Multi-robot Information Sampling Using Deep Mean Field Reinforcement Learning

Authors

Tuffa Said, University of North Florida
Jeffery Wolbert, University of North Florida
Siavash Khodadadeh, University of North Florida
Ayan Dutta, University of North FloridaFollow
O. Patrick Kreidl, University of North FloridaFollow
Ladislau Boloni, University of Central Florida
Swapnoneel Roy, University of North FloridaFollow

Document Type

Conference Proceeding

Publication Date

1-1-2021

Abstract

We study the problem of information sampling of an ambient phenomenon using a group of mobile robots. Autonomous robots are being deployed for various applications such as precision agriculture, search-and-rescue, among others. These robots are usually equipped with sensors and tasked with collecting maximal information for further data processing and decision making. The studied problem is proved to be NP-Hard in the literature. To solve the stated problem approximately, we employ a multi-agent deep reinforcement learning framework and use the concepts of mean field games to potentially scale the solution to larger multi-robot systems. Simulation results show that our presented technique easily scales to 10 robots in a 19 × 19 grid environment, while consistently sampling useful information.

Publication Title

Conference Proceedings - IEEE International Conference on Systems, Man and Cybernetics

First Page

1215

Last Page

1220

Digital Object Identifier (DOI)

10.1109/SMC52423.2021.9658795

ISSN

1062922X

ISBN

9781665442077

Citation Information

T. Said et al., "Multi-robot Information Sampling Using Deep Mean Field Reinforcement Learning," 2021 IEEE International Conference on Systems, Man, and Cybernetics (SMC), 2021, pp. 1215-1220, doi: 10.1109/SMC52423.2021.9658795.