July 13, 2023 at 14:30 CEST in Sala Consiglio DAER - Building B12, Second Floor, Via La Masa 34 - Milano
Join the event online at the following link:
Prof. Roberto Furfaro, University of Arizona (Tucson, AZ, USA)
Over the past few years, there has been an explosion of machine learning techniques involving the use of deep neural networks to solve a variety of problems ranging from object detection to image recognition and natural language processing. The recent success of deep learning is due to concurrent advancement of fundamental understanding on how to train deep architectures, the availability of large amount of data and critical advancements in computing power (use of GPUs). One can ask how such techniques can be employed to provide integrated and closed loop solutions for space autonomy as well as Guidance, Navigation and Control (GNC). In this talk, we will provide an overview of deep reinforcement learning and meta-learning (learn-to-learn) that have been recently developed by my research team in the context of space GNC , including planetary landing, close proximity operations around small bodies, asteroid intercept and spacecraft docking and rendezvous in relative motion. We will show how deep recurrent policies trained on a distribution of Markov Decision Processes can be devised to provide accurate and robust autonomous agents that rapidly adapt to unseen conditions (e.g. uncertain dynamics, sensors and actuators failures) generally not encountered during the training phase. Importantly, we will show how to integrate the trained policies with data-driven hazard detection to precisely guide the spacecraft to hazard-free areas on the planetary surface.
Dr. Furfaro is currently Full Professor, Department of Systems and Industrial Engineering, Department of Aerospace and Mechanical Engineering, University of Arizona. He is also the Deputy Director of the Space, Security, Safety & Sustainability Center (Space4 Center) and Director of the Space Systems Engineering Laboratory. He obtained a Laurea Degree (M.S. equivalent) in Aeronautical Engineering (1998, University of Rome La Sapienza) and a Ph.D. in Aerospace Engineering (2004, University of Arizona). He has a broad range of expertise and research interests and has been working on a numerous and diverse projects including development of guidance navigation and control of planetary landers, systems engineering for close-proximity operations on small bodies, machine learning applications to space situational awareness and G&C for hypersonic vehicles. He has served as PI and Co-PI of numerous high- impact research and development grants with a total amount of funds received by NASA, AFRL and other DoD agencies exceeding $80M. He published more than 90 peer-reviewed journal papers and more than 250 conference papers and abstracts. Since 2013, he has been appointed technical member of the American Astronautical Society (AAS) Space Flight Mechanics committee and he served as Technical Chair of the 2015 AAS/AIAA Spaceflight Mechanic Meeting (Williamsburg Virginia). He is currently technical member of the AIAA Astrodynamics Committee and the AAS Space Surveillance Committee as well as Associate Editor for IEEE Transactions on Aerospace and Electronic Systems. During phase B-D (2011- 2016) of the OSIRIS REx Asteroid Sample Return Mission, he was the systems engineering lead for the Science Processing and Operations (SPOC). He is currently the Target Follow-up lead for the NASA NEO Surveyor Mission. For his contribution to the OSIRIS REx mission, the asteroid 2003 WX3 was renamed 133474 Roberto Furfaro. Recently, Prof. Furfaro has been elected the 2021 Da Vinci Fellow at the College of Engineering, University of Arizona. Additionally, Prof. Furfaro has been elected AIAA Associate Fellow, Class 2022 and AAS Fellow, Class 2021.
Prof. Daniele Mortari from Texas A&M University (College Station, TX, USA)
This lecture summarizes what the Theory of Functional Connections (TFC) is and presents various applications of optimization problems in aerospace engineering. The TFC performs analytical functional interpolation. This allows to derive analytical expressions with embedded constraints, expressions describingall possible functions satisfying a set of constraints. TFC has been extended to Multivariate domains and to a wide class of constraints, including points, functions, and derivatives constraints, relative constraints, linear combination of constraints, component constraints, and integral constraints. An immediate application of TFC is on constrained optimization problems as the whole search space is reduced to just the space of solutions fully satisfying the constraints. This way a large set of constrained optimization problems are turned into unconstrained problems, allowing more simple, fast, and accurate methods to solve them. For instance, TFC allows to obtain fast and accurate solutions of linear and nonlinear ordinary differential equations. Various examples of optimization problems in aerospace engineering are provided.
Daniele Mortari is a professor working in the field of attitude and position estimation, satellite constellation design, sensor data processing, and various topic in linear algebra and numerical algorithms. He has taught at the Aerospace Engineering School of University of Rome, and at Electronic Engineering of the University of Perugia. He received in 1982 his doctor degree in Nuclear Engineering from the University of Rome. He has been widely recognized for his work, including receiving best paper Award from AAS/AIAA and from journal Mathematics, three NASA’s Group Achievement Awards, 2007 IEEE Judith A. Resnik Award, and 2016 AAS Dirk Brouwer Award. He has published 1 book, 125 journal articles, 255 conference papers, and delivered 113 seminars. He is a Member of the International Academy of Astronautics, IEEE, AAIA, and AAS Fellow, AIAA Associate Fellow, Honorary Member of IEEE-AESS Space System Technical Panel, and former IEEE Distinguish Speaker. He is Editor-in- Chief of section Functional Interpolation of the Journal Mathematics.