A review on reinforcement deep learning in robotics

Abstract

Since the world is experiencing the industrial revolution 4.0, robotics is one of the many instruments that is making a significant effect. Reinforcement Learning (RL) has been emerged as one of the promising techniques in recent years to significantly improve control over this technological wonder. RL allows robots to become self-aware and self-directed toward completing a certain goal, which is then followed by user actions. This scientific field has seen multiple significant advancements over decades of hard work, and it will continue to do so in the future. As a result, this paper fills a need in the scientific community by providing a systematic assessment of research papers published in the last decade. In relation to the study issue, this paper raises and answers several relevant questions. Future scholars will have a good understanding of RL-based robotics after reading this study, which they will be able to apply into their own research.