Green-way traffic control pre-emption for emergency vehicles

Abstract

This project aimed at using machine learning approach for traffic prioritizing of emergency vehicles in traffic light junctions. The project also aimed at knowing the exit route of emergency vehicle upon reaching the junction so as to efficiently use other routes. To achieve this, we built a detection system that utilizes, siren detection, image recognition and route selection mechanisms. Then an overriding system to the traffic management system. Siren detection circuit was built to detect the presence of an emergency vehicle by its siren sound to initiate the rest of the process. For object detection, we trained a model using 2,348 images were 1,400 images were training datasets, 801 were validation data sets and 147 were testing data sets. These were obtained from available open source datasets and manually taking pictures of ambulances from places around the campus. Google Colab was then used to train our model. Transceiver antennas were also integrated into the object detection model to achieve a two- way communication between the emergency vehicle and the traffic management system. We then used Arduino boards to simulate he performance of the model.