Reward Shaping Study for Sub-Goal Based Robotic Tasks

• Type: Article in International Conference Proceedings Book
• Date: 2022

Authors

Antão, L (Author) Other; Santos, L (Author) Other; Reis, J (Author) Other; Gil Manuel Gonçalves (Author) FEUP

Conference proceedings International

Title: 2022 2nd International Conference on Robotics, Automation and Artificial Intelligence, RAAI 2022

Pages: 74-79

2nd International Conference on Robotics, Automation and Artificial Intelligence, RAAI 2022

9 December 2022 through 11 December 2022

Indexing

Scopus

Other information

Authenticus ID: P-00Y-C07

DOI: 10.1109/raai56146.2022.10092972

Abstract (EN): Robotics has been providing essential tools to improve the Industry's flexibility and reconfiguration. Nowadays, the tasks are highly complex, making it mandatory to renew programming techniques and craft sophisticated solutions. This is where Reinforcement Learning comes in handy: it broadens the horizon for new and disruptive methodologies, incorporating tolerance to uncertainties using exploration. However, those solutions can be challenging when it comes to building the reward function. So, this work proposes a ''divide and conquer'' approach, which transforms a long and complex task into smaller and easier achieved goals that boosts training efficiency and performance. This study tested different combinations of parameters and proposes a set of guidelines that used a Pickand-Place environment. The observed performance shows higher success rates and faster convergence for two and three sub-goals when comparing the sub-goal approach versus a sequential and traditional reward. © 2022 IEEE.

Language: English

Type (Professor's evaluation): Scientific