The oil&gas enterprises have their own safety concerns list. Poor maintenance of equipment is one of the top safety concerns for many enterprises. There are a lot of methods of increasing of safety, but its shouldn’t be used alone. A complex approach and addition the modern technologies in safety ensuring make possible not only increase the quality but increase the efficiently of production. There are a lot of kinds of quality improving of maintenance of electromechanical equipment, one of them are the augmented reality system. This is a new technology and there are no methods and effective estimation for developing and using this technology in industrial production. Developing the special unified modules of augmented reality for maintenance of electromechanical equipment is the actual task. The aim of this research-to identify possible uses, to estimate the effectiveness in ensuring the safety of production and to develop a way of integrating into typical process control systems of the augmented reality system for maintenance of electromechanical equipment.

The paper presents a methodology for training neural networks for vision tasks on synthe-sized data on the example of steel defect recognition in automated production control systems. The article describes the process of dataset procedural generation of steel slab defects with a symmetrical distribution. The results of training two neural networks Unet and Xception on a generated data grid and testing them on real data are presented. The performance of these neural networks was assessed using real data from the Severstal: Steel Defect Detection set. In both cases, the neural networks showed good results in the classification and segmentation of surface defects of steel workpieces in the image. Dice score on synthetic data reaches 0.62, and accuracy—0.81.

The article discusses the structure of autonomous electrical systems with an electric drive and an electric energy generation system. An approach is described for assessing the overall efficiency of technological cycle equipment with specification of parameters that must be considered when calculating it. The necessity of taking into account energy efficiency and operability indicators of autonomous electric complexes is substantiated. The effectiveness of the use of autonomous systems decreases throughout the entire life cycle due to wear and tear. As the operability of electromechanical equipment, an indicator of the residual resource is selected. The calculation of the residual resource, with the definition of its boundary values and recommendations for further use. Nominal energy efficiency parameters of autonomous electrotechnical complexes with electric drive deteriorate during operation, which also needs to be taken into account. To simplify the assessment of the operational state of electromechanical equipment, a state diagram is presented that allows you to track the dynamics of degradation of individual nodes. Taking into account the described indicators and means of control, it is possible to increase the efficiency of using autonomous complexes.