Problem Statement

Active open-pit mines lack a continuous, integrated, and explainable system that predicts high-risk highwall sectors, detects rockfall in real time, and issues location-specific, actionable alerts to protect personnel and equipment. Current approaches rely on periodic surveys, fragmented sensors, or single-technology detection, which cannot reliably answer three critical questions at the moment they matter: (1) Which wall sector is becoming unstable right now? (2) Has a rockfall actually started, and where? (3) What immediate action should the mine take—inspect, restrict access, or close a zone? Without such a system, mines continue to operate with blind spots in their safety monitoring, exposing workers and assets to preventable risk and incurring avoidable production losses.

Solution

RockSentinel AI is an integrated, low-cost early-warning platform that fuses DEMs, drone/LiDAR surveys, geotechnical sensors, weather and blast data, and live Doppler radar plus overlapping cameras to forecast high-risk highwall sectors and confirm active rockfall in real time. An explainable AI ris

In Detail

RockSentinel AI is an integrated, AI-powered rockfall prediction and early-warning system for open-pit mines. It combines Digital Elevation Models (DEMs), periodic drone/LiDAR surveys, overlapping fixed cameras, Doppler radar, geotechnical sensors, weather data, and blast/vibration information to assess highwall stability continuously. The system analyses parameters such as wall geometry, displacement, strain, pore pressure, rainfall, temperature, wind, vibrations, blast timing, historical incidents, and real-time visual or radar movement. Its AI risk engine identifies patterns associated with increasing instability and assigns each mine sector a probability-based risk level, such as low, moderate, high, or critical. Doppler radar and overlapping cameras provide real-time detection and confirmation of falling material. The radar identifies moving debris and estimates its location, velocity, trajectory, and runout, while cameras provide visual verification and event evidence. Blast schedules and vibration data help distinguish planned blasting from unexpected rockfall or abnormal post-blast movement. Periodic drone/LiDAR surveys update the mine’s 3D surface model, measure unexpected changes in wall geometry, identify newly detached material, estimate rockfall volume, and improve runout-zone mapping. Planned excavation, blasting, stockpiling, and bench advancement are recorded so that intentional changes are not incorrectly classified as slope failure. The platform presents all information through a user-friendly web or mobile dashboard. It displays a live 3D risk map showing vulnerable highwall sectors, sensor coverage, nearby haul roads, equipment, personnel zones, and potential runout paths. When risk increases or a rockfall is detected, the system sends alerts through the dashboard, SMS, email, or mine-control-room communication systems. Each alert includes the affected sector, confidence level, triggering evidence, estimated danger zone, and recommended action, such as increased inspection, restricting access, stopping haul-truck movement, closing a haul road, or notifying the geotechnical team. RockSentinel AI is designed as a decision-support system rather than a replacement for qualified geotechnical engineers. Its main objective is to reduce monitoring blind spots, improve response time, minimise false alarms through multi-sensor verification, and enable safer, more resilient, and cost-effective mining operations.