Automatic monitoring the risk coupling of foundation pits: integrated point cloud, computer vision and Bayesian networks approach

Abstract

Purpose

Manual monitoring is a conventional method for monitoring and managing construction safety risks. However, construction sites involve risk coupling - a phenomenon in which multiple safety risk factors occur at the same time and amplify the probability of construction accidents. It is challenging to manually monitor safety risks that occur simultaneously at different times and locations, especially considering the limitations of risk manager’s expertise and human capacity.

Design/methodology/approach

To address this challenge, an automatic approach that integrates point cloud, computer vision technologies, and Bayesian networks for simultaneous monitoring and evaluation of multiple on-site construction risks is proposed. This approach supports the identification of risk couplings and decision-making process through a system that combines real-time monitoring of multiple safety risks with expert knowledge. The proposed approach was applied to a foundation project, from laboratory experiments to a real-world case application.

Findings

In the laboratory experiment, the proposed approach effectively monitored and assessed the interdependent risks coupling in foundation pit construction. In the real-world case, the proposed approach shows good adaptability to the actual construction application.

Originality/value

The core contribution of this study lies in the combination of an automatic monitoring method with an expert knowledge system to quantitatively assess the impact of risk coupling. This approach offers a valuable tool for risk managers in foundation pit construction, promoting a proactive and informed risk coupling management strategy.

Citation

Li, X., Yang, X., Feng, K. and Liu, C. (2024), "Automatic monitoring the risk coupling of foundation pits: integrated point cloud, computer vision and Bayesian networks approach", Engineering, Construction and Architectural Management, Vol. ahead-of-print No. ahead-of-print. https://doi.org/10.1108/ECAM-02-2024-0149