There are always these prodigious steps in our evolution that wreaks changes on our current way of life on a titanic scale. Be it the first time species were living on land instead of water or the Cambrian explosion that caused an outbreak of new species. In human time frame, we have evolved from hunter-gatherers to take leaps on to our planetary neighbors.
With every sector beaming with the boons of AI, Space exploration is no longer an exception. In fact, AI use in Space Exploration might be the most out of any other sector. With the constant influx of data, you cannot completely rely on human intelligence solely, as they are prone to many distractions. AI, however, gives outcomes with negligible errors which is why it makes sense to create a greater reliance on it.
Overview of AI
AI or Artificial intelligence is an umbrella term that entails algorithms to teach machines to think and predict as humans do (actually even better than the output expected of a human ingrained with the knowledge of their fields). These activities can include Speech recognition, speech to text, problem-solving or even detecting the distance of an object from another.
Daily interactions with AI applications are more common than we realize. If you have used OK Google to inquire about the temperature of your city or have unlocked your phone just by holding it up your face, you have already met an AI application.
Past achievements with AI
Wouldn’t it be cool to have your very own TARS on your trip to the barren soils of our celestial red neighbor? Okay! That was my way of praising Interstellar. Now coming back to the article, well, the day might be closer than many of us think.
AI has already revolutionized space exploration for quite some time now. The decommissioned satellite EO-1 (Earth Observing 1) has been helpful in gathering images of natural calamities such as hurricanes or volcanic eruptions. The AI onboard the EO-1 helped optimize the analysis and response for those natural disasters and even could relay the message to the satellite to take images of the occurrences before the ground crew would be alerted about the event.
Same goes with SKICAT (Sky Image Cataloging and Analysis Tool) which assisted researchers to discover objects during the second Palomar Sky Survey where it classified at least a thousand objects in low resolution. That’s a lot more than a human could identify.
Similar applications have helped astronomers identify over 50 probable cases of gravitation lensing. These play a crucial role in the discovery of dark matter (An elusive hypothetical form of matter that is believed to account for 85% of the matter found in the universe).
Recent Achievements with AI
By the end of 2017, AI helped NASA and Google in making the discovery of two obscure planets. One is a minuscule sized planet (compared to the other planets) in Kepler-90, now called Kepler-90i which is a super-earth-sized rocky planet. It revolves around its star closer than Mercury is to our sun. The star is located roughly 2500 light years away. Another planet found in Kepler 80, now called Kepler-80g, which is the 6th planet from its star.
This was exciting in the field of astronomy because the mountain of data stored in the Kepler telescope was fed into the AI systems. The scientists trained their neural networks using data sets that already contained detected planets. The goal was to see if it could differentiate between Real planets and false indications (could be nonplanetary bodies around their stars like dwarf planets, exoplanets or even asteroids) and turns out it exceeded the expectations of the team of researchers.
CIMON aboard the ISS
Envisioning a football-shaped buddy that floats around in space and interacts with astronauts aboard the ISS (International Space Station) while playing their favorite playlist?
Abbreviation for Crew Interactive Mobile CompaniON, it weighs around 5 kg and looks like a spherical shaped printed plastic ball. It is designed to test the interactions machines can have with humans in outer space. The building and developer company for this project was Airbus (on behalf of German Space Agency, DLR) and the location, Friedrichshafen and Bremen in Germany. The AI blended with the machine is developed by IBM Watson.
The applications of CIMON includes Search for objects, inventory management, documenting experiments, photography, and videography. It is also perfectly capable of hearing and seeing and thus it can act as a personal assistant like Cortana or Siri, just in Outer space!
Autonomous Exploration for Gathering Increased Science or AEGIS has been in use since 2016. This AI Algorithm has been on the Martian Curiosity rover. This has been helpful in selecting targets for its Chemistry and camera Complex remote geochemical spectrometer instrument. This is done in order to check the chemical composition of the soil in that region, with the use of laser-induced breakdown spectroscopy and high-resolution images of the targeted areas within the martian troposphere.
Since then the performance output has skyrocketed to 93% compared to 24% in the past When it didn’t have such an advanced AI algorithm.
Next Big Thing in AI for Space Exploration
Out of the two celestial neighbors; Venus and Mars, the latter provides the best-suited surface for humans to colonize. Venus having a temperature, scorching enough to melt lead, might not be the best fit to look for our next home. On the other hand, a smaller deserted ball of red soil and evidence of water is the closest thing we have for a new home.
When the seemingly red dot is closest to earth, the total journey (one way alone) can take up to 8 months. Not only would it be a very tempestuous voyage through space involving deadly radiation but even on the surface, explorers would be bombarded with dangers. This will include radiation from the sun, deadly dust storms that block the sun completely for weeks and the severely low gravity. So you’ll need a companion which is not affected by these factors. At least not that severely, compared to our fragile species.
Rollin Justin is a robot infused with AI applications. Researchers plan to ship Justin with a few humans to Mars to assist the latter in their studies regarding the further analysis of the Martian atmosphere, whenever we decide to visit up there!
Justin looks like a humanoid that can detect the differences between smooth and coarse plastic surfaces. It is also a great companion if you want to throw around a ball and have a game of catch on Mars. Also, adjusting the robot to the gravity of Mars will be easier than it would be for a human as lower gravity would distort the normal growth of human bones and cause other physical issues.
Also, the data received on Mars would be enormous and dependency on AI might help avoid errors while analyzing data and making predictions based on it.
GPS in Space
Getting lost on the surfaces of Rocky planets could lead one to a lot of problems. We have become exceptionally dependent on GPS today. It would be excruciatingly difficult to explore when venturing into other exoplanets or even the planets and moons in our solar system.
NASA Frontier Development Lab (FDL) has been developing an AI application that will make it easy to wander across the surface of Titan, Mars or the moon. Researchers Philippe Ludivig, Andrew Chung, Ross Potter, and Benjamin Wu are have worked on simulating the lunar surface to a local environment.
To create the surface of the moon here on earth, The team fed almost 2.4 million images of the moon captured, taken by a lunar rover. This helped simulate the surface of the moon inside the machine. This was sufficient for the moon to navigate through the moon’s topography with surprising accuracy.
The team is now planning to do the same with the Martian atmosphere, to see how effective the AI would be on other celestial bodies. If this could happen, Martian exploration might just ease up for scientists planning to go there.
Like all the applications of AI, we can conclude nothing concretely. Every new step to newer uses of AI opens a new door to gather fresh information and work it to human advantage. As we plan to send our first team of researchers to MARS, AI applications are only bound to expand.
We still do not understand the possibility of AI in the distant future. However, AI in space will make navigating through the vast empty nothing, the next biggest evolutionary step in our quest to gain a better understanding of the universe.
And who knows? Maybe be a substantial ally in our aspiration to colonize other planets!