Autonomous navigation and 2d mapping using SONAR

Abstract

This paper deals with a small compact robot that can autonomously navigate in an unknown indoor environment and at the same time construct a real time 2D map of the environment. The entire system consists of a robot and an application that is used for monitoring and mapping. The map is constructed based on 3 SONAR readings and the heading of the robot is determined by a single axis digital compass. Sonar readings are sent by the robot to the application which processes the data for mapping the path traversed. The robot is a standalone system and the autonomous navigation is done by the on-board controller. In this method, we presented the grid based mapping system and by the end of the robot's run, we obtained a 2D occupancy matrix of the path followed by the robot and its environment. Sonar readings were filtered using the Kalman filter and thus accuracy was better. The technique also implemented a stack of the path followed which can be backtracked. It can also operate in manual mode.