Influence of consumer attitudes and social interactions in electric vehicle purchasing: integrating agent-based modelling and machine learning

Abstract Accepted by: Prof. Aris Syntetos Understanding consumer attitudes towards electric vehicle (EV) purchasing is essential for addressing the slow adoption rate. Traditional aggregated models of EV adoption employ a top-down approach, yet often fail to capture individual-level attitudes. In contrast, agent-based modelling (ABM) enables a bottom-up approach that reflects the heterogeneity in consumer decision-making and simulates social interactions. This study introduces an integrated model to analyze consumer attitudes towards EV adoption, incorporating empirical data and synthesized social interactions through ABM. The model undergoes micro-validation and optimization through parameter variation experiments and supervised machine learning (SML) methods. Results indicate that consumer attitudes towards EV purchasing are positively influenced by early adopters and environmental factors. These attitudes are further shaped by observing EVs in residential areas and receiving positive feedback from social circles. Perceptions of EVs as an environmentally friendly alternative also significantly enhance these attitudes. These findings suggest that marketers should develop targeted strategies for specific consumer segments, and policymakers should prioritize environmental awareness campaigns to drive positive public EV attitudes in the UK. This study emphasizes the importance of incorporating consumer heterogeneity and social interactions in attitude formation, which offers insights into EV promotion within Rogers’s Diffusion of Innovations Theory.