DEVELOPMENT OF A METHODOLOGY FOR THE EVALUTATION OF UAV-BASED PHOTOGRAMMETRY: IMPLEMENTATION AT AN UNDERGROUND MINE
Bibliographic record
Abstract
Autonomous systems in underground mining are increasingly being implemented as tools to collect data in inaccessible areas and improve the safety of mine personnel. There are many areas in the underground mining environment that cannot be accessed by personnel due to the high potential for ground fall and insufficient ground support. By combining unmanned aerial vehicles (UAVs) with technologies such as photogrammetry and LiDAR (Light Detection and Ranging) scanners, 3D point clouds can be created for inaccessible sites. A 3D digital point cloud can provide valuable geotechnical information such as the ability to measure discontinuities, inspect rock conditions, generate accurate volume estimates, and obtain a georeferenced geometry of the inaccessible opening. There are many challenges to operating UAVs and collecting high-quality imagery in underground environments including poor lighting and visibility, dust, water, confined spaces, air turbulence, and a lack of GPS coverage for navigation and stability. Due to the difficult flying conditions and GPS-denied environment, several companies are developing UAVs with LiDARbased simultaneous localization and mapping (SLAM) to enhance the obstacle detection and avoidance capabilities of the platforms and minimize the potential for a collision. The objective of this research was to develop a methodology that can be used to evaluate UAV-based imaging tools designed to fly in underground environments. A series of demonstrations was designed to test the functionalities of available UAVs and to identify the most effective platforms for collecting UAV-based photogrammetric imagery in an underground mine. Each of the four participating teams was challenged to fly their UAV-based systems (Hovermap, Elios, M2, Ranger/Batonomous) in underground drifts and long-hole stopes while capturing high-quality imagery that could be used to create a 3D digital photogrammetric model of the opening. The demonstrations were held at Barrick Gold Corporation’s Golden Sunlight Mine (GSM) in Whitehall, MT. The systems were evaluated based upon the performance of the collision avoidance (or recovery) system in the underground environment and the quality and accuracy of the data provided. By successfully completing the underground flights and demonstrating well-developed SLAM-based collision avoidance, the Hovermap system proved to be the most reliable, robust, and easily controllable system. The Elios system, relying on collision recovery rather than avoidance, is an affordable alternative for flights in difficult environments. The imagery collected by each system was used to generate photogrammetric point clouds using three software packages: Agisoft PhotoScan, Bentley ContextCapture, and Pix4Dmapper. The point clouds were qualitatively compared based on completeness and detail and quantitatively evaluated for accuracy by comparing the geometry of the point cloud to LiDAR scans of the stopes. Based on the results of the qualitative comparison, the point clouds considered in the accuracy evaluation were built using the photogrammetry software Bentley ContextCapture. When the photogrammetric point clouds were compared with the LiDAR point clouds (assumed to be an accurate baseline reference), the mean error values ranged between 0.47 and 2.86 feet. Despite the different conditions and locations in which the imagery was collected for each model, the observed error varies by less than one order of magnitude. Improvements in the coverage and overlap of the imagery as well as in the method used for georeferencing could further increase the accuracy of the photogrammetric point clouds.
Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.
How this classification was reachedexpand
Full frame distilled prediction
Teacher imitationNot calibrated prevalence, not ground truth. Human validation pending. Learned from the 10,348 direct Codex labels and 10,348 direct Gemma labels. Candidate is the union of thresholded teacher heads; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels or direct frontier model labels.
Codex and Gemma teacher scores by category
| Category | Codex | Gemma |
|---|---|---|
| Metaresearch | 0.001 | 0.000 |
| Meta-epidemiology (narrow) | 0.000 | 0.000 |
| Meta-epidemiology (broad) | 0.000 | 0.000 |
| Bibliometrics | 0.000 | 0.001 |
| Science and technology studies | 0.000 | 0.000 |
| Scholarly communication | 0.000 | 0.000 |
| Open science | 0.001 | 0.000 |
| Research integrity | 0.000 | 0.000 |
| Insufficient payload (model declined to judge) | 0.000 | 0.000 |
Machine scores (provisional)
The two teacher heads of the student model, read on this work. A score orders the frame for review; it never asserts a category, and the validation status ships verbatim with every row.
Baseline scores from an immature model (maturity gate not passed, 7 training rounds). Scores rank; they never assert a category.
score_only:v0-immature-baseline · verbatim from the scoring run: score_only means the number may rank works, and no category label ships from itClassification
machine, unvalidatedMachine predicted; a candidate call from one teacher head, not a consensus.
How this classification was reached, model by model and score by score, is at the end of the page under "How this classification was reached".