Work

Forecasting Road Traffic Volume Using Temporal-Only Time Series Models

Year: 2025

Type: conference-paper

Abstract: Predicting road traffic in road networks is essential for alleviating congestion and enhancing traffic flow efficiency. This study presents a traffic volume prediction model trained using time-series data collected from loop detectors installed at three junctions in Istanbul, Turkey. The detectors record vehicle counts, providing the historical data used to train a residual LSTM model designed to forecast vehicle volumes 15 minutes into the future. The proposed model achieved an average mean absolute error of 0.26 across multiple loop detectors, demonstrating its effectiveness in accurately predicting short-term traffic volumes.

Source: IEEE International Conference on Pervasive Computing and Communications Workshops and other Affiliated Events (PerCom Workshops)

DOI: 10.1109/percomworkshops65533.2025.00094

Authors: Araya Kibrom Desta, Ismail Arai

Institution: Nara Institute of Science and Technology

Language: English

Topic: Traffic Prediction and Management Techniques

Subfield: Building and Construction

Field: Engineering

Domain: Physical Sciences

SDGs: Sustainable cities and communities, Industry, innovation and infrastructure

Open Access status: green

Funder: Japan Science and Technology Agency