NASA's latest AI will navigate the Moon using landmarks
(2022/12/20)
- Reference: 1671561007
- News link: https://www.theregister.co.uk/2022/12/20/nasa_ai_moon_navigation/
- Source link:
NASA scientists are developing an AI to help future astronauts traverse the Moon's surface without the aid of satellite fixes, instead relying on landmarks to pinpoint lunar location.
Anyone who has participated in a land navigation exercise using a compass and topographical map will be familiar with the process NASA is developing its AI to perform: locate an object on the horizon, shoot an azimuth, and duplicate the process to get a triangulated location.
Of course, this being a NASA project to triangulate and navigate the Lunar surface, the endeavor is a bit more complicated.
[1]
"While a ballpark location estimate might be easy for a person, we want to demonstrate accuracy on the ground down to less than 30 feet (9 meters)," [2]said NASA Goddard Space Flight Center research engineer Alvin Yew, who is developing the system.
[3]
[4]
"It's important for explorers to know exactly where they are as they explore the lunar landscape," Yew said.
Getting the deets from LOLA
The Lunar Reconnaissance Orbiter, which has been snapping photos of the Moon's surface for [5]more than a decade , hosts a sensor bundle known as [6]LOLA that's a critical component of Yew's project. More precisely known as the Lunar Orbiter Laser Altimeter, the device has been snapping topographical images of the Moon's surface for as long as the LRO has been in orbit, and Yew's AI's first project is turning those images into surface-level horizon maps.
To quickly analyze the images generated by Yew's AI, an additional software tool built at the Goddard Center, known as the Goddard Image Analysis and Navigation Tool (GIANT) will step in. Armed with a 3D view of the surrounding lunar landscape, Yew's AI and GIANT, a lost explorer could scan the horizon and be given a precise location, along with directions back to safety.
[7]NASA's Mars InSight uploads its (probably) final image, shares it in a tweet
[8]Plaice in spaaace: NASA boosts astronauts' cognition with piscine diet
[9]NASA starts assessing Orion capsule for refurb
[10]KaRIn to hitch a ride on NASA's water-tracking sat
"In contrast to radar or laser-ranging tools that pulse radio signals and light at a target to analyze the returning signals, GIANT quickly and accurately analyzes images to measure the distance to and between visible landmarks," NASA said.
And this is the backup?
Yew said that portable devices equipped with local maps would enable the AI to navigate a variety of missions, but this visual Lunar navigation system isn't even a primary system for future Moon missions; it's envisioned as a backup.
"It's critical to have dependable backup systems when we're talking about human exploration," Yew said, and that's precisely what his AI would be: a backup for more the traditional networking solution NASA is planning for the Moon, [11]LunaNet .
Lunar navigation is a central part of LunaNet's planned services, though NASA hasn't described how LunaNet's navigation system would work in much detail.
[12]
The space agency said that LunaNet navigation, like the rest of the LunaNet, would maintain "operational independence from data processing on Earth while maintaining high precision," but otherwise will make navigation on the Moon's surface and orbit more practical.
If and when LunaNet goes down or a surface mission ends up out of range, Yew's landmark navigation service would step in. NASA said it also sees the system having applications on Earth in situations where explorers can't obtain a GPS satellite fix. ®
Get our [13]Tech Resources
[1] https://pubads.g.doubleclick.net/gampad/jump?co=1&iu=/6978/reg_offbeat/science&sz=300x50%7C300x100%7C300x250%7C300x251%7C300x252%7C300x600%7C300x601&tile=2&c=2Y6I@j@Zp0DIfHGWv69CLSQAAAAA&t=ct%3Dns%26unitnum%3D2%26raptor%3Dcondor%26pos%3Dtop%26test%3D0
[2] https://www.nasa.gov/feature/steering-by-landmarks-on-the-moon
[3] https://pubads.g.doubleclick.net/gampad/jump?co=1&iu=/6978/reg_offbeat/science&sz=300x50%7C300x100%7C300x250%7C300x251%7C300x252%7C300x600%7C300x601&tile=4&c=44Y6I@j@Zp0DIfHGWv69CLSQAAAAA&t=ct%3Dns%26unitnum%3D4%26raptor%3Dfalcon%26pos%3Dmid%26test%3D0
[4] https://pubads.g.doubleclick.net/gampad/jump?co=1&iu=/6978/reg_offbeat/science&sz=300x50%7C300x100%7C300x250%7C300x251%7C300x252%7C300x600%7C300x601&tile=3&c=33Y6I@j@Zp0DIfHGWv69CLSQAAAAA&t=ct%3Dns%26unitnum%3D3%26raptor%3Deagle%26pos%3Dmid%26test%3D0
[5] https://www.theregister.com/2009/07/06/lro_moon_snaps/
[6] https://lunar.gsfc.nasa.gov/lola/
[7] https://www.theregister.com/2022/12/20/nasa_mars_insight_farwell_tweet/
[8] https://www.theregister.com/2022/12/16/nasa_boosts_cognition_with_fishy/
[9] https://www.theregister.com/2022/12/15/orion_capsule_post_flight_analysis/
[10] https://www.theregister.com/2022/12/14/karin_nasa_water_satellite/
[11] https://www.nasa.gov/feature/goddard/2021/lunanet-empowering-artemis-with-communications-and-navigation-interoperability
[12] https://pubads.g.doubleclick.net/gampad/jump?co=1&iu=/6978/reg_offbeat/science&sz=300x50%7C300x100%7C300x250%7C300x251%7C300x252%7C300x600%7C300x601&tile=4&c=44Y6I@j@Zp0DIfHGWv69CLSQAAAAA&t=ct%3Dns%26unitnum%3D4%26raptor%3Dfalcon%26pos%3Dmid%26test%3D0
[13] https://whitepapers.theregister.com/
Anyone who has participated in a land navigation exercise using a compass and topographical map will be familiar with the process NASA is developing its AI to perform: locate an object on the horizon, shoot an azimuth, and duplicate the process to get a triangulated location.
Of course, this being a NASA project to triangulate and navigate the Lunar surface, the endeavor is a bit more complicated.
[1]
"While a ballpark location estimate might be easy for a person, we want to demonstrate accuracy on the ground down to less than 30 feet (9 meters)," [2]said NASA Goddard Space Flight Center research engineer Alvin Yew, who is developing the system.
[3]
[4]
"It's important for explorers to know exactly where they are as they explore the lunar landscape," Yew said.
Getting the deets from LOLA
The Lunar Reconnaissance Orbiter, which has been snapping photos of the Moon's surface for [5]more than a decade , hosts a sensor bundle known as [6]LOLA that's a critical component of Yew's project. More precisely known as the Lunar Orbiter Laser Altimeter, the device has been snapping topographical images of the Moon's surface for as long as the LRO has been in orbit, and Yew's AI's first project is turning those images into surface-level horizon maps.
To quickly analyze the images generated by Yew's AI, an additional software tool built at the Goddard Center, known as the Goddard Image Analysis and Navigation Tool (GIANT) will step in. Armed with a 3D view of the surrounding lunar landscape, Yew's AI and GIANT, a lost explorer could scan the horizon and be given a precise location, along with directions back to safety.
[7]NASA's Mars InSight uploads its (probably) final image, shares it in a tweet
[8]Plaice in spaaace: NASA boosts astronauts' cognition with piscine diet
[9]NASA starts assessing Orion capsule for refurb
[10]KaRIn to hitch a ride on NASA's water-tracking sat
"In contrast to radar or laser-ranging tools that pulse radio signals and light at a target to analyze the returning signals, GIANT quickly and accurately analyzes images to measure the distance to and between visible landmarks," NASA said.
And this is the backup?
Yew said that portable devices equipped with local maps would enable the AI to navigate a variety of missions, but this visual Lunar navigation system isn't even a primary system for future Moon missions; it's envisioned as a backup.
"It's critical to have dependable backup systems when we're talking about human exploration," Yew said, and that's precisely what his AI would be: a backup for more the traditional networking solution NASA is planning for the Moon, [11]LunaNet .
Lunar navigation is a central part of LunaNet's planned services, though NASA hasn't described how LunaNet's navigation system would work in much detail.
[12]
The space agency said that LunaNet navigation, like the rest of the LunaNet, would maintain "operational independence from data processing on Earth while maintaining high precision," but otherwise will make navigation on the Moon's surface and orbit more practical.
If and when LunaNet goes down or a surface mission ends up out of range, Yew's landmark navigation service would step in. NASA said it also sees the system having applications on Earth in situations where explorers can't obtain a GPS satellite fix. ®
Get our [13]Tech Resources
[1] https://pubads.g.doubleclick.net/gampad/jump?co=1&iu=/6978/reg_offbeat/science&sz=300x50%7C300x100%7C300x250%7C300x251%7C300x252%7C300x600%7C300x601&tile=2&c=2Y6I@j@Zp0DIfHGWv69CLSQAAAAA&t=ct%3Dns%26unitnum%3D2%26raptor%3Dcondor%26pos%3Dtop%26test%3D0
[2] https://www.nasa.gov/feature/steering-by-landmarks-on-the-moon
[3] https://pubads.g.doubleclick.net/gampad/jump?co=1&iu=/6978/reg_offbeat/science&sz=300x50%7C300x100%7C300x250%7C300x251%7C300x252%7C300x600%7C300x601&tile=4&c=44Y6I@j@Zp0DIfHGWv69CLSQAAAAA&t=ct%3Dns%26unitnum%3D4%26raptor%3Dfalcon%26pos%3Dmid%26test%3D0
[4] https://pubads.g.doubleclick.net/gampad/jump?co=1&iu=/6978/reg_offbeat/science&sz=300x50%7C300x100%7C300x250%7C300x251%7C300x252%7C300x600%7C300x601&tile=3&c=33Y6I@j@Zp0DIfHGWv69CLSQAAAAA&t=ct%3Dns%26unitnum%3D3%26raptor%3Deagle%26pos%3Dmid%26test%3D0
[5] https://www.theregister.com/2009/07/06/lro_moon_snaps/
[6] https://lunar.gsfc.nasa.gov/lola/
[7] https://www.theregister.com/2022/12/20/nasa_mars_insight_farwell_tweet/
[8] https://www.theregister.com/2022/12/16/nasa_boosts_cognition_with_fishy/
[9] https://www.theregister.com/2022/12/15/orion_capsule_post_flight_analysis/
[10] https://www.theregister.com/2022/12/14/karin_nasa_water_satellite/
[11] https://www.nasa.gov/feature/goddard/2021/lunanet-empowering-artemis-with-communications-and-navigation-interoperability
[12] https://pubads.g.doubleclick.net/gampad/jump?co=1&iu=/6978/reg_offbeat/science&sz=300x50%7C300x100%7C300x250%7C300x251%7C300x252%7C300x600%7C300x601&tile=4&c=44Y6I@j@Zp0DIfHGWv69CLSQAAAAA&t=ct%3Dns%26unitnum%3D4%26raptor%3Dfalcon%26pos%3Dmid%26test%3D0
[13] https://whitepapers.theregister.com/
Re: Can any apps do the non-AI part?
Paul Hovnanian
"You'd actually have to use a paper map."
A two dimensional database of significant landmarks with their geographic positions. Same thing.
"It's important for explorers to know exactly where they are as they explore the lunar landscape,"
Neil Barnes
They know where they are. They're on the moon. It's the conspiracy theorists who need the maps.
but I can see a slight flaw in this method
Anonymous Coward
I know that the lunar landscape is fairly stable...but you'd only need a few new craters (and accompanying high side wallls) and all of a sudden the landscape has changed and it might then appear to be like another area of the moon.
Plus the actual surface is pretty much the same colour all over, so over the entire 3.793×107 km2 surface area, there are bound to be "similarities" (in terms of landscape) with multiple places all looking the same...so, a lost astronaut could easily find himself going in the wrong direction.
Surely a "cheap as chips" localised GPS system, from just a few lunar satellites would be much better...plus to begin with you'd only need to place them in orbits so they just cover the side of the moon facing the Earth. And it's not like the lunar atmosphere will block any of the signals.
Can any apps do the non-AI part?
locate an object on the horizon, shoot an azimuth, and duplicate the process to get a triangulated location.
I thought about it, and there aren't any apps that could take a bearing from two landmarks and show the resulting fix. You'd actually have to use a paper map.
If I'm wrong, I'd sure like to have that app... assuming it's Android compatible, of course.