Smart Journey in Istanbul: A Mobile Application in Smart Cities for Traffic Estimation by Harnessing Time Series

TL;DR

This paper presents an AI-powered mobile app for Istanbul that forecasts traffic congestion using time series models like LSTM, Transformer, and XGBoost, demonstrating the Transformer’s superior accuracy for practical smart city traffic management.

Contribution

It introduces a novel traffic forecasting mobile application utilizing advanced AI models and evaluates their performance, highlighting the Transformer model's effectiveness in real-world traffic prediction.

Findings

Transformer achieved the highest accuracy in traffic prediction

The app can assist citizens in daily traffic management

Time series models effectively forecast traffic congestion

Abstract

In recent decades, mobile applications (apps) have gained enormous popularity. Smart services for smart cities increasingly gain attention. The main goal of the proposed research is to present a new AI-powered mobile application on Istanbul's traffic congestion forecast by using traffic density data. It addresses the research question by using time series approaches (LSTM, Transformer, and XGBoost) based on past data over the traffic load dataset combined with meteorological conditions. Analysis of simulation results on predicted models will be discussed according to performance indicators such as MAPE, MAE, and RMSE. And then, it was observed that the Transformer model made the most accurate traffic prediction. The developed traffic forecasting prototype is expected to be a starting point on future products for a mobile application suitable for citizens' daily use.