Artificial neural network-based risk assessment for occupational accidents in the shipbuilding industry in Turkey

Accidents in the workplace are critical issues that necessitate attention from environmental safety and health perspectives to enhance operational safety. The elevated rate of workplace accidents in Turkey underscores deficiencies in reliability analysis studies. Effective planning of preventive measures, considering factors such as location and timing, is pivotal in accident prevention. Sociological and regional disparities, alongside technical factors such as service type, working hours, and age, contribute significantly to accident causation. Reliability analysis studies aim to identify potential causes of accidents, enabling early detection of hazardous situations and combinations. This proactive approach allows responsible personnel to implement measures that mitigate or eliminate specific types of accidents, thereby safeguarding both lives and business assets. This study utilizes Artificial Neural Network (ANN) to forecast potential occupational accidents in the shipbuilding industry before they occur. Analyzing data from 146 occupational accidents involving ship electricians between 2012 and 2016, out of a total of 1165 occupational accidents in the industry, the study estimates potential accidents for 2017. The results demonstrate that ANN achieves high accuracy in predicting occupational accidents.