Proposal of a Low-Cost System Architecture for Prototyping and Learning SLAM

Abstract

Most SLAM enabled robotics system require expensive hardware, which makes it difficult f or students to experiment with them. This work proposes a low cost mobile robotic system for implementing and testing SLAM, allowing students to experiment with its various implementations in the real world and better understand each one. We present experimental results that show sensor data acquirement, such as odometry and images, and communication of those data to a central computer, where the SLAM algorithm can be implemented using whatever technique and programming language the student chooses.