Predictive Analytics of Human Errors in the Fireworks Industry

The town of Sivakasi in Tamil Nadu's Virudhunagar district in India produces majority of the country's consumption of firework items. Handling numerous chemicals is a necessary part of the firework industry's manufacturing process. As a result, the firework industry is commonly reported to be highly dangerous because of the hazardous nature of the chemicals used to create the sparkling effects during the ignition of firework crackers. Previous research have focused on harmful behaviors and hazardous conditions, pointing to human error as the primary cause of many accidents. According to the findings of this study, the majority of explosions were caused by the improper handling of hazardous chemicals and carelessness when making fireworks. Therefore, a method was presented in this paper which aims to examine the likelihood of human error in the fireworks industry. The proposed method uses task analysis and prediction of human error to shape the performance elements. The presented model can also be used to examine potential accident scenarios. The results show that the presented greedy-based process compared with the rule mining-based approach gives better accuracy and outcomes for the prediction of human error possibilities in the fireworks industry.