Predictive Analysis of Carbon Dioxide Emissions in Heterogeneous Urban Traffic using Neural Networks

India has emerged as the primary cause of greenhouse gas emissions globally, primarily attributable to a rapid surge in motorized vehicle usage, surpassing the population growth rate. The emissions are resulting in various problems related to deterioration in air quality, thereby creating an unnecessary increase in carbon footprint. The transportation sector is a major source, accounting for approximately 29% of the nation’s total emissions. Notably, over 90% of carbon dioxide (CO₂) emissions emanate from this sector, with road transportation being the predominant contributor. Despite the critical nature of this issue, obtaining comprehensive data on the extent of these emissions remains challenging. To address this information gap, our study focuses on predicting CO₂ emissions from diverse vehicle types operating on various fuels within specific busy urban road segments in Ahmedabad city. Employing neural networks, our research delves into the intricate network of global road transportation to extract crucial insights for quantifying and comprehending CO₂ emissions under different traffic conditions in the reduction of carbon footprint. This neural network-driven approach establishes a framework for assessing and mitigating CO₂ emissions in mixed-traffic environments across multiple Indian cities, extending its applicability beyond Ahmedabad.