DEVELOPMENT OF A SOCIAL REINFORCEMENT LEARNING BASED AGGREGATION METHOD WITH A MOBILE ROBOT SWARM

2022-9-09
Gür, Emre
In this thesis, the development of a social, reinforcement learning-based aggregation method is covered together with the development of a mobile robot swarm of Kobot- Tracked (Kobot-T) robots. The proposed method is developed to improve efficiency in low robot density swarm environments especially when the aggregated area is difficult to find. The method is called Social Reinforcement Learning, and Landmark-Based Aggregation (SRLA) and it is based on Q learning. In this method, robots share their Q tables inside the aggregated area. The developed robot swarm uses mostly off-the-shelf, open source, additive manufactured components and is extendable, and easily maintainable by design. Kobot-T is intended to be used together with a three-part heterogeneous swarm system capable of accomplishing tasks from foraging to aggregation for long hours. To prove and report the success of SRLA several aggregation methods from the literature are tested and the result of the comparison is given

Suggestions

DEVELOPMENT OF AN EXTENSIBLE HETEROGENEOUS SWARM ROBOT PLATFORM
Bilaloğlu, Cem; Turgut, Ali Emre; Şahin, Erol; Department of Mechanical Engineering (2022-1-13)
This thesis introduces Kobot -- an extensible heterogeneous swarm robot platform. Kobot platform uses a common hardware and software architecture based on off-the-shelf components, 3-D printing, and open-source software that evolves with state of the art. Robots built using this common architecture range from wheeled to flying robots and formed a heterogeneous swarm. The common architecture enabled developing and testing systems for the lightweight flying robots on resourceful ground robots. As a result, Ko...
Design of a Micro Piezo Actuated Gripper for a Swarm Robotic System
Kaygusuz, Batuhan; Turgut, Ali Emre (2016-07-31)
In this paper, we introduced a novel micro swarm robotic platform, called the AttaBot, designed specifically for complex systems and swarm robotics research. The novelty of the AttaBot platform is two-folds. (1) Artificial pheromones are implemented, (2) Each AttaBot has a gripper to grip small discs on the ground. In this way, AttaBot platform can be used in many different swarm scenarios including foraging, collective motion and collective transport. We introduced the design of the AttaBot, focusing mainl...
Evolving aggregation behaviors in a swarm of robots
Trianni, V; Gross, R; Labella, TH; Şahin, Erol; Dorigo, M (2003-01-01)
In this paper, we study aggregation in a swarm of simple robots, called s-bots, having the capability to self-organize and self-assemble to form a robotic system, called a swarm-bot. The aggregation process, observed in many biological systems, is of fundamental importance since it is the prerequisite for other forms of cooperation that involve self-organization and self-assembling. We consider the problem of defining the control system for the swarm-bot using artificial evolution. The results obtained in a...
Perpetual Robot Swarm: Long-Term Autonomy of Mobile Robots Using On-the-fly Inductive Charging
Arvin, Farshad; Watson, Simon; Turgut, Ali Emre; Espinosa, Jose; Krajnik, Tomas; Lennox, Barry (2018-12-01)
Swarm robotics studies the intelligent collective behaviour emerging from long-term interactions of large number of simple robots. However, maintaining a large number of robots operational for long time periods requires significant battery capacity, which is an issue for small robots. Therefore, re-charging systems such as automated battery-swapping stations have been implemented. These systems require that the robots interrupt, albeit shortly, their activity, which influences the swarm behaviour. In this p...
A self-organized collective foraging method using a robot swarm
Karagüzel, Tugay Alperen; Turgut, Ali Emre; Department of Mechanical Engineering (2020)
In this thesis, a collective foraging method for a swarm of aerial robots is investigated. The method is constructed by using algorithms that are designed to work in a distributed manner, by using only local information. No member in the swarm has access to global information about positions, states or environment. The environment, that robots are planned to operate in, contains a virtual scalar field which consists of grids containing constant values. The grid values indicate desired regions of the environ...
Citation Formats
E. Gür, “DEVELOPMENT OF A SOCIAL REINFORCEMENT LEARNING BASED AGGREGATION METHOD WITH A MOBILE ROBOT SWARM,” M.S. - Master of Science, Middle East Technical University, 2022.