Nebula Public Library

CubeSats have the potential to revolutionise Earth observation from space by exploring different operating points in terms of spatial, temporal, and spectral resolution. However, physical limitations, mostly in terms of size, prevent them from obtaining high-quality images. OPS-SAT, while characteriaed by a small form factor, provides an impressively capable and flexible hardware/software platform for executing demanding machine learning algorithms (inference stage).

In this activity, Mission Control Space Services deployed a low-level implementation of the OPS-SAT SmartCam model using a Field Programmable Gate Array (FPGA), comparing against a high-level CPU model using Tensorflow Lite. Experiments showed that the FPGA implementation reproduced the precision and accuracy of the high-level model, while running at a slower speed. Further optimisations of the FPGA are expected to close the gap in timing and unlock new methods for deploying deep learning on spacecraft.

In this project the authors looked at event-based sensing and processing for space situational awareness (SSA). Many advantages exist with the new paradigm of neuromorphic engineering, the authors investigate whether event-based optical data is suitable for the task and if processing with Spiking Neural Networks (SNN) provides advantages in terms of efficiency and efficacy. With the implementation of a novel heterogeneous LIF SNN, authors claim to surpass the state of the art in event-based SSA with a 15% increase in accuracy.

In this activity, a proof of concept of an attitude control system based on artificial intelligence, as an alternative to current state-of-the-art systems based on model approaches, was developed and tested in a simulator and in the engineering model of ESA's OPS-SAT.