Sunday, August 24, 2008

Competing with the Human Element

When the U.S. Corona satellites produced their first imagery, the imagery products had cloud cover over most of their target of interest rendering the image useless. This simple fact that I keep on re-discovering for other applications like finding a lost boat in the ocean or simply taking photos over New Mexico from a high altitude balloon, started a Space race in the early 1960's between the U.S. and the U.S.S.R. I only recently learned about it when I was watching a PBS/NOVA show called Astrospies (a show also viewable as Des espions dans l'espace in France and as Die Weltraumspione in Germany). This story is fascinating in many respects. I had heard about the U.S. Manned Orbiting Laboratory (MOL) but had never heard about its Russian equivalent the Almaz program. While I knew the U.S MOL did not go anywhere, I did not realize it was internally in competition with the U.S. NRO spysats: A situation that featured for the first time a clear fight between human and robotic spaceflights. The robotic projects eventually won with the KH spacecrafts in the U.S. In Russia, however, humans won and the program actually flew successfully three manned missions. The Almaz spacecrafts in which these flights took place looked like Salyut stations:


The inside of these Almaz stations look like any military airplane dedicated to observation.

But the most amazing part are the technical specifications of the camera called Agat weighting more than 2 tons which had a lens housing that is more than 6 meters wide. It had a magnification capability of about 80 times. The final resolution was said to be 8 cm.


After 3 billion dollars spent on the U.S. MOL, the equivalent program got canceled in the U.S. on June 10th, 1969 while the first russian Almaz flight took off, ironically, on July 4th, 1974.

Why am I telling this story ? well, I am stricken by the fact that the cloud cover was such a pain to weed out that it became a mission requirements to have a human in the loop thereby yielding very large funding expenditures. It also showed that progress in a ten year time frame (from the early sixties to the early seventies) in communication and computing power have essentially removed the Human element in the loop requirement.

Second, the most important part of these observation program requirements continuously balances the resolution vs the Field of View. If your sensor has high resolution, then on-board power and bandwidth put a hard constraint on your footprint on the ground (such is the case of EO-1 and the hyperion hyperspectral camera for instance). This is why the DARPA LACOSTE program using coded aperture in the IR seems to be far reaching: high resolution and large field of view are requirements of the program. Yet, the program does not remove the Human element: even after having pushed tons of data to the ground, none of these programs can figure out the intent and are therefore likely to yield operator overload. It is one thing to see everything, it is another to make sense of it (TM).

The Astrospies DVD can be bought here or from Amazon.

Photos: Astrospies documentary.

No comments:

Printfriendly