IoT-BASED SMART AMBULANCE SYSTEM WITH REALTIME TRAFFIC CONTROL

Abstract

Road accidents are a serious dilemma in cities that require swift and effective countdown system for traffic situations. This project introduces an innovative dual-component system that is designed for smart accident detection and prioritization of ambulance passage. The software uses CNN or Convolutional Neural Network which was trained on “Accident” and “Non-Accident” datasets to analyze live webcam feeds and uploaded images for real-time traffic incident detection. The trained CNN model is saved as ‘model.h5’ and used in a Python application for both live video and image-based detection. Previous versions of crash detection and emergency vehicle prioritization are too slow and challenging to implement. Our proposed project combines accident detection with an RFID-triggered ambulance priority mechanism. By detecting an accident through a webcam or an uploaded image using CNN, the software creates alerts. Whenever an emergency vehicle crosses an intersection, the Arduino detects the vehicle using the RFID tag. In order to prioritize the emergency vehicle's passage, the traffic signal controller will turn the lights green; other vehicles will be stopped by turning the lights red. In parallel to it, the Arduino UNO microcontroller sends a SMS message to the hospital via Twilio, which gives an early warning about the incoming emergency. The technology when combined with AI, offers a well-planned implementation, where the emergency responses are likely to be faster and better.

Author

Mrs. PAVITHRA S, Ms. GOKILAVANI D, Mr. GOKULKANNAN C, Mr. KANISK G R, Mr. JAIPRATHESH M