Computer Vision Techniques for Autonomic Collaboration between Mobile Robots

Autonomic Computing is concerned with improving the self-management and adaptability in software. In this paper, we propose an Autonomic concept that would allow robots to interact with each other using an alternative communication protocol in the event that direct wireless digital communications fail. The aim of the research is to create enhanced levels of autonomous behavior that enable the robots to still communicate via a more primitive mechanism. The alternative communication protocol uses a form of textual semaphore. The visual means of communication involves processing textual images that contain meaningful messages or instructions. These textual images are transmitted via a display screen mounted on the robot. The robot receiving the message does so by taking a photo and performing image processing. The research focus is on how robots could adapt to a hardware communications failure by using the technique outlined here. Our experiments have proven that a robot can read text from a screen and respond with an action, e.g., to rotate. Results also indicate that further work is needed in the area of image processing in order to improve the accuracy of the character recognition. It is hoped that these results will lead to further research that can build upon the ideas presented and seek to establish a two-way conversation as opposed to a monologue.