Awesome AI4CAD Hub Papers · Datasets · Projects
← Back to papers

AutoMate

arXiv 2021 51.8 method

TLDR

Proposes SB-GCN for predicting CAD assembly mates using parametric BREP representations, with a new dataset and 72.2% accuracy in a commercial CAD tool.

Reasoning

Strengths include a novel representation learning on BREPs for mate prediction, a large-scale dataset release, and integration with a commercial CAD system. Weaknesses are the narrow focus on mate prediction rather than full assembly generation and moderate accuracy.

Read-first score

Read-first score 51.8, weighted from topical fit, citation, graph, method, reproducibility, and recency signals. Original total remains 70.

Methodology quality 25%
80

Screens visible abstract and analysis fields for experiment, dataset, baseline, metric, and limitation evidence. markers=benchmark,dataset,evaluation,metric

Recency 8%
49

Uses a gentle age decay so recent papers surface without erasing older foundations. 2021

Topical relevance 42%
43.8

Uses existing LLM keyword relevance scores normalized to 0-100. AI for CAD,computer-aided design,neural CAD,generative CAD,parametric CAD,B-Rep,boundary representation,constructive solid geometry,CSG,sketch extrusion,CAD generation,CAD reconstruction,text-to-CAD,image-to-CAD,point cloud to CAD,CAD program

Reproducibility 25%
38

Screens links and visible text for paper, code, dataset, artifact, and repository signals. pdf=True; code=False; dataset=False; markers=dataset

Field roles

Methodology anchor

Rank sensitivity

Stability: volatile; rank range: 53.

Keyword Scores

B-Rep
10
boundary representation
10
computer-aided design
9
parametric CAD
9
AI for CAD
8
neural CAD
7
CAD program
5
generative CAD
2
CAD generation
2
CAD reconstruction
2
constructive solid geometry
1
CSG
1
sketch extrusion
1
text-to-CAD
1
image-to-CAD
1
point cloud to CAD
1

Deep Analysis

Innovations

  • SB-GCN: a representation learning scheme on BREPs that retains topological structure of parts
  • Using learned BREP representations to predict CAD type mates defined relative to BREP topology
  • First large-scale dataset of BREP CAD assemblies with benchmark mate prediction tasks
  • Integration with a commercial CAD system for mate suggestion, achieving 72.2% accuracy

Methodology

The paper proposes SB-GCN, a graph convolutional network that learns representations on parametric boundary representations (BREPs) preserving topological structure, and uses these representations to predict pairwise mate constraints. They compiled a large dataset of BREP assemblies for training and evaluation, and built a tool that suggests mate completions in a commercial CAD system.

Key Results

The mate suggestion tool integrated with a commercial CAD system achieves 72.2% accuracy in predicting mate completions.

Tags