SHIVA — Structural Health Intelligence via Visual Analytics

Faculty: Deepak Subramani (CDS) and Debraj Ghosh (Civil Engineering)

Abstract:
Large-scale civil infrastructure such as dams and bridges requires continuous monitoring to ensure public safety and structural integrity. Current Structural Health Monitoring (SHM) approaches rely on periodic manual inspection or sparse sensor networks, leaving critical gaps in real-time awareness. This project develops an AI-driven SHM framework that is physically grounded and uncertainty-aware.

We couple physics-based numerical models of structural mechanics with modern machine learning, enabling automated inference of structural condition from heterogeneous data sources including drone imagery, vibration sensors, and strain gauges. SHM is fundamentally a challenging inverse problem; our approach addresses this through probabilistic data assimilation that accounts for uncertainty from material variability, sensor noise, and modeling error.

The outcome is a deployable system that updates structural assessments continuously as new observations arrive, enabling proactive rather than reactive infrastructure management. This framework offers government agencies and industry partners a scalable tool for risk management of critical national assets.

Blurb for Students:
Imagine being able to tell whether a dam or bridge is healthy, just by analyzing data from drones and sensors. That is exactly what this project is about.

Bridges and dams don’t fail overnight. They degrade slowly, and catching that early can save lives. We will build an AI system that continuously watches over large infrastructure, learns from real-world data, and flags potential problems before they become dangerous.

You will work at the intersection of physics, machine learning, and real-world impact. No prior experience in all areas is expected. Curiosity, rigor, and a willingness to learn are what matter most.