Enhancing Energy Consumption in Automotive Component Manufacturing: A Hybrid Autoregressive Integrated Moving Average–Long Short-Term Memory Prediction Model

Abstract

The automotive industry faces continuing challenges with regard to advancing sustainability and reducing energy consumption and vehicle emissions. South Africa accounts for half of the total CO2 emissions in Africa and is the world’s 12th-largest CO2 emitter. In this study, we aimed to develop a model combining autoregressive integrated moving averages (ARIMAs) with long short-term memory (LSTM) to determine the best fit for prediction using the lowest root mean square error configuration and enhance energy consumption in automotive component manufacturing. The ARIMA model dissects time-series data into the components of level, trend, and seasonality, while the automatic ARIMA function refines the model parameters. Simultaneously, utilizing historical data, the LSTM model uses specific algorithms to predict future electricity generation and carbon emissions for the automotive component’s manufacturing sector. According to our results, the predicted variables’ interdependence revealed an enhancement in energy intensity for vehicle body part products equal to 29%, a cumulative energy savings of 7.22%, and an increase in energy efficiency equal to 16.25%. Our model’s predictive fitness holds significant potential for allowing automotive component manufacturers to make informed economic and technical decisions toward the development of low-carbon products. Critically, improved energy efficiency in automotive component manufacturing activities is critical for lowering energy consumption, greenhouse gas emissions, sustainable transportation, and production costs.