Funding the future of European space through OSIP

ESA set up the Open Space Innovation Platform (OSIP) to discover and invest in new unconventional ideas that could greatly benefit and advance European space industry and academia.

Here, the minds behind five projects funded during this period tell us about their research, motivations and goals, as well as how ESA Discovery funding is helping them take their activities to the next level.

An AI Foundation Model to better understand our planet

As AI is revolutionising the way we analyse data, a team from École Polytechnique Fédérale de Lausanne (EPFL), Switzerland is studying how to create the first language-powered, sensor-agnostic AI Foundation Model that would be able to analyse independently remote sensing images from different Earth Observation (EO) satellites.

The model has been trained to extract information from multiple satellites (for example Sentinel-1 and Sentinel-2) over the same area, showing it can interpret new input efficiently and perform tasks beyond its training data sets.

Gencer Sümbül, scientist at EPFL’s Environmental Computational Science and Earth Observation Laboratory highlights the novelty of the project: “Developing an AI model that can understand and interpret Earth observation images from different satellite sensors in a truly “sensor-agnostic” way has not been done before. It’s critical (yet challenging) because the data varies widely in resolution, spectral ranges covered, and revisit time. We want a model that can integrate all of that seamlessly to help users get more meaningful insights.”

“The next phase of the project will focus on enabling the model to fuse visual and language semantic information. If successful, it could retrieve images from different sensors based on descriptions provided in human language or provide specific answers about the content of the images in human language, facilitating interaction between humans and EO data,” says Gabriele Meoni, Innovation Officer at ESA.

“The funding and expert guidance from ESA Discovery gave us the momentum to push our research toward a large-scale model and its actual use for Earth observation scenarios. The support of ESA is vital to focus our efforts on developments that really have the potential to impact the way we do multi-sensor remote sensing today,” adds Gencer.

ESA has been conducting extensive research on Foundation Models for EO over the past years, and in collaboration with NASA organised the “International Workshop on AI Foundation Model for EO” in May this year.

Testing for Moon and Mars on Earth: a groundbreaking lab

Researches from the Danish Technological Institute (DTI) and Aarhus University (AU) are studying together how to build and operate a full scale test and research facility for large space assets, including Lunar and Martian roving vehicles and exploration robots under driving conditions.

The partnership builds on the expertise of both institutions, with DTI leading the operations and providing start-up incubation support, while AU spearheads planetary research.

The SpaceSite Lab aims to address critical gaps in ground-based testing facilities by studying a full-scale (Ø30m diameter, 7m high) Dusty-Windy Thermal Vacuum Chamber (DWTVC), capable of simulating Moon and Mars environments with realistic pressure, temperature, and mobilised dust conditions. These testing capabilities are essential for exploration programmes such as ESA’s Terrae Novae Programme or Argonaut, Europe’s lander programme.

The current study looks at key factors crucial for informed decision-making, such as the technical configuration, architectural design, supporting infrastructure, public-private partnerships, the estimated cost, and a comprehensive stakeholder survey.

“The support from ESA Discovery via OSIP has been instrumental in conducting the feasibility study, including the creation of an architectural design for a facility specifically tailored to meet the demands of testing large space assets. It has enabled collaboration with researchers from Aarhus University and brought in the globally renowned architects from BIG - Bjarke Ingels Group, all working in partnership with ESA to bring this visionary idea forward,” says Christian Dalsgaard, Senior Consultant at the Danish Technological Institute.

The commercial potential of the lab goes beyond space applications. “The facility offers value in terrestrial applications such as measuring aerosol emissions from vehicles and heating appliances, testing large ventilation systems, and studying the response of plants and crops to specific atmospheric conditions,” says Christian.

How will AI guide objects in space for in orbit servicing

As in orbit servicing becomes a reality, how will spacecraft be able to navigate close to each other? A team of researchers from Politechnico di Milano, Italy are looking into autonomous AI-aided relative navigation to improve sustainability for space exploitation.

In this scenario, a "Chaser" spacecraft approaches and interacts with a "Target" spacecraft in close proximity, under the risk of collisions and facing uncertainties in the Target’s shape and optical properties, which are never well known in advance and could even change once in orbit. The Chaser has to continuously estimate and control its position and attitude relative to the Target. The AI-based relative navigation algorithm is specifically designed to handle these uncertainties.

The algorithm will be tested in two ways: on a space-certified processor at ESA to check it can be easily integrated with spacecraft hardware, and at Thales Alenia Space in Cannes to ensure it can handle unforeseen changes in the Target’s appearance.

There is a strong commercial potential for the near future, as many companies are now developing solutions to remove space debris and support in-orbit refuelling and repairs. “Demonstrating that the algorithm is ready to operate on existing hardware is a key milestone, as it would confirm its reliability and effectiveness for commercial use,” says Paolo Panicucci, Assistant Professor at the Department of Aerospace Science and Technology,  Politecnico di Milano and part of the DART lab.

About the ESA Discovery programme, Paolo comments: “In addition to the valuable funding, OSIP facilitated interactions with ESA experts, helping to shape the project and ensuring high technical quality. We gained access to space-grade equipment not otherwise available to us. Co-funding the research with an industrial partner provided further access to essential testing facilities. This is a fruitful partnership between academia and industry, enhancing the project’s overall impact.”

On ESA’s side, “the ongoing research represents a key step in bringing AI “in-the-loop” of our Guidance Navigation and Control (GNC) subsystems. It addresses a very relevant scenario for ESA – servicing with a partially known target – while tackling the challenges of the limited computational resources available onboard,” says Francesco Capolupo, GNC Systems Engineer at ESA.

A pencil to clean dust in space

As humans prepare to return to the Moon and explore other planets, a challenging enemy encountered by the Apollo missions is back in focus: dust. Fine, abrasive, and electrostatically charged dust poses risks to equipment, habitats, and astronaut health.

To solve this issue, a research team from the Romanian National Institute for Laser, Plasma and Radiation Physics (INFLPR) is studying how to efficiently remove different types of dust from surfaces using a pulsed electron beam.

The novelty of the idea is that the pen could work on any type of dust, without the need to pretreat affected surfaces, overcoming the limitations of current methods. The removal technique employs a collimated pulsed electron beam (e-beam) with relatively high energy (~13 keV) to effectively expel dust particles from surfaces. “It directly transfers momentum from energetic electrons to the dust and does not require specialised surface coatings, embedded conductive wires, or differentiation based on dust properties, making it a versatile and efficient solution,” explains Prof. Dr. Cătălin Ticoș, Senior Scientist at the Data Science & Plasma Physics Group, INFLPR.

Part of its space exploration objectives, ESA is looking into dust-resistant materials and understanding the effects of dust on relevant surfaces. “This activity adds a new tool to lunar and planetary dust mitigation strategies for future exploration missions. Impact of dust can be detrimental on optical surfaces, thermal control surfaces, mechanisms or even astronaut suits due to its charging and abrasiveness,” says Malgorzata Holynska, Materials and Processes Engineer at ESA.

As the technique provides an effective solution for removing dust from surfaces in low-pressure environments, it could have exciting terrestrial applications as well, with great commercial potential. The team is currently exploring ways to scale the project, planning to develop a prototype tailored for specific applications, including spacesuits, solar panels, and optical sensors. “Moving forward, we are pursuing two key directions: filing an international patent and exploring commercialization through a startup,” says Cătălin.

“We received valuable feedback from ESA specialists, highlighting specific scientific challenges we had not previously considered. Expanding our research scope to address these issues proved highly beneficial, enhancing both the depth and impact of our work,” adds Cătălin.

Better sensing satellites: the next generation of magnetic sensors

Measuring magnetic fields in space helps us study a planet’s interior structure, track space weather, navigate spacecraft and keep them safe from charged particles coming from the Sun and cosmic rays. Current space magnetometers are limited by size, weight, power needs and harsh space conditions. They also suffer from interference and stability issues.

Making magnetometers smaller and more efficient makes it easier to fit them on compact spacecraft. New designs also aim to stay stable over long periods while measuring the full magnetic field direction.

To address this, a team from Hasselt University, Belgium is looking into a new type of magnetometer using diamond-based quantum sensors (NV centers), providing a compact, accurate, and long-term stable sensing solution.

The goal is to develop a compact magnetometer with high accuracy and long-term reliability, making it well suited for future space missions, including deployment on nanosatellite constellations. “Based on lessons learned from operating the first diamond quantum sensor aboard the International Space Station within the OSCAR-QUBE mission, we are addressing the identified limitations by innovative methods, while also taking the first steps toward on-chip miniaturization of this technology with future Earth observation missions in mind,” says Prof. dr. Jaroslav Hruby, Visiting Professor at Hasselt University.

“Magnetometers are key instruments across many ESA missions, from studying Earth’s magnetic field to exploring other planets and supporting fundamental physics experiments. Nitrogen-vacancy (NV) diamond magnetometers are a promising new approach: they offer vector measurements, robustness, and wide operating ranges in a compact form. This activity looks at how such sensors could be miniaturized for space use, de-risking different design concepts early so ESA can focus future investments on the most effective architectures,” says Aaron Tilmann Strangfeld, Quantum and Emerging Sensing Technologies Engineer at ESA.

About participating in an ESA Discovery activity, Jaroslav comments: “It enabled us to explore new and disruptive concepts with direct and fast feedback. As a student-driven team, it also gave us the chance to sustain opportunities for future engineers and scientists to step into the field of quantum and space research, which can strengthen Europe’s position in these emerging fields in the longer timescales.”

Discover more about ESA's Open Space Innovation Platform, including how you can submit your own ideas, via our dedicated webpage.