Simultaneous localization and mapping (SLAM) is a process in which a mobile robot can build an environment map where it is located and meanwhile use this map to calculate its location, using its sensors, for the obstacles distances measurements, objects and landmarks detection, odometry, environment images, etc. The big majority of the surveys in this topic have been focusing on the computing efficiency meanwhile in the assurance of consistent and accurate estimations for the robot map and orientation. Besides, there has been many research in topics such as nonlinearity, data association and landmarks characterization, that are vital to reach a practical and robust SLAM implementation. This research project presents a methodology based on Extended Kalman Filter in an embedded architecture in robots from the Robotics and Automation Research Laboratory of the Electrical Engineering Department. Through the applications, it is intended to accomplish autonomous navigation in internal environments, comparing its performance to the simulated trials and other correlated works.
News published in Agência FAPESP Newsletter about the scholarship: