Degree: Doctor of Philosophy Abstract: Offsite construction has become a viable alternative to traditional construction methods by establishing controlled and automated manufacturing environments for construction products. To reap the benefits of industrialized construction, manufacturing execution and quality play an important role. However, most quality control processes remain manual, relying on visual inspection and operator’s expertise. Automated manufacturing processes in offsite construction facilities and their corresponding machinery would benefit from smart manufacturing, information, and communication technologies to control quality and, ultimately, reduce defects and optimize production. This study presents a framework for the automatic inspection and quality assessment of BIM-based construction products as a knowledge-based cyber-physical system that bridges Industry 4.0 principles with zero-defect manufacturing and lean techniques. The proposed methodology is based on the well-known 5C cyber-physical architecture, adapted to a building information modeling environment. In this research, light-gauge steel frame assemblies manufacturing is selected as the case study. At first, a knowledge model for steel frame assemblies manufacturing is proposed to link, at the design stage, product information, manufacturing operations, and quality specifications. By enhancing building information models with the developed ontology model, offsite practitioners can access quality information at the design stage and prepare adequate quality control strategies beforehand. Then, a vision-based cyber-physical inspection system is proposed that generates quality-related information as required by the BIM model. The designed CPS system employs visual sensors (cameras) to provide real-time inspection and quality control of the frame assembly manufacturing process, in addition to providing a platform for data storage and future analysis of quality-oriented data. Finally, in an effort to integrate human input and knowledge into the system, cognitive and supervisory roles are discussed. The implementation of the proposed system enables quantification of quality issues, as well as analysis of the source of defects in steel frame assemblies manufacturing.