AI Agents for Automating the Analysis and Design of Structural Systems
Project Description
The recent advent of large language models allows machines to understand natural human language. Building upon this advancement, this project examines whether autonomous AI agents can perform multiple core tasks in structural engineering—from interpreting conceptual sketches to conducting structural analysis, code-compliant design, and result visualization. In particular, this project aims to prototype AI-agentic workflows that could automatically convert hand-drawn sketches or schematic drawings (e.g., beams, frames, and truss systems) into finite-element models and generate preliminary designs in accordance with existing codes of practice (e.g., Eurocode, American Concrete Institute, etc.). This study will evaluate the accuracy, reliability, and efficiency of this AI-agentic workflow and provide the practical guidelines for deploying these AI agents.
Supervisor
DIMITRAKOPOULOS Ilias
Quota
6
Course type
UROP1000
UROP1100
UROP2100
UROP3100
Applicant's Roles
The students will
(1) build AI-agent workflows that solve structural systems from schematic sketches.
(2) design structural systems using AI agents through existing codes of practices. (3) generate plots and visualizations of structural behavior (e.g., shear force, bending moment diagrams, deflections, etc.)
(4) produce technical reports summarizing the design of structures via AI agents.
(1) build AI-agent workflows that solve structural systems from schematic sketches.
(2) design structural systems using AI agents through existing codes of practices. (3) generate plots and visualizations of structural behavior (e.g., shear force, bending moment diagrams, deflections, etc.)
(4) produce technical reports summarizing the design of structures via AI agents.
Applicant's Learning Objectives
Through this project, the students will
1. Design, implement, and evaluate an end‑to‑end AI‑agent workflow that converts hand‑drawn sketches of beams, frames, or truss systems into validated finite‑element models.
2. Apply code‑compliant design rules from Eurocode, ACI, and other relevant standards to automatically generate preliminary structural designs using the AI agents.
3. Generate and interpret key structural response plots (shear force, bending moment, deflection diagrams) produced by the AI agents and compare them against benchmark solutions.
4. Critically assess the accuracy, reliability, and computational efficiency of the AI‑agent workflow through systematic testing on a diverse set of case studies.
5. Document best‑practice guidelines for deploying AI‑agents in routine structural engineering workflows, including data preparation, model validation, and error handling procedures.
1. Design, implement, and evaluate an end‑to‑end AI‑agent workflow that converts hand‑drawn sketches of beams, frames, or truss systems into validated finite‑element models.
2. Apply code‑compliant design rules from Eurocode, ACI, and other relevant standards to automatically generate preliminary structural designs using the AI agents.
3. Generate and interpret key structural response plots (shear force, bending moment, deflection diagrams) produced by the AI agents and compare them against benchmark solutions.
4. Critically assess the accuracy, reliability, and computational efficiency of the AI‑agent workflow through systematic testing on a diverse set of case studies.
5. Document best‑practice guidelines for deploying AI‑agents in routine structural engineering workflows, including data preparation, model validation, and error handling procedures.
Complexity of the project
Challenging