Summary of A Comparative Study on Machine Learning Approaches For Rock Mass Classification Using Drilling Data, by Tom F. Hansen et al.
A comparative study on machine learning approaches for rock mass classification using drilling data
by Tom F. Hansen, Georg H. Erharter, Zhongqiang Liu, Jim Torresen
First submitted to arxiv on: 15 Mar 2024
Categories
- Main: Machine Learning (cs.LG)
- Secondary: Computer Vision and Pattern Recognition (cs.CV)
GrooveSquid.com Paper Summaries
GrooveSquid.com’s goal is to make artificial intelligence research accessible by summarizing AI papers in simpler terms. Each summary below covers the same AI paper, written at different levels of difficulty. The medium difficulty and low difficulty versions are original summaries written by GrooveSquid.com, while the high difficulty version is the paper’s original abstract. Feel free to learn from the version that suits you best!
| Summary difficulty | Written by | Summary |
|---|---|---|
| High | Paper authors | High Difficulty Summary Read the original abstract here |
| Medium | GrooveSquid.com (original content) | Medium Difficulty Summary The paper presents an innovative approach to automating the translation of Measure While Drilling (MWD) data into actionable metrics for rock engineering in drill and blast tunnelling. By leveraging a large dataset of 500,000 drillholes from 15 tunnels, the researchers introduce machine learning models that accurately classify MWD data into Q-classes and Q-values, which describe the stability of the rock mass. The paper explores both conventional machine learning and image-based deep learning approaches to achieve accurate rock mass quality classification. The results show that a K-nearest neighbours algorithm in an ensemble with tree-based models using tabular data achieves a cross-validated balanced accuracy of 0.86 for classifying rock mass into different Q-classes, while a CNN with MWD-images achieves a balanced accuracy of 0.82 for binary classification. Regressing the Q-value from tabular MWD-data also yields high performance, with a cross-validated R2 score of 0.80 and MSE score of 0.18. The study demonstrates the value of MWD data in improving rock mass classification accuracy and advancing data-driven rock engineering design. |
| Low | GrooveSquid.com (original content) | Low Difficulty Summary The paper uses special sensors to collect lots of information about the rocks while they’re being drilled into. This information can help engineers decide how to make tunnels safely. Right now, engineers mostly rely on their own observations to make these decisions. The researchers want to use computers to analyze this sensor data and turn it into useful information that engineers can use. They used a lot of data from 15 different tunnel projects and found that they could get very good results by using special computer programs. This is important because it could help make tunnels safer and reduce the amount of work engineers have to do. |
Keywords
* Artificial intelligence * Classification * Cnn * Deep learning * Machine learning * Mse * Translation
