Dominique Low, MDA Systems; Christos Koulas, MDA Systems; Domenico Di Giovanni, MDA Systems
Keywords: AI, artificial intelligence, neural network, machine learning, Space Domain Awareness, Space Situational Awareness, GPU, on board processing, orbit determination, simulator
Abstract:
The amount of space debris and satellites in orbit is increasing, making the on-board detection of Resident Space Objects (RSOs) ever more important for spacecraft situational awareness and collision mitigation. On-board low-cost optical sensors are increasingly being used to produce image data for Space Situational Awareness (SSA). The ability to accurately interpret optical sensor image data for RSOs and propagate their orbits rapidly is a fundamental building block for applications like automated flight dynamics for collision avoidance.
The RSO Detector Convolution Neural Network (CNN) is a notable detection model developed by MDA with support of the Canadian Space Agency (CSA). The model extracts the position information for RSOs in optical images of star streaks, and was trained with simulated images based on Sapphires principal payload. Verification of the model was based on the proximity of predicted RSOs to the position of the truth mask RSOs. The model was able to achieve 83% accuracy with fully simulated on-orbit data. Two-thirds of the detection errors are attributed to missed RSOs, and one third attributed to false positives. Detection performance improved when reducing features associated with non-nominal image data, suggesting that broader on-orbit optical image datasets and greater temporal coverage would improve model accuracy. The RSO Detector CNN was then run on a configurable Graphical Processing Unit (GPU) in an effort to mimic a space qualified SoC, making RSO location predictions with an average speed of 23ms.
The success of the RSO Detector CNN demonstrates that this technique would be suitable for closed loop applications such as automated flight control and other time-critical operations. This capability will prove to be key when integrated with an onboard SSA system.
Date of Conference: September 27-20, 2022
Track: Machine Learning for SSA Applications