Towards advanced robotic manipulation

TL;DR

This paper evaluates the limitations of current robotic manipulation techniques and explores reinforcement learning approaches, including Hindsight Experience Replay and reward shaping, to improve real-world robotic control.

Contribution

It introduces reinforcement learning alternatives to Hindsight Experience Replay and identifies key research questions and directions for advancing robotic manipulation.

Findings

Hindsight Experience Replay has limitations in real-world applications

Reinforcement learning with reward and goal shaping shows promise

Several research questions for future work are identified

Abstract

Robotic manipulation and control has increased in importance in recent years. However, state of the art techniques still have limitations when required to operate in real world applications. This paper explores Hindsight Experience Replay both in simulated and real environments, highlighting its weaknesses and proposing reinforcement-learning based alternatives based on reward and goal shaping. Additionally, several research questions are identified along with potential research directions that could be explored to tackle those questions.