Factors predicting green behavior and environmental sustainability in autonomous vehicles: A deep learning-based ANN and PLS-SEM approach

Abstract

With their cost-effective performance, potential to encourage environmentally friendly behavior, and increased sustainability, autonomous vehicles (AVs) are expected to lead to significant changes in the economy, society, and the environment. This study investigates factors predicting green behavior and environmental sustainability in AVs. The study developed a research model based on the “Innovation Resistance Theory” (IRT). The proposed model was evaluated with data obtained from 1266 participants through a deep learning-based “artificial neural network” (ANN) and the “partial least squares structural equation modeling” (PLS-SEM) approach. The findings indicated a positive relationship between green behavior and environmental sustainability with AVs. A positive relationship is also found between green behavior and motivators, including environmental benefits, environmental concerns, economic benefits, and technophilia. In contrast, cost barriers, along with security and privacy concerns, negatively predict green behavior. The sensitivity analysis using the ANN approach revealed that economic benefits were the most crucial factor in predicting green behavior. These results offer important insights into understanding the key barriers and drivers predicting the acceptance of AVs. The findings contribute to stakeholders making informed decisions, developing effective strategies, and contributing to AVs' sustainable and successful integration into social life.