TY - GEN
T1 - Automating reinforced concrete design through ontologies
AU - Jędrzejewska, Agnieszka
AU - Leonardi, Maria Laura
AU - Granja, José
AU - Paszek, Agata
AU - Oliveira, Daniel V.
AU - Azenha, Miguel
N1 - Publisher Copyright:
© 2026, fib. The International Federation for Structural Concrete. All rights reserved.
PY - 2026
Y1 - 2026
N2 - Current trends show the future of structural concrete design at the intersection of sustainability, automation, and emerging artificial intelligence-based tools. Achieving this vision requires formal knowledge representations capable of capturing not only engineering logic, but also the regulatory and material-specific constraints embedded in design standards. In this context, this contribution builds upon the Eurocode Core Ontology (ECO), a semantic framework that formalises the normative logic of Eurocode-based structural design, and investigates its extension toward reinforced concrete engineering. The primary contribution of this work is the development of a concrete-specific extension of ECO, introducing material semantics, reinforcement concepts, cracking-related quantities, and verification logic used in reinforced concrete design. This extension is presented as a proof of concept for the extensibility of ECO, demonstrating how domain-specific knowledge can be integrated without modifying the ontology core. A theoretical case is discussed of hydration-induced cracking risk assessment for a mass concrete wall, illustrating how regulatory knowledge and verification criteria can be represented and coordinated within the proposed framework. In addition, the paper outlines a conceptual reference architecture for ontology-enabled design assistance, in which language-based agents provide a user-oriented interface to ECO-based knowledge and verification services through a standardised integration layer inspired by the Model Context Protocol (MCP). The architecture is intended to support structured querying, orchestration of reasoning services, and traceable explanation of design decisions, while all deterministic engineering checks remain external to the language model. While grounded in Eurocode practice, the conceptual foundations of the proposed approach are compatible with the fib Model Code, reflecting their shared limit-state-oriented design philosophy. The presented framework provides a basis for further development of interoperable and regulation-aware tools for reinforced concrete design.
AB - Current trends show the future of structural concrete design at the intersection of sustainability, automation, and emerging artificial intelligence-based tools. Achieving this vision requires formal knowledge representations capable of capturing not only engineering logic, but also the regulatory and material-specific constraints embedded in design standards. In this context, this contribution builds upon the Eurocode Core Ontology (ECO), a semantic framework that formalises the normative logic of Eurocode-based structural design, and investigates its extension toward reinforced concrete engineering. The primary contribution of this work is the development of a concrete-specific extension of ECO, introducing material semantics, reinforcement concepts, cracking-related quantities, and verification logic used in reinforced concrete design. This extension is presented as a proof of concept for the extensibility of ECO, demonstrating how domain-specific knowledge can be integrated without modifying the ontology core. A theoretical case is discussed of hydration-induced cracking risk assessment for a mass concrete wall, illustrating how regulatory knowledge and verification criteria can be represented and coordinated within the proposed framework. In addition, the paper outlines a conceptual reference architecture for ontology-enabled design assistance, in which language-based agents provide a user-oriented interface to ECO-based knowledge and verification services through a standardised integration layer inspired by the Model Context Protocol (MCP). The architecture is intended to support structured querying, orchestration of reasoning services, and traceable explanation of design decisions, while all deterministic engineering checks remain external to the language model. While grounded in Eurocode practice, the conceptual foundations of the proposed approach are compatible with the fib Model Code, reflecting their shared limit-state-oriented design philosophy. The presented framework provides a basis for further development of interoperable and regulation-aware tools for reinforced concrete design.
KW - design automation
KW - large language model
KW - model context protocol
KW - ontology
KW - standards
UR - https://www.scopus.com/pages/publications/105051143889
M3 - Conference contribution
AN - SCOPUS:105051143889
SN - 9782940643318
T3 - fib Symposium
SP - 5809
EP - 5817
BT - Structural Concrete - 2050 Towards Carbon Neutrality, AI Design, and Robotic Construction - Proceedings of the 7th fib Congress, 2026
A2 - Júlio, E.
A2 - Fernandes, P.
A2 - Bogas, J.
A2 - Costa, H.
A2 - Gonilha, J.
PB - fib. The International Federation for Structural Concrete
T2 - 7th fib Congress on Structural Concrete 2050: Towards Carbon Neutrality, AI Design, and Robotic Construction, 2026
Y2 - 15 June 2026 through 19 June 2026
ER -