Vision-augmented robotic fabrication (V-aRF): systematic review on contemporary approaches and computational methods in architectural fabrication and assembly using machine vision

This systematic review examines the integration of vision systems into robotic fabrication workflows in architecture, emphasizing current methods, computational strategies, and both the potential and limitations of these technologies. The emergence of the Vision-augmented Robotic Fabrication (V-aRF) paradigm represents a significant shift, promising greater precision, adaptability, and creativity in architectural materialization. This is the first paper in the field of architectural robotics to comprehensively review the literature on vision-augmented robotic fabrication. The paper discusses the primary motivations for integrating vision systems in robotic workflows, which include adapting to dynamic robotic workcell conditions and improving the interaction between robotic systems, human agents, and the immediate operational environment. It also underscores critical challenges such as the reliance on controlled operational environments, the computational demands of processing large volumes of data in real time, the difficulties in generalizing systems across diverse materials and techniques, and obstacles in knowledge dissemination and replicability of research findings. By addressing these challenges and harnessing advanced computational methods, this paper presents an implementation-centric methodological framework for V-aRF workflows, which holds the potential for more adaptable fabrication practices. Moreover, it outlines potential future research directions to enhance system autonomy, computational efficiency, and the capability for human–machine co-creation.