A Stochastic State-transition and Safety Assessment for Marine Accidents Based on Human Factors

The man-machine systems on the marine safety, such as a ship navigation system, should be designed to consider the fact that the thinking faculty and operating ability of ship operators tend to be affected by their mental work-load under the pressure from navigation environment. For this purpose, it is desirable that the relation between stress degrees of operators and human errors during operating a man-machine systems is clarified by analyzing the state-transition gotten to a marine accident.This paper about this problem consists of (1) the analysis of state-transition by the stochastic process model based on Markov chains using the accident incident probability estimated by the fault tree analysis, and (2) the prediction of effectiveness to reduce the risk by taking countermeasures in specified functional system. As a practical example, the stochastic state-transition analyses and the suitable measures for preventing from “the ship collision” are investigated using the proposed method.