Описание проекта
Steel Belt Monitoring (SBM) is an advanced machine vision system implemented at the sinter plant №1 (UPO-1) of Donskoy GOK. The system is designed for monitoring the condition of the steel belt in sintering furnaces and improving production efficiency through:
Overheating prediction: Early detection of overheating risks with an accuracy of up to 93.94%, which helps prevent emergency situations. Defect detection: Identifying damage to the belt at early stages with an accuracy of 79.72%, reducing repair costs and preventing unplanned downtimes. The system is based on neural networks and machine learning algorithms integrated with the plant's technological processes. It provides operators with recommendations for optimizing furnace parameters, reducing energy consumption, and minimizing human error. SBM combines innovative technologies with the real needs of heavy industry.
Технологии, использованные в проекте:
Ultra-preciseneural networks for image analysis.
Machine learning for overheating prediction.
TensorFlow and OpenCV platforms.
Integration with Industrial Process Control Systems (IPCS).
Стадия проекта
Working solution
Рынки и сферы применения
Metallurgy: Monitoring the condition of steel strips and improving the efficiency of roasting kilns.
Iron ore pellet production: Predicting overheating and detecting defects during the sintering process.
Heavy industry: Adapting for monitoring industrial equipment using machine vision.
Construction materials production: Implementing monitoring technologies in similar high-temperature processes.
Ключевые достижения
Reduction of emergency downtime by 93.94%.
Increase in defect detection accuracy to 79.72%.
Reduction of false alarms by nearly 4 times, significantly improving operator efficiency.
Direct economic benefit of $327,331 in the first three months of operation.
Измеримые результаты
Economic impact: Due to the reduction in downtime and emergency repairs, ferrochrome production increased. The calculation based on the difference between 2023 (before system implementation) and 2024 (after implementation) showed a significant reduction in losses.
Social and environmental impact: Improved employee safety through automated monitoring, and reduced energy consumption due to optimized operational parameters of roasting kilns.
Уникальность проекта
Steel Belt Monitoring (SBM) is the first system in Kazakhstan that integrates machine vision with real-world manufacturing processes. The system not only automates monitoring but also provides predictive maintenance, minimizing the risks of emergency situations and reducing financial losses. SBM sets a new standard for industrial automation, enabling businesses to effectively leverage the capabilities of artificial intelligence.
Планы на будущее
Expansion of the system to other sections of Donskoy GOK and ERG enterprises.
Development of new modules for integration with ERP and predictive maintenance systems.
Scaling to international markets for application in metallurgy, construction, and other industries.
Партнеры или инвесторы
DGOK, BTS, ERG.