SUMMARY
Context. The article discusses approaches to anticipation identification being an essential part of the decision-making process done by the operator by using the example of a sea captain in ergatic systems of critical infrastructures in the sea transport management. The mentioned above aspect of anticipation of operators can be regarded as being a complex form of human-machine interaction and, certainly, claims for further elaboration of information and tools to be used. Objective. The way to approach development is taken as being based on an information analysis of the full range of trajectories of decision-making by operators at the time of performing complex multi-stage actions. These items are rooting out of their adopted strategy of human-machine interaction. Besides, it leads to the formation of a metric being able to algorithmically represent the enormous number of variants. It can be done taking into account conditions of combinatorial representation in terms of the geometric theory of groups on the Cayley graph. Method. Being a part of the approach elaboration the having been obtained during the analysis of the database of navigation simulators mathematical model of experimental data collecting and processing succeeded to be constructed. To confirm the formalalgorithmic approach a simulation was challenged to be carried out helping to form the trajectory of the operator’s decision making in critical situations. It was felicitously performed basing on the three-factor ERO-AEA-EAPI model. Thus, the algebraic and software representation of the metric decision space is noticed to uncover approximate complex human-machine interactions in uncertain environments. As a result, the converting process of data of the main subject of critical infrastructure (i.e. the operator) into knowledge is able to be coped with. In addition, factors possible to be gauged in the proposed metric are able to be uncovered. Results. In order to carry out the feasibility assessment of the developed approach as well as formal-algorithmic ones, an experiment was performed by using the Navi Trainer 5000 navigation simulator (NTPRO 5000). During having one of the most troublesome operations i.e. mooring we wanted the server data to be analyzed. As a result, data about anticipation being shaped as triangular constructs in the quasi-isometric space of Cayley graph is reported to have been obtained. The automated neural networks being used for result obtaining led to delivering of the possibility to get multiple data regression and to analyze the relationships of many independent variables. It is considered to be clear evidence due to having found out results of scattering and reliability diagrams. Conclusions. The having been presented in the investigations formal-algorithmic approach together with the developed software tools and the approaches of converting data into knowledge about operator anticipation are said to embrace the possibility to classify and to identify individual decision-making strategies when managing a vessel and to predict the likelihood of poor consequences. With regarding to the cogency of the proposed approach and models these issues happen to have been successfully justified by means of the automated processing of experimental data.