Injury prediction based on safety climate questionnaire score using artificial neural networks

According to the statistics of the Ministry of Labor in Taiwan, there are over 54,000 occupational accidents each year from 2011 to 2015. One third of the accidents happened in the manufacturing industry, of which the steel manufacturing industry is a part. Steel industry employees are undoubtedly working in an extremely vulnerable environment. Therefore, it is important to improve workers' safety. Predicting accidents may be one solution for managers to understand the dangers ahead of time, so that they can implement protective actions. Artificial Neural Networks (ANNs) is a widely used method for prediction. It has been used in many areas and objectives. In this study, we use an ANNs model to predict whether an operator in the steel industry will encounter an accident. The prediction factor is the score from safety climate questionnaire, which is an easy way to assess the safety of the organization. The results showed that the prediction accuracy is 61.72%. Therefore, it's a potential method to predict accidents in steel industry.