A solution for automated monitoring and diagnostics of photovoltaic modules of industrial solar power plants is proposed. The solution is based on the use of an unmanned aerial vehicle with a specialized payload and a ground-based intelligent information and control system to detect problem areas of the station, in particular partial shading and pollution. To perform the detection procedures, a neural network based on the Fast R-CNN architecture with the learning algorithm – Inception v2 (COCO) was used. The results of preliminary tests showed that the accuracy of detecting problem areas is at least 92%. The article presents a mathematical model that allows calculating the installed power monitored by the complex, depending on the type of station and UAV, meteorological parameters, and the performance of computing equipment. Numerical calculations have shown that when using the FIMI X8SE UAV and a computing device based on the RTX2080 GPU, the installed monitored power will be up to 7.5 MW.

Advances in diagnostics and monitoring using aerial surveillance drones can provide a clear overview of the operational status of solar arrays. The authors have developed an automated system for solar power plant intelligent monitoring and maintenance. The mentioned system, which is intended for detecting shades and dust covering cells, consists of a drone and intelligent control system situated on the ground. Preliminary test results have shown that accuracy of faulty parts is 92 % if the weather is clear. In order to assess efficiency of the developed system, the authors have built a mathematical model for counting all contributing factors like a solar power plant’s established capacity, UAV type, computational performance, and weather conditions.

One of the most important conditions for the efficient operation of solar power plants with a large installed capacity is to ensure the systematic monitoring of the surface condition of the photovoltaic modules. This procedure is aimed at the timely detection of external damage to the modules, as well as their partial shading. The implementation of these measures solely through visual inspection by the maintenance personnel of the power plant requires significant labor intensity due to the large areas of the generation fields and the operating conditions. Authors propose an approach aimed at increasing the energy efficiency of high-power solar power plants by automating the inspection procedures of the surfaces of photovoltaic modules. The solution is based on the use of an unmanned aerial vehicle with a payload capable of video and geospatial data recording. To perform the procedures for detecting problem modules, it is proposed to use “object-detection” technology, which uses neural network classification methods characterized by high adaptability to various image parameters. The results of testing the technology showed that the use of a neural network based on the R-CNN architecture with the learning algorithm – Inception v2 (COCO)—allows detecting problematic photovoltaic modules with an accuracy of more than 95% on a clear day.

Agriculture is considered to be one of the most important sectors of the economy in different countries. Presently, this industry, is witnessing a transition to mass digitalization of business processes, which makes it possible to effectively implement elements of strategic development and proactive management. This problem solving requires the creation of advanced monitoring systems for agriculture. The authors have developed an automated complex for monitoring and diagnostics of vineyards based on the use of an unmanned aerial vehicle (UAV) and specialized software. The proposed solution makes it possible to assess the phytosanitary condition of the vineyard using the procedures of neural network classification of grape diseases based on leaves images. To perform the detection procedures, a neural network based on the Fast R-CNN architecture with the InceptionV2 learning algorithm was used. Preliminary testing results of the technology effectiveness have demonstrated that the accuracy of infected leaves is at least 91%, while using a training sample containing 2,500 images of both healthy and damaged leaves. The developed mathematical model showed that the complex is capable of monitoring up to 2.5 hectares of vineyard during daylight hours.

Relevance. Proactive management of the processes of effective realization of the varietal potential of grapes is associated with the need to introduce innovative digital technologies for automated monitoring of heterogeneous data sources characterizing agro-climatic conditions and degradation processes of the biological state of plants. Currently, there is a steady trend aimed at digitalization of the viticulture and winemaking industry. There is a whole complex of scientific, practical, technical, technological tasks associated with the introduction of digital technologies for collecting the necessary information, aggregating it and creating a pre-processing technique for implementing procedures for multifactorial data analysis with their further use in decision support systems. The solution of the above-described tasks of a systemic nature requires the creation of scientific and methodological foundations for the implementation of intelligent adaptive automated monitoring of various objects and processes of agricultural enterprises.Methods. The above technology is based on the complex use of methods of technical vision, neural network classification and detection of grape leaves, evaluation of the quality of training neural network algorithms, video recording methods when using unmanned aerial vehicles (UAVs).Results. The results of the development of information technology for automated neural network detection of signs of deterioration of grape plantations for proactive management of the processes of effective realization of the varietal potential of grapes are presented. The technology allows the vineyard service personnel to promptly receive information about signs of deterioration of the condition of grape plantations based on video recording data of grape plants obtained using UAVs in static and dynamic mode. The results of testing the accuracy of detecting affected leaves showed that the mAP value of the trained neural network is at least 91%, which is sufficient to identify problem areas.