An Analysis of Simultaneous Localization and Mapping (SLAM) Algorithms

Abstract

This paper provides an introduction to two Simultaneous Localization and Mapping (SLAM) algorithms: EKF SLAM and Fast-SLAM. SLAM allows an autonomous robot to accurately map an unknown environment as well as locate itself within the environment. These algorithms work iteratively, by moving about the environment and extracting and observing various landmarks in the environment. EKF SLAM and Fast-SLAM solve the SLAM problem by using probabilities to control for errors in the robot's sensors. This paper provides a discussion of these two algorithms and compares their run times and the accuracy of the maps they produce.