Better terrain maps of Earth... and beyond
The Defense Advanced Research Projects Agency (DARPA) thinks that today's computer maps of the Earth are inaccurate for its needs. So it recently awarded a grant to a Rensselaer Polytechnic Institute (RPI) researcher to develop better computer terrain maps of the surface of Earth and even our moon or Mars. RPI will develop new methods of compressing the data gathered by radar and laser scanning. Even if there is a specific focus to improve navigation of unmanned aerial vehicles (UAVs) and to position soldiers where they have the best visibility of their surroundings, other applications are envisioned, such as radio-rowers on the moon or Mars.
Here is what says W. Randolph Franklin, an associate professor of electrical, computer, and systems engineering at RPI.
"I’m studying better ways to compress the massive amounts of terrain data now available from radar and laser scans of the Earth’s surface," Franklin says.
Current methods often produce unacceptable terrain maps, giving rise to errors that are clearly visible in any commercial mapping product, according to Franklin. For example, one common mapping software renders Niagara Falls as a gentle slope, while another has 50-foot elevation contours crossing a shoreline.
And because DARPA has specific goals, such as improving the navigation of its UAVs, Franklin will develop several terrain maps.
"I will be researching and developing three different terrain representations," Franklin says. "I will also be studying some important applications of terrain data." One application is geared toward identifying the best sites to position a group of soldiers to allow them to see as much terrain as possible.
Below are two pictures illustrating other facets of the research work by W. Randolph Franklin (Credit: RPI/Franklin). On the top one, you can see "60 observers in the Adirondacks, viewed from above with a surface scan."
And the second one shows "the same 60 observers, placed with a constraint of 'intervisibility,' so that each observer can see at least one other observer.
These two pictures have been extracted from a previous work by Franklin and his colleagues, "Multiple Observer Siting on Terrain With Intervisibility or Lo-Res Data." Here is a link to this paper from 2004 (PDF format, 6 pages, 816 KB).
And for more information about this Geospatial Representation and Analysis (GEO*) project at DARPA, here are two links to the original solicitation and to the the project page.
Finally, here is Franklin's conclusion about future applications in space.
"A far-out application for radio towers would occur when the moon or Mars are settled," Franklin says. "Both have no ionosphere to enable long-distance radio, and the moon has no stable satellite orbits for potential communication satellites." He suggests that ground-based radio relays, visible to each other, could be the best way to communicate on these surfaces.
Sources: Rensselaer Polytechnic Institute news release, October 31, 2005; and various web sites
You'll find related stories by following the links below.