THE INNOVATIVE POTENTIAL OF GENERATIVE PRE-TRAINED TRANSFORMERS (GPTs) FOR QUALITY INSPECTIONS IN SWEDISH CONSTRUCTION PROJECTS

Abstract

Approaching quality inspection plans in Swedish construction projects as mere checklists and minimizing the clients’ involvement, can reduce their value. We propose improving this process through a general cloud service concept for clients, designers, and contractors, utilizing generative pre-trained transformers (GPTs). Methodologically, we synthesize literature insights on GPT uses for construction, and empirical inquiries on developing a quality self-inspection service. We posit that through such a service, project knowledge, known quality defects and lessons-learned from previous cases can be better accessed and shared – potentially leading to time savings, suggesting best practices, and improving the collaboration among clients, designers, and contractors.