OUR VISION
The XAIBIM Ecosystem: A Knowledge Distillation Process
XAIBIM introduces a dual ecosystem for civil engineering: the XAIBIM Studio for expert knowledge curation and the XAIBIM Nexus, a neuro-symbolic inference engine that operates with total transparency and auditability.
[1] P. Guo et al., '"Advancing BIM information retrieval with an LLM-based query-domain-specific language and library code function alignment system,"' Automation in Construction, vol. 178, p. 106374, 2025. DOI: 10.1016/j.autcon.2025.106374.
[2] A. M. Khan et al., '"BIM Integration with XAI Using LIME and MOO for Automated Green Building Energy Performance Analysis,"' Energies, vol. 17, no. 13, p. 3295, Jul. 2024. DOI: 10.3390/en17133295.
THE CRITICAL CHALLENGE
The Unacceptable Risk of 'Black Box' AI
The new EU AI regulatory framework and the risks of opaque AI make the use of non-auditable systems in engineering unfeasible. The XAIBIM Nexus is designed from its core to overcome this challenge, guaranteeing transparency.
[3] European Parliament and Council of the EU, '"Regulation (EU) 2024/1689 (Artificial Intelligence Act),"' Official Journal of the European Union, L, 2024/1689, Jun. 2024. https://eur-lex.europa.eu/eli/reg/2024/1689
[4] L. Chen et al., '"Can large language models replace human experts? Effectiveness and limitations in building energy retrofit challenges assessment,"' Building and Environment, 2025. DOI: 10.1016/j.buildenv.2025.112891.
KNOWLEDGE CURATION
CESA: The Audited Knowledge Base
The XAIBIM Studio generates the ecosystem's knowledge base: the CESA (Curated, Explainable, Structured Asset). Each CESA is an asset validated by an expert and sealed with immutable cryptographic integrity to ensure the reliability of the XAIBIM Nexus.
[5] P. Guo et al., '"Advancing BIM information retrieval with an LLM-based query-domain-specific language and library code function alignment system,"' Automation in Construction, vol. 178, p. 106374, 2025. DOI: 10.1016/j.autcon.2025.106374.
[6] I. Aurangzeb and J. H. Yoon, '"Artificial intelligence- and blockchain-enabled carbon emissions ledger system (AB-CELS) for sustainable construction processes,"' Automation in Construction, vol. 176, p. 106286, 2025. DOI: 10.1016/j.autcon.2025.106286.
EXPLAINABLE ARCHITECTURE
The XAIBIM Nexus Neuro-Symbolic Pipeline
The XAIBIM Nexus operates with an auditable pipeline: it starts with semantic requirements extraction (NLP), performs geometric interpretation (CNNs), analyzes relational context (GNNs) and executes deductive normative validation (SWRL).
[7] Y. Zhou et al., '"A Design for Safety (DFS) Semantic Framework Development Based on Natural Language Processing (NLP) for Automated Compliance Checking Using BIM: The Case of China,"' Buildings, vol. 12, no. 6, p. 780, 2022. DOI: 10.3390/buildings12060780.
[8] N. Montás-Laracuente et al., '"A systematic review of artificial intelligence for capturing real-world structures into building information modelling,"' Journal of Building Engineering, vol. 113, p. 114093, 2025. DOI: 10.1016/j.jobe.2025.114093.
[9] J. Ko and D. Lee, '"Graph neural networks for classification and error detection in 2D architectural detail drawings,"' Automation in Construction, vol. 170, p. 105936, 2025. DOI: 10.1016/j.autcon.2024.105936.
[10] W. Solihin et al., '"BIM-based automated rule-checking in the AECO industry: Learning from semiconductor manufacturing,"' Automation in Construction, vol. 162, p. 105406, 2024. DOI: 10.1016/j.autcon.2024.105406.