An integrated approach to analyse shiphandling expertise in a full mission bridge simulator

The aim of this doctoral research is to increase and deepen the understanding of the concept of Expertise in marine pilotage. The research introduces a methodology that combines a suite of variables to evaluate levels of “shiphandling performance”. Ten Marine Pilots participated in the research using a Full Mission Bridge Simulator, however the number of pilots does not give a good understanding of the scale of the data collected. Four different manoeuvres performed by each participant were used as experimental conditions, resulting in 6-8 hours of continuously collected data for each pilot. Those manoeuvres were controlled on three factors: the level of difficulty (easy and difficult), familiarity with the port (homeport and foreign port) and manoeuvring phase (approach, swing and closing). The dependant variables included performance, physiological and behavioural measurements. Before performing the manoeuvres in the simulator, pilots were required to plan each vessel movement. The plan was required to identify the expected use of propulsion power, vessel positions and speeds. Through a face to face interview this information was translated in numerical values. Dependent variables were obtained through a direct comparison between planned values and actual outcomes in the manoeuvres, as recorded by the simulator. Those variables were related to vessel’s XTD (Cross Track Distance), speed and propulsion power used. Statistically significant results were recorded in the factor phase, showing how the swing offered important challenges in the positioning of the vessel while rotating it. Pilots estimated the vessels speed differently in the two ports (underestimating the vessel’s speed in the foreign port) and for the two levels of difficulty (probably since dealing with two different type of propellers). Pilots underestimated the need of propulsion power in the swing phase for the more difficult manoeuvres. Such underestimation suggested further considerations with reference to safety, when manoeuvres are performed close to operative limits and maximum propulsion power available. Physiological measurements, comprising EEG (power spectrum density distribution in the bands B1 and B2), ECG (heart rate and LH/HF index) and pupil dilation, were compared against self-reported measurements (Likert scale on seven levels and NASA TLX), to better appreciate the relationship between different techniques to assess mental workload. Results obtained from measuring ECG, EEG, and pupil dilation provided some indications that physiological variables correlated to scores obtained from self- assessment scales. Light correlations were identified between the self-assessment Likert scale, heart rate and pupil dilation. Increasing the level of difficulty induced a significant increment in the levels of responses, particularly in the HR. The use of eye trackers facilitated the measurement pilots’ gaze, and voice recordings identified variations in speech behaviour. When viewing these results through the lens of Smith and Hancock’s Perceptual Cycle model of situational awareness, participant pilots were able to consistently direct their attention to specific and more relevant sources of information, depending on manoeuvring conditions. Significant results showed how gaze active search, was dependant and adaptive to the specific shiphandling tasks elicited by the manoeuvres. Eye trackers were able to document significant interactions between the subjects and their working environment, through the accounting of pilots’ orders. Results showed how the frequency of those orders were significantly higher when more critical shiphandling conditions were experienced. This research offers its contribution to the broader field of Expertise as applied to the Maritime Industry. It demonstrates how a set of empirical techniques can be used to assess a specific, yet multivariate skill – that of shiphandling. The study explores and aims to make a contribution in the validation and use of physiological measures as relatively unobtrusive proxies for mental workload. Results in this area can improve and increase the number of tools available to the shipping industry for the prevention of accidents. This research has moved our understanding of where the “red lines” of workload for marine pilots might be, levels beyond which the safety of the operations could be compromised. This study provides a tangible example of how more complex theoretical constructs such as situation awareness, can be unpacked in their constitutive elements (e.g., perception, attention). This research shows how the latest eye tracking technology can be profitably used to highlight differences and characteristics of those processes, and how those processes do change depending on specific contexts and goals. In sum, it depicts expertise as a combination of human and ship performance, embedded in a context of the difficulty of the task, the familiarity of the environment and the phase of the manoeuvre. Defining a set of standard and unobtrusive measurements contributes to address a gap identified in Industry assessment standards and, more broadly to better understand and define the nature of Expertise in such specific environment. This research was conducted in a full mission bridge simulator, and the aim for the future would be to adopt similar techniques in the real world. Such analysis would be able to highlight areas of improvement where pilots’ approach to manoeuvres could be discussed, reconsidered and modified