Trustworthy AI and robotics: Implications for the AEC industry

Abstract

Human-technology interaction is concerned with trust as an inevitable user acceptance requirement. As the applications of artificial intelligence (AI) and robotics emerge in the architecture, engineering, and construction (AEC) industry, there is an immediate need to study trust in such systems. This paper presents the results of a systematic review of the literature published in the last two decades on (1) trust in AI and AI-powered robotics and (2) AI and robotics applications in the AEC industry. Through a thorough analysis, common trust dimensions are identified and the connections to the existing AEC applications are determined and discussed. Furthermore, major future directions on trustworthy AI and robotics in AEC research and practice are outlined. Findings indicate that although AEC researchers and industry professionals increasingly study and deploy AI and robotics, there is a lack of systematic research that studies key trust dimensions such as explainability, reliability, robustness, performance, and safety in the AEC context