Human reliability analysis of conventional maritime pilotage operations supported by a prospective model

Abstract

One of the most critical phases of a ship voyage is the port arrival and departure. In addition to the conventional bridge team, these activities also employ maritime pilots with advanced local knowledge to support the safe navigation. Recent works addressed the risk assessment of maritime pilotage operations based on expert opinions, AIS data, and accident reports. Despite these efforts, it is still hard to determine the extent of this contribution and the main factors driving the operation success. Looking forward to advance in the direction of better understanding the contribution of the human and organizational factors in maritime pilotage operations, this paper proposes a methodology for human reliability analysis (HRA) supported by the prospective Technique for Early Consideration of Human Reliability (TECHR) combined with Bayesian Networks (BNs). For a generic operation, the results suggest that employing one additional pilot onboard reduces the accident probabilities by about 2.38 to 3.49 times, while removing a pilot increases the accident probability by about 5.15 to 8.92 times (both compared to a base scenario with one pilot onboard). Additionally, the sensitivity analysis revealed the most important performance shaping factors (PSFs), which include domain knowledge, training & experience, and safety culture.