A probabilistic model of human error assessment for autonomous cargo ships focusing on human-autonomy collaboration

Despite the use of automation technology in the maritime industry, human errors are still the typical navigational risk influencing factors in autonomous ships with the third degree of autonomy. However, there is an urgent need for new human error probability assessment focusing on the autonomous cargo ships with human-autonomy collaboration. Hence, to assess these human errors during the emergency response process, a probabilistic model is proposed in this paper. Firstly, the risk factors are identified and classified by analysing the operational process of the Shore Control Centre (SCC). This is followed by the establishment of an event tree model delivered from human errors using Technique for Human Error Rate Prediction (THERP). Furthermore, Bayesian Networks (BNs) model is utilized for the three stages of perception, decision, and execution. Finally, the human errors probabilities are calculated for the mentioned three stages focusing on human-autonomy collaboration. Moreover, the importance of human error factors is quantified with sensitivity analysis, which can provide flexible references for the theoretical construction of the SCC and training of operators. The process was applied to assess the probabilities of human errors focusing on human-autonomy collaboration under the remote navigation mode of an autonomous cargo ship (test ship) in the city of Wuhan, China.