Collaborative navigation of UAV and UGV using vision and LIDAR sensors

This paper describes our on-going research efforts toward a collaborative navigation system of an Unmanned Ground Vehicle (UGV) and an Unmanned Aerial Vehicle (UAV). Since the range of sight of a UGV is very limited, the UGV cannot obtain a complete map of its surrounding area to a given destination before it travels if the map information is not provided a priori. Thus, the UGV is not able to generate an optimal path from the source to the destination before its mission. Although a complete map could be provided to the UGV a priori, GPS signals should also be available for the UGV to localize its position in the given map. If the UGV is in a GPS-denied environment, it would be challenging for the UGV to travel to the destination along an optimal path. In this paper, we utilize a UAV to assist a UGV when complete map information is not provided and/or GPS signals are not available to the UGV. In particular, we present the applied algorithms of image processing and path planning to find an optimal path for the UGV based on the vision data obtained from the UAV. The UGV navigates along the obtained optimal path using its Light Detection and Ranging (LIDAR) sensor. The collaborative navigation system consists of a UAV, a UGV, and a processing server. The UAV is equipped with a vision camera and a Wi-Fi transmitter, and the UGV is equipped with a LIDAR, an actuator control processor, and a Wi-Fi transmitter. The processing server is designed to process and relay navigation information between the UAV and the UGV. This collaborative navigation system is experimentally validated in an open field with structured paths with an intersection.