An Internet of Things (IOT)-Enabled Enhanced Smart Transportation System for Traffic Management and Congestion Control

Traffic congestion is a major impediment to efficient transportation, resulting in increased travel time, fuel consumption, emissions, and road use. This study presents an Internet of Things (IoT) enabled enhanced smart transportation system for traffic management and congestion control using distributed IoT sensing, edge computing, cloud computing, adaptive signal control, and Long Short-Term Memory (LSTM) congestion prediction. Traffic data including vehicle count, speed, density, queue length, road occupancy, and congestion index were simulated using IoT sensing devices and processed at the edge layer to enable fast traffic monitoring and control. An LSTM congestion prediction model was trained at the cloud layer to learn traffic patterns and predict congestion, while an adaptive controller was used to manage signalized intersections based on real-time traffic and congestion status. Google Colab and Simulation of Urban Mobility (SUMO) were used to train the LSTM congestion prediction model and simulate traffic under light, moderate, heavy, and peak-hour traffic intensities. The LSTM congestion prediction model achieved an R² of 0.9813, an RMSE of 0.02110, an MAE of 0.01678, and an MSE of 0.000445 on the reconstructed evaluation dataset. The proposed adaptive signal control system reduced waiting time, queue length, travel time, and communication delay by up to 48.9, 47.9, 28.9, and 67.5%, respectively, compared to a conventional fixed-time traffic control system, while increasing traffic throughput by 23.7% under peak-hour conditions (16.5% on average across all traffic conditions). The results, reported as means over ten simulation runs per traffic condition, show that the proposed IoT-edge-cloud enhanced smart transportation system can provide an intelligent traffic management system that reduces congestion and improves traffic flow.