At École Polytechnique Fédérale de Lausanne (EPFL)
At Massachusetts Institute of Technology Lincoln Laboratory (MIT LL)
Unmanned aircraft can range in size from tiny hand-launched varieties like the Raven to large airplane sized varieties like the Predator or the Global Hawk. Many UAS may not have very much spare room on-board for a sense and avoid system due to other equipment. As a result, ground-based sense and avoid (GBSAA) solutions are being developed. With GBSAA, all or most of the equipment necessary for sense and avoid is located on the ground rather than on the aircraft.
The particular GBSAA project my colleagues and I worked on was funded by the US Army. In conjunction with two other organizations, we developed the system over several phases. The first phase was very restrictive with subsequent phases allowing more types of missions to be performed safely in the US national airspace. One highlight of this project was a successful demonstration of many of the GBSAA phases at Dugway Proving Ground in Utah, including one scenario where two live UAS flew at each other and the GBSAA system prevented a collision.
I performed several duties for this project. I was heavily involved in algorithm development for all phases of the GBSAA algorithms, which involved coding in MATLAB, C, and C++. I was also heavily involved in algorithm evaluation and data analysis. I also co-led Lincoln Laboratory’s part of the schema development for the interface between various portions of the system. Finally, I lead the requirements process for Lincoln’s portion of the system.
While no papers have yet been published about this project, several news articles have been published about the demonstration at Dugway Proving Ground. Additionally, one phase of the algorithm is heavily based upon the new ACAS-X, the replacement to ACAS (TCAS in the US).
A Small Sampling of News Articles about this Project:M. J. Kochenderfer and J. P. Chryssanthacopoulos, “Partially-Controlled Markov Decision Processes for Collision Avoidance Systems,” in International Conference on Agents and Artificial Intelligence, Rome, Italy, 2011. J. P. Chryssanthacopoulos and M. J. Kochenderfer, “Accounting for State Uncertainty in Collision Avoidance,” Journal of Guidance, Control, and Dynamics, vol. 34, iss. 4, pp. 951-960, 2011.
The aim of this research was to develop an analytical definition for well clear which could be used in self-separation algorithms for UAS. This research was performed in conjunction with two of my colleagues at the Lincoln Laboratory. We recognized that a well clear threshold was a separation standard as defined by the International Civil Aviation Organization (ICAO). As such, we used a collision-risk approach to derive a boundary for well clear which could be used in the rigorous safety assessments required by aviation authorities.
My role in this project was to set up and run Monte Carlo simulations of millions of aircraft encounters to produce data. I then analyzed the data and produced results. This work resulted in two published papers and was used heavily by regulatory committees responsible for defining well clear.
Weibel, R. E., Edwards, M. W. M., Fernandes, C., Establishing a Risk-Based Separation Standard for Unmanned Aircraft Self Separation, AIAA Aviation Technology, Integration, and Operations conference, Virginia Beach, Virginia, 2011.
Weibel, R. E., Edwards, M. W. M., Fernandes, C., Establishing a Risk-Based Separation Standard for Unmanned Aircraft Self Separation, Ninth USA/Europe Air Traffic Management Research & Development Seminar, Berlin, Germany, 2011.
As the desire to use unmanned aircraft (UAS) has increased over the years due to the wide variety of applications and users, the need for access to the US National Airspace has also grown. One of the main barriers to UAS airspace access is the ability of UAS to comply with regulations. In addition to needing to perform self separation, UAS also need to perform collision avoidance. These two functions together are called sense and avoid.
Unmanned aircraft can range in size from tiny hand-launched varieties like the Raven to large airplane sized varieties like the Predator or the Global Hawk. For those UAS with room for sense and avoid equipment, air-borne sense and avoid (ABSAA) is being developed. ABSAA is similar to GBSAA except that all or almost all of the equipment is deployed on the aircraft itself.
The particular ABSAA project my colleagues and I worked on was funded by the US Navy for the Triton UAS. In particular, I was responsible for modelling, simulation, and system analysis. I was also the modelling and simulation lead for the project. This involved developing models of various system components in MATLAB’s Simulink, developing the accreditation plan, and performing project management tasks.
At Clemson University
Advisor: Dr. Jeong-Rock Yoon
Project Title: “Tumor Detection Using Magnetic Resonance Elastography”
Description: The aim of the project was to detect possible tumors. It is generally accepted that most tumors are much stiffer than surrounding healthy tissue. In a recent experiment, it is possible to use MR techniques in quantifying the stiffness of tissues. In particular, this research project is based on the article MR Elastography of the Liver: Preliminary Results found in Radiology: Volume 240:Number 2-August 2006 . The goal is “to develop a method for measuring liver stiffness with magnetic resonance (MR) elastography.” The eventual goal is to be able to detect tumors non-invasively using MR elastography. The research was performed as part of the MACOBE creative inquiry group, a group focusing on integrating mathematical and computational science into Bioengineering modelling and design problems.
Graduate Advisor: Dr. Taufiquar Khan
Thesis Title: Sparse Representation for Detection of Transients Using a Multi-Resolution Representation of the Auto-correlation of Wavelets
Abstract: This thesis seeks to detect damped sinusoidal transients, specifically capacitor switching transients, buried in noise and to answer the following questions: 1.) Can the transient s(t; q) be sparsely represented from sδ (t) = s(t; q) + ε(t) using sparsity methods, where ε(t) is white Gaussian noise? 2.) Does computing the local auto-correlation of the signal around the transient improve detection? 3.) How does the auto-correlation shell representation compare to the wavelet representation? 4.) Which basis is ”best”? 5.) Which method and representation is best? This thesis explores detection schemes based on classical meth- ods and newer sparsity methods. Classical methods considered include reconstruction via wavelets and reconstruction in the novel multi-resolution representation based on the auto- correlation functions of compactly supported wavelets. For simplicity, only four bases are considered: Haar, Daubechies 2, Daubechies 4, and Symlets 2. Sparsity methods include the iterative soft, hard, and combined thresholding algorithms.
Advisors: Prof. Dr. Peter Maass and Dr. Dennis Trede
Title: Regularisation and Inverse Problems
Description: While studying for my Master’s degree, I participated in a study abroad program at Universitaet Bremen in Bremen Germany. My research in Germany focused on a theoretical introduction to inverse problems and Tikhonov regularisation. This research served as a foundation for my thesis work.