Using AI To Count And Map Craters On The Moon

by Alan Davison | December 31, 2020 4:34 am

It is no secret that technology is always advancing to become more efficient. This is particularly true for space tech.

Outer space has always been a vast and mysterious place that researchers from all over the world have dedicated countless hours to studying.

This is specifically relevant to the craters on the moon. Throughout the years, various tech has been used to study the moon’s craters.

However, recent A.I[1]. that has been implemented has proven to be by far the most remarkable.

The newest development of A.I. technology in relation to space study revolves around counting the moon’s craters. Previously, researchers had been counting the craters by hand.

This daunting process meant mapping out craters with telescope photos that were transferred to other maps or moon globe systems.

While telescope photos from space are generally high quality, this method simply wasn’t as efficient as it could be. In order to really change things up, a new system of mapping and counting had to be discovered.

Currently, the artificial intelligence that is being used to be map out the moon’s craters was studied by researchers from China, Italy, and Iceland.

Their findings, which are outlined in the journal titled Nature Communications, revealed that hand-counting and mapping simply weren’t doing the trick.

In fact, it is estimated that this technology found an additional 109,000 new craters on the moon’s surface. This is an incredible difference compared to physical counting.

But how does artificial intelligence programming work? Basically, it boils down to a super-smart algorithm that is trained with various data that is collected by lunar orbiters.

Simply put, the algorithm allows the machine to collect more information and imaging from the moon’s surface.

Chinese researcher Chen Yang who is an Earth Science Professor at Jilin University[2] in China has been the driving force behind the project.

Ultimately, Yang found that a specifically deep neural network had the ability to analyze previous moon crater data. This process set the bare-bones for the system to identify craters accurately allowing it to discover even more moon craters.

The newest algorithm has really paved a clear path in the way that researchers can study these magnificent dents on the moon’s surface.

While counting the number of craters is important, so is studying their size. The craters were said to form when the moon and Earth were created and are a result of the impact of meteor strikes.

Recent data from the A.I. has discovered that the indents created by the meteor strikes can span anywhere from 0.6 miles to over 60 miles (or 1 to over 100 kilometres).

This is really quite impressive and truly showcases the massive nature of impact from some of the meteors.

What is even remarkable is how this data compares to previously recorded data from the old method of mapping out the craters.

The old method was only able to record a cluster of around 5,000 craters at 12 miles in size. Moreover, this data reveals findings that weren’t originally noticed in relation to the mid and low latitudes of the moon’s surface.

This significant difference really gives researchers a more reliable way to study the moon’s surface[3]. The old method of doing everything by hand was not only tedious but simply not as reliable.

The naked eye can only see much and the telescope imaging is, well, imaging. Too much of the previous data could be left up to interpretation whereas the new algorithms provide a more detailed report of size and overall numbers.

The A.I. programming that goes into this more valuable technology is breaking barriers in the way that researchers can study outer space.

The technology behind this system could even be applied to future projects as well. In fact, the artificial intelligence behind this project has been able to help researchers determine the age of the craters themselves by using them like they would a fossil.

It seems crazy to think about, but researchers have reported that the data collected from the A.I.

has been able to detail craters within craters. However, the A.I. data isn’t the only factor in figuring out the ages of craters.

The Earth’s moon has no known atmosphere and certainly doesn’t have plate tectonics or water like the Earth does.

Because of the moon’s unique surroundings and composition, the craters that are being found with the A.I. algorithms are estimated to be around 4 billion years old.

Applying this technology can provide researchers with valuable and priceless information regarding other areas of the solar system as well.

In recent years, there have been rumors of using the technology in an attempt to scope out the surfaces of Mars, Ceres, and the moons of Jupiter and Saturn.

What makes this so incredible is that these surfaces are further away from Earth. Researchers are confident that technology will be able to withstand traveling at those distances.

Considering what the project has uncovered on the moon’s surface, it is exciting to think about what it might show on Mars[4], Ceres, or the moons of Jupiter[5] and Saturn[6].

These surfaces are just as uneven and roughly composed as the moon, so there’s no doubt that artificial intelligence will be able to count and map these surfaces as well.

Jupiter and Saturn have icy moons so the imaging data that is analyzed from those areas will certainly have much more to tell about the moons.

Much like how the programming has provided details surrounding the age of the moon’s surface, it could possibly do the same for Jupiter and Saturn’s moons. Once again, this is because of the overall conditions that the objects are in.

The combination of better technology and the right circumstances has definitely proven that this newer A.I. counting and mapping system is super-efficient.

As technology, in general, continues to progress, especially in concerns to space travel, there seems to be some promise of seeing so much more of outer space.

Outer space is a vast place and the utilization of programming like this can only move planetary studies even further.

Endnotes:
  1. A.I: https://en.wikipedia.org/wiki/Artificial_intelligence
  2. Jilin University: https://global.jlu.edu.cn/
  3. moon’s surface: https://www.space.com/19582-moon-composition.html
  4. Mars: https://en.wikipedia.org/wiki/Mars
  5. Jupiter: https://en.wikipedia.org/wiki/Jupiter
  6. Saturn: https://en.wikipedia.org/wiki/Saturn

Source URL: http://www.newsinsider.org/537/using-ai-to-count-and-map-craters-on-the-moon/