The DARPA of geospatial
- By Patrick Marshall
- Sep 14, 2008
When you think of NURI, think of the Defense Advanced Research
Projects Agency...but only for geospatial projects.
NURI stands for NGA University Research Initiatives. NGA, of
course, is the National Geospatial-Intelligence Agency, a combat
support agency at the Defense Department.
The NURI project has been operating since 1997 and has already
awarded 103 grants to researchers at more than 60 universities to
investigate topics related to geospatial intelligence. Each grant
is for three years and generally amounts to $150,000 per year.
This year's awards (GCN.com/1201) range from
one for a research project on 'Spaceborne Magnetic
Gradiometry after Swarm: Novel Approaches to Mapping the
Earth's Magnetic Field Employing Nonlinear Magneto-Optical
Rotation Sensors' to a project titled 'Purpose- Aware
Dynamic Graph Models for Representing and Reasoning about
Networks.'
Seem like a long way from digital mapping? Actually, no:
Geospatial intelligence involves a lot more than mapping. NGA has
distilled its research road map into these six broad topic
areas:
- Acquire: Including sensor networks, detection of moving
objects.
- Identify: Spatiotemporal data mining.
- Integrate: Image data fusion, reuse and preservation of
data.
- Analyze: Visualization, process automation vs. human
cognition.
- Disseminate: Multilevel security.
- Preserve: Grid computing for geospatial data, reuse and
preservation of data.
NGA's InnoVision Directorate, tasked with 'the focal
point in NGA to address the future,' oversees the NURI
program. Ernie Reith, deputy director of the InnoVision
Directorate, said a lot of basic scientific research needs to be
done to improve geospatial intelligence. For example, the agency
looks to the first project mentioned above ' the one
involving mapping Earth's magnetic field ' to
eventually help researchers improve the accuracy of Global
Positioning System devices and control orbits of satellites. The
research could also help provide more accurate electronic targeting
capabilities.
A more detailed road map ' which Reith said is not public
' deals with a broad spectrum of specific challenges. They
include finding ways to track criminals or terrorists, detecting
weapons of mass destruction and finding ways of discovering
relevant data in a large database. Reith said the program's
priorities cover a lot of ground, 'from actual intelligence
problems to helping the analyst out to display devices.'
Each year, a panel of senior NGA scientists ranks proposed
topics against the road map and then finalizes a list of specific
topics. Universities then write proposals on those lists of topics
are made with the input of other intelligence agencies.
The four topic areas for 2008 were:
- The 'geopotential' of gravity and the magnetic
field.
- Developing efficient target detection and tracking techniques
using image data from multiple sources.
- Research on improving techniques for automatically extracting
features from remotely sensed imagery by using contextual
cues.
- Research to develop analytic tools and techniques that track,
monitor and predict natural or anthropogenic activities and provide
estimates of the causes of visual scenes.
Some trends can be discerned in the evolution of research topic
areas. Previous years were more heavily weighted toward areas of
data acquisition, for example, while this year's topics lean
toward basic research affecting the tracking of objects.
As for the future? 'There are challenges out there that I
suspect we're going to see along the lines of data storage
and data discovery,' Reith said. 'Something we hear
back from people in the field is that, with all of the data sources
that are being provided, there's an awful lot of difficulty
finding the data that they need.'