Copyright © 2026 Authors retain the copyright of this article. This article is an open access article distributed under the Creative Commons Attribution License which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.

@article{193084,
        author = {Mayuri Hinge . and Mrs. Harshita Jain and Arpan Multeli and Gurvesh Dhomne and Vivek Sawwalakhe and Hatim Noor and Shrihari Chakole},
        title = {Autonomous Driving Using Deep Reinforcement Learning},
        journal = {International Journal of Innovative Research in Technology},
        year = {2026},
        volume = {12},
        number = {9},
        pages = {3868-3873},
        issn = {2349-6002},
        url = {https://ijirt.org/article?manuscript=193084},
        abstract = {Autonomous driving represents a major application of artificial intelligence, with the potential to improve safety and traffic efficiency. This paper presents a Deep Reinforcement Learning approach based on vehicle control trained in realistic simulation environments such as AirSim. Sensor information collected from cameras, LiDAR, and vehicle telemetry is first fused and transformed into a compact state representation. This structured state is then provided to a Deep Reinforcement Learning framework consisting of a CNN for spatial feature extraction, an LSTM for temporal dependency modeling, and a policy network trained using Double Deep Q-Network (DDQN). The integrated architecture enables the agent to learn optimal driving decisions based on both current observations and historical context. The learned agent will decide to steering, acceleration, and braking commands guided by safety-focused reward functions. The experimental findings indicate that the proposed approach to follow stable lane keeping, obstacle avoidance, and strong adaptability across varying environmental conditions. These results emphasize the potential of Deep Reinforcement Learning as a scalable and reliable solution for autonomous driving systems.},
        keywords = {Soft Actor-Critic (SAC), Actor-Critic Algorithms, Sensor Fusion, AirSim Stimulator, End-to-End Control, Motion Planning, Vehicle Control., Autonomous Driving, Reward Function Design, Deep Reinforcement Learning, Proximal Policy Optimization (PPO)},
        month = {February},
        }