Human-Related Uncertainty Analysis for Automation-Enabled Façade Visual Inspection: A Delphi Study

Abstract

Given that traditional façade visual inspection entails laborious, dangerous, and inefficient manual work, automation-enabled façade visual inspection has become a prevailing trend in both academia and industry. However, automation-enabled applications often encounter uncertainty problems. For automation-enabled façade visual inspection, uncertainty in reliability and efficiency is an important factor that determines the value of introducing automation to façade visual inspection. During automation-enabled façade visual inspection, human efforts play important roles throughout the whole process and compose a human–cyber–physical system. Therefore, human-related activities and human factors are prominent causes of uncertainty in automation-enabled façade visual inspection. To understand human-related uncertainty, this work designed a Delphi study with an expert panel to quantitatively evaluate human-related activities and human factors. Also, an optimized fuzzy Delphi method was adopted to process the collected evaluation opinions. Based on the results, the most critical activities and human factors influencing uncertainty were extracted. Additionally, a structure of uncertainty generation was developed to analyze the evaluation results and provide recommendations for uncertainty control. This research contributes to facilitating understanding of the uncertainty problem in human–cyber–physical systems and to providing effective recommendations for uncertainty control in automation-enabled façade visual inspection.