Data fusion localization for an autonomous mobile robot

Autonomous mobile robots are able to navigate in an environment by using merely the computing capabilities they have on-board. Among the aspects of autonomous navigation, localization plays a very important and challenging role. Without knowing its position the robot would not be able to perform tasks such as path planning and motion control. Map- based probabilistic approaches are a widely used solution to the problem of mobile robots localization. However, they are not efficient when the robot is deployed outdoors due to the lack of natural landmarks. The aim of this document is to experiment an approach where the localization provided by a map-based localization algorithm is fused with a GNSS system by means of a Kalman filter to obtain a more robust and accurate localization. This work is part of a broader project developed by InnoTech company in collaboration with California State University, Los Angeles.