BridgeNet: A Dataset of Graph-based Bridge Structural Models for Machine Learning Applications

TL;DR

BridgeNet is a comprehensive, publicly accessible graph-based dataset of 20,000 bridge structures designed to advance machine learning applications in structural engineering, including form-finding, reconstruction, and generative design.

Contribution

It introduces a novel, multimodal dataset of bridge models with diverse data representations to support ML research in structural design and analysis.

Findings

Supports tasks like edge classification and parameter inference

Enables cross-modal reconstruction between graphs, meshes, and images

Facilitates development of ML-based surrogate models for form-finding

Abstract

Machine learning (ML) is increasingly used in structural engineering and design, yet its broader adoption is hampered by the lack of openly accessible datasets of structural systems. We introduce BridgeNet, a publicly available graph-based dataset of 20,000 form-found bridge structures aimed at enabling Graph ML and multi-modal learning in the context of conceptual structural design. Each datapoint consists of (i) a pin-jointed equilibrium wireframe model generated with the Combinatorial Equilibrium Modeling (CEM) form-finding method, (ii) a volumetric 3D mesh obtained through force-informed materialization, and (iii) rendered images from two canonical camera angles. The resulting dataset is modality-rich and application-agnostic, supporting tasks such as CEM-specific edge classification and parameter inference, surrogate modeling of form-finding, cross-modal reconstruction between graphs, meshes and images, and generative structural design. BridgeNet addresses a key bottleneck in data-driven applications for structural engineering and design by providing a dataset that facilitates the development of new ML-based approaches for equilibrium bridge structures.