Human Decision research project
The product creation process cannot be viewed in isolation. The current socio-economic situation, which is based on the assumption that natural resources are infinitely available, makes the integration of circular economy (CE) concepts an indispensable factor right from the planning phase of a product. From a policy perspective, the EU-wide implementation of a digital product passport forces companies to consider CE in product creation, generate data and prohibit the destruction of usable goods. Therefore, disassembly and the ability of the assembly system to perform it must be considered in the early stages of the product creation process in the future, as required by VDI2021. The consideration of product sustainability and circular economy confronts engineers and product designers with a multitude of options that must be made based on a complex interplay of existing products, processes, reusability and economic framework conditions. The aim of this project is to explore solutions to the above challenges in order to support engineers in decision-making processes using human-centred hybrid neural-symbolic decision support systems.
Aims of the research project
- Building a decision-making process that integrates expert knowledge, benefits users and expands the knowledge base through user interaction.
- Exploring insights and solutions for symbolic modelling, information sharing and efficient closed-loop product creation.
- Investigation of neural-based multi-objective optimisation with symbolic models and reinforcement learning for contextual solutions in circular product creation.
- Development and integration of a decision support system to optimise the use of resources and economic KPIs in modular, assembly line-free assembly systems using virtual CAD prototypes.
- Verification and validation of pilots of the decision support system using a human-centred approach and expert knowledge as well as derivation of development requirements
→ Development of a human-centred hybrid neural-symbolic decision support system
Focus of the research project
- Integration of new products and reconfiguration in assembly line-free (dis)assembly systems: Remanufacturing of components in a circular economy.
- Human-centred design: methodological framework that facilitates the integration of people/end users into the development process.
- Knowledge-based engineering: formulation of symbolic representations for various complex systems with the aim of achieving material circularity through design methods.
- Neurosymbolic multi-objective optimisation: application of reinforcement learning to the optimisation of multiple objectives in the field of product creation.
The project results will improve product creation in the future, ...
... by improving our understanding of human decision making in circular production processes and investigating and evaluating solutions for human-AI interaction in the context of DS/AI application.
Research team
Publications & relevant previous publications
| Title | Author | DOI |
|---|---|---|
| Pipeline for ontology based modelling and automated deployment of digital twins for planning and control of manufacturing systems. Journal of Intelligent Manufacturing | A. Göppert, L. Grahn, J. Rachner, D. Grunert, S. Hort, & R. H. Schmitt | doi.org/10.1007/s10845-021-01860-6 |
| Open research knowledge graph: next generation infrastructure for semantic scholarly knowledge | M. Y. Jaradeh, A. Oelen, K. E. Farfar, M. Prinz, J. D'Souza, G. Kismihók, S. Auer | doi.org/10.1145/3360901.3364435 |
| Collaborative and Cross-Stakeholder Ontology Engineering | F. Khan, F. Engel, N. Krdzavac, S. Auer | doi.org/10.15488/16295 |
| Mappings Between Ontologies in NFDI4ING Terminology Service | N. Krdzavac, F. Engel | doi.org/10.5281/zenodo.16735777 |
| A Deep Latent Factor Graph Clustering with Fairness-Utility Trade-off Perspective | S. Ghodsi, A. Seyedi, T. L. Quy, F. Karimi, E. Natoutsi | doi.org/10.48550/arXiv.2510.23507 |