A Radically New Approach to Safety
New technologies have made load-haul-dump machine operations safer at the Severny mine, operated by Nornickel subsidiary Kola Mining and Metallurgical Company. The process is now monitored through a specialized control system.
The deployment is particularly notable in the context of earlier projects at the same mine. Previously, the site introduced the “Antinaezd” system, which detects approaching workers and warns the operator of a load-haul-dump machine (LHD). The new system expands that safety framework: instead of simply warning about proximity between workers and equipment, it goes further by automatically detecting specific safety violations.
One of the most common and dangerous violations involves transporting workers in vehicles not intended for passenger use – for example, inside an LHD bucket. When such a violation is detected, the system immediately sends an alert to the mine’s dispatch center. The project is designed to strengthen industrial safety culture at mining operations, where human error remains one of the leading causes of hazardous incidents.
Underground mining, heavy equipment operation, limited visibility and blind spots are universal challenges across the mining industry. Experts believe that if the system proves both effective and scalable, the technology could see broad adoption across Russian mining companies. The first large-scale deployments, however, are expected inside Nornickel’s own operations in Norilsk.
Understanding the Challenge
The project clearly demonstrates the transition from traditional video surveillance and safety briefings toward intelligent industrial monitoring systems in which AI identifies risks in real time.
As a result, industrial safety, machine vision, edge computing – where data is processed as close as possible to the point of generation rather than inside centralized cloud data centers – and predictive analytics are emerging as a standalone category of import-independent IT solutions for the mining industry.
Kola MMC is one of the Murmansk region’s key industrial assets. The company, which serves as the primary employer for three settlements, has a workforce of around 16,000 people. Industrial safety has therefore long remained a top priority. Its main mining asset, the Severny mine, supplies ore to the concentrator plant in Zapolyarny.
Since 2019, the company has operated the AS KUB industrial safety and occupational health management system, along with an electronic medical screening platform, the “Flotation Assistant” data analysis software for concentrator operations and computer vision systems used to monitor personal protective equipment compliance. The company is also known for deploying drones to inspect underground workings at the Severny mine.
A Combined IT Team
The project was developed by specialists from Nornickel’s IT department with support from the Severny mine’s automation division, the self-propelled equipment maintenance department and Kola MMC’s industrial safety division.
The pilot version of the system was developed within a compressed timeline while complying with all corporate requirements and methodologies. Following testing, the company decided to move ahead with a full-scale IT deployment. Importantly, the new system uses cameras that were already installed as part of the mine’s video recording infrastructure.
The project’s main technical challenge involved processing video streams directly onboard the mining equipment itself. To address that issue, engineers used onboard computing units – servers adapted for operation in confined spaces, including directly inside the LHD cab. Another major issue involved transmitting data to the surface because not all areas of the mine have communications coverage. In such cases, the data is stored until the system detects a stable signal, after which the information is uploaded sequentially to system servers located in the site’s administrative complex.
A mannequin was used to train the AI system. After the video analytics model was created, the algorithm was tested under multiple operating conditions, including poor lighting, heavy dust and vehicle movement. Full-scale deployment began in March. This type of applied AI – designed not only to improve productivity, but also to protect workers’ lives – could become one of the most in-demand areas of industrial IT development in Russia.
