Purpose: Few of the established risk identification methods refer to low-severity yet high-frequency safety risks data that may lead to several safety risks being ignored, thus reducing the potential of learning from a considerable number of cases. The purpose of this study is to explore a new valid method based on preaccident safety supervision data to identify these minor construction safety risks during routine construction operations.Design/methodology/approach: A total of 329 official construction safety supervision reports containing 5,159 safety problem records from Shanghai between 2016 and 2018 served as raw material for in-depth analysis. Given the characteristics of the data collected, text mining integrated with natural language processing was applied to review the supervision reports and group safety risks automatically.Findings: This study clarifies the way in which the supervision data should be employed to analyze high-frequency–low-severity safety risks. From these data, seven unsafe-act-related and nine unsafe-condition-related risks are identified. Regarding unsafe-act-related risks, inappropriate human behaviors could usually occur in personnel management, contract management, expense management, material management and acceptance work. For unsafe-condition-related risks, hoisting, scaffolding and reinforcement works are the main generators of onsite safety hazards during construction operations.Practical implications: The study includes implications for project managers and supervisors to facilitate more effective proactive risk management by paying more attention to collecting and employing the supervision data established in each routine inspection.Originality/value: Whereas previous research focused on analyzing severe accidents, this study seeks to identify the high-frequency–low-severity construction safety risks using the preaccident supervision data. The findings could provide a new thought and research direction for construction safety risk management.
1 May 2025 in Journal of Management in Engineering
Importance of little objects in cars such as door handles is obvious, both in daily lives and in industrial manufacture. However, since the lack of the distinctive appearance and feature, obtaining the location of them is still remaining a challenge. This paper proposes an effective approach for the detection of the door handles of cars. The method innovatively combines frequency and spatial domains' algorithms to detect the location of little objects of cars in a relatively large image. To illustrate the method more concisely, our method localizes door handles through image segmentation and visual saliency detection. First, by segmenting the image we can remove the unnecessary area to improve the speed and accuracy of our approach. After finding the region of interest, our approach uses a visual saliency detection algorithm named Spectral Residual Approach which can get the location of door handles accurately. At last, the approach is tested by different kinds of images of vehicles. The results of the experiments show that our approach is obvious and practical.
1 October 2025 in Journal of Construction Engineering and Management