On November 13, Justin, an associate professor at the National University of Singapore, delivered a keynote lecture titled "Digitalization of Singapore's Building Codes" at our university's "Frontiers in Intelligent Construction and Smart City Research" forum. The presentation vividly showcased Singapore's systematic approach to integrating structural design into intelligent compliance review systems. Professor Justin highlighted that through the development of the ICSG verifier and AI-assisted code inspection system, Singapore's construction industry is accelerating its transition to digitalized review processes. However, the sector still faces multiple challenges, including industry collaboration and technical standardization.
Professor Justin's research team has developed a next-generation ICSG validator for BIM compliance verification. By comparing generated scripts with reference scripts, the system demonstrates superior performance in model parameter checks and spelling error detection. The validator effectively identifies parameter anomalies and naming conventions in models, significantly improving early-stage review efficiency. Meanwhile, the automated model inspection system addresses long-standing industry pain points. To tackle challenges like discrepancies between beam design and analysis models and difficulties in automatically identifying structural elements, the system incorporates an engineering rule library and checklist, enabling automated verification of structural models and reducing human review oversights.
Professor Justin's report also highlighted three core challenges in implementing intelligent review systems: first, the steep learning curve for industry practitioners; second, insufficient data collaboration between contractors and consultants; and third, limited system compatibility due to inconsistent market modeling standards. In response, Singapore's regulatory authorities have launched a systematic response plan, including organizing tiered technical training programs, issuing unified modeling standard documents and operational guidelines, and driving comprehensive digital transformation across the industry.
During the expert Q&A session, Professor Justin explored the application potential of large language models in construction management scenarios. He noted that on-site dynamic management involves high complexity, and technical implementation still requires breakthroughs. Regarding the functional boundaries of the IFC validator, Professor Justin responded that the current system has limitations in attribute-level verification, and the next-generation product will enhance deep validation capabilities. As for the issue of BIM model quality evaluation standards, the industry has yet to establish a unified assessment system, which has been listed as a key research topic.
Professor Justin emphasized that the development of intelligent compliance review systems is a vital component of the construction industry's digital transformation strategy. The next phase will focus on advancing AI applications during construction phases and collaborating with industry associations to establish a model-based quality evaluation system, thereby providing a robust technical foundation for smart construction.
