The Space Reviewin association with SpaceNews

CSS-4 launch
The launch of a Minotaur rocket carrying TacSat-2 in December 2006. (credit: NASA)

Some ORS for ORS

During World War 2, the famed physicist Freeman Dyson was a freshly-minted college graduate who went to work in the Operational Research Section (ORS) of the Royal Air Force’s Bomber Command. At one point Dyson determined that if Bomber Command removed two machine gun turrets and two crewmembers from its bombers, the aircraft would be able to fly 80 kilometers per hour faster, spend less time over enemy territory, and thus be less vulnerable to fighters. “But our proposal to rip out the turrets went against the official mythology of the gallant gunners defending their crewmates,” Dyson wrote, and his superior never raised the issue with the head of Bomber Command. Dyson figured that thousands of British airmen lost their lives because his boss would not take their calculations to the commanding general.

So far, most of the well-known problems have involved the launch vehicle side of the equation, where ORS still lacks an inexpensive and reliable launcher. But that is really less than half the equation.

Although operations research was successful during World War 2, it earned a bad reputation during Vietnam, when Robert McNamara took it to extremes, trying to prove that when two lines on a graph crossed, the United States would “win.” But the basic concept is sound: statistical analysis can be applied to some military operations, at the very least to shed light on effectiveness and to aid in decision-making. Today in the space business it is called Mission Utility Analysis, and that is what is needed for another ORS, Operationally Responsive Space.

In his article responding to my article, Dr. James R. Wertz made the case for ORS (“It’s time to get our ORS in gear,” The Space Review, January 7, 2008). Dr. Wertz and I don’t really disagree. He is a believer in ORS who has tried to answer the question of what ORS is supposed to accomplish, whereas I was pointing out that so far, ORS has not yet accomplished most of its initial goals, and the jury is therefore still out on whether ORS can do what its proponents claim. For example:

  • TacSat-1, which public sources state was originally scheduled for 2004 launch, still has not flown
  • TacSat-2 finally flew, but was not fully operational for several months due to what should have been a foreseeable policy issue
  • The launch side of the ORS equation has not been solved
  • ORS still needs to demonstrate exactly what requirements it can meet and how

As of 2006, several independent organizations noted that ORS had not made its case. The Government Accountability Office reported that ORS was not proceeding as fast or orderly as planned. Similarly, RAND produced its National Security Space Launch Report that found that “ORS full-scale development should not be undertaken until an operational concept, a family of candidate payloads, and launch vehicles and infrastructure are aligned.” It has been nearly two years since those reports were produced, but there appears to be no public, independent assessment indicating that these criticisms have been met. TacSat-2 was not a definitive answer to these issues.

So far, most of the well-known problems have involved the launch vehicle side of the equation, where ORS still lacks an inexpensive and reliable launcher. But that is really less than half the equation. The RAND study noted that they heard from many persons who stated that the payload challenge—developing a useful satellite within the weight limitations—was greater than the launch vehicle challenge. RAND claimed (p. 42) that the ORS demonstration satellites cannot achieve the kind of optical quality that military commanders need. They based this on a 2004 study by the Air Force’s Scientific Advisory Board (SAB). Maybe a lot of progress has been made since 2004 and this is no longer the case. More likely, this is a definitional problem, and perhaps the Scientific Advisory Board was demanding performance that ORS advocates believe is unnecessary to fulfill many military requirements.

Yes, ORS will be able to do some things that UAVs, and big satellites, cannot do. But are those things sufficient to justify their cost?

Dr. Wertz stated that the goal for ORS is $20–25 million for the launch, payload, spacecraft bus, and one year of operations (excluding the non-recurring development cost). The Government Accountability Office stated that TacSats 2-4 cost approximately $40 million each, and the only currently available booster is the Minotaur, which cost $25–31 million apiece. So the demonstration flights all will cost $65–71 million apiece, excluding the operations costs. That’s very little in space terms, although for a one-year lifetime satellite, it looks less attractive at second blush. More to the point, the cost target is less than a third of these demonstration projects, and how confident are ORS advocates that those goals can be achieved? Certainly part of the answer hinges on what happens with the Falcon 1 rocket, which could potentially cut up to $23 million off the topline cost. But first Falcon 1 has to successfully reach orbit. Then it has to reach its cost goal. And the satellites have to come down in price as well, something that RAND said is going to be tough. The best way to bring the costs down is to build a lot of satellites, but that may run counter to the goal of matching what the military builds to what it needs over a long period of time, so that spares don’t sit on the ground for a decade gathering dust and becoming obsolete.

No matter what they end up costing, ORS will have to demonstrate its cost superiority to other relatively cheap alternatives like UAVs (a Global Hawk UAV costs approximately $26 million, whereas a Predator costs about $4 million) whose costs can be amortized over decades of operation, compared to the nominal one-year lifetime of an ORS satellite. Yes, ORS will be able to do some things that UAVs, and big satellites, cannot do. But are those things sufficient to justify their cost?

Since I wrote my earlier article, Space News has reported that the Air Force is investigating the possibility of modifying TacSat-1 to carry a more useful payload. If they can do that for minimal cost, that would be a better solution than letting the satellite sit on the ground forever. But that experience demonstrates one of the potential pitfalls of ORS: building satellites for a future need that may evaporate and being faced with the difficult choice of scrapping the investment or trying to make lemon custard out of lemons. It may be possible, but it will require careful planning. And there are other examples from the national security space field where these kinds of things happened, for instance, buying too many GPS satellites and seeing the cost savings evaporate as the satellites sat in storage for far longer than planned.

But, on the other hand, it might be worth doing simply to have an insurance policy against enemy attack on existing American space assets. Robert Butterworth, president of Aries Analytics and a fellow at the George C. Marshall Institute, recently wrote a short policy paper about ORS, arguing that it may make sense even at the higher cost of the Minotaur rockets. Butterworth argues that even the possession of this capability might serve as a deterrent to attacking American satellites, as an adversary may determine that the United States could quickly recover. His view is that the most useful satellites would be synthetic aperture radars, but for complex policy reasons, it may not be possible to procure them as part of an ORS effort.

On the other hand, it might be worth doing simply to have an insurance policy against enemy attack on existing American space assets.

The last point on my above bullet list is the crux of the issue. My own concern is that unless there is a well-thought plan for implementing ORS, we may end up developing what in Washington people call “self-licking ice cream cones”—or satellites that satisfy only their own makers and not the community of potential users. A $25-million satellite may be brutally cheap, but how many people will use it? We cannot afford to give every tactical commander their own satellite, so how can we ensure that the most users get access to this capability, yet still make it flexible and not loaded up with requirements? There must be a way to statistically demonstrate the utility of ORS at least for certain applications. Certainly this would have to be a classified exercise, but the factors in any calculation should be things like cost, lifetime, number of users that can be serviced, data bits produced, and some kind of measurement of the quality of the data. A big legacy system like a high-resolution imagery satellite may score badly on the cost side of the equation, but when divided by the amount of data returned, the lifetime of the satellite (perhaps a decade, compared to a year for an ORS satellite), and the quality of the data, the ORS system may not be the clear winner. Similarly, considering that you can buy many Predator drones for the cost of a single ORS satellite, it may not prove to be the obvious choice for tactical commanders either. If so, will the few things that ORS can uniquely accomplish—assuming that it can be made to work—be worth the costs?

The launch of TacSat-3 this summer could be a good way of providing some hard data to fill in the blanks in the equation. Freeman Dyson at least had the data in hand, but a boss too timid to take it to the leadership. Hopefully, a future ORS advocate will not be so reticent, and if the data is good enough, it will be impossible for the people in power to ignore.