Patricia Wollstadt, Mariusz Bujny, Satchit Ramnath, Jami Shah, Duane Detwiler, Stefan Menzel,
"CarHoods10k: An Industry-grade Data Set for Representation Learning and Design Optimization in Engineering Applications",
IEEE Transactions on Evolutionary Computation Special Issue on Benchmarking Sampling-Based Optimization Heuristics: Methodology and Software (BENCH), 2022.
Research and development of cutting-edge optimization frameworks that exploit the advances in novel machine learning methods is dependent on the availability of large data sets resembling the targeted application. Especially in the engineering domain such high quality data sets are rare due to confidentiality concerns and generation costs, be it computational or manual efforts. Here, we introduce the OSU-Honda Automobile Hood Dataset (CarHoods10k), an industry-grade 3D vehicle hood dataset of over 10000 shapes along with mechanical performance data that were validated against real-world hood designs by industry experts. CarHoods10k offers researchers and practitioners the unique opportunity to develop novel methods on realistic data with relevance to real-world vehicle design. Illustrating central use cases, in this paper, we, first, apply methods from geometric deep learning to learn a compact latent representation for design space exploration. Second, we use machine learning models to predict mechanical hood performance from the learned latent representation, thus demonstrating the effectiveness of machine learning for building metamodels, which are used in design optimization whenever possible to replace costly engineering simulations. Third, we integrate CarHoods10k in a topology optimization approach based on evolutionary algorithms to demonstrate its capability to search for high-performing structures, while maintaining manufacturability constraints.
Download Bibtex file