AI-powered Traffic Lights

With the ever-increasing population, automobiles production and sales have seen a massive percentage increase. With the increase in production, there has also been an increase in traffic congestion. Most of the traffic occurs because of the inefficient traffic light systems in intersections where several streams of traffic need to cross each other. Present-day traffic signals are operated in three ways, either the signals are based on timers or a manual controller who regulates the signal or an actuated signal. Traffic jams cause many environmental issues such as air pollution which leads to Global warming as well as many respiratory problems. The proposed system gives a solution to traffic congestion by using the cameras in signals, determining the traffic density in each lane, and assigning priority to each lane to be cleared. The traffic density of each lane is determined by pre-trained machine learning algorithms. This system is also cost-effective since we need not install any additional hardware technologies* (IR sensors, proximity sensors) instead use the existing cameras that are already installed. The system assigns lanes with emergency vehicles (Ambulances, Firetrucks, etc.) with the highest priority irrespective of the lane's traffic density. The system focuses on reducing the travel time of vehicles and optimizing traffic flow, which in turn reduces the usage of natural resources.