Linkify: Learning from Interface-Augmented Assembly Graphs
2026-07-01 • Computer Vision and Pattern Recognition
Computer Vision and Pattern Recognition
AI summaryⓘ
The authors created Linkify, a system that helps find the right parts for mechanical assemblies by focusing on the connections between parts, which are important for how they work together. They improved an existing dataset by fixing the details of where parts touch and used 3D point clouds to represent these contact areas. Their method uses a special graph model that considers both part shapes and their contact regions to predict missing parts better than simpler methods. They show that understanding the exact interfaces and using attention in their network is key to making accurate predictions. They also shared their improved data and tools for others to build on.
assembly graphinterface geometrypoint cloudGraph Attention NetworkGATv2Fusion 360 Gallery Assembly datasetmasked part predictioncontact computationpart retrievalgenerative design
Authors
Anushrut Jignasu, Daniele Grandi
Abstract
We present Linkify, a framework for learning from interface-augmented assembly graphs to enable context-aware part retrieval in mechanical assemblies. While recent generative AI methods for CAD have focused largely on isolated parts or monolithic assemblies, the rich geometric information at the interfaces between parts, where function is realized, remains underexplored. We address this gap by recomputing high-fidelity interface geometry for the Fusion 360 Gallery Assembly dataset, correcting missing and erroneous contacts, and generating point-cloud representations of local contact regions. Using this data, we construct assembly graphs whose nodes encode part geometry and whose edges encode interface geometry via a pretrained point-cloud encoder. On top of this representation, we train a Graph Attention Network based on GATv2 to solve a masked part prediction task: given an assembly with one part held out, the model predicts the class of the missing component from a large vocabulary of geometrically clustered parts, thereby approximating a realistic part-retrieval scenario. Compared to non-graph baselines such as logistic regression and k-nearest neighbors operating on aggregated node features, Linkify achieves higher Top-K accuracy and F1 scores. Ablation studies on graph connectivity, edge attributes, and attention mechanisms demonstrate that accurate contact computation and dynamic attention over interfaces are critical for performance. Our corrected interface dataset and training pipeline, released publicly, provide a foundation for future interface-aware models for assembly retrieval, validation, and generative design.