AI-StructVisionActive

AI-StructVision: Teaching AI to Recognize Dangerous Structural Damage

Can AI learn what ‘dangerous damage’ looks like in critical infrastructure?

AI-Powered System Engineering
Get in touch
Students
5
Bhatnagar, Boynton, Chet, Henock, Ruiz
Focus
Safety
Damage severity AI
Method
CV
Computer vision
Output
Model
Severity classifier
01

Overview

This project explores whether AI can be guided to recognize when structural damage is truly dangerous.

02

Novel contributions

  1. 01

    Severity-aware classification

    Goes beyond binary damage detection to classify structural damage severity.

  2. 02

    Engineer-guided AI training

    Training data labeled using criteria engineers apply during real inspections.

  3. 03

    Ethical AI in safety-critical systems

    Students reflect on implications of AI in high-stakes decisions.

Interested in this research?

For code access, collaboration or questions about AI-StructVision, contact the principal investigator, Dr. Tadesse Wakjira.

Contact the PIOpen positions