Search for: All records

Abstract Designing for manufacturing poses significant challenges in part due to the computation bottleneck of Computer-Aided Manufacturing (CAM) simulations. Although deep learning as an alternative offers fast inference, its performance is dependently bounded by the need for abundant training data. Representation learning, particularly through pre-training, offers promise for few-shot learning, aiding in manufacturability tasks where data can be limited. This work introduces VIRL, a Volume-Informed Representation Learning approach to pre-train a 3D geometric encoder. The pretrained model is evaluated across four manufacturability indicators obtained from CAM simulations: subtractive machining (SM) time, additive manufacturing (AM) time, residual von Mises stress, and blade collisions during Laser Power Bed Fusion process. Across all case studies, the model pre-trained by VIRL shows substantial enhancements in generalizability, as measured by R2 regression results, with improved performance on limited data and superior predictive accuracy with larger datasets. Regarding deployment strategy, case-specific phenomenon exists where finetuning VIRL-pretrained models adversely affects AM tasks with limited data but benefits SM time prediction. Moreover, the efficacy of Low-rank adaptation (LoRA), which balances between probing and finetuning, is explored. LoRA shows stable performance akin to probing with limited data, while achieving a higher upper bound than probing as data size increases, without the computational costs of finetuning. Furthermore, static normalization of manufacturing indicators consistently performs well across tasks, while dynamic normalization enhances performance when a reliable task dependent input is available. 

Abstract Recent advances in computer-aided design tools have helped rapidly advance the development of wireframe DNA origami nanostructures. Specifically, automated tools now exist that can convert an input polyhedral mesh into a DNA origami nanostructure, greatly reducing the design difficulty for wireframe DNA origami nanostructures. However, one limitation of these automated tools is that they require a designer to fully conceptualize their intended nanostructure, which may be limited by their own preconceptions. Here, a generative design framework is introduced capable of generating many wireframe DNA origami nanostructures without the need for a predefined mesh. User-defined objectives that guide the generative process are input as either single- or multi-objective optimization problems. A graph grammar is used to both contextualize physical properties of the DNA nanostructure and control the types of generated design features. This framework allows a designer to explore upon and ideate among many generated nanostructures that comply with their own unique constraints. A web-based graphical user interface is provided, allowing users to compare various generated solutions side by side in an interactive environment. Overall, this work illustrates how a constrained generative design framework can be implemented as an assistive tool in exploring design-feature trade-offs of wireframe DNA nanostructures, resulting in novel wireframe nanostructures.