[323] AI system application practice for protection of pumps at hazardous plants
A Kostyukov and S Boychenko
SPC Dynamics, Russia
To protect machinery of hazardous productions from breakdowns and accidents, the most efficient practice would be an application of AI systems, ensuring early detection of defects and lower costs for their elimination. Modern AI-based systems provide assessment of machinery health even if the information on machinery construction and operational procedures is unavailable or insufficient, thus ensuring high reliability of the results. The paper considers the examples of practical application of AI systems for condition monitoring of centrifugal pumps at oil refineries. The data presented shows that the systems recommend the personnel to take all necessary actions at an early stage of defect development, thus preventing accidents and ensuring the lowest possible cost of maintenance services.
