Maritime accidents are complex processes in which many factors are involved and contribute to accident development. In order to capture underlying factors in accidents, countries adapted an accident investigation system with the aim of learning from these rare events and prevent similar occurrences in the future. Often these accident investigation reports are converted into databases, which lack a concise and user-friendly classification system, as a result there are a lot of inadequacies in data-collection and tagging procedures. Therefore, the authors propose to apply an approach to classify human factors (HFs) appeared in past maritime accidents, aiming to develop a set of HFs categories which can be used for accidents learning. For this purpose, an accident database was obtained and a two-stage approach is adapted to conduct analysis: first, an open card-sorting case study is organised to group the HFs extracted from an historical accident database. Second, a hybrid card-sorting method is utilized to fully achieve the classification of HFs. Our study revealed issues where HFs are weakly defined and similar factors are duplicated by investigators who populate the database. High level categories were developed and presented which covers great majority of HFs concerns involved in accidents.
- maritime accidents
- human factors