Skip to content
Home » Blog » Data Science in Space: ready for launch!

Data Science in Space: ready for launch!

Mix Space Exploration with Artificial Intelligence and you already have the plot for a science-fiction story. But reality may not be that far away from this picture as we might think. In this post we write about actual applications of Data Science in Space: from satellite data applications, to already existing solutions for satellite operations, and space mission assistants.

If you are a fan of science-fiction stories, when you think of Artificial Intelligence (AI) in Space, characters such as TARS, the sarcastic spacecraft assistant from Interstellar; Marvin, the depressed robot from the Hitchhiker’s Guide to the Galaxy, or the scary Hal 9000 from 2001: A Space Odyssey, may come to your mind. While all these stories sit somewhere in the future, what is the actual role Artificial Intelligence plays in Space Science today?

From Apollo missions to the XXI century, the other Space Odyssey 

Space technology and exploration are rapidly growing with signs resembling what once was a golden era. Both private and public organizations are investing in a sector foreseen to have a huge economical return in the next decades [1, 2].

Besides the economical reasons, for us scientists, space is the ultimate frontier we have always dreamt to explore*.

Our parents’ generation saw the first man stepping on the Moon’s surface. The 60s and the 70s were two decades where the sense of wonder emerged and inspired many young people to pursue scientific and technological careers. The futuristic stories science-fiction writers once imagined in the 40s and 50s were becoming something real. The future was already here.

However, in the 80s, there was a sudden drop of interest in space exploration, linked to the end of the Cold War. The rapid growth of the space industry seen in previous decades came to a stop.

In the 90s and early 2000s many successful scientific missions were launched, such as the incredible international and collaborative challenge represented by the International Space Station (ISS). But the general interest for space exploration kept dropping. The canvas of the XXI century with settlements on the Moon and Mars was no longer in the public imagination. Space did not turn out to be what once was promised.

Space Industry in the XXI century

Today, the space industry has experienced a rebirth, and it is again a source of inspiration and motivation for many. Thanks to the rise of new companies such as SpaceX, United Launch Alliance, Blue Origin, or Virgin Galactic, among many others, and the interest of public agencies such as NASA and ESA, without forgetting the Chinese Space program, the space industry is in its best moment since the space race era in the 60s and 70s.

Synchronized landing of the SpaceX Falcon Heavy two side boosters (Image credit: SpaceX).

Technological advances and cost reductions led to new companies entering into a sector that was reserved only to public agencies in the past. A NewSpace community has been created with the aim to search for new commercial opportunities in the exploration and use of space technology. The focus is oriented into specific services derived from low-cost satellites, and the data obtained from them.

The CubeSat revolution

One of the key ingredients in manufacturing low-cost satellites has been the development of CubeSat standards [3]. With them, companies and research institutes can nowadays manufacture their own satellites in a flexible manner, and launch them to space at a low cost.

CubeSats nanosatellites (Image credit: NASA)

All these incentives have translated into an explosion of new small companies dedicated not only to build satellites, but also to build new launch and propulsion systems that bring these small satellites to Low Earth Orbit. Satellite and rocket development are a multidisciplinary effort that merge multiple technologies, from robotics, and 5G, to Artificial Intelligence.

AI meets Space

While artificial intelligence has experienced a huge growth of applications on Earth during the last years, this has not yet been the case in the space industry. Fully integrating space technology and artificial intelligence has the potential to revolutionize both sectors and bring them another step forward. A new domain full of possibilities.

There is still a long way to go until we see advanced AI tools extensively implemented in the space industry, on space missions and beyond. Some of these are currently in development, or in the first stages of their deployment in space.

Satellite Data Processing

Impactful applications of AI in Space are available for satellite data processing and intelligence. Satellites have a privileged view of the planet Earth. Thousands of Low Earth Orbit objects are continuously collecting data from our planet. This data contains very valuable information about many aspects of our planet.

Sentinel-2 satellite (Image credit: ESA)

An example of such applications is the Copernicus project launched by the European Space Agency. Copernicus is served by a set of satellites (including the Sentinel families), and in situ stations (ground, air, and sea), with the mission to better understand our planet and sustainably manage the environment we live in. As part of this mission, Copernicus has launched a call for the development of innovative AI solutions (AI4Copernicus) to extract information from the collected data [4]. It focuses on domains of high economic and societal impact, such as Health, Agriculture, Energy, Security, and others like Maritime or Aviation. At PickleTech, we are working ourselves on one of these solutions, we will talk about it in a future post.

Satellite Operation and Space Debris

Another area where AI can have a significant impact is in satellite operations. In particular, AI can support the operation of large satellite constellations, including its relative positioning, communication with Earth, and end-of-life management. Coordinating thousands of objects is a task that might not be possible to carry out by a human. In contrast, AI solutions can quickly analyze and react to input signals received, vital for satellite operations.

One of the consequences of misoperation are collisions between satellites. The direct product of these collisions is space debris. In order to avoid the increasing amounts of space debris, the ESA and the German Research Center for Artificial Intelligence (DFKI) launched a technology transfer lab that works for satellite autonomy and collision avoidance [5].

Space Missions

Whether Mars or the Moon, huge amounts of data come from each space mission. Autonomous spacecrafts equipped with artificial intelligence may be very useful to reduce mission costs and to explore uncharted regions of the Solar System.

In that direction, some AI tools have already been developed, as it is the case with the last Mars rovers [6]. In order to provide autonomy to the vehicle when exploring the terrain composition and geochemistry, the rovers are equipped with computer vision solutions. They may not find traffic jams in Mars, but risks such as dangerous cliffs or sand banks must be avoided. These are edge computing tools, beneficial to avoid the up to 20 minutes communication delay between the Earth and the red planet.

CIMON assistant a the International Space Station (Image credit: Airbus)

The ISS has an onboard AI assistant in their everyday procedures called CIMON (Crew Interacting Mobile Companion) [7]. Although it does not look as fancy as the onboard assistants found in the movies, its job is essential to reduce astronauts’ stress. CIMON allows the astronaut to keep both hands free, with no need to manually operate a computer. A fully voice-controlled access to documents and media, allows the astronaut to navigate through operating and repair instructions and procedures for experiments and equipment. CIMON serves as a complex database of all the necessary information for working on the ISS, and can also be used as a camera for documentation purposes.

NASA is also developing a companion for astronauts aboard the ISS, called Robonaut [8], to work alongside the astronauts, or take on tasks that are too risky for them. 

As these examples show, the space industry is evolving towards a full autonomous approach to missions.

PickleTech goes for launch!

At PickleTech we love innovation, and we are constantly thinking of ideas and developing projects in sectors that excite us. Space is one of them, because Space is part of who we are.

Space exploration was an important booster of our motivation to pursue scientific careers years ago. In these careers, we have experience analyzing data coming from satellite imagery to develop impactful applications on Earth for past projects.

Reconstructed maps for different elements from Sentinel data (Image credit: ESA)

And currently, we are developing ideas aligned with the spirit of the AI4Copernicus program, as well as engaging with experts in the Space industry to exploit the possibilities of AI in Space.

Space exploration combined with Artificial Intelligence opens the door to a new world only limited by our imagination. An exciting future awaits ahead of us, at PickleTech we take part in it.


* As kids, space was the ultimate frontier we dreamt of. Until the ultimate frontier became a collider facility underground…

References

[1] https://hbr.org/2021/02/the-commercial-space-age-is-here

[2] https://www.cnbc.com/2020/10/02/why-the-space-industry-may-triple-to-1point4-trillion-by-2030.html

[3] https://www.cubesat.org/

[4] https://ai4copernicus-project.eu/

[5] https://www.dfki.de/en/web/news/esa-dfki-transferlab0

[6] https://www.nature.com/articles/d41586-017-00626-6

[7] https://www.airbus.com/en/newsroom/press-releases/2020-04-cimon-2-makes-its-successful-debut-on-the-iss[8] https://www.nasa.gov/audience/forstudents/5-8/features/nasa-knows/what-is-robonaut-58.html

EN