The time is coming where humans will start to explore the deeper reaches of space on a much more consistent level. NASA has already used unmanned shuttles, probes, rovers, and exploratory ships to find out more information about galaxies near and far. We have found out more about the size, scope, and shape of the universe because of technological advancements. Now, AI is going to help us do even more.
There is only so much we can do with our own minds, which is why artificial intelligence has been helping us for years with analyzing and interpreting the data that we get from space. Up until a few years ago, it was completely common to lose all information we had due to human error, lost bandwidth, and even computer viruses. Now, artificially intelligence machines are doing most of the heavy lifting and we have lost significantly less data since it has taken over.
The amount of data that our space programs get is insane. It is estimated that one satellite will produce 400 terabytes of data per year, which is crazy to think anyone would be able to collect, analyze, and implement all of it in a timely manner. That isn’t even all that large! Some telescopes and rovers may transmit up to 800 terabytes per day!
AI is able to get through all of that data quickly. It gets broken down onto individual computers and then put back together in a way that scientists can then understand and implement.
One of the biggest problems with the systems that we use for space exploration is that everything gets controlled from one place: Earth. It doesn’t always work out that well because we lose access to things all the time – millions of dollars are thrown away because we lose connection with a rover or a probe and we can never get it back.
Artificial intelligence will look to make these devices more autonomous. The hope is that AI and machines will learn to pick up patterns and be able to anticipate solutions to problems without having to wait for the answer from earth. This is already happening, so it shouldn’t be too far away.
In that same vein, there has been proof that we haven’t succeeded in space exploration as much as we want because of the lack of adaptability that our systems have. Science is great, but the technology we have now and the capabilities we use are fairly rigid. We don’t know everything about space yet, so it is difficult for us to understand that things can be slightly off from what we predict. We just recently saw that with Philae, a program that cost millions of dollars and it was unable to operate properly and get useful samples because gravity was slightly different than it was calculated to be.
Some of our ships, probes, and rovers are using technology that is over a decade old – AI will change that.
Currently, NASA has two programs that will help on this front. The first is the ASPEN network that is based on techniques we already use with AI to set up different systems that can react in different ways depending on what can happen. Of course, this is still somewhat rigid in comparison to what we need. However, there is also the creation of a temporal reasoning system, which will allow the machines to think and operate for themselves somewhat.
NASA is also working with the Autonomous Sciencecraft Experience (ASE), which has been helping with onboarding science algorithms to help with AI identification using image data, execution management software to allow for low-level autonomy, and Continuous Activity Scheduling, Planning Execution, and Re-planning (CASPER) to help re-plan activities after something does go wrong. This has all been in effect since 2003, so great strides are being made in this particular program.
Artificial Intelligence is going to change the world and the universe because it is going to take things that should be impossible and make them possible. NASA relies on artificial intelligence to process data, help with navigation, save time, and in some cases even save lives. Without AI, our dreams of living on Mars, making it beyond Pluto, and even learning what else is out there would be nonexistent.