PATTERN FORMING ROBOTS AND LOCALIZATION USING AUGMENTED REALITY TAGS

Abstract

In recent advances in the field of robotics, pattern formation of robots has gained serious attention by researchers for surveillance applications among others. Multiple unmanned aerial vehicles (UAV’s) and unmanned ground vehicles (UGV’s) can be used for such applications. In order to achieve coordination between the multiple robots for performing task collectively, centralized control is one the most optimal technique. In this paper, we propose pattern formation of autonomous robots using Augmented Reality (AR) Tags with localization. In our approach, the system consists of leader robot and follower robot working on centralization technique. The autonomous robots are working and controlled using ROS platform. The master robot generates the coordinates according to the given pattern and moves using Odometry, after which the master searches for multiple slaves with AR-Tags attached, compute their distances and coordinates using image processing and assign the respective generated coordinates. Furthermore, in order to increase stability and reduce error, elliptical shape robot hardware is fabricated with more accuracy and perfection. This paper introduces a new technique for pattern formation using AR-Tags with single point reference approach. The attained results are very accurate and stable with a tolerance of 2 cm variation in distance measurement and robots are able to arrange themselves according to the desired pattern. The described method and technique are completely portable and generated results are more accurate than other expensive techniques.