Augustin Marks de Chabris

Laurentian University

• •Research Focus: AI-driven energy prediction and operational cycle detection for underground battery-electric and diesel vehicles.
• •Status: MASc→PhD conversion approved.

Laurentian University

• •Awards: 2x Ontario Graduate Scholarship (OGS), 2024–2025 and 2025–2026; Graduate Academic Excellence Award 2025; Hatch Graduate Scholarship 2025.
• •Research Focus: Developing AI models to predict fuel and energy consumption of Load-Haul-Dump (LHD) vehicles in underground mining operations.

Laurentian University

• •Graduated with “Magna Cum Laude” honours.
• •Awards: Academic Excellence Entrance Scholarship (2019–2020), Athletic Bursary (2019–2024).
• •Thesis: “Applying Natural Language Processing to Detect Depression Severity using Social Media.”

CanmetMINING (Natural Resources Canada)

• •Built duty-cycle segmentation pipeline for underground battery-electric LHD telemetry (time-series sensors); unsupervised approaches (e.g., clustering) to separate Loading / Hauling / Dumping / Transiting with explicit validation on labelled windows.
• •Developed energy-prediction models (classical ML), reporting results in kWh/shift; compared against Canmet's Energy Consumption Model (ECM) as a physics baseline to quantify difference in performance and error bounds.
• •Implemented reproducible experiments (versioned features, ablations on physics-informed vs. raw features, window sizes).

Hard-Line Solutions

• •Leveraged Generative AI and prompt engineering to accelerate application development, creating 6 internal tools in 2 months and significantly increasing deployment speed.
• •Developed a data transfer tool to migrate development team wiki pages to Confluence, enhancing collaboration and documentation efficiency.
• •Designed a PDF-to-JSON conversion program, improving the hardware team's ability to analyze and utilize data effectively.

Hard-Line Solutions

• •Increased threading software efficiency by 50% through algorithmic optimizations.
• •Explored integration of Large Language Models (LLMs) into workflows, identifying avenues for enhanced productivity.
• •Evaluated algorithms for hardware output analysis, contributing to improved decision-making processes.