This project aims to understand and analyse the current Australian policy and legal framework in light of joining the Artemis Accords and, consequently, the US-led international block of space powers.

This project will research a novel deep learning pipeline for achieving robust and reliable forecasting of natural disaster events with hyperspectral satellite imagery. The majority of the existing machine learning algorithms do not possess the ability to forecast the occurrence of natural disasters in advance and they are only able to detect their occurrence when the disaster event happens. To address this limitation, we propose novel research which leverages the spatiotemporal modelling capability of deep learning to forecast the occurrence of natural disasters in advance using a satellite onboard execution environment. The research will be conducted using the example use cases from the application area of bushfire event forecasting considering the practical significance to Australia. The developed framework will be deployable on a CubeSat Kanyini-type mission with Hyperscout-2 payload and will use onboard hardware to execute the algorithm.

The aim of this project is to research, develop and test goal-oriented algorithms and software that will grant a spacecraft autonomous capability to undertake its mission robustly and adaptively in real-time. The activities will focus on coupling optimisation and machine learning techniques to orbital and sensing prediction models, such that when sensing data is obtained in real time the next most optimal action can be determined.

The autonomy will be experimentally tested using the DSTG Buccaneer Main Mission (BMM) spacecraft scheduled to launch in 2023. BMM features the MANTIS payload with a controllable, deployable arm for self-inspection imaging. The project output will be a set of algorithms, methodologies, and approach to grant the spacecraft the ability to take an optimal image of itself against a backdrop of Australia in real-time. The outcomes will be learnings of the relationship between on-board and off-board autonomy that will be applicable to other spacecraft missions.

Spacecraft autonomy has been recognised as a key enabler of the next-generation space systems that aim at increasing responsiveness and continuity of space-based observations, covering large areas with higher resolutions, minimizing communication and data access latencies, and reducing costs of both the space and ground segments.

Spacecraft autonomy encompasses onboard autonomous decision-making capabilities that enable the space segment to continue mission operations and to survive critical situations without relying on ground segment intervention. It relates to all aspects of spacecraft operations, including continuous mission planning and execution on board, real-time spacecraft control outside ground contact, maximisation of mission objectives in relation to the available onboard resources and capabilities of other spacecraft, and system robustness in presence of on-board failures and external uncertainties.

This project aims at addressing the above requirements by developing novel concepts, methods and technologies to provide new AI-based spacecraft autonomy capabilities for the next-generation space systems, such as dynamically networked formations of heterogeneous satellites. It focuses on high impact areas of spacecraft autonomy and onboard AI as identified and prioritised with the industry and defence partners, including:

• WP1: Onboard processing and actionable intelligence
• WP2: Small spacecraft and constellation resilience
• WP3: Dynamic optimisation of constellation resources
• WP4: Real-time tasking and resource allocation

The output of this project is a set of autonomous algorithms, demonstrating their capability through software simulations with use-cases provided by the industry partners.

This project will leverage and contribute to the IPC Capability Demonstrator with an aim to be demonstrated on DST RMS STaR Shot as a pathway to commercialisation of the developed technology solutions for industry partners.

This project will develop brand new capabilities for onboard AI processing and analysis of hyperspectral imagery on smart satellite platforms. In particular, the project will tackle the key modules of calibration, coarse and fine-grained segmentation, and joint space-ground inference of onboard AI processing of hyperspectral data. New capabilities in these areas will transform the ability of a satellite to automatically make sense of the rich and multidimensional spectral modalities in an end-to-end manner onboard the satellite itself. This will create new opportunities to enable accurate, efficient, and reliable automated detection and classification of natural phenomena and human activities over a wide area on Earth.

At the heart of the project, the research team will develop a novel multi-task learning framework for hyperspectral data. This framework will be employed to create a Panoptic Segmentation network: an approach which unifies object detection, semantic segmentation and instance segmentation in a single network to simultaneously predict a dense pixel-level segmentation across multiple spectral channels from space. In addition to this, the project will develop a lightweight deep learning based atmospheric correction network which can also be deployed onboard; and explore how joint learning between satellite and ground-based sensors can be used to support the inference of detailed information in areas not covered by ground sensors.

This Phase 1 project will develop a proof-of-concept demonstrator system, developing the key techniques for later optimisation and integration to run onboard the Kanyini satellite (SASAT1).

This research aims to provide a solution for energy-efficient AI-based on-board processing of hyperspectral imagery supporting automated early detection of fire smoke. We propose using modified and resampled MODIS imagery data that emulates the swath as well as spectral, spatial, and radiometric resolution of HyperScout-2 channel 1 hyperspectral imagery. In doing so, we intend to provide a solution that meets on-board processing limitations and up/downlink data transfer restrictions of the Kanyini – HyperScout-2/ with Intel’s Myriad X VPU chip.

Based on a semi-automatically created fire smoke training dataset, our proposed AI processing approach is performed at two levels: a) on-board feature and band selection, and b) ground AI neural network tasks – in order to optimize on-board processing and downlink data transfer. Expected outputs include on-board and ground AI algorithms for fire smoke detection, applicable for various hyperspectral imagery datasets.