Search for: All records

Human remote-control (RC) pilots have the ability to perceive the position and orientation of an aircraft using only third-person-perspective visual sensing. While novice pilots often struggle when learning to control RC aircraft, they can sense the orientation of the aircraft with relative ease. In this paper, we hypothesize and demonstrate that deep learning methods can be used to mimic the human ability to perceive the orientation of an aircraft from monocular imagery. This work uses a neural network to directly sense the aircraft attitude. The network is combined with more conventional image processing methods for visual tracking of the aircraft. The aircraft track and attitude measurements from the convolutional neural network (CNN) are combined in a particle filter that provides a complete state estimate of the aircraft. The network topology, training, and testing results are presented as well as filter development and results. The proposed method was tested in simulation and hardware flight demonstrations. 

We propose an implementation of an LQR con- troller for the full-state tracking of a time-dependent trajectory with a multirotor UAV. The proposed LQR formulation is based in Lie theory and linearized at each time step according to the multirotor’s current state. We show experiments in both simulation and hardware that demonstrate the proposed control scheme’s ability to accurately reach and track a given trajectory. The implementation is shown to run onboard at the full rate of a UAV’s estimated state. 

To grant unmanned aerial vehicles (UAVs) greater access to the National Airspace System (NAS), a reliable system to detect and track them must be established. This paper combines multiple radar systems into a single network to provide tracking of UAVs across a wide area. Each radar detects the UAV’s path and those detections are combined using a recursive random sample consensus (R-RANSAC) algorithm. Outdoor flight experiments show the ability of the system 

This paper presents a fixed-wing autopilot code base called ROSplane. ROSplane utilizes the ROSflight board, firmware, and driver, which was developed to make autopilot development faster, easier and cheaper. By leveraging a textbook and university course content, the autopilot facilitates education and accelerates research and development. The textbook provides high-level documentation for the code. The code is structured to facilitate learning by providing a framework for student assignments. The addition of ROSplane software and documentation make ROSflight closer to a plugand- play solution while maintaining simplicity and usability for researchers and students. ROSplane has been used in a graduate level flight dynamics class, demonstrated through test flights, and modified for research purposes. 

This paper demonstrates a feasible method for using a deep neural network as a sensor to estimate the attitude of a flying vehicle using only flight video. A dataset of still images and associated gravity vectors was collected and used to perform supervised learning. The network builds on a previously trained network and was trained to be able to approximate the attitude of the camera with an average error of about 8 degrees. Flight test video was recorded and processed with a relatively simple visual odometry method. The aircraft attitude is then estimated with the visual odometry as the state propagation and network providing the attitude measurement in an extended Kalman filter. Results show that the proposed method of having the neural network provide a gravity vector attitude measurement from the flight imagery reduces the standard deviation of the attitude error by approximately 12 times compared to a baseline approach.