Maritime accident statistics show that the majority of accidents/incidents are attributed to human errors as the initiating cause. Some studies put this as high as 95% of all accidents (collision, grounding, fire, occupational accidents, etc). The traditional way to investigate human factors in maritime industry is the statistical analysis of accident data. Although this analysis can provide key findings, it cannot capture the causal relationship between performance shaping factors and human performance in the everyday routine work, and is not suitable to be used in the individual assessment of cadets.
To reveal the effects of human factors in maritime and assess the performance of cadets, a full-mission simulator is widely used. Different scenarios such as bad weather, day and night environment, different traffic load, etc. can be simulated. The fine details of the cadet performance can be recorded in the simulator during the assessment. As a result, other than performance failure, the near misses can also be detected. Additionally, a number of cadets can go through the same scenarios at the same time and between-subjects comparison is enabled.
Besides the operations recording provided by the simulator, biosignal-based tools can additionally help in the human factors study in maritime. The existing methods include palmer perspiration, electrocardiography, etc. However, the psychophysiological states that can be recognized by these methods are limited. Electroencephalogram (EEG) biosignals can be used to directly assess the “inner” mental states of subjects. Nowadays, since the EEG devices become portable, easy to setup, and affordable in price, EEG-based tools can be used to assess psychophysiological state of subjects. Using the sensors during performing the task we can recognize the cadet/captain’s emotions, attentiveness/concentration, mental workload, and stress level in real time. In this work, we propose a real-time brain state recognition system using EEG biosignals to monitor mental workload and stress of cadets during simulator-based assessment. Currently, the proposed and implemented system includes stress and mental workload recognition algorithms. The EEG-based mental state monitoring can reflect the true “inner” feelings, stress level and workload of the cadets during the simulator-aided assessment. The time resolution is up to 0.03 second. As a result, we can analyze the recognized brain states and the corresponding performance and behavior recorded by the simulator to study how human factors affect the subject’s performance. For example, we can check is there any correlation of the cadet’s stress level and performance results. Finally, the proposed EEG-based system allows us to assess whether a cadet is ready to perform tasks on the bridge or needs more training in the simulator even if he/she navigated with few errors during the assessment.