Restricted Research - Award List, Note/Discussion Page
Fiscal Year: 2023
1842 The University of Texas at El Paso (143730)
Principal Investigator: Velez-Reyes,Miguel
Total Amount of Contract, Award, or Gift (Annual before 2011): $ 350,000
Exceeds $250,000 (Is it flagged?): Yes
Start and End Dates: 5/2/22 - 5/1/24
Restricted Research: NO
Academic Discipline: Electrical & Computer Engineer
Department, Center, School, or Institute: Electrical & Computer Engineer
Title of Contract, Award, or Gift: Innovative Analysis of Spectro-Temporal Signatures using Machine Learning for Ground-Based Remote Sensing ofUnresolved Resident Space Objects
Name of Granting or Contracting Agency/Entity:
UNIVERSITIES SPACE RESEARCH ASSOCIATION
CFDA: 12.56
Program Title: DOD, NDEP, DOTC-STEM Education Outreach Implementation
Note:
Advancing Space Domain Awareness (SDA) to provide tactical, predictive, and intelligence information on resident space objects (RSO) will rely on successfully extracting information from ground-based remote sensing assets. Assets such as hyperspectral (HSI) and polarization imaging systems are becoming available and the geographical spatial diversity provided by assets such as the Falcon Telescope Network are bringing large amounts of multi-optical data that provides further insights into aspects of RSO that were inaccessible previously for SDA. Accurate interpretation of these signatures may allow us to perceive, predict, comprehend, and react appropriately to changing situations in the space domain. The unresolved RSO (URSO) problem arises because optical remote sensing assets cannot spatially resolve RSO that are far away (e.g. GEO) or that are small (e.g. CubeSats). Furthermore, the system costs are motivating the study of small aperture COTS telescopes that are not able to spatially resolve objects even in LEO but can collect data with high spatial diversity. HSI systems collect spectro-temporal signatures of URSOs. The high-spectral resolution allows resolving the URSO in the spectral domain even though it cannot be resolved in the spatial domain. The signature contains information on properties/parameters of the URSO that can be exploited by machine learning (ML) approaches to solve the related inverse problem. While many recent advances in ML apply to the extraction of information from URSO signatures, those related to variational autoencoders (VAEs) and convolutional neural networks (CNN) are most relevant. The main goal of this project is to provide meaningful research opportunities in collaboration with AFRL to undergraduate and graduate students. A second goal is to strengthen collaborations between UTEP, and AFRL/USSF to develop a long-term partnership in remote sensing for SDA.
Discussion:
Withdrawn by institution.