The goal of this project is to develop a smart traffic lights system that could sense real time changes in volume and speed of traffic. Using BT and WiFi signals, the system extracts statistical information on the road. According to that, a policy of traffic light would be determined, in order to receive a wardrop (equilibrium).
Nowadays, traffic lights systems have preset programs of control, regardless of the real state of traffic. Thus they cannot resolve unpredicted heavy traffic or emergency situations such as accidents. The monitoring of traffic that does exist today is based on traffic cameras and image processing algorithms which are expensive, or user based applications which are not necessarily reliable.
The goal of this project is to provide a simple, reliable and efficient system which would be able to extract traffic characteristics in real time, and subsequently could control the traffic light system based on said characteristics.
The idea of the project relies on the assumption that almost every vehicle contains electronic devices that transmit Wi-Fi and Bluetooth, such as smart phones or speakers. Each traffic light constitutes an independent entity which is controlled by a raspberry pi (a credit-card-sized single-board computer), and each two neighboring traffic lights are connected in order to transfer data between them. Once a traffic light, or to be precise a raspberry pi that controls the traffic light, senses a Wi-Fi or BT transmission, it saves the mac address and timestamp from the transmission in local database, and also transfers this data to its neighbors. The moment one of the neighbors’ sense transmission with the same mac address, it could calculate the time that took to a certain vehicle to pass the traffic section between the two traffic lights.
Repeating this process, creates a database from which traffic characteristics can be extracted, upon which the proper traffic light sequence can be selected and executed.