Atta-Bot: A Do-It-Yourself approach for swarm robotics

Abstract

Swarm robotics research often relies on simulated environments to test algorithms and understand cooperative behaviors. The "reality gap"—the disparity between simulations and real-world conditions—highlights the necessity of using physical robots to validate swarm behaviors. This paper introduces Atta-Bot, a cost-effective and DIY-friendly robot manufactured from common components, designed to bridge this gap. The performance of Atta-Bot in swarm exploration is evaluated using a random walk algorithm, considering variables such as movement distance, initial positions, robot quantity, and execution time. Atta-Bots use this algorithm to explore, detect obstacles, and relay information to a central server, ensuring data accuracy through verification with computer vision. The results show that greater coverage is achieved with increased execution time and a larger number of robots, especially when starting from the corners of the scenario. The Atta-Bot performs effectively in swarm exploration and adapts well to various environments. Validation with computer vision confirms the reliability of the data, affirming the functionality of the prototype.