Customized Car Seat Design with GAN-Based Generative Models and Random Forests for Comfort Evaluation

Abstract

An automobile seat is an essential part that reduces vibration in addition to provide comfort and restraint to its passenger. Automotive seat designs place a high value on dynamic comfort since the seat is in continuous touch with the passenger of the vehicle. Personalized comfort and ergonomic adjustability are becoming crucial differentiators in contemporary automobile design. In order to optimize automobile seat design utilizing 3D printing technology, this research suggests an intelligent framework that blends Random Forests (RF) and Generative Adversarial Networks (GAN) using python. Data were preprocessed using min-max normalization and then reduced using PCA for feature extraction. GAN created unique seat designs, while RF classified and adjusted comfort levels. The GAN-RF method performs better, according to quantitative data, with 97.5% precision, 97.8%recall, 98.0% accuracy and an F1-score of 97.6% in comfort prediction. The framework integrates anthropometric data collection, computational ergonomics, and additive manufacturing to achieve user-specific seat geometries and material distributions. By employing multi-objective optimization, vibration absorption, pressure distribution, and long-duration comfort are simultaneously enhanced. Prototypes fabricated using 3D-printed lattice and gradient structures demonstrate superior adaptability compared to conventional foam-based seats. The proposed approach not only improves passenger well-being and reduces fatigue during extended driving but also establishes a pathway toward sustainable, on-demand, and modular seat production in the automotive industry.