CyberTraining: Pilot: Cognitive and Generative AI-driven Cyber-Infrastructure Training Platform for Construction Education

The construction industry has the lowest digitalization level among all major industries in the US and suffers from long-standing challenges, such as low productivity, difficulty in effective collaboration and communication, and lack of digital skills in the construction workforce. Therefore, broadening the adoption of digital technologies and cyber-infrastructures (CI) in the construction industry is a key to its future success and will revolutionize its operations.

Building information modeling (BIM) is considered as the most significant and promising CI in construction, which refers to a sophisticated process of generating digital representations of the physical and functional characteristics of facilities. Due to the cross-disciplinary nature of BIM, the lack of effective BIM education resources, such as instructors and curriculum materials, has become a major challenge for BIM CI training in the US.

Recently, generative artificial intelligence (AI) has attracted huge attention from the public for its ability to understand and respond to human language, suggesting its potential to revolutionize education and intelligent tutoring. This pilot project aims to develop a cognitive and generative artificial AI-driven CI training platform by integrating advanced skills and knowledge of BIM into the undergraduate curriculum in construction programs.

The new knowledge gained from this project will build a solid foundation for understanding how generative AI can be used for intelligent tutoring of engineering education including construction engineering, civil engineering, and electrical engineering in the US, which could help mitigate the nation's labor shortage.

This project will pioneer a new paradigm of integrating generative AI into the Nation’s educational curriculum to train the future scientific community. The project will develop a BIM CI training framework for construction education by synergizing generative AI and cognitive science theories. Specifically, this project will 1) develop undergraduate-level BIM-related curriculum materials by incorporating the knowledge space theory from cognitive science to fuse these materials.

Expert discussion will be employed to ensure the quality and suitability of developed materials for undergraduate construction programs in the US; 2) develop a generative design method by integrating parametric modeling and computational techniques to automatically produce BIM CI worked examples with various levels of difficulties. These worked examples will serve as a significant supplement to curriculum materials, helping instructors prepare tailored problems for BIM CI training; and 3) explore the feasibility of Generative AI for BIM CI training.

This project will develop a ChatGPT-based web application as the CI training platform. The training platform developed from this project can provide personalized, adaptive, and interactive training to students, leading the transformation of the educational ecosystem across all engineering fields. Extensive experiments will be conducted to validate the developed curriculum materials, worked examples, and the training platform.

By providing a personalized learning platform, this project will contribute to the establishment of deeper engagement with various institutions and underrepresented groups in Michigan, the Midwest, and the US.

This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.