Updated on 4 September 2025
LLM4CVC - AI based planning for automated disassembly
About
Circular value creation requires that products and components can be efficiently dismantled and reused. However, disassembly processes are still largely manual, undocumented, and cost-intensive — especially in high-variance product environments.
LLM4CVC addresses this gap by developing an AI-based system that generates structured disassembly instructions from existing product documentation. The solution will combine state-of-the-art Large Language Models (LLMs) with a modular software architecture that connects to various industrial data sources (e.g. Digital Product Passes, BOMs, CAD).
The core of the project is a robust LLM-powered instruction engine that produces step-by-step disassembly sequences, including skill and tool requirements – forming a digital basis for repair, remanufacturing or automation.
Stage
- Early idea
- Already defined
Topic
- Data technologies | Data ecosystems for the realisation of circular value creation exploiting the full potential of digitalisation – e.g., harnessing existing, purpose-built platform solutions.
- Data technologies | Interoperability of CVC-relevant data ecosystems, quality assurance and traceability across systems
- Data technologies | (AI based) recognition systems (e.g. image recognition) to evaluate materials, components and products and determine the best use paths
- Data technologies | (AI based) process and system control technologies
- Data technologies | (AI based) Material and Product Design, Decomposition and Separation
- Data technologies | Approaches to support SME fully exploit the value of existing CVC-related data
- Data technologies | Design of an adaptable Digital Product Pass:
- Enabling technologies | Manufacturing and machine learning, e.g., to increase the flexibility of industrial processes, modular approaches, reduce use of materials, quality assurance and certification of products)
- Enabling technologies | Robotic / handling - and assistance systems
- Enabling technologies | Life cycle assessment / Product life cycle management – e.g., Digital Twin / Digital Product Passport
- Enabling technologies | Reverse Manufacturing (e.g. adaptive automation for high variance, sorting, sophisticated logistic systems)
Type
- Consortium seeks Partners
Attached files
Organisation
Similar opportunities
Project cooperation
Digital Twins for Remanufacturing
- Already defined
- Consortium seeks Partners
- Enabling technologies | Robotic / handling - and assistance systems
- Data technologies | (AI based) process and system control technologies
- Data technologies | (AI based) Material and Product Design, Decomposition and Separation
- Data technologies | Algorithm that shows the (positive) impact of a Circular Economy process or Circular Economy product
- Enabling technologies | Life cycle assessment / Product life cycle management – e.g., Digital Twin / Digital Product Passport
- Data technologies | Simulation models and predictive analytics to assess the scalability of circular processes across industries
- Enabling technologies | Reverse Manufacturing (e.g. adaptive automation for high variance, sorting, sophisticated logistic systems)
- Enabling technologies | AI-driven diagnostic systems, e.g., for assessing the viability of reused, remanufactured, and recycled components
- Data technologies | (AI based) recognition systems (e.g. image recognition) to evaluate materials, components and products and determine the best use paths
- Data technologies | Data ecosystems for the realisation of circular value creation exploiting the full potential of digitalisation – e.g., harnessing existing, purpose-built platform solutions.
Adrian Barwasser
Research Fellow at Fraunhofer IAO
Stuttgart, Germany
Project cooperation
- Already defined
- Consortium seeks Partners
- Data technologies | Design of an adaptable Digital Product Pass:
- Data technologies | (AI based) process and system control technologies
- Data technologies | (AI based) Material and Product Design, Decomposition and Separation
- Data technologies | (AI based) recognition systems (e.g. image recognition) to evaluate materials, components and products and determine the best use paths
Christian Schröder
Research Associate at Fraunhofer IWU
Chemnitz, Germany
Project cooperation
Digital ecosystem for circular value creation
- Already defined
- Consortium seeks Partners
- Data technologies | Assistance and Expert systems
- Data technologies | Design of an adaptable Digital Product Pass:
- Data technologies | (AI based) process and system control technologies
- Data technologies | (AI based) Material and Product Design, Decomposition and Separation
- Data technologies | Approaches to support SME fully exploit the value of existing CVC-related data
- Enabling technologies | (Advanced/smart) Sensors, e.g., enabling materials, components and product flows measurement
- Data technologies | Interoperability of CVC-relevant data ecosystems, quality assurance and traceability across systems
- Enabling technologies | Industry 4.0 technologies (IoT, big data analytics) for monitoring and managing circular value chains
- Enabling technologies | Life cycle assessment / Product life cycle management – e.g., Digital Twin / Digital Product Passport
- Data technologies | Simulation models and predictive analytics to assess the scalability of circular processes across industries
- Enabling technologies | AI-driven diagnostic systems, e.g., for assessing the viability of reused, remanufactured, and recycled components
- Data technologies | Data ecosystems for the realisation of circular value creation exploiting the full potential of digitalisation – e.g., harnessing existing, purpose-built platform solutions.
- Enabling technologies | Manufacturing and machine learning, e.g., to increase the flexibility of industrial processes, modular approaches, reduce use of materials, quality assurance and certification of products)
Yoav Nahshon
Team Leader Materials Informatics at Fraunhofer IWM
Freiburg, Germany