Space domain awareness requires advanced sensing, data processing, and algorithms. Such sensing requires large amounts of data collection, which in turn requires automated reduction of data to highlight for a space guardian the most relevant activities in space. The emergent areas of investigation into smart sensing machine learning for automated reduction of data are areas that could potentially revolutionize how ground-based remote sensing of satellites and space debris is accomplished.
Characterization of space objects is a remote sensing problem. By using photons reflected or emitted from the space object one seeks to understand the properties of that space object. Properties can be as a function of time, either over long durations such as months and weeks, or seconds to minutes. Non-traditional sensing methods, multispectral, infrared, and fusion of information can all reveal characteristics of a space object. Many times determining the characteristics of a space object from observations is a classical inverse problem. Machine learning techniques and novel sensing techniques have been explored in the past (references available). However, Machine Learning algorithms have advanced dramatically in the past several years and we seek advanced techniques to solve remaining challenges in space object characterization. Additionally, new modalities for sensing, to include “smart sensors” which permit fusion of diverse sensing methods promise to provide new types of data, potentially enabling enhanced characterization of space objects.
Sign up now to be invited to the webinar series where we’ll discuss the context behind each problem statement and answer questions from startup and university teams.