Covering satellite automation software, ground-station network access and Copernicus Sentinel-1 image processing, the latest ESA InCubed co-funded activities tap into market needs from across the space economy value chain.
“InCubed supports the entire breadth of the commercial EO arena, from space assets and the ground segment to downstream applications, a fact that is nicely demonstrated by these initiatives with AIKO, Spaceit and KappaZeta. All three companies have identified areas where they can add remarkable value through innovative solutions, and ESA co-funding, technical and commercial de-risking support is helping them achieve a faster time to market – a critical factor for commercial success” – Giuseppe Borghi, Head of ESA Φ-lab.
AIKO: MiRAGE autonomous satellite management
Satellite manufacturers and operators need to optimise their resources and be able to respond rapidly to in-flight events, but the necessity of human intervention often presents bottlenecks and inefficiencies in mission management. To address these issues, Turin-based company AIKO has developed MiRAGE (Mission Replanning through Autonomous Goal gEneration), a Machine Learning-based onboard automation package that renders satellites much less dependent on ground control. MiRAGE analyses in-orbit and operational data to enable the spacecraft to identify and react to unexpected events, thereby lowering operating costs and improving the quality of delivered services.
“MiRAGE empowers satellites to operate autonomously, overcoming the limitations of human-centric mission control,” says AIKO CEO Lorenzo Feruglio. “The software paves the way for benefits such as improved cost efficiencies, increased activity lifespan and boosted resource utilisation. MiRAGE is also highly scalable and will be able to support constellation architectures in which hundreds or even thousands of satellites work collaboratively to reach mission goals.”
The two-year MiRAGE InCubed initiative includes completing product development and acquiring flight hours in order to train and refine the Machine Learning models.
Spaceit: Aggregated Marketplace for Ground Station Services
A further challenge for satellite operators is access to the terrestrial networks that underpin both mission control and data downlinking. Commercial and institutional operators alike need a low-cost, low-latency ground infrastructure at their fingertips, but currently there is a lack of one-point-of-entry communication solutions that are integrated with multiple ground station networks.
Spaceit OÜ, an Estonian start-up founded in 2015, saw a gap in the market for connecting satellite operators with ground stations. Silver Lodi is the company’s CEO and co-founder: “We realised there was a need for a single platform for ground station services, a solution that we were well placed to provide as part of our cloud-based Mission Operations Platform.” The new platform is called Aggregated Marketplace for Ground Station Services, and development under InCubed co-funding started in December 2021. Standalone ground stations, together with virtual and physical networks, will be brought together into one unified ecosystem, allowing users to book services, manage contacts and communicate with satellites in one cross-network environment.
The user interface for Aggregated Marketplace for Ground Station Services, along with the activity to integrate the solution with the ground station segment, are currently being developed, with the Mid-Term Review due in June. In parallel, commercial negotiations with partners and customers are underway, and the product is expected to be ready for launch by the end of this year.
KappaZeta: analysis-ready Sentinel-1 data
The Copernicus Sentinel-1 mission gathers frequent Earth observation data (currently every three to six days across Europe), using radar to generate medium-resolution images day and night with no interference from cloud cover. Sentinel-2 by contrast produces high-resolution optical images by day, but the presence of clouds may mean that capturing frequent, useful data at a given location is problematic. Crucially for example, cloud-free vegetation images in Autumn may be available only once a month from Sentinel-2, which is far too sporadic for efficient farm management.
The KappaOne solution seeks to provide the best of both worlds by using Artificial Intelligence (AI) modelling to fuse radar and optical imagery. Simulated Sentinel-2 images will effectively be created based on Sentinel-1 data and organised into data layers such as natural-colour satellite images and biomass estimate.
The solution is being developed by KappaZeta, which like Spaceit is based in Tartu in Estonia. CEO Kaupo Voormansik explains the significance of KappaOne: “Providing Sentinel-1 analysis-ready data for commercial and governmental entities is a landmark activity for KappaZeta. We see that Sentinel-1 imagery is largely underutilised and in its raw format is too complex for the majority of users. KappaOne will make data usage very easy – in fact the name derives from our aim to deliver one-click integration of calibrated, pre-processed information.”
Development of KappaOne kicked off in October 2021 with partial funding from InCubed. The first data layer services are due for release in the next few months.
To know more: MiRAGE, Aggregated Marketplace for Ground Station Services, KappaOne
Copernicus Sentinel-1 image courtesy of ESA/ATG medialab
Share