Lane Fuller, Advanced Scientific Concepts; Robert Karl, Jr., Advanced Scientific Concepts; Bruce Anderson, Advanced Scientific Concepts; Max Lee-Roller, Advanced Scientific Concepts
Keywords: RSO, ASO, Flash LiDAR, Space Situational Awareness, SSA, simulation, modeling, BRDF, Deep Learning, Machine Learning, organized point cloud
Abstract:
Due to the increasing amount of Resident Space Objects (RSOs) in the Earth’s orbit, the ability to quickly and accurately extract information about them has become an important element of Space Situational Awareness (SSA). The Global Shutter Flash LiDAR (GSFL) has been proven to be an effective and highly reliable sensor to acquire multi-dimensional data about RSOs over a wide range of in-orbit distances from less than a meter to 5km, with the potential to extend this range to 50km. Various information products are available from advanced GSFL embedded processing depending on the mission needs and concept-of-operations (ConOps). This information is available on-platform in real or near real-time, or for transmission off-platform for further processing and analysis. In order to support development of the GSFL and associated system interaction, a comprehensive simulation testbed has been developed. With this tool, scientists and engineers can work through concept development and feasibility studies, mission ConOps, and system design decisions at the highest levels. The simulation testbed fits into the Model Based System Engineering (MBSE) approach in a number of ways, including functional/behavioral and performance modeling. Additionally, the simulation testbed is an important part of Machine Learning (ML) and Artificial Intelligence (AI) workflows to generate the large organized point cloud datasets needed for Deep Learning Convolutional Neural Networks (DL-CNN), for example. As a tool for software development, the simulation testbed is used for unit and regression testing during development and sustainment, with realistic real-time dynamic scenarios and accurate GSFL organized point cloud data. The simulation testbed has been constructed with layers of abstraction and developed in an object-oriented manner for continuing capability extension and refinements as its use expands.
Date of Conference: September 27-20, 2022
Track: Machine Learning for SSA Applications