C2C Assist: A Knowledge Graph-Constrained LLM for Cradle-to-Cradle Product Design

Meno Said Haddad; Alexander Roith; Arthur Seibel

doi:10.1016/j.procir.2026.05.237

C2C Assist: A Knowledge Graph-Constrained LLM for Cradle-to-Cradle Product Design

Meno Said Haddad^*
, Alexander Roith
, Arthur Seibel

^*Corresponding author for this work

Research output: Journal contributions › Conference article in journal › Research

Abstract

This paper introduces a cradle-to-cradle design assistant that combines a Neo4j knowledge graph with large language model (LLM) reasoning to guide sustainable product development. The graph encodes materials, end-of-life pathways, and design rules, while the LLM generates product-specific recommendations grounded in this structured knowledge. Expert evaluation across five product cases (toothbrush, sunglasses, smoothie maker, coffee machine, and air fryer) shows strong coverage of material health, circularity, and creative design options, but weaker performance in water and soil stewardship, air and climate protection, and overall conceptual consistency. Results demonstrate that graph-LLM integration improves the concreteness and usability of circular design guidance, positioning such systems as valuable partners in sustainable product innovation.

Original language	English
Journal	Procedia CIRP
Volume	142
Pages (from-to)	139-144
Number of pages	6
ISSN	2212-8271
DOIs	https://doi.org/10.1016/j.procir.2026.05.237
Publication status	Published - 2026
Event	36th CIRP Design Conference - CIRP Design 2026 - Tokyo, Japan Duration: 16.03.2026 → 18.03.2026 Conference number: 36 https://tore.tuhh.de/entities/event/c93f6d97-da30-43f1-a5b3-ec0ff75b0eaf

Bibliographical note

Publisher Copyright:
Copyright © 2026. Published by Elsevier B.V.

UN SDGs

This output contributes to the following UN Sustainable Development Goals (SDGs)

SDG 9 Industry, Innovation, and Infrastructure
SDG 12 Responsible Consumption and Production

Research areas and keywords

Cradle to Cradle
Design
Knowledge Graphs
Large Language Models
Product design
Product (mathematics)
New product development
Product lifecycle
Product design specification
Informatics

ASJC Scopus Subject Areas

Control and Systems Engineering
Industrial and Manufacturing Engineering

Access to Document

10.1016/j.procir.2026.05.237Licence: CC BY-NC-ND

Cite this

@article{3839d60b79f74a7c9d94d1ebe218a8b0,

title = "C2C Assist: A Knowledge Graph-Constrained LLM for Cradle-to-Cradle Product Design",

abstract = "This paper introduces a cradle-to-cradle design assistant that combines a Neo4j knowledge graph with large language model (LLM) reasoning to guide sustainable product development. The graph encodes materials, end-of-life pathways, and design rules, while the LLM generates product-specific recommendations grounded in this structured knowledge. Expert evaluation across five product cases (toothbrush, sunglasses, smoothie maker, coffee machine, and air fryer) shows strong coverage of material health, circularity, and creative design options, but weaker performance in water and soil stewardship, air and climate protection, and overall conceptual consistency. Results demonstrate that graph-LLM integration improves the concreteness and usability of circular design guidance, positioning such systems as valuable partners in sustainable product innovation.",

keywords = "Cradle to Cradle, Design, Knowledge Graphs, Large Language Models, Product design, Product (mathematics), New product development, Product lifecycle, Product design specification, Informatics",

author = "Haddad, \{Meno Said\} and Alexander Roith and Arthur Seibel",

note = "Publisher Copyright: Copyright {\textcopyright} 2026. Published by Elsevier B.V.; 36th CIRP Design Conference - CIRP Design 2026, CIRP Design 2026 ; Conference date: 16-03-2026 Through 18-03-2026",

year = "2026",

doi = "10.1016/j.procir.2026.05.237",

language = "English",

volume = "142",

pages = "139--144",

journal = "Procedia CIRP",

issn = "2212-8271",

publisher = "Elsevier B.V.",

url = "https://tore.tuhh.de/entities/event/c93f6d97-da30-43f1-a5b3-ec0ff75b0eaf",

}

TY - JOUR

T1 - C2C Assist

T2 - 36th CIRP Design Conference - CIRP Design 2026

AU - Haddad, Meno Said

AU - Roith, Alexander

AU - Seibel, Arthur

N1 - Conference code: 36

PY - 2026

Y1 - 2026

N2 - This paper introduces a cradle-to-cradle design assistant that combines a Neo4j knowledge graph with large language model (LLM) reasoning to guide sustainable product development. The graph encodes materials, end-of-life pathways, and design rules, while the LLM generates product-specific recommendations grounded in this structured knowledge. Expert evaluation across five product cases (toothbrush, sunglasses, smoothie maker, coffee machine, and air fryer) shows strong coverage of material health, circularity, and creative design options, but weaker performance in water and soil stewardship, air and climate protection, and overall conceptual consistency. Results demonstrate that graph-LLM integration improves the concreteness and usability of circular design guidance, positioning such systems as valuable partners in sustainable product innovation.

AB - This paper introduces a cradle-to-cradle design assistant that combines a Neo4j knowledge graph with large language model (LLM) reasoning to guide sustainable product development. The graph encodes materials, end-of-life pathways, and design rules, while the LLM generates product-specific recommendations grounded in this structured knowledge. Expert evaluation across five product cases (toothbrush, sunglasses, smoothie maker, coffee machine, and air fryer) shows strong coverage of material health, circularity, and creative design options, but weaker performance in water and soil stewardship, air and climate protection, and overall conceptual consistency. Results demonstrate that graph-LLM integration improves the concreteness and usability of circular design guidance, positioning such systems as valuable partners in sustainable product innovation.

KW - Cradle to Cradle

KW - Design

KW - Knowledge Graphs

KW - Large Language Models

KW - Product design

KW - Product (mathematics)

KW - New product development

KW - Product lifecycle

KW - Product design specification

KW - Informatics

UR - https://www.scopus.com/pages/publications/105042725306

U2 - 10.1016/j.procir.2026.05.237

DO - 10.1016/j.procir.2026.05.237

M3 - Conference article in journal

AN - SCOPUS:105042725306

SN - 2212-8271

VL - 142

SP - 139

EP - 144

JO - Procedia CIRP

JF - Procedia CIRP

Y2 - 16 March 2026 through 18 March 2026

ER -

C2C Assist: A Knowledge Graph-Constrained LLM for Cradle-to-Cradle Product Design

Abstract

Bibliographical note

UN SDGs

Research areas and keywords

ASJC Scopus Subject Areas

Access to Document

Fingerprint

Cite this