Post by Nishantha Ruwan

Autonomous and semi-autonomous driving technologies offer a promising pathway toward safer and more efficient transportation, with the potential to reduce both energy consumption and emissions. However, realizing these benefits requires carefully designed optimization strategies and purpose-built datasets to rigorously evaluate performance. In this work, researchers analyze real-world data capturing both driver behavior and vehicle dynamics to develop a novel reward function based on the Deep Deterministic Policy Gradient (DDPG) framework. The proposed reward function is designed to minimize energy usage and excess travel time while ensuring that acceleration remains within physical constraints. This approach achieves an average energy reduction of 7% across multiple vehicle powertrains—including Internal Combustion Engine Vehicles (ICEVs), Plug-in Hybrid Electric Vehicles (PHEVs), and Battery Electric Vehicles (BEVs). These gains are obtained while maintaining realistic acceleration profiles, preserving safe following distances, and limiting increases in trip duration to less than 1%. https://lnkd.in/gVVNxm_x