How Can AI Assist Humans in Space Exploration? NASA* Astronauts Tell Us What They Think

Part 2: Intelligent Robots Can Help Us Colonize Mars and Worlds Beyond

By MaryLiz Bender 

AI Astronauts
From left to right: Nicole Stott, Tom Jones, Story Musgrave. Credit: NASA*

In the first article of this series, we heard from experienced NASA* astronauts, Nicole Stott, Thomas Jones, and Story Musgrave on “Human biology and helpful robots.” Today, we will continue that discussion, turning our focus to how AI can help us colonize Mars and worlds beyond.

NASA* Astronaut Nicole Stott starts at the beginning: 

“Immediately and maybe from the simplest application, I would see AI working on a near-continuous basis both before and after humans arrive to work or settle somewhere — primarily at first to build the necessary infrastructure (e.g. structures, habitats, pipelines, power arrays, mining) … [and] in situ resources and materials that have to be sent in advance.”

Robonaut Robonaut uses several novel techniques for establishing remote control of its subsystems and enabling the human operator to maintain situation awareness. Credit: NASA*

NASA* Astronaut Tom Jones, having just returned from a series of workshops on exploration telepresence at the Keck Institute for Space Studies, has some fresh take-aways to add: 

“I think telepresence will be a big part of that too. You’ll have autonomous robots that will be doing routine camp setup chores, but you’ll also have interactive robots where you put your mind down on the surface of Mars from orbit and have your brain and hands present in the form of the robots’ manipulators and their vision systems and so forth. ... Even when you’re down on the surface, because of the radiation environment, you’re not going to be outside very much. You’re living in your subterranean habitat, and looking out through a periscope or a bunch of video sensors and you’re actually doing most of your field work through the proxy that you sent out there — the robot.”

If we supply these spacefaring robotic emissaries with enhanced computer vision, it could eliminate the need for so much human telepresence interaction. Coupled with object detection, they would be equipped to avoid hazardous objects and terrain, while also pinpointing in situ resources that could be used to build a colony on another world.

In discussing exploration telepresence with NASA* Astronaut Story Musgrave, he asks: “If we have a huge experience of Mars, do we need to put a human there? How much will putting a human there enrich the experience?” Our AI emissaries will certainly help us decide if another world is worth exploring in human form. Exoplanets are certainly no exception. In discussing the colonization of worlds beyond our solar system, Story suggests, 
“There’s all different kinds of ways to do it. You can send frozen sperm and frozen eggs and put them together and you just send a system to incubate them when you get there. You can think about hibernation.” He agreed that using exploration telepresence will be beneficial, adding, “But again, I think we need to explore it with AI, I think we need to explore it with communications and find out what’s there, before we set out.”

We’ve already made incredible strides in sending intelligent robots to space. However, they still rely on commands sent from Earth, which are subject to communication delays. With more purposefully-designed AI integrated into our robotic systems, the more effective and efficient our missions will become. For instance, we could map the surfaces of other worlds to identify the best landing sites for future spacecraft and search for resources. This type of work is currently being done using Intel® deep learning technology to build detailed maps of the lunar surface. Object detection could be used to automate the spacecraft’s trajectory, keeping it safe from the obstacle course of asteroids and space debris. 

It comes as no surprise that there are seemingly endless applications for AI in exploring worlds beyond our own. The major challenge is that it will require powerful processing of large datasets. Thankfully, Intel offers free access to the DevCloud so you can begin experimenting with the right tools for the right job, right away.

How will you use AI to explore and colonize other worlds?

About the Author:

MaryLiz Bender

MaryLiz Bender is a user experience designer, digital content coordinator at The Planetary Society, and associate producer for the Planetary Radio podcast. She dedicates her life to space outreach and education, engaging the public through music, art, and her mobile observatory.

Website: http://maryliz.me

For more complete information about compiler optimizations, see our Optimization Notice.