AI agent generates structural designs from natural language prompts
Can AI Co-Design Structures? An AI Agent for Early-Stage Structural Design
Artificial Intelligence
Summary
Creating safe and efficient structural designs often requires a lot of detailed work and expertise. The authors showed a system where an AI can understand simple text instructions and turn them into early-stage building structures using a specialized computer model. This method lets users interact with the design process by giving prompts that the AI turns into editable, graph-based structural models. While still in early stages, this approach points toward AI helping architects and engineers co-create building frameworks more intuitively.
What this means in practice
- •For structural engineers: Generate and iteratively refine structural models from text prompts integrated with structural modeling software.
- •For construction project managers: Connect early-stage structural data models produced by AI agents to downstream construction planning workflows for improved coordination.
Authors
Ricardo Maia Avelino, Mennatallah El-Assady, Tom Van Mele, Philippe Block
Abstract
Recent advances in large language models and agentic workflows offer a new paradigm to early-stage structural design translating design intent into structural models. Although text-to-3D approaches have demonstrated that natural language can be used to produce visual, mesh-based geometric models, structural design requires domain-specific geometric arrangements and semantically rich data types. This paper introduces a proof-of-concept workflow for prompt-to-structure generation using an AI agent connected to a COMPAS-based structural modelling API through a Model Context Protocol (MCP) server. Through controlled modelling operations, the agent creates and modifies a graph-based structural data model composed of explicit building elements. The resulting artifacts can be inspected, modified, serialised, and further connected to downstream AEC workflows. The approach is demonstrated through selected prompt-to-structure examples and by progressive and editable design sequences that resemble typical iterative structural design workflows. The results are preliminary, but demonstrate the possibility of combining natural language, agentic workflows, and domain-specific structural data structures as a first step toward AI-assisted structural co-design.