Modeling the diffusion of Electric Vehicles by a robust explanation of the Bass diffusion model: An application to Pan-European Policy Analysis

Accurate prediction of the diffusion of new transportation technologies such as electric vehicles (EVs) is critical for defining policy, infrastructure planning, and anticipating impacts on energy, emissions, and mobility systems. The Bass diffusion model is widely used to forecast technology adoption, but its parameters traditionally do not explicitly account for market-specific factors driving diffusion. This research proposes a generalized Bass model incorporating the effects of policy, economic, technological, and social variables derived from expert surveys. However, limited survey sample sizes introduce uncertainty that must be addressed. We develop a novel approach using extreme value theory to robustly estimate the Bass parameters while accounting for errors from imperfect survey data. Our approach links forecasting models with market factors from multiple data sources while rigorously handling uncertainties, supporting the design, evaluation, and impact assessment of transportation policies. The methodology is applied to forecast the adoption of regional electric vehicles throughout Europe in different policy scenarios related to factors such as charging infrastructure, purchase incentives, battery costs, and environmental awareness.