Cooperative and Coordinated Localization of Swarm Robots using Adaptive Boids Rules

Trevor Smith, Eduardo Gutierrez, Jonas Amoama Bredu, Yu Gu, Jason Gross

Peer Reviewed

Abstract:	This study seeks to improve the localization estimate of swarm robotics through cooperation, agents sharing information, and coordination, multiple robots altering their movements to work together, by utilizing boids rules. Since swarm robots are often constructed of simple and low-quality sensors, due to the increase in cost with swarm size, complex perception tasks, such as localization, become difficult challenges to solve. However, robot swarms have the advantage of large numbers. Therefore, the robots can coordinate their movements, to perform more localization measurements with each other to improve their localization estimates. Boids rules model complex biological flocking behavior, that allows the swarm to coordinate their movements. Thus, by incorporating flocking behavior, into the localization strategy, the robots can group together and perform more localization updates at necessary. In addition, this study expands boids rules with the inclusion of homing and task rules, along with adapting the gains on each rule based on the local stimuli of neighboring robots and the current localization estimate. As a result, the robots are encouraged to perform more favorable emergent behaviors such as clustering when the agents’ localization estimates are poor and diffusion behaviors when the estimate is sufficient. The proposed algorithm is decentralized and independent of the localization strategy, thus it can be easily incorporated into existing swarm robotic frameworks. This method was tested with multiple localization frameworks such as decentralized EKF, Covariance Intersection, and Dead Reckoning in a randomly generated way-point bounded environment. As a result of this study, at the cost of increasing task completion time, the proposed algorithm was able to improve the average localization estimate of the swarm, with the greatest improvement occurring at low-density swarms with poor estimators.
Published in:	Proceedings of the 35th International Technical Meeting of the Satellite Division of The Institute of Navigation (ION GNSS+ 2022) September 19 - 23, 2022 Hyatt Regency Denver Denver, Colorado
Pages:	2927 - 2940
Cite this article:	Smith, Trevor, Gutierrez, Eduardo, Bredu, Jonas Amoama, Gu, Yu, Gross, Jason, "Cooperative and Coordinated Localization of Swarm Robots using Adaptive Boids Rules," Proceedings of the 35th International Technical Meeting of the Satellite Division of The Institute of Navigation (ION GNSS+ 2022), Denver, Colorado, September 2022, pp. 2927-2940. https://doi.org/10.33012/2022.18560
Full Paper:	ION Members/Non-Members: 1 Download Credit Sign In