Embedded GPU based autonomous robot use cases

Abstract

Recent advances in embedded GPU solutions allow the use of these devices for onboard mobile robotics applications. The focus of this paper is on the implementation and validation of navigation tasks for various autonomous mobile robot platforms. These platforms include ground vehicles, aerial vehicles, and mobile manipulators. The main challenge for these implementations was the feasibility of the adopted solutions on embedded platforms with dedicated GPUs. As each of our solutions contains deep neural networks, the existence of an embedded GPU is a must. We evaluated several different boards for our solutions with various experimental setups. The code and datasets for the demos are available on author’s home page.