Cloud-based Real Time Traffic Assessment through Artificial Neural Networks and Server less Integration

At present, road traffic is the greatest challenge facing due to several elements including increased number of vehicles, limited road capacity, and uncoordinated traffic signals. Because of this issue, people are wasting fuel, causing negative effects on the environment. In this research, a traffic dataset is collected from Kaggle. Machine learning (ML) and deep learning (DL) methods are applied to solve the traffic congestion problem. Support Vector Machines (SVM), Random Forests (RF), decision trees (DT), Naive Bayes (NB), and artificial neural networks (ANN) are implemented to forecast the direction of traffic over a particular day. The implemented model’s accuracy is analyzed and results evidence that the ANN model is outperformed compared to other basic ML models. The results are stored using a cloud Storage Service for permanent records. Users can access the cloud to examine the flow of traffic and it helps to avoid traffic congestion.