Hazard Recognition App for Underground Mines
- Principal Investigator (PI): Angelina Anani, PhD, Mining and Geological Engineering (MGE).
- Co-PI's: Nathalie Risso, PhD, MGE; Edward Wellman, MGE.
- Research Assistants: Pedro Lopez; Carolina Gamez Gonzales
The goal of the Hazard Recognition in Underground Mines application or “HUMApp” is to improve safety by automatically detecting structural hazards in underground mines. Developing the app involved creating a labeled images dataset for underground mines hazards, an ML (Machine Learning) computer vision-based model to identify geotechnical hazards, and a prototype for a mobile app for real-time prediction.
Completed tasks include data collection, labeling, training, and app demonstrations. Next steps include increasing the dataset, recording a training video, and writing a literature review paper. The app has been demonstrated at the annual Society for Mining, Metallurgy, and Engineering (SME) conferences in the U.S. and at the 2023 Canadian Institute of Mining, Metallurgy, and Petroleum (CIM) convention.