18:58, 22 March 2026

AI Risk-Based Systems Deployed in ALROSA Mines

Executive outlines how artificial intelligence and computer vision are being used to improve safety across Siberia’s mining operations

Artificial intelligence and computer vision are no longer buzzwords but practical tools that help save lives in industrial settings, according to Sergey Sementsov, head of the transport safety monitoring center at ALROSA. In an interview with Sib.fm, he described how digital technologies are used to combat driver fatigue in haul trucks, prevent incidents in mines, and support the transition away from foreign software.

Predictive Maintenance and “Antison”

“We devote most of our time to developing new technologies and integrating them into production processes. The main goal is to reduce injuries, prevent traffic accidents, and avoid workplace incidents,” Sementsov said.

He described the approach as risk-based. Algorithms are used to predict equipment failures and the likelihood of traffic accidents. These models feed into dashboards that highlight vulnerabilities in real time, allowing teams to take preventive action before incidents occur.

“We have completed a major project to monitor driver fatigue detection systems. All of our mining equipment, as well as transport infrastructure used for passengers and hazardous cargo, is equipped with these systems. They operate independently of internet or cellular connectivity. When the system detects an abnormal condition in a worker, it provides prompts that help us prevent accidents,” Sementsov said.

Cameras in Helmets

As part of its broader development strategy, the company is implementing a large-scale digital transformation project in underground mines. This includes video recorders embedded in miners’ protective helmets. The devices help monitor the use of personal protective equipment, detect support lagging that does not meet mining standards, identify rock fractures, and flag the absence or misuse of tools during underground work. Neural networks also verify workers’ protective gear before they descend into the mine.

“Today, many companies have stepped up their efforts. We regularly receive proposals to replace segments that were previously managed by foreign vendors. We have already replaced part of our software, with additional transitions planned for 2026–2028. Overall, our digital transformation and localization roadmap extends through 2030. We are working to replace everything with domestic software,” Sementsov said.

Earlier, it was reported that AI is being used to monitor the loading of mining trucks at a diamond deposit in Russia.

Extractive industry