ResQ Shield is a real-time emergency monitoring system designed to detect and flag fake SOS alerts using behavioral analysis and anomaly detection techniques.
The system is inspired by real-world patterns derived from the Montgomery County 911 Calls dataset and simulates how an emergency operations center can intelligently filter genuine alerts from suspicious or malicious ones.
- The system models normal emergency behavior using patterns such as time distribution, location density, and alert frequency.
- When an SOS alert is triggered, it extracts key features like:
- Alert frequency
- Location repetition
- User trust level
- The alert is processed through a three-layer detection pipeline:
- Rule Engine (deterministic logic)
- Statistical Anomaly Detection (Isolation Forest-inspired)
- Behavioral Profiling (user history)
- These scores are combined into a final anomaly score and classified as:
- 🟢 Genuine
- 🟡 Suspicious
- 🔴 Fake
- 🚨 Interactive SOS trigger system
- 📊 Real-time operations dashboard
- 🧠 Multi-layer anomaly detection pipeline
- 📈 Live alert simulation and visualization
- 🎯 Behavior-based fake alert detection
- 🌙 Dark-themed Emergency Operations Center UI
This project is inspired by patterns from the Montgomery County 911 Calls dataset, which provides real-world emergency call distributions used to model normal behavior.
- HTML, CSS, JavaScript
- Frontend-based simulation
- ML-inspired anomaly scoring logic
This is a prototype system built for demonstration and hackathon purposes. It uses a lightweight approximation of anomaly detection rather than a fully trained machine learning model.
- Integration with real-time backend APIs
- Deployment with live geolocation tracking
- Full Isolation Forest model implementation
- Map-based visualization of alerts
- Advanced evaluation metrics (precision, recall)
Developed as a hackathon project to explore intelligent emergency response systems.