A system for evaluating the status of future human error prevention activities in manufacturing sites based on information noticed during daily work

In factories such as aircraft manufacturing, where the number of products manufactured is so small that the defective product rate must be reduced to zero, a great many human error countermeasures are taken. However, the burden on workers due to too many countermeasures has exceeded the limit, and there is an urgent need to optimize the management of error countermeasures as a whole. To that end,(1) Collection of detailed human factor information on problematic events such as nonconformance events(2) Collection of human factor information during normal operations(3) Structural analysis of human factors(4) Propose guidelines for human error countermeasures based on the analysis results and present multiple specific measures(4) Propose guidelines for human error countermeasures based on the results of the analysis and present multiple specific countermeasures.In this study, with regard to (2), we attempted to examine a method for collecting human factor information from on-site conversations, based on the activation of constructive communication at the work site. We called “hints that lead to constructive communication” “positive words,” and designed a system to evaluate this information from multiple perspectives. In the collection phase, the system uses Google Form, which allows workers to input information at any time from their own information terminals, and simple summaries are fed back to the site as a preliminary report the next morning.In the evaluation phase,Phase I: Establishment of a safety management systemPhase 2: Strategic safety management: Leadership, followership, and information gathering/sharing system in the field.Third step: Reputation management: educational curriculum for the next generation, development of generalist staff with diverse knowledge.Fourth step: Strategic management: Future production including social contribution, regional contribution, investment in the future, and fostering future value.We have developed a system that can present the satisfaction level of on-site communication regarding the four levels of communication, as well as advice on communication content that should be activated in the future. We have developed a generative AI that has been specially trained using past literature data on safety management, reputation management, and brand management, and this generative AI can provide “advice on communication content” in response to on-site data. The prototype application was tested at an actual factory. The prototype application is now being tested at an actual factory, and improvements are being made in order to deploy it in activities aimed at preventing nonconformance events.