Enhancing Road Safety: Detecting Aggressive Driving Behaviors on Highways Using 1D Convolutional Neural Networks

Abstract

Road traffic accidents pose a significant global public health challenge, resulting in millions of deaths and injuries each year. This paper investigates the use of artificial intelligence (AI) and machine learning (ML) techniques to analyze and classify driver behaviors, with a particular focus on detecting aggressive driving style on highways. A one-dimensional convolutional neural network (1D CNN) was employed to process and analyze driving data from the UAH-DriveSet dataset and independently collected real datasets for safe driving, with aggressive driving data simulated from this safe driving dataset. The model developed in this research demonstrated good generalization capabilities across different drivers. The integration of this model into real-time driver monitoring systems has the potential to significantly enhance road safety by alerting drivers to dangerous behaviors and encouraging safer driving practices