Good structural design depends on judgment that is hard to write down: how a designer reads a design brief, conceptualizes a scheme, and knows when a structure feels right. We present a human-AI co-design framework that keeps the designer in the loop throughout training and deployment. The model learns implicit design knowledge directly from human demonstrations and collaborates with the designer through iterative back-and-forth on a shared artifact.

→ Project Overview
We developed GooGym2D, an open-source 2D structural design testbed based on the Gymnasium API. It frames truss design as graph construction in continuous space, with terminal finite-element analysis checking connectivity, mechanisms, yielding, and buckling. The pieces form a single human-in-the-loop cycle — teaching becomes teaming.
Demonstration
Designers build bridges in the browser. 323 seed demonstrations across three spans capture different ways of solving the same structural problem.
Correction
A flow-matching policy learns a distribution of plausible design actions. The AI acts first; the designer replays its rollouts and takes over only where it fails.
Co-design
The trained policy returns to the Bridge Designer as a partner, offering ghost-bar continuations while the designer keeps the final say.



→ Publications
- Capturing Structural Intuition: Human-Gated Imitation Learning for Structural Design with Flow Matching
Tao Sun, Shaoyi Wang, Simon Schleicher, Ramon E. Weber · IASS 2026 - From Teaching to Teaming: Human-AI Co-Design Environment for Architectural and Structural Design
Tao Sun, Shaoyi Wang, Simon Schleicher, Ramon E. Weber · ACADIA 2026
Explore the full interactive project page, best viewed on desktop: luposun.github.io/GooGym2D

Leave a Reply