DesignQA: A Multimodal Benchmark for Evaluating Large Language Models’ Understanding of Engineering Documentation

Anna C. Doris1,*, Daniele Grandi2, Ryan Tomich3, Md Ferdous Alam1, Mohammadmehdi Ataei4, Hyunmin Cheong4, Faez Ahmed1

1Massachusetts Institute of Technology, Cambridge, MA
2Autodesk Research, San Francisco, CA
3MIT Motorsports, Cambridge, MA
4Autodesk Research, Toronto, ON, Canada

MIT Logo Autodesk Logo

DesignQA is a novel benchmark aimed at evaluating the proficiency of multimodal large language models in comprehending and applying engineering requirements in technical documentation. Developed with a focus on real-world engineering challenges, DesignQA uniquely combines multimodal data—including textual design requirements, CAD images, and engineering drawings—derived from the Formula SAE student competition.

DesignQA Overview

Extraction

Comprehension

Compliance