AI and IoT Based Innovative Solutions for Traffic Management System in Smart Cities

Traffic forecast has grown more challenging in the modern era due to the rapid population growth and regular traffic congestion. Increasing the level of interactions in vehicles leads to the degradation of time and energy, gasoline, and natural resources, as well as the potential for fatalities from people caught in traffic. Since there are not numerous researchers exploring the problem of traffic forecasting and regulation systems, the accuracy could potentially be lower. Thus, for a smart city, this research developed an effective traffic prediction system combining the AI algorithm and IoT devices. Data gathering using IoT gadgets, extraction of characteristics, categorization and the regulation of traffic signals system comprise the five significant phases of the proposed system. First, the dataset is used to gather the IoT traffic statistics. These gathered data are then sent into the Attention based Deep Recurrent Neural Networks (ADRNN) classifier, which determines the spot that has the most traffic. Ultimately, a microcontroller is used to control traffic. In comparison to the previous approaches, ADRNN achieves 98.2% accuracy and a 97.9% F-score.