NIMA and Commercial Imagery
Q&A with Robert D. Steele,
President of OPEN SOURCE SOLUTIONS Inc.
What's good about NIMA's approach?
The good news here is that NIMA is finally making a real effort to
completely recast its system to receive and process commercial imagery,
which is today the *primary* source of geospatial data despite the
fact that NIMA's legacy architecture is totally dependent on national
imagery sources that are extremely expensive and largely unreliable for
mapping and wide area surveillance needs.
What's bad about it?
The bad news here is that NIMA is still not getting the support
it needs from the CINCS or SecDef for a hard-dollar commitment to out-year
procurements
of commercial imagery. NIMA is telling the one meter people they have
to come up with a system for delivering their stuff into an archaic
legacy process, without giving them a firm commitment to buy their
offerings. Doug Smith, Deputy Director of NIMA and a good man, has
publicly stated that NIMA needs to spend at least $250M a year, and
closer to $500M a year, buying commercial imagery. But OSD, NIMA and
the CINCs are all unwilling to make this happen now. Until DoD budgets
for this level of out-year
procurement, NIMA must be seen as still going through the motions and
paying lip service to commercial imagery sourcing while being financially
unable to actually buy
commercial imagery.
The other bad thing about this is that NIMA is assuming that
U.S. one meter will be commercially viable, and is also assuming that
the armed forces can do without maps until the year 2003 and beyond.
I have real doubts about the viability of the U.S. one meter industry
in the absence of firm out-year funding commitment from Congress. At the
same time I am still very upset with NIMA and OSD for their
continued refusal to buy French 10 meter imagery in bulk, so as to
resolve quickly all of the urgent deficiencies for 1:50,000 (ten meter)
maps with contour lines for the Third World. NIMA has told both the
White House and the Secretary of State (EARTHMAP Report, October 1995)
that it is not going to address White House and "rest of government"
needs for 1:50,000 maps urgently needed to support the Administration
in its sustainable development policy initiatives world-wide--and at
the same time, OSD is misleading the CINCs into thinking NIMA can meet
their needs for contingency maps, which is out and out not true. It
takes 60-90 days to create maps under optimal conditions. Thus, OSD
continues to leave both our armed forces and our diplomats out on a limb.
Is this too late or right on time?
Way too late. This should have been part of the original
Congressional mandate to put one meter into the private sector. I
support the one meter U.S. space capability but the one meter industry is
being asked to
do the impossible: to create an incredibly expensive commercial space
system which in the past was only affordable by the U.S. Intelligence
Community using tax payer dollars--and to do so with absolutely no
assured marketplace or even the real prospect of selling their product.
Could NIMA go farther? Is NIMA still too timid when it comes to getting
imagery from commercial sources?
To give NIMA its due, NIMA is not the problem. Inside NIMA, the
leadership understands the value of French 10 meter (which is the ONLY
industrial-strength commercial imagery system in place, and likely to
not only remain so, but become even more dominant with the forthcoming
launches of the five meter satellites). The problem is at the OSD level,
where bureaucrats are deliberately telling the Hill that we do not need
money for commercial imagery (limiting us to $10 million this year, when
NIMA knows it needs $250M), and with the CINCs, who are literally
oblivious of the fact that virtually all of their contingency plans are
completely unsupportable from the mapping side of the house--and in the
absence of maps, policy options are severely constained in the early
days of the crisis. I would say NIMA is being much too timid about
speaking truth to power and giving our mapping shortfalls, and the utility
of non-U.S. sources, a fair hearing in public forums.
What are the drawbacks from relying too much on federal sources for
imagery?
National imagery was designed to look at great detail--in fact,
it was NOT designed to do maps (wide area surveillance) but was jury-
rigged after the fact when we cancelled the originally planned mapping
satellites. National imagery for maps is like taking tens of thousands
of photographs through a straw, and then trying to paste them together
to create a mosaic of a larger area that one French snap-shot would cover
at a cost of $6-10 per square kilometer (probably about a 1000X cheaper
than national approach). Then add to this the fact that these little
postage stamps are taken over years and seasons, so that you have
multiple seasons, mutliple times of day, multiple angles of look, and you
can imagine what kind of monster process we have put in place.
The other drawback is that national imagery is responsive to national
needs for strategic intelligence, and secondarily to CINC needs for
targeting intelligence. Mapping is a poor third--during the Gulf
War the head of DMA, Phil Nuber, had to appeal personally to Stormin'
Norman to get time for collecting to meet mapping needs.
Why would anyone need one-meter imagery anyway?
One meter is useful in meeting perhaps 10% of the requirements,
but like the Air Force with its golf courses (the enlisted folks have
to play or the officers can't justify a golf course at every air force
base), OSD is saying that one meter will meet all of our needs because
they can't afford it otherwise. One meter is important in looking at
details--aircraft types, nuclear facilities, terrorist training camps.
It is not only inappropriate for wide area surveillance, but largely
unaffordable--one meter requires 10X to 100X more computer storage,
more communications bandwidth, more TIME to transmit and process, and more
dollars than 10M, and this is the "fact of life" that OSD continues to obscure and the
CINCs still do not understand.
I think one meter should not be in the private
sector. It should be a government system bid under competitive
contract, and it should be used to trade with the French, the Indians,
the Canadian, the Russians and others who have built superior wide
area surveillance systems.
Who in the commercial sector is selling such imagery?
Other than a few local companies that use Cessna's to take
one meter from an airplane, no one. A few companies that want to
create one meter satellites are having funding problems (one major
player is $600M short in capitalization), launch problems, and
future processing problems--one meter will not be "industrial-
strength" before the year 2010 and perhaps not even before the
year 2015.
As far as we know, how much one-meter imagery does the government
generate on its own?
Quite a lot in terms of collection, but where the government
falls apart is in processing. There is a hugh collection funnel, and
we only process about 10 per cent of what we collect.
What else is significant or relevant to NIMA's plans as stated in the
following CBD announcement?
It is a good and necessary announcement, but it flies in the
face of political and fiscal and operational reality. Politically, we
have to make a commitment to buying $250M (if 10 meter/1 meter mix)
to $500M (if only 1 meter) a year in the out-years beginning with next
year. Fiscally, it continues to overlook the cost to the taxpayer
of fore-going the savings and the efficiencies from using 10 meter,
which is more than adequate for 90% of our mapping needs. Operationally,
it continues to gloss over the fact that our armed forces and diplomats
and business world need the world at 1:50,000, that 90% of the world is
NOT available at 1:50,000 (ten meter), and that French commerical
imagery can easily solve 30-50% of our needs in the next TWO YEARS,
if OSD would simply wake up and smell the coffee that's already on the
burner.