TY  - JOUR
AB  - High-value circular strategies such as reuse depend on architecture decisions made before embodiment is fixed. However, reuse-relevant information is dispersed across Systems Modeling Language (SysML)/Product Lifecycle Management (PLM) artefacts, rule logic, standards and regulations, and external datasets (e.g., Lifecycle Assessment factors). This paper presents a Model-Based Systems Engineering (MBSE) framework and Artificial Intelligence (AI)-supported evidence workflow that enables traceable, reproducible assessment of reuse-oriented circularity for system structure variants at an architecture decision gate. The approach separates inputs by provenance product-specific SysML v2 variant models linked to PLM configurations, a versioned circularity rule library, and curated collections and quantitative datasets and connects them through explicit, reviewable mappings. Deterministic Key Performance Indicator (KPI) pipelines compute engineer-assessable indicators (reuse-yield proxies, non-destructive disassembly, interface standardization/interchangeability) directly from the model backbone. A Retrieval Augmented Generation (RAG) component supports evidence acquisition and semantic alignment for externally dependent KPIs (e.g., avoided embodied carbon through reuse) and for justification from standards, while maintaining human accountability. The framework is instantiated on two alternative 3D printer printhead architectures. Results highlight the integration–circularity tradeoff.
AU  - Juresa, Yannick
AU  - Mollahassani Madjdabadi, Damun
AU  - Göbel, Jens Christian
DO  - 10.17619/UNIPB/1-2648
PB  - Universitätsbibliothek
DP  - Universität Paderborn
LA  - eng
PY  - 2026
SP  - 1 Online-Ressource (Seite 41-50) : Diagramme
T2  - 1st International Symposium: March 24 – 26, 2026, Heinz Nixdorf Institute, Paderborn University
TI  - Framework for model-based systems engineering circularity assessment
UR  - https://nbn-resolving.org/urn:nbn:de:hbz:466:2-58859
Y2  - 2026-10-10T06:17:34
ER  -