Data Visualisation and Machine Learning for Truck Performance Prediction and Analysis

The Challenge

The transition to battery-electric haul trucks represents an important step in decarbonising Western Australia’s mining sector. However, widespread adoption is currently constrained by uncertainties surrounding vehicle performance, reliability, and maintenance requirements. Unlike traditional diesel-powered trucks, battery-electric vehicles exhibit distinct operational characteristics, necessitating the development of new predictive tools for charge level estimation and operational planning.

Proposed Solution

Electric Power Conversions Australia (EPCA) proposes to develop a machine learning model trained on operational data from EPC’s 777D battery-electric haul trucks. The model aims to accurately predict the battery’s state of charge (SOC) within a ±5% error margin under consistent operating conditions, thereby enhancing operational efficiency and reliability.

Proposed Benefits to Western Australia

This project is supported by the MRIWA Mining Technology, Equipment and Services (METS) Innovation Program, which fosters industry-led research to stimulate the development of innovative products, companies, and markets within the Western Australian mining and minerals sector. EPCA’s project is expected to deliver a validated SOC prediction system that empowers mine operators to make informed decisions, optimise fleet performance, and maximise the value derived from each battery-electric haul truck.

Financial Assistance DOI

DOI 10.71342/430042142281