High Performance Deep Learning GPR Feature Detector Model for Potash Mining
Bibliographic record
Abstract
Ground Penetrating Radar (GPR) has been an essential nondestructive geophysical tool in Saskatchewan's potash mines for over six decades. This innovative technology, used in conjunction with an active boring machine, facilitates realtime data collection, particularly for imaging the immediate clay seam (414-clay seam) above the mine roof. Its reliability is demonstrated by the accuracy with which the roof beam thickness (mine roof to 414-clay seam) is interpreted in realtime, crucial for making informed safety decisions during mining operations. The imperative for a robust auto-picking algorithm tailored to handle complexities in potash mine GPR data is emphasized. The previously developed algorithm called the Clustered Ratio Derivative (CRD) algorithm, employing unsupervised machine learning for realtime GPR interpretation showcased promising results. However, the CRD algorithm faces limitations due to potential sensitivity to variations in input data, especially in the presence of noise or anomalies. Geological variations, such as the presence of “stray clays” within the roof beam, pose challenges to the algorithm's performance and accuracy. In response to these challenges, this paper proposes a novel deep learning-based algorithm leveraging two distinct Convolutional Neural Network (CNN) architectures. These CNNs are designed to navigate the intricacies of the GPR data pattern specific to potash mines. The presented results indicate promising levels of accuracy, with the new method achieving 95.4% accuracy in detecting the 414-clay seam and an average of 88% accuracy in finding “stray clays” seen in the dataset when compared to a geophysicist interpretation. The overall analysis suggests that this new approach has the potential to detect mining room roof features with a high degree of accuracy.
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 machine prediction
Teacher imitationNot calibrated prevalence, not ground truth. Human validation pending. The Gemma side is a direct model label for every work in the frame, read from the title-only record. The Codex side is a classifier learned from the 10,348 direct Codex labels and calibrated to design-weighted sample rates; fields without enough sample support carry no Codex call. Candidate is the union of the two sides; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels.
Distilled classifier scores by category (both heads)
| Category | Codex | Gemma |
|---|---|---|
| Metaresearch | 0.000 | 0.001 |
| Meta-epidemiology (narrow) | 0.001 | 0.000 |
| Meta-epidemiology (broad) | 0.001 | 0.001 |
| Bibliometrics | 0.000 | 0.000 |
| Science and technology studies | 0.000 | 0.000 |
| Scholarly communication | 0.000 | 0.001 |
| Open science | 0.001 | 0.001 |
| Research integrity | 0.001 | 0.001 |
| Insufficient payload (model declined to judge) | 0.002 | 0.001 |
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 source (direct Gemma or distilled Codex), 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".