Применение видеоаналитики для мониторинга и повышения безопасности технологических процессов

Abstract

Представлены результаты применения видеоаналитики, основанной на алгоритмах искусственного интеллекта и компьютерного зрения, для мониторинга нештатных ситуаций и повышения безопасности технологических процессов. Предложены инновационные подходы к анализу видеоданных, позволяющие в режиме реального времени отслеживать соблюдение правил безопасности при выполнении работ, выявлять нестандартные события и предотвращать потенциальные аварийные ситуации. Сформулированы рекомендации по использованию результатов исследований при развитии безопасных технологий и производств

In conditions of the rapid development of industry and the complexification of production processes, traditional methods of safety control become inefficient. Therefore, studies in the field of the use of computer vision technologies to control compliance with the industrial safety requirements for production facilities are critical for any industry and field of human activities. Results of the use of video analytics based on algorithms of artificial intelligence and computer vision to monitor contingencies and improve the safety of workflows have been stipulated. The analysis of the current solutions in the sphere of video analytics and their integration with production process management systems have been provided. Innovative approaches to video data analysis enabling control the compliance with safety rules, detecting non-standard events, and preventing potential accidents in real-time mode have been proposed. An extended description of the data collection preparatory process for model training and verification of its accuracy has been provided. The scientific novelty of the study is the development of a system able to identify violations occurring in conditions of a complex workflow and disturbing external factors, for example, limited visibility within the monitoring area or multiple objects within the registration area potentially causing false alarm system activations, with high accuracy. This approach helps to significantly expand the scope of application of the system for production facilities. The efficiency of the use of the developed methods for real production facilities has been demonstrated. The recommendations on the use of the study results for the development of safe technologies and productions have been formulated