• Priyesh Patel
• Meet Nandu
• Saifil Momin
• Neepa Shah

People all around the world are enthusiastic about the advent of autonomous vehicles for the public. There are many paradigm shifts taking place in this sector due to information explosion and the advancements in Artificial Intelligence. Although self-driving technology has permeated the commercial sector, its potential is not being fully utilized in the military sector where such autonomy can save millions of lives and significantly reduce the training costs. Existing autonomy in the military is limited to the use of remotely controlled automatic vehicles. Self-navigating autonomous vehicles can optimize this functioning by inducing the power of decision making that enables self-learning. In this paper, we extensively studied the existing Unmanned aerial Vehicle (UAV), Unmanned ground vehicle (UGV), and Unmanned underwater (UUV) vehicle systems and their drawbacks where we compared the performances of the existing algorithms based on vital parameters. We further state scope of Reinforcement Learning (RL) alongside with Convolutional Neural Network in autonomous military vehicles, which overcomes the drawbacks on existing algorithms. Consequently, we propose a reward-based learning model using RL to train the agent to navigate in a dynamic simulation environment by means of obstacle detection and global path optimization.

Autonomous vehicles, self-driving vehicles, UAV, UGV, UUV, Reinforcement Learning, Convolutional Neural Network

"Military Adaptation of Self-driving System using Reinforcement Learning", International Journal of Emerging Technologies and Innovative Research (www.jetir.org), ISSN:2349-5162, Vol.5, Issue 12, page no.431-437, December-2018, Available :http://www.jetir.org/papers/JETIR1812460.pdf

"Military Adaptation of Self-driving System using Reinforcement Learning", International Journal of Emerging Technologies and Innovative Research (www.jetir.org | UGC and issn Approved), ISSN:2349-5162, Vol.5, Issue 12, page no. pp431-437, December-2018, Available at : http://www.jetir.org/papers/JETIR1812460.pdf