Generative AI-driven framework for estimating future electric vehicle usage

In the rapidly advancing field of electric vehicle (EV) adoption and infrastructure planning, the scarcity of detailed and comprehensive datasets poses significant challenges for effective decision-making. This study introduces a cutting-edge framework leveraging generative AI, specifically, Sequential Generative Adversarial Networks, to synthesize realistic survey and travel sequence data. The framework’s Data Fusion module integrates socio-demographic attributes with travel behaviors, creating enriched synthetic datasets that capture multi-dimensional insights. Using Indiana as a case study, the research demonstrates key applications, including future EV adoption projections and charging demand estimation for residential and public stations. Results indicate an 18-fold increase in EVs by 2031 under optimistic scenarios, with existing charging stations meeting less than 50 % of demand. By addressing critical data gaps, this generative AI-driven approach provides actionable insights for strategic EV infrastructure development, enabling informed policy-making and promoting sustainable mobility solutions.