This paper presents a novel real-time dispatching algorithm using Deep Reinforcement Learning (DRL) designed to optimise autonomous haulage trucks' operations in open-pit mining. Our DRL model, simulated within an environment that accurately replicates autonomous haulage truck behaviours, accounts for vehicle interactions - an aspect typically challenging for conventional mathematical optimisation approaches. Further distinguishing this work is the model's ability to adapt to varying fleet sizes without requiring retraining the model, an improvement over other DRL approaches. Experimental results highlight the superior performance of our model against industry-standard methods, including those combining Linear Programming (LP) and heuristics, reflecting a 15% to 20% increase in transportation efficiency and a decrease in fuel consumption. These advancements position us closer to environmentally friendly mining operations. This study offers a methodological advancement for the mining industry and other sectors employing centralised autonomous vehicle systems, with potential for future applications in different transportation contexts.


    Access

    Check access

    Check availability in my library

    Order at Subito €


    Export, share and cite



    Title :

    Real-time Dispatching for Autonomous Vehicles in Open-pit Mining Deployments using Deep Reinforcement Learning


    Contributors:


    Publication date :

    2023-09-24


    Size :

    619667 byte





    Type of media :

    Conference paper


    Type of material :

    Electronic Resource


    Language :

    English




    Real-time dispatching of delivery vehicles

    Minkoff, Alan Seth | DSpace@MIT | 1985

    Free access

    Dynamic shared autonomous vehicles dispatching algorithm

    Rong, Shaoling / Tang, Chongwu / Chen, Min | VDE-Verlag | 2022



    Tackling Real-World Autonomous Driving using Deep Reinforcement Learning

    Maramotti, Paolo / Capasso, Alessandro Paolo / Bacchiani, Giulio et al. | IEEE | 2022