Real-Time AI and CFD Hazard Detection System
Project Overview
This pipeline trains ML models on CFD simulations to detect anomalies early, trigger alerts, and visualize hazard propagation in a real-time dashboard.
Tech Stack
Team
Mentors
- Kshama Rai
- Garima Yadav
- Shanul Haque
Mentees
- Pranjali Vishwakarma
- Greeshma Prabhu
- Nandini Eswaran
Problem Statement
Chemical plants require proactive anomaly detection and propagation prediction to avoid incident escalation.
Objectives
- - Detect hazards from integrated sensor streams
- - Predict hazard evolution with CFD-trained AI
- - Trigger early warning alarms
- - Visualize events in 3D operations dashboard
Methodology
The project follows a structured implementation approach that includes Run CFD for leak and flow-disturbance scenarios, Build labeled normal and hazard datasets, Train ML models for signature detection and severity, and Deploy real-time Streamlit monitoring dashboard. These steps are executed iteratively to validate assumptions, improve performance, and ensure reliable delivery of the final solution.
Expected Outcome
By the end of this project, the team is expected to deliver Working AI hazard detection stack, Interactive 3D hazard visualization, and Integrated CFD plus AI inference demo. Together, these outcomes reflect both technical feasibility and practical value for demos, evaluation, and future scaling.
Future Scope
- - Generalize to refineries and gas storage
- - Increase geometry and process complexity
- - Integrate real plant sensor feeds