Utilization Of Google Maps Data And Machine Learning For Traffic Congestion Prediction In Medium-Sized Urban Areas

Amelia, Sisil Azizah

Innovative Research in Civil and Environmental Engineering (IRCEE)
Vol. 2 No. 1 (2025): IRCEE - April

This study explores the use of real-time data from Google Maps and machine learning algorithms to predict traffic congestion in medium-sized urban areas. By applying various machine learning models, including Long Short-Term Memory (LSTM), Neural Networks, and Random Forests, this research aims to evaluate the accuracy and effectiveness of congestion predictions based on data such as weather conditions, time of day, road type, and special events like accidents or public gatherings. The results indicate that the LSTM model provides the most accurate predictions, with an accuracy rate of 89.4%. The study also identifies key factors influencing congestion, such as time of day, weather conditions, and local events. These findings can be used to improve traffic management in medium-sized cities by employing data-driven prediction systems to reduce congestion and enhance traffic efficiency.