Prototype of a safety management support tool using the results of a safety activity awareness survey

Abstract: In recent years, enhanced safety management has been implemented in various sectors. In particular, it has been aimed at voluntarily evaluating the culture, awareness, and organization of safety on an ongoing basis with the cooperation of employees.Until now, the evaluation of motivation regarding safety activities has been carried out using a questionnaire-based survey method. However, it is difficult even for those with sufficient experience and skills in safety management to decipher the background of the collected results and future issues. For this reason, many companies ask experts and safety consultants to conduct assessments, but the reality is that they are not able to analyze the results down to the detailed departmental level due to the high cost involved. Therefore, the objective of this study was to develop a system that automatically indicates issues related to safety management and their solution policies by analyzing the results of an awareness survey on safety activities from various perspectives.Since 2007, this laboratory has conducted a total of 200,000 people at more than 70 businesses, and has data on their awareness of safety activities. These big data were processed using variance-covariance structural analysis to create a structured model of the influence of the questionnaire items. Based on this structural model, all survey data were multilayered. Furthermore, the relationship between the multilayered data and the past findings of the experts (assessment results) was formulated using machine learning (deep learning). Finally, we developed an application that implements the results obtained in this way. In other words, we developed a system that, upon inputting the results of an awareness survey on safety activities, outputs “Current issues regarding safety activities,” “Priorities for solving issues,” and “Advice for solving issues. The system was developed to output “current issues regarding safety activities,” “priorities for resolving issues,” and “advice for resolving issues.The system is now being used to examine the content of the “advice for on-site management” obtained from this system for the businesses that conducted the safety awareness survey in 2024 (approximately 15,000 subjects). In the future, we plan to have businesses in the railroad, manufacturing, IT service, and medical industries join the project to verify the validity of the system and to improve its practicality.

Keywords: Human factors, Safety management, Safety activity awareness survey

DOI: 10.54941/ahfe1006569

Cite this paper:

​