Abstract
Aging of equipment and systems represents a critical challenge in process industries, as it significantly increases the likelihood and severity of industrial accidents. In this study, the Association Rule Mining (ARM) technique was adopted in the analysis of 546 accident reports from process industries. The high co-occurrence factors and association rules related to system aging were extracted to explore latent relationships among aging mechanisms, accident types, hazardous substances, accident consequences, and post-accident actions. Results highlight the value of historical accident data in characterizing aging-related accident patterns through considering relationships among multiple dimensions, and stress the need for more comprehensive perspectives in accident analysis to inform decision-making and safety management.
