Dr. Chee-Kheng Chng and Dr. Sofia McLeod have developed a new novel crater-based navigation system called STELLA (Spacecraft craTer-basEd Localization for Lunar mApping) for long-duration lunar mapping missions.
"We already have detailed maps of the lunar surface, where craters serve as the main landmarks," Dr. Chng said.
"Our system helps a spacecraft detect the craters it sees through its camera, match them with the craters on the lunar map, and use that information to determine its position."
The evaluation of STELLA's work was published in the journal Astrodynamics.
"We tested its effectiveness on a comprehensive and realistic synthetic dataset, featuring more than 15,000 images, which emulated a year-long lunar orbit and featured a range of conditions, including challenging lighting and different angles," Dr. McLeod said.
"STELLA consistently demonstrated an ability in most scenarios to achieve meter-level position accuracy and sub-degree altitude accuracy."
"This serves as a promising baseline, laying the groundwork for further developments and deployments of vision-based navigation solutions for long-duration orbital missions."
With a growing focus on lunar exploration globally, there is increasing need for accurate location data.
"Because it's not possible to put people in space easily or cheaply, there needs to be a way to find information before you go, and that's where AI comes in," Dr. McLeod said.
"The current systems available for lunar-based navigation are not accurate enough for the kinds of missions where you need really precise locations."
"Say you send up a satellite to search for water resources on the Moon, which could then later be referenced to find a good base camp location on the Moon’s surface."
"If your satellite positioning system isn’t accurate, then you might end up putting your base camp kilometers away from a water source, which would be disastrous."