Despite the maritime industry’s significant role in the global economy, human error is still the leading cause of maritime accidents. This study presents an assessment of human factors in 31 collision accident reports in the USA from 2010 – 2017, involving 50 ships. In this study, the categorization of Error-Producing Conditions (EPCs) from the Human Error Assessment and Reduction Technique (HEART) methodology to the 4M (man, machine, media, and management) framework, EPC–4M was carried out. This study aims to develop maritime accident assessment and reduction techniques in terms of technical understanding of the risk and safety linked to human factors. The most categorization of the collision data is a complex task requiring a high level of comprehension and skill. A total of 154 EPC – 4M categorizations were identified. The average of human error probability for USA collision accidents was determined to be 59%. The results of the study reinforce the idea that human and management interactions are the most common factors that lead to accidents.
This research aimed to investigate differences in collision avoidance judgments made by navigators who had different levels of experience and were restricted to using either radar or landscape information, but not both. Collision avoidance judgment was assessed for four groups using metrics of timing and degrees of bearing relating to course alteration. Experiments were conducted in a simulator-based navigation room using a comparatively complex traffic situation in which multiple collisions needed to be avoided in sequence.
The results of course change timings indicated that less experienced navigators required less time than experienced navigators when restricted to using radar information. Results pertaining to the degree of bearing change relating to course alteration indicated that navigation using landscape information provided decisions that were more consistent, regardless of the experience level, whereas navigation using radar information led to a wide range of decisions regardless of the experience level. This suggests that access to landscape information contributes to prevent navigators from making diverse judgments for collision avoidance. Results thus show that the use of mainly landscape information appears to be crucial in maintaining safety when navigators avoid each other.