Evaluation of indoor mobile robot localization techniques

Abstract

Mobile robot localization is an important operation used in several applications. For instance, the location of the mobile robot can be used to associate environment data captured by the robot with the location where this data is captured, or to issue robot commands dependent on location and remaining battery energy. We explore three localization methods for indoor robots: optical wheel encode, ultrasonic sensors, and received WiFi signal strength. We experiment with these localization methods on a National Instruments wheeled DANI robot, compare their accuracies, and evaluate their merits.