Write a comment
Advanced Space, LLC has successfully deploying machine learning tools for space navigation technology in cislunar space. The CAPSTONE spacecraft embarked on pioneering software tests, establishing a foundational shift towards autonomous orbital navigation.
This advancement, known as SigmaZero, employs a Neural Network (NN) to identify and address navigational challenges, such as detecting by Clarence Oxford
Los Angeles CA (SPX) Mar 22, 2024
Advanced Space, LLC has successfully deploying machine learning tools for space navigation technology in cislunar space. The CAPSTONE spacecraft embarked on pioneering software tests, establishing a foundational shift towards autonomous orbital navigation.
This advancement, known as SigmaZero, employs a Neural Network (NN) to identify and address navigational challenges, such as detecting subtle accelerations that could otherwise disrupt the spacecraft's course. Crucially, SigmaZero facilitates key operations like orbit determination in scenarios lacking immediate ground control support, a critical development for future space exploration and settlement.
The initiative is part of a collaboration with the Intelligence Advanced Research Projects Activity (IARPA) under the MicroE4AI (Microelectronics for Artificial Intelligence) program. SigmaZero's capabilities reflect Advanced Space's mission to pioneer technologies enhancing spacecraft autonomy through machine learning (ML).
IARPA's director, Dr. Catherine Marsh, expressed enthusiasm about the SigmaZero experiment outcomes, highlighting them as key to the MicroE4AI program's objective of fostering resilient computing tools that propel microelectronics and software innovation.
SigmaZero was tested onboard the CAPSTONE spacecraft, demonstrating anomaly classification capability. Initial tests involved simulated data mimicking unaccounted spacecraft maneuvers, with the software's analysis aligning perfectly with ground-based computations. A subsequent trial expanded SigmaZero's application across nine distinct anomaly scenarios, confirming its efficacy in accurately identifying and classifying navigation errors, a testament to its potential in both cislunar and interplanetary missions.
Dr. Nathan Re, Principal Engineer and SigmaZero's Project Lead, emphasized the software's capacity to instantly process and understand navigation data, a task traditionally demanding expert human analysis. Tyler Hanf, a Software Engineer, highlighted the accelerated software development and testing enabled by ML, underscoring the importance of automation in ensuring the safety and reliability of burgeoning spaceflight activities.
Beyond navigation, the SigmaZero tests also offer insights into optimizing flight software management and deployment, promising further advancements in spacecraft operation and mission flexibility.
Related Links
Advanced Space
Space Tourism, Space Transport and Space Exploration News