CADGENIUS: An AI-Based Assistant for AutoCAD

This paper surveyed the research behind CADGENIUS and its five features, and proposed a layered architecture and workflow that connects them. The review suggests a hybrid strategy: deterministic checks for correctness, learned models for flexibility, language models for orchestration and explanation, and provenance logging as a shared backbone. Future work includes building a prototype, evaluating each feature with designers on realistic parts, creating a benchmark for conversational CAD, and studying on-premise deployment for sensitive designs.

[1]   K. D. D. Willis et al., “Fusion 360 Gallery: A dataset and environment for programmatic CAD construction from human design sequences,” ACM Trans. Graph., vol. 40, no. 4, Art. 54, 2021. [2]   R. Wu, C. Xiao, and C. Zheng, “DeepCAD: A deep generative network for computer-aided design models,” in Proc. IEEE/CVF Int. Conf. Comput. Vis. (ICCV), 2021, pp. 6772-6782. [3]   A. Vaswani et al., “Attention is all you need,” in Adv. Neural Inf. Process. Syst. (NeurIPS), 2017, pp. 5998-6008. [4]   Y. Ganin, S. Bartunov, Y. Li, E. Keller, and S. Saliceti, “Computer-aided design as language,” in Adv. Neural Inf. Process. Syst. (NeurIPS), 2021. [5]   T. B. Brown et al., “Language models are few-shot learners,” in Adv. Neural Inf. Process. Syst. (NeurIPS), 2020. [6]   S. Joshi and T. C. Chang, “Graph-based heuristics for recognition of machined features from a 3D solid model,” Comput.-Aided Des., vol. 20, no. 2, pp. 58-66, 1988. [7]   P. Lewis et al., “Retrieval-augmented generation for knowledge-intensive NLP tasks,” in Adv. Neural Inf. Process. Syst. (NeurIPS), 2020. [8]   A. Seff, Y. Ovadia, W. Zhou, and R. P. Adams, “SketchGraphs: A large-scale dataset for modeling relational geometry in computer-aided design,” in ICML Workshop Object-Oriented Learn., 2020. [9]   D. Ha and D. Eck, “A neural representation of sketch drawings,” in Proc. Int. Conf. Learn. Represent. (ICLR), 2018. [10] A. Dosovitskiy et al., “An image is worth 16x16 words: Transformers for image recognition at scale,” in Proc. Int. Conf. Learn. Represent. (ICLR), 2021.