Reinforcement Learning-Based Fleet Dispatching for Greenhouse Gas Emission Reduction in Open-Pit Mining Operations

Published in Resources, Conservation & Recycling, (under review), Da Huo, Yuksel Asli Sari, Ryan Kealey, Qian Zhang, 2022

Abstract: In typical mining operations, more than half of the direct greenhouse gas (GHG)emissions come from haulage fuel consumption. Conventional haul truck dispatching approaches rely on fixed scheduling based on heuristics, linear programming, or previous experience, which often causes low efficiency that wastes resources and elevates GHG emissions. Artificial intelligence (AI) advancements show promising applications in smart dispatching systems. However, few studies have evaluated the GHG reduction potential of AI-powered real-time dispatching in mining operations. This research uses reinforcement learning to improve fleet performance and reduce GHG emissions. The proposed algorithm trains the fleet to make decisions based on loading, traffic, queueing, and maintenance conditions. Simulations showed that the reinforcement learning approach could reduce GHG emissions by over 30% while achieving the same production levels as compared to fixed scheduling. The proposed method also shows advantages in handling operational randomness and balancing fleet size, productivity, and emissions for climate targets.

Download paper here