There is a high risk of accidents onboard ships despite the implementation of technologically advanced subsystems whose basic function is protecting people and reducing material damage. Research of human error analysis indicate that human factor substitution with program agents can achieve significant improvements in decision-making processes in distress ship situations. This research will carry out a functional analysis of ship's distress situations and analysis of available agent technologies. By means of analyses input indicators will be identified according to which appropriate types of program agents will be selected. According to the preselected agent technology, a decision model will be developed in the selected class of distress ship's situations.