We use language models to automatically match customer requirements with internal manufacturing data and general knowledge (specifications, standards, open knowledge databases). This minimizes errors, avoids unplanned costs, and increases the efficiency of your quoting and manufacturing processes.

R&D sercives:

  • Automatic extraction of customer requirements from specifications
  • Matching requirements with internal manufacturing data
  • Verification against relevant standards and open knowledge databases
  • Flagging of discrepancies and missing documentation
  • Early warning of potential conflicts before the start of a project

Knowledge base for materials development: literature and company-specific documents

We build a material- and process-specific knowledge base, drawing both from scientific literature with citation networks and from your company’s own test reports, data sheets, drawings, and standards, and make this knowledge accessible via natural language using AI. This enables you to make faster, more informed decisions by accessing dispersed technical and experiential knowledge, receive fact-based answers instead of guesswork, and preserve corporate know-how across generations.

R&D services:

  • Indexing of scientific literature, including citation networks
  • Integration of in-house test reports, data sheets, and drawings
  • Development of a material- and process-specific knowledge base
  • Semantic search with source links for every answer
  • Ongoing expansion of the knowledge base with new internal documents

Knowledge graph-based cause-and-effect analysis

AI-supported evaluation and neurosymbolic AI are current areas of research aimed at identifying previously unknown factors influencing KPIs such as quality or scrap, as well as providing decision support for material selection or process routes. They thereby reduce costs by utilizing existing data instead of conducting new tests and enable lower-risk, data-driven decisions.

R&D services:

  • Modeling cause-and-effect relationships in knowledge graphs
  • Identification of unknown factors influencing quality and scrap KPIs
  • Decision support for material selection and process route selection
  • Use of Neurosymbolic AI to Link Data and Domain Rules
  • Derivation of prioritized hypotheses for targeted follow-up experiments

Research project: 

  • Digital Modeling and Simulation for Design, Processing and Manufacturing of Advanced Materials
    EU Project DiMat

Research topic: Cross-organizational, IP-secure learning

We are conducting research on the use of federated learning for IP-secure learning across organizational boundaries, with Fraunhofer as a trusted partner. In the long term, this will allow you to protect your intellectual property while simultaneously optimizing cross-organizational value chains.

R&D services:

  • Design of federated learning architectures for multiple corporate partners
  • Sensitive raw data remains within each company
  • Exchange of only suitable model information instead of raw data
  • Fraunhofer as a trusted, neutral mediator in the learning process
  • Assessment of joint optimization potential along the value chain

Research project: