Mapping sea ice in the Arctic is essential for maritime navigation, and growing vessel traffic highlights the necessity of the timeliness and accuracy of sea ice charts. In addition, with the increased availability of satellite imagery, automation is becoming more important. The AutoICE Challenge investigates the possibility of creating deep learning models capable of mapping multiple sea ice parameters automatically from spaceborne synthetic aperture radar (SAR) imagery and assesses the current state of the automatic-sea-ice-mapping scientific field. This was achieved by providing the tools and encouraging participants to adopt the paradigm of retrieving multiple sea ice parameters rather than the current focus on single sea ice parameters, such as concentration. The paper documents the efforts and analyses, compares, and discusses the performance of the top-five participants’ submissions. Participants were tasked with the development of machine learning algorithms mapping the total sea ice concentration, stage of development, and floe size using a state-of-the-art sea ice dataset with dual-polarised Sentinel-1 SAR images and 22 other relevant variables while using professionally labelled sea ice charts from multiple national ice services as reference data. The challenge had 129 teams representing a total of 179 participants, with 34 teams delivering 494 submissions, resulting in a participation rate of 26.4 %, and it was won by a team from the University of Waterloo. Participants were successful in training models capable of retrieving multiple sea ice parameters with convolutional neural networks and vision transformer models. The top participants scored best on the total sea ice concentration and stage of development, while the floe size was more difficult. Furthermore, participants offered intriguing approaches and ideas that could help propel future research within automatic sea ice mapping, such as applying high downsampling of SAR data to improve model efficiency and produce better results.
As Earth Observation (EO) is undoubtedly one of the industries that will benefit the most from the 4th industrial revolution driven by Artificial Intelligence (AI), the European Space Agency (ESA) Centre for Earth Observation (also known as the European Space Research Institute or ESRIN)) has contracted a team of contractors led by SpaceTec Partners to accompany its staff in exploiting opportunities of rapprochement between the AI and EO communities to the largest possible extent. To implement this ambition, the Artificial Intelligence for Earth Observation (AI4EO) initiative builds a community that fosters the interaction between AI and EO experts, developed a custom AI4EO platform, and host three challenges on the platform to tackle grand societal issues. The objectives of the challenges such as air quality or food security are to combine EO with an innovative AI method enabling the development of novel solutions. This paper introduces the AI4EO project and the first challenge.
The space industry is currently witnessing two concurrent trends: the increased modularity and miniaturization of technologies and the deployment of constellations of distributed satellite systems. As a consequence of the first trend, the relevance of small satellites in line with the "cheaper and faster" philosophy is increasing. The second one opens up completely new horizons by enabling the design of architectures aimed at improving the performance, reliability, and efficiency of current and future space missions. The EU H2020 ONION project ("Operational Network of Individual Observation Nodes") has leveraged on the concept of fractionated and federated satellite systems (FFSS) to develop and design innovative mission architectures resulting in a competitive advantage for European earth observation (EO) systems. Starting from the analysis of emerging needs in the European EO market, the solutions to meet these needs are identified and characterized by exploring FFSS. In analogy with terrestrial networks, these systems envision the distribution of satellite functionalities amongst multiple cooperating spacecrafts (nodes of a network), possibly independent, and flying on different orbits. FFSS are considered by many as the future of space-based infrastructures, as they offer a pragmatic, progressive, and scalable approach to improve existing and future space missions. This paper summarizes the main results of the ONION project and the high-level design of the marine weather forecast mission for polar regions.
