EEG-based stress recognition in competency assessment using maritime navigation simulator

Abstract

Human factors receive increasing attention in maritime operations as it is found that human errors cause up to 90% of maritime incidents. One of the solutions to enhance safety in maritime navigation is to consider both technical and non-technical skills during training and assessment. Technical skills are always the first and foremost step in training or assessment. However, it may not be enough to solely conduct the navigation safely and efficiently by technical skills. Non-technical skills such as situational awareness, workload and stress management, and decision making also play a critical role in safe navigation and should not be overlooked. In addition, the seafarers spend a long time on board as part of their typical work pattern, and due to the current pandemic, they might be stranded at sea for even longer. It has been urged that ship owners and managers take the right actions to care for seafarers' mental well-being, such as stress and fatigue. To fulfil the needs of non-technical skill assessment and monitoring of mental well-being, electroencephalogram (EEG) is utilized to recognize the seafarers' mental stress when performing specific tasks in a full-mission Advanced Navigation Research Simulator (ANRS). The EEG-based recognition has unique advantages over the other biosensors. For example, it has high temporal resolution and reflects the mental states that are not identifiable from facial expressions with acceptable accuracy. A simulator allows the researchers to design and implement navigation scenarios with different weather conditions, traffic density, vessel types, emergency alarms, etc. This is ideal for conducting experiments for human factors study. An experiment has been designed and conducted in ANRS to validate the use of EEG for maritime navigational competency assessment. Participants with varied maritime backgrounds and roles were invited for data collection, e.g., navigating officers and experienced marine Pilots. Demanding events such as engine failure, close quarter situation, and other bridge equipment failure were marked during the EEG data recording. The results show that the EEG based stress recognition correlates with the demanding events in the experiment and can reflect the difficulty level of the challenging circumstances. The proposed EEG-based recognition can be used to study the human element in competency assessment using a maritime navigation simulator to ensure objective competency assessment.