Workshop Description
Self-localization is critical for the safe and effective navigation for autonomous mobile robots. However, localization failures in highly dynamic, feature-poor, or ambiguous environments can lead to localization inaccuracy and ultimately failure. In this workshop session, we will present a data-driven approach to predict localization failure of autonomous robots in warehouse environments. We introduce the Flowcean framework and explore its ROS2 integration to provide an overview on the data collection process and the training of a convolutional neural network. Participants will have the opportunity to familiarize themselves with the framework in a hands-on session that demonstrates the capabilities of the Flowcean framework in a simple robotic usecase.
