Green technology automotive shape design based on neural networks and support vector regression

Purpose – The green technology cars have received much attention due to the air pollution and energy crisis. The purpose of this paper is to increase automotive designers’ understanding of the affective response of consumers about automotive shape design. Consumers’ preference is mainly based on a vehicle's shape features that are traditionally manipulated by designers’ intuitive experience rather than by an effective and systematic analysis. Therefore, when encountering increasing competition in today's automotive market, enhancing car designers’ understanding of consumers’ preferences on the shape features of green technology vehicles to fulfil customers’ demands, has become a common objective for automotive makers. Design/methodology/approach – In this paper, questionnaires were first used to gather consumer evaluations of certain adjectives describing automobile shape. Then, automotive styling features were systematically examined by numerical definition-based shape representations. Finally, models were individually constructed using support vector regression (SAR), which predicted consumer's affective responses, based on the adjectives selected, and which also incorporated the relationship between consumer's affective responses and automotive styling features. Findings – In order to predict and suggest the best automotive shape design, the results of this experiment of SVR can provide a basis for the future development of automobiles, particularly for green vehicle design, and support automotive makers in ensuring that automotive shape design to satisfy consumer needs. Originality/value – SVR is a valuable choice as an evaluation method to be applied in the design field of green vehicles.