top of page

Artificial Intelligence in Space Exploration

By TST Astronomy Department


Artificial intelligence (AI) has been making strides in various industries, and space exploration is no exception. AI technology is increasingly being used to assist with space exploration, from launching rockets to managing space stations. In fact, AI has become a crucial tool for space exploration, helping to enhance efficiency and accuracy while reducing costs and risks. Exploration of this void cannot be done solely by the minds of humans and we, indeed, needed help to explore this space further. This does not mean we cannot explore space without the help of AI, but it has only accelerated our step toward the unknown.


AI has served as a dominating tool in the exploration of space. While AI was commonly used before in aviation, say autopilot mode in airplanes, and industries, it serves a much more prominent role in this segment. For Example -

  1. Helps to optimize the use of fuel in space flight.

  2. Ensure satellites don’t collide with each other by creating intricate trajectories.

  3. Navigation and capturing essential images of objects by rovers, etc.

AI is surely an intricate structure and understanding it is in itself a job. We surely are not here to understand AI but how AI is for space exploration. Above mentioned points are just a few positives to give you insight into the power and usefulness of this tool. Read below to have a detailed understanding of this dominating intelligence in space.


The Ways AI Has Dominated in Space Exploration

  • Space Flight: AI is used extensively in this area particularly from coordinating the balance of a space shuttle during launch and landing to optimizing the use of fuel. Space exploration wouldn’t be much blooming if Space flight wasn’t a factor in it. The use of AI in space flight is not something new - SpaceX, Nasa, and other industries have been using it for some time now.

For example:

  • SpaceX used an AI autopilot system to enable its Falcon 9 craft to carry out independent operations such as docking with the International Space Station where it carries out cargo deliveries for NASA.

  • SpaceX’s autonomous reusable rockets calculate the trajectory of the rocket before impact, taking into account fuel usage, atmospheric interference, and “sloshing” from liquids within the engine.

  • CIMON 2, a robot designed by Airbus, navigates using fans to propel itself within spacecraft interiors.

  • Robotics In Planetary Exploration: AI is exploitably used in robotics for interplanetary exploration. The robots they use are predominantly termed and designed as rovers, as it allows easy movement of the rover on the rocky surface of this unknown planet. Rovers navigate and create trajectories whilst avoiding craters and potholes for their routes aided by the intricate mind of AI. They analyze data and learn the results of the samples collected from the surface and high-resolution images. Perseverance, the most powerful rover launched by NASA on Mars, is one such example. It accommodates AEGIS, a computer-vision-based detection system, to find captivating rocks to sample and eye out for threats. Another experiment by ESA (European Space Agency) is the creation of Hopper bots that can navigate through legs like animals and perform leaps.

Curiosity, the predecessor of Perseverance, is another autonomous rover drifting over the surface of Mars for over a decade and the fun fact is - It’s still operational!

  • Data Analysis: Artificial intelligence (AI) has revolutionized data analysis in space exploration. With the enormous amount of data collected by space probes and telescopes, AI algorithms can quickly and accurately identify patterns and anomalies that may be missed by human analysts. AI can also help in predicting future events and providing decision support for space missions.

For example, NASA's Mars rover, Curiosity, uses AI to analyze images of the Martian terrain to determine where to collect soil samples. In addition, NASA's Kepler telescope uses AI to analyze data from over 150,000 stars to identify potential exoplanets. AI also plays a vital role in space traffic management, helping to prevent collisions between satellites and space debris.

Perseverance 1st attempt to collect soil samples of Mars


  • Mission Design and Operations: AI has begun to improve the way we plan for space expeditions accounting for risks, safety precautions, and making real-time adjustments. Take the Perseverance Rover, for example. During the mission to get Perseverance to Mars, the Entry, Descent, and Landing (EDL) flight dynamics team relied on AI for its complex scheduling and mission planning to ensure a safe landing from the top of the atmosphere of Mars to the ground, which takes about 7 minutes. The AI can schedule the times for the spacecraft that has been supporting the rover ejects off. It can also plan for the employment of the heat shield when the rover enters the Martian atmosphere, which now flies itself, it can plan for the deployment of the parachute, and use all the past data to compute its landing on Mars's surface. During landing time the AI has to remove the heat shield and use the Terrain Relative Navigation system to successfully land.

AI can take in data collected from scientists in the past and apply it, in real-time, to independently carry out tasks and help with scheduling, which would otherwise need large amounts of manpower.


  • Exoplanet Discovery: AI can also be used to discover if a frequency picked up by the Kepler Telescope, a telescope launched by NASA in 2009, is either a false positive or an actual planet outside the solar system. The convolutional neural network (CNN) consists of artificially made neural networks that can be applied to analyzing visual images. AstroNet K2, developed by Google, is a CNN that works to identify if signals emitted from the Kepler Telescope are false positives or actual exoplanets. This autonomous deep-learning machine had expanded the capabilities of K2 by being able to systematically remove the noise that prohibited the K2 from discovering exoplanets at a faster rate, and spot exoplanets that experienced astronomers had missed.


Hold on, is there a catch?

  • Extreme Temperatures and Pressure: A huge challenge presented to AI, such as rovers, is extreme temperatures and pressure changes. The temperatures in space are extremely different from temperatures on Earth, which is why it is key to design the AI in a way where it can adapt to the change in temperature and atmosphere when being launched into space from Earth. Improper testing and research may lead to damage to the AI’s material or malfunctions to the systems, which could then cause corruption of data, solar flares, or the destruction of spacecraft electronics. Depending on where you are in space the temperature and pressure can shift. Take Mars for instance. Mars is the fourth planet from the Sun, while Earth is the third planet from the Sun, which means that it is colder on Mars than it is on Earth. To give you an example of how colder it would be think of it as minus 200 degrees Fahrenheit or minus 129 degrees Celsius. The AI equipment needs to be designed so that its systems do not freeze.


  • Sound: In addition to temperature and pressure, sound can also damage AI equipment such as a rover. When designing a rover scientists have to keep in mind that during the launch, it will encounter an enormous amount of sound - about 204 decibels. A detrimental volume for a human’s hearing will also be detrimental to the AI itself. Note that this challenge only appears during launch time because in space there is no sound due to the lack of vibrating air particles that transmit sound on Earth. However, it is still a problem because this sound could cause some fasteners and electrical cables to loosen during liftoff.


The Future of AI in Space Exploration

There are many possibilities for the future of AI in space exploration. Every day scientists are working to discover new uses and applications of AI and how it can help humans combat challenges in space. NASA is currently working on designing AI equipment for its mission to the Sun. Nasa’s Parker Solar Probe is currently being designed to explore the Sun’s atmospheric corona, which is the outermost part of the Sun’s atmosphere. As far as the integration of AI in space exploration goes, we can already see these advancements being done in this particular area with SpaceX and other industries constantly working on improving space exploration and the AI that goes within it. From making reusable shuttles that dock back after the launch to creating the trajectories that obviously require AI, we have come a far way and this is just the beginning of an era!


Sources

  1. https://ai.jpl.nasa.gov/public/projects/ops-for-autonomy

  2. https://www.springboard.com/blog/data-science/ai-space-exploration

  3. https://www.forbes.com/sites/bernardmarr/2023/04/10/artificial-intelligence-in-space-the-amazing-ways-machine-learning-is-helping-to-unravel-the-mysteries-of-the-universe/?sh=58386e1f7b60

  4. https://mars.nasa.gov/news/8671/nasas-perseverance-rover-goes-through-trials-by-fire-ice-light-and-sound/

  5. https://www.bbc.com/future/article/20230306-just-how-loud-is-a-rocket-launch#:~:text=Nasa's%20measurements%20at%20the%20time,for%20longer%20than%2030%20seconds.

  6. https://www.esa.int/ESA_Multimedia/Images/2018/04/Comparing_the_atmospheres_of_Mars_and_Earth#:~:text=Mars%20is%20about%20half%20the,rich%20in%20nitrogen%20and%20oxygen.

  7. https://www.jpl.nasa.gov/videos/7-minutes-to-mars-nasas-perseverance-rover-attempts-most-dangerous-landing-yet

  8. https://www.aiacceleratorinstitute.com/ai-in-space-exploration/#mission-design-and-operations

  9. https://www.technologyreview.com/2019/04/01/136239/deep-learning-has-found-two-exoplanets-that-human-astronomers-missed/


Comentarios


bottom of page