TraceBot is a low-cost, AI-powered disaster search and rescue robot designed to assist rescue teams in locating survivors trapped under debris during natural disasters such as landslides, earthquakes, and structural collapses.
The system combines thermal sensing, machine learning, GPS tracking, and IoT communication to detect human presence and transmit real-time location data to rescuers through a web-based monitoring dashboard.
- Human detection using MLX90614 thermal sensor
- Embedded AI inference using a Decision Tree model
- Real-time GPS location tracking
- WiFi-based data transmission
- Live monitoring dashboard hosted on ESP32
- Portable and cost-effective design
- Designed for disaster response and search-and-rescue operations
- ESP32 Microcontroller
- MLX90614 Infrared Thermal Sensor
- NEO-7M GPS Module
- 4WD Robotic Chassis
- DC Geared Motors
- Embedded C / Arduino IDE
- Python
- Machine Learning (Decision Tree)
- HTML
- IoT Communication
The system follows the workflow below:
- Robot navigates through disaster zones.
- Thermal sensor collects temperature readings.
- ESP32 processes sensor data.
- Decision Tree-derived threshold classifies human presence.
- GPS coordinates are captured.
- Detection results are displayed on a live dashboard.
- Rescue teams receive victim location information.
- Classification Accuracy: 95.54%
- GPS Localization Error: < 3 meters
- Average Telemetry Latency: 150 ms
- ESP32-CAM integration for live video streaming
- CO₂ sensor integration
- Acoustic sensor integration
- Multi-sensor fusion techniques
- Swarm robotics implementation for large-scale disaster response
| Name | Role |
|---|---|
| Aleesha Sujith Abraham | AI/ML Development, Research |
| Amna Sajid | Hardware & Integration |
| Annpriya George | Development & Testing |
| Joyal Maria Johnson | Development & Documentation |