1. Open-Data-Driven Smart Mine Haulage System for Northern Mining Operations
Mine haulage is one of the most critical and costly components of surface mining, but smart transportation research often depends on confidential fleet-management data. This project will develop an open-data-driven framework for weather-aware mine haulage analysis in northern Canada. Two undergraduate interns will build a reproducible prototype that integrates public geospatial, topographic, remote-sensing, and climate datasets to assess haulage difficulty and productivity risk. The case studies will focus on Canadian surface mining regions, such as the Fort McMurray oil sands area and selected Québec mining districts. Mine locations will be obtained from Natural Resources Canada’s "Principal Mineral Areas, Producing Mines, and Oil and Gas Fields" dataset. Alberta oil-sands spatial layers will be collected from Alberta Energy Regulator shapefiles, and Québec mining/geoscience information will be accessed through SIGÉOM. Road networks will be extracted from OpenStreetMap and processed as haulage-route graphs. Terrain variables, including elevation change, average slope, and ruggedness, will be derived from a Digital Elevation Model (DEM), such as Copernicus DEM GLO-30 or GLO-90 products. Sentinel-2 imagery will be used to inspect land disturbance, exposed ground, snow cover, and surface context around selected mine areas. Weather variables, including temperature, snowfall, snow on ground, precipitation, wind speed, visibility, and wind chill, will be obtained from Environment and Climate Change Canada historical climate stations and, where station coverage is limited, complemented by NASA POWER or Open-Meteo/ERA5-Land reanalysis data. The project will produce two connected outputs: an interpretable Haulage Difficulty Index based on terrain, route, and weather factors, and a synthetic haulage productivity-risk dataset for machine-learning experiments. Final deliverables will include Python scripts, GIS layers, maps, a small dashboard, a technical report, and a poster suitable for undergraduate research presentation.
Research area, student roles & skills
Research area: Mine haulage is a major cost driver at mine sites, yet research is often constrained by limited access to proprietary fleet data. This project develops an open-data-driven framework to assess haulage difficulty and productivity risks in northern mining operations. The intern will integrate geospatial, topographic, remote sensing, and weather datasets to characterize terrain, road conditions, and seasonal climate impacts. A haulage difficulty index will be developed and used to generate a synthetic dataset for interpretable machine learning models. Outcomes include an open-source data pipeline, a geospatial dashboard prototype, and a Canadian mining case study supporting smart, sustainable, and weather-resilient mine
Student roles:
Two interns will work on complementary subtopics within the same project.
Student 1 will lead "Open Geospatial Haulage Difficulty Mapping." This student will focus on data acquisition, GIS processing, and spatial analysis. The main tasks include collecting mine-location data, extracting road networks, processing DEM and Sentinel-2 data, integrating historical weather records, and calculating route length, slope, elevation gain, terrain ruggedness, snow exposure, and cold-temperature indicators. Student 1 will develop the Haulage Difficulty Index and produce maps showing spatial and seasonal variations in haulage difficulty. Student 2 will lead "Synthetic Haulage Productivity-Risk Modelling." This student will use the features generated by Student 1 to construct a synthetic haulage dataset. The tasks include defining realistic truck-haulage scenarios, generating productivity-risk labels, training interpretable machine-learning models, comparing weather and terrain effects, and identifying the most influential factors using feature importance or SHAP analysis. The two students will collaborate closely to ensure that the geospatial features, synthetic data, models, maps, and dashboard are connected within one reproducible workflow.
Week 1: literature review, software setup, and case-study selection. Week 2: collect NRCan, AER/SIGÉOM, OpenStreetMap, DEM, Sentinel-2, and climate data. Week 3: clean spatial data and define mine-centred study areas. Week 4: extract road networks and build route graphs. Week 5: derive terrain and road features from DEM. Week 6: process weather, snow, and cold-temperature indicators. Week 7: Student 1 constructs the Haulage Difficulty Index, while Student 2 defines synthetic haulage assumptions and variables. Week 8: generate synthetic productivity-risk records and validate feature ranges using literature and engineering judgement. Week 9: train baseline models, such as random forest or gradient boosting, and compute model interpretability metrics. Week 10: compare mining sites, seasons, and weather scenarios. Week 11: develop dashboard figures, maps, and technical documentation. Week 12: finalize code, report, poster, presentation, and future industry-data validation plan.
Skills required:
The student should have a background in mining engineering, civil/geological engineering, computer engineering, data science, geomatics, or a related field. Basic programming experience in Python is preferred, especially using pandas, NumPy, matplotlib, or scikit-learn. Familiarity with GIS, QGIS, remote sensing, statistics, or machine learning would be an asset but is not mandatory. The student should be comfortable learning open-source tools, processing spatial datasets, reading technical documentation, and working independently. Prior experience with mine haulage systems, truck-shovel operations, road design, or weather data would be helpful, but the project is designed to be feasible without proprietary mining datasets.