Supplemental Funding: Design and Demonstration of Multi-agent System for Underground Mine Rescue Missions

Abstract Pillar stability is essential for the safety of underground mining operations, especially during emergencies such as seismic events, rockbursts, or roof falls. This project will develop a rapid-response assessment system using a drone equipped with a multi-return LiDAR sensor and a high-resolution 360-degree camera to capture detailed three- dimensional geometry and panoramic surface imagery of pillars without exposing personnel to hazardous areas. The collected LiDAR and imagery data will be processed to create a colorized 3D model, from which image processing and geometric analysis will detect cracks, spalling, deformation, and roof/?oor convergence. When baseline scans are available, change detection will quantify displacement and material loss. The results will be summarized in a Pillar Stability Index, supported by visual overlays, heatmaps, and automated reports. Field validation in underground experimental mine environments will demonstrate the system’s ability to provide fast, accurate, and non-contact stability assessments to support safer, more informed decision-making in emergencies.