The Space Reviewin association with SpaceNews

NPOESS satellite
The NPOESS program has been saddled with cost overruns and schedule delays in large part because of a desire for technological leaps in capability rather than step-by-step progress. (credit: NOAA)

NPOESS: another example of technological overreach?

In 1994, the Clinton Administration decided that the distinction between civilian and military weather satellites was no longer relevant and abolished it. In its place, it created the National Polar-orbiting Operational Environmental Satellite System (NPOESS) managed by an Integrated Program Office (IPO) made up of representatives from the Defense Department, NASA, and the National Oceanic and Atmospheric Administration (NOAA). Delays and cost overruns have pushed the launch of the NPOESS Preparatory Project (NPP), which will carry prototype versions of the sensors, back to 2009 at the earliest. The definitive versions of the spacecraft will, perhaps, be launched in 2012. The system as a whole will not become fully operational before 2013 or 2014 at best.

On November 16, 2005, in testimony before the full House Science Committee, a representative of the prime contractor, Northrop Grumman Space Technology, said that when finished, “NPOESS will give civilians more precise advance warning of hurricanes and severe weather… and will revolutionize battlefield situational awareness with timely knowledge of the weather for use by the military to its advantage during conflicts and operations. Observation to delivery time will be just 15 minutes compared to the hours that are needed today.” This gives some idea as to the extraordinary goals involved in the program. It also explains why the delays and cost overruns should surprise no one.

Why should it take fifteen or more years to build something as relatively simple as a weather satellite?

One of the main sensors for this project is the Visible Infrared Imager/Radiometer Suite (VIIRS). At about 200 kilograms it’s the heaviest NPOESS instrument, and its ability to gather high-resolution sea surface temperature and other environmental data will be unmatched. It is an exceptionally complex piece of hardware that uses technology derived from the military’s Defense Meteorological Satellite Program (DSMP) and from civilian weather and Earth observation satellites. Its development process has been a troubled one, to say the least. The failure of the cooling system during a test last year seems to have set off an effort to rework the whole effort. Today, according to one source, the subcontractor has been successfully dealing with its problems, at least recently.

Why should it take fifteen or more years to build something as relatively simple as a weather satellite? America’s first weather satellite, TIROS, went from concept to orbit in less than three years. One credible explanation is that the government’s way of doing business, and that of its major contractors, have become so encrusted with procedures and regulations that it is now impossible to replicate the creative feats of engineering that made the 1950s the “Golden Age of Prototypes.”

On the other hand, it may be that the specifications of this new generation of satellites require that the builders operate at the extreme limits of what is possible given today’s capabilities. According to a 2004 GAO report, seven out of the 13 instruments on NPOESS require new technology. This is contrary to the well-known principle that only one or—at most—two new technologies should be incorporated into any new space system. It was NASA’s decision to ignore this principle that led to the demise of the X-33 program.

In the 1990s, there was a belief in Washington and on Wall Street that all technological problems would inevitably be solved by the “Information Revolution.” The amazing growth of the computer industry and the introduction of the Internet did, indeed, revolutionize many aspects of the economy. These changes did not replace the need for sound engineering judgment. The tendency of the US government to demand the most technologically advanced “cutting edge” systems from industry led to a set of extremely expensive failures, and a few equally expensive successes.

In those days, the fact that two of the most powerful men in Washington, Al Gore and Newt Gingrich, were both technophiles and true believers in the digital revolution made it easy for proponents of extremely ambitious projects to sell their ideas. “Pushing the envelope,” to use the cliché popularized by Tom Wolfe in The Right Stuff, may be exciting, but it often produces unpleasant and expensive surprises.

If more of our technology development efforts aimed for step-by-step advances instead of giant leaps, cost overruns and programmatic crises might become less frequent.

NPOESS is managed by an unusual interagency organization, but its difficulties cannot really be blamed on the way the government has structured the program. The anger that some House members directed at Conrad Lautenbacher of NOAA and Ron Sega of the Pentagon was aimed at the wrong targets. The complexity and sensitivity of sensors, like the VIIRS and the Conical Microwave Imager/ Sounder (CMIS), are the real source of the overruns and delays, and that is something that neither of those gentlemen had anything to do with. The ongoing restructuring of the whole program, which looks as if it will take more than a year, is an indication of how serious the situation is. A recent GAO report called NPOESS “A Program in Crisis”.

While it is always a good idea to take GAO reports with a few grains of salt, the management upheavals at the contractors’ shops show that something more than just the normal development screwups are happening. Northrop Grumman officials say that they are going to supervise their subcontractors more closely than ever and they promise to bring additional corporate resources to bear. This may solve the problems, but Congress has a right to be skeptical and will, almost certainly, want to keep a close eye on this program.

Like the SBIRS missile launch detection satellite program, NPOESS is something the US government cannot live without. The delays are already causing both NOAA and the military to devise ways to overcome a possible future gap in coverage. It would be useful if someone in Congress were to ask why it is that such essential programs are so often given technological goals that are barely achievable. Why do so many US government technology development efforts aim at revolutionary improvements in capability, instead of settling for incremental progress?

If more of our technology development efforts aimed for step-by-step advances instead of giant leaps, cost overruns and programmatic crises might become less frequent. This will mean a reform of the process by which the government determines what its requirements are and how ready industry is to fulfill them. This may be the key to long-term and effective acquisition reform.