Results of Implementing Operator Fatigue Monitoring Systems

Operating heavy mining equipment is monotonous work that requires maximum concentration throughout a 10–12 hour shift. Statistics show that up to 65% of traffic accidents at mining enterprises are caused by driver fatigue, and using a mobile phone while driving increases the risk of an accident fourfold. Solving this problem requires shifting from administrative measures to automated hardware-based monitoring.

In his presentation, Eduard Bondarev details a large-scale case study from Kuzbassrazrezugol regarding the implementation of an operator fatigue monitoring system (developed by OKO Systems). The project covered 683 haul trucks and was later expanded during the rollout phase to include passenger buses, crew transport vehicles, semi-trailer trucks, and fuel tankers. Today, approximately 3,500 drivers are under continuous monitoring, a significant portion of whom operate on public roads.

Hardware Base and IT Landscape Integration

The system operates on a non-contact optical method for tracking gaze direction. Optical sensors, an audible alarm, and a vibration motor are installed in the cabin. When signs of drowsiness or distraction are detected (e.g., eyes closed for more than 3 seconds), sound and vibration signals are activated, which turn off automatically as soon as the driver returns their gaze to the road.

An important step was integrating the solution with existing enterprise systems. Linking it with the dispatch system allowed for real-time access to waybill data (full name, employee ID), simplifying the work of field specialists. Integration with the surround-view camera system eliminated the need to mount a second monitor in the cabin—all information is now available in a single window.

Maintaining the functionality of such a fleet requires a structured technical support service. A distributed engineering team ensures that malfunctions are resolved within 96 hours, maintaining a technical readiness factor of 95%. Furthermore, a portion of service requests is resolved remotely. Occupational safety instructions strictly mandate that operating a vehicle with a non-functional monitoring system is strictly prohibited.

Discipline, Sabotage, and Accident Reduction

Round-the-clock monitoring by company dispatchers and the developer's monitoring center has yielded impressive results. Accident rates for technological transport have decreased by 50%. During the reporting period, the system prevented over 6,000 instances of falling asleep at the wheel and detected over 28,000 cases of distraction from the road.

The speaker openly discusses the issue of personnel resistance during the initial stages. The system recorded 160 cases of intentional sabotage (driver interference with equipment). All incidents were proven using video archives, and the violators were held disciplinarily accountable, which significantly improved the driving culture.

Expanding Functionality: From Gaze Tracking to Mining Analytics

Beyond the primary task of monitoring alertness, the accumulated video data (about 20 GB per vehicle monthly) has enabled the resolution of related production tasks. Cameras are used to monitor the condition of haul roads, excavator faces, body loading, and compliance with work and rest schedules.

Neural networks are used to automatically monitor the use of seat belts by passengers in crew transport vehicles and to detect instances of people moving around the cabin while the vehicle is in motion. The company plans to use the terabytes of accumulated data to train AI. Specifically, algorithms are being developed to analyze platform loading profiles and automatically detect spillage of mining mass on roads, which directly impacts the service life of large-diameter tires.

What You Will Learn from This Presentation

• How is the integration of the fatigue monitoring system with dispatch software and surround-view cameras technically implemented?
• What protocol is applied when equipment malfunctions or attempts at sabotage by drivers are detected?
• How is 24/7 monitoring of 3,500 operators of mining and auxiliary equipment organized?
• How is video data from vehicles used for incident investigation and quality control of mining operations?
• How does machine vision help monitor the use of PPE and seat belts in passenger transport?