Localization System Using Artificial Landmarks for Indoor Low-Cost Mobile Robots

Abstract

Localization is a prerequisite for land navigation of mobile robots. It is usually implemented with multiple sensors and high performance computers, so it is not usually a low-cost mobile robot capability. In this paper we present the architecture and design of a prototype embedded system which reports its location using a minimum-performance hardware and artificial landmarks. In order to evaluate accuracy, experimental tests were performed in an indoor environment, obtaining repeated locations from the prototype system. Based on the results, it is concluded that the proposed system is able to localize itself, with ±12 cm of accuracy, at a frequency of 2 Hz. Conclusions raise the possibility, in future works, for the inclusion of land navigation using maps into low-cost robots, that until today do not usually present this characteristic.