Recent advancements of robotics in construction

In the past two decades, robotics in construction (RiC) has become an interdisciplinary research field that integrates a large number of urgent technologies (e.g., additive manufacturing, deep learning, building information modelling (BIM)), resulting in the related literature being both fragmented and vast. This paper has explored the advances in RiC in the past two decades using a mixed quantitative-qualitative review method. Initially, 940 related articles (170 journal articles and 770 conference papers) were identified by keyword-searching in Scopus and then fed into a bibliometric analysis to build science maps. Following this, a qualitative discussion highlights recent achievements in RiC across three dimensions: tasks, algorithms, and collaborations. Moreover, four future research directions are proposed: 1) in-depth integration of BIM and robotics; 2) near-site robotic fabrication; 3) deep reinforcement learning for flexible environment adaption; and 4) high-level robot-to-robot collaboration. The contributions of this research are twofold: 1) identifying the latest research topics and trends concerning robotic technologies in construction; and 2) providing in-depth insights into the future direction of RiC. The findings from this research can serve both academia and industry in terms of promoting robotic algorithms, hardware, and applications in construction industry.