The system of preventive maintenance for various types of equipment is considered. To identify the need for managerial influence on the processes of maintenance and repair of the product, the coefficient of variation of the maintenance (repair) cycle is proposed. This coefficient is formed by the values of operating time or work between stops for maintenance or repair. To determine its numerical value, it is recommended to use such concepts as "maintenance (repair) cycle" and "frequency of maintenance (repair)". The values of the coefficient characterizing the stability of the process of repair and maintenance of the product are determined. These values of the coefficient are proposed to be used to justify the need to analyze the causes of failures or, in other words, the instability of the repair and maintenance process. The main managerial influences on the process of maintenance and repair are determined, which can be taken as a result of the analysis of the causes of failures. These include: the decision to search for a technical service provider, expanding the competencies of employees, updating the material base. Modern industry 4.0 tools allow you to get the required level of quality when working with a large data stream. The use of neural network technology in various technical service management tasks is due to some advantages over classical methods of statistical processing of results. The experiments of various authors are presented in which it is shown that neural network technologies are able to interpret graphic information, justify its use in technical diagnostics at the stage of analyzing diagnostic information in the form of oscillograms, which allows the use of complex diagnostic equipment in the absence of experts. A neural network simulation was carried out to recognize defects in bearing assemblies. As a result, the created neural network recognized the type of defect in accordance with the matrices of input signals and the output matrix of types of defects, which makes it possible to develop a mechanism for improving technical service using neural networks as a tool for analyzing diagnostic information in the absence of experts in this field.
