Home /Research /UnrealNavigation: Simulation Software for testing SLAM in Virtual Reality
PERCEPTION

UnrealNavigation: Simulation Software for testing SLAM in Virtual Reality

Anne Bettens, Benjamin Morrell, Mauricio Coen, Neil McHenry, Xiaofeng Wu, Peter W. Gibbens, Gregory Chamitoff

Year
2020
Citations
11

Abstract

We present UnrealNavigation; an integrated software platform for testing Simultaneous Localization and Mapping (SLAM) algorithms for autonomous robotic navigation. The frame- work, which is part of a Virtual Reality simulation in space environments is designed to allow rapid testing of autonomous navigation algorithm components. The architecture is designed with a plug-and-play style adaptability in mind, allowing for a wide range of uses and rapid development. Two approaches to SLAM will be demonstrated in the novel simulation frame- work, utilizing a game development engine, Unreal, connected by the Robotic Operating System (ROS). Comparison of these two algorithms was enabled through the simulation framework created. The different approaches authenticate the ability to integrate a holistic algorithm system or separate components into the framework.

Keywords

Computer scienceFrame (networking)Simultaneous localization and mappingAdaptabilitySoftwareVirtual realityGame engineHuman–computer interactionPlug-inArchitecture

Related papers

Browse all PERCEPTION papers