AFA Policy Forum
"The Role of S&T in Acquisition"
The Honorable Marvin R. Sambur, Assistant Secretary of the Air Force for Acquisition
General Gregory S. Martin, AFMC Commander
and Dr. Mark J. Lewis, Chief Scientist of the Air Force
Air & Space Conference 2004—Washington, DC
September 15, 2004
[Click here for printer-friendly version.]
Dr. Sambur: The topic is about S&T, so the issue is the hypothesis that budget pressures, foreign
competition and consolidation of the defense industry and workforce threatens future American dominance in
From my title slide I guess you would assume that I don't believe that statement. How many people here
actually believe that statement, that we're threatened? Two people. I'm not going to be a contrarian here.
I don't believe that statement at all. As a matter of fact, if we actually look at the budgets with respect
to S&T, you'll see that the budgets are going up in S&T. You'll notice that in S&T the appropriated budget was
actually greater than what the President's budgets were, consistently for the last 12-14 years. And that is
because the S&T part of the budget is the most protected part of the budget. It has many people who have a
vested interest and good political connections, if I can use that expression, and that part of the budget is
actually very robust and very immune to budgetary declines.
I also have a little footnote here which I think is worth paying attention to and that is that the United
States DoD S&T budget is bigger than not just the S&T budgets, but the entire defense budgets of all but ten
other countries in the world. If you compare what our S&T budget is against those other ten countries, we are
significantly greater than them. So the hypothesis that budget pressures will cause the S&T part of the budget
to decline I don't think is in reality going to happen. I think the S&T part of the budget is very robust.
The next question is with respect to whether or not the S&T budget is really taking on the important issues.
Are we really doing the things in S&T that are intended to give us dominance? True dominance over the next
The new threats that are facing the military are very, very different than in the past. In the past, we had
a solitary enemy and we knew in many ways what that enemy was capable of doing and we knew in many ways how to
plan for it, how to develop technologies that were necessary, but if you look at this picture you can see the
threats that are emerging are very difficult to predict. Terrorists wearing bomb belts or IEDs that are very
difficult to predict.
If you were doing the S&T budget four years ago and allocating funds, how many people would have thought
about putting S&T money aside for any of those activities? So the issue here is that these threats now are very
difficult to predict and when they do come, the issue is not a long-term type of research and development
activity, but the issue is really how to get it out into the field, the solutions, as soon as possible.
If you read one of the articles that was in the Pentagon Early Bird, I forget the journal, they talked about
the fact that the IED development is something akin to the Manhattan Project. Anybody read that article? Where
all of the significant amount of funding is being poured into solving the improvised explosive devices? So the
issue in S&T is really how to come up with quick solutions to problems that are very difficult to predict. Not
long-term research and development cavities, although there are many of those things that are still captured in
The issue really is about the ability to respond quickly. That is becoming the main thrust of a lot of S&T
activities. There still is a core of research and development that is growing and very robust, but our issue is
quickly responding, and in the Air Force we've done that quite effectively.
We were able to arm a Predator and do something that the normal acquisition community said would take at
least two years. We were able to do that in a matter of months. So the issue right now is reducing the cycle
time to get this S&T into the field very quickly.
In order to make S&T more effective, we have to utilize and multiply the effectiveness of our budgets with
that out in industry. The question that a lot of people in industry come to me with is, “we're doing all these
novel things. How do we let the Air Force know what we're doing? Where is the entry point, so to speak, to let
the Air Force know about our development efforts?” Many years ago, there was a process in which the military
would evaluate your R&D and determine whether or not you're headed in the right direction. We don't have that
now. So the second question people ask me is, “how do I know that my R&D is hitting on things that are of
interest to the military?”
I think General Martin might be talking about this with respect to AFRL, how to get in to them to make them
know about what you're doing and potentially get funding for that and also potentially to get an alignment with
your R&D program and what the Air Force is doing.
There's a program called Tech Connect that enables you to let us understand what you're doing, to have it
demoed and to potentially give you funding if we think it's a good idea. There are also battle labs. If all of
these things are not familiar and you have this question with respect to “how I can make my own S&T more
effective and aligned with that of the Air Force and the military,” you can call me in my office and we can give
you the particulars on those issues.
The bottom line is that I think we will maintain dominance. I don't think there's anything unhealthy about
the S&T budget. We are constantly working on that. We're constantly aligning it with the capabilities that are
necessary for the Air Force. We have this elaborate Capability Review and Risk Assessment (CRRA) process. We
challenge ourselves to see if we're really developing the technologies that meet the capability needs of the Air
Force. We're also very interested in making sure that industry's R&D is aligned with our needs. And also, most
importantly, interested in making sure that if there are good ideas out there that you have an entry point, that
you have an ability to tell the Air Force about it and get it evaluated.
Dr. Lewis: Thank you. I want to start out with a little confession which is that I've only been the
Chief Scientist of the Air Force for about a week, so with that in mind, I actually come from academia and in
true academic tradition, I won't provide any answers today, but I will pose a bunch of questions that I hope will
address the topic that we're discussing.
I want to start with a lesson from history and this comes courtesy of the great aviation historian, Dick
Hallion. He points out that it's important to maintain your S&T lead in the acquisition process. For example,
in 1903 the Wright Brothers invented the airplane. In 1912, nine years later, they built their Speed Scout,
their high-speed version of the airplane which was essentially a modified version of their earlier airplane. For
the historians in the audience, you know that the Wright Brothers spent a lot of time suing other people who were
trying to invent the airplane and didn't spend so much time developing their concepts.
On the other side of the Atlantic, in 1912 they built an airplane with a top speed of 108 miles an hour, which
is at least a generation beyond what the Wright Brothers had at the same time. A monoplane airplane, it has all
the elements of a modern aircraft. There's an important lesson there from the beginning of aviation.
One of the S&T challenges is that sometimes new technologies are not compatible with existing technologies.
Let me talk a little bit about what I think we would argue is the mission of Air Force S&T and there are of
course many elements in the Air Force S&T role. One thing it will do is create new options to meet the emerging
needs of the warfighter. Sometimes that has to be done on a very rapid scale.
Solve existing technical challenges. Develop evolutionary growth of existing capabilities. But also shape
the future using revolutionary new capabilities. So we need a broad portfolio. We need to be doing evolutionary
work. We also must be seeking those revolutionary capabilities.
Position the US industry to develop, deliver and sustain superior systems. This is also an important role for
Air Force S&T. And also encouraging academia to pursue not only Air Force relevant problems, but also keep the
next generation going, keep the next generation workforce up to speed. Finally, maintaining an in-house expertise
for smart acquisition decisions. That's extremely important.
Our Air Force personnel need to know, frankly, a dumb idea when they see a dumb idea. They need to be able to
recognize the wheat and the chaff. I would argue that all these Air Force S&T missions are related to acquisition.
There of course are several key components in our Air Force S&T structure. The Air Force Research Laboratory
is our primary S&T organization. In-house research and expertise. They're responsible for many, many activities.
Within AFRL, the Air Force Office of Scientific Research sponsors researchers throughout the world, actually,
doing basic research activities that are relevant to the Air Force. Relevant to the Air Force mission and,
again, relevant to acquisition.
We also have the Air Force Scientific Advisory Board, which is essentially a panel of outside experts who
provide guidance and ties to the external scientific community. Then my office, the Air Force Chief Scientist,
and our role is to be the principal scientific advisor to the Chief of Staff and to the Secretary of the Air
I think the key point we want to make here is that there is a very extensive structure for S&T in the Air
Force, very keenly aware of the needs of S&T and how it relates to acquisition.
I want to pose a series of challenges that I see as being issues that we're faced in trying to match S&T to
acquisition. Challenge number one is embodied by this plot of time required for concept for initial operation
of selected fighters starting in 1945 with the P-80 and running up to the present. The P-80 had about a two-year
development time from initial concept to operation. That's about an order of magnitude smaller than a modern
fighter aircraft, so S&T has to fit into a much longer development schedule, and frankly, there are many good
reasons for that schedule having been drawn out. We're building very, very complex systems that require lots of
people, lots of man hours to make them operational.
Of course, that's in a world that's constantly changing. If we take that 20-year development time for a
fighter and then superimpose it, we'll see we get six presidential terms in that 20 years, at most three
administrations. Four senatorial terms, 12 representative terms, and eight terms for service chiefs. Remember,
the S&T portion actually starts before this 20-year cycle. So we very often have long lead times and have to
operate within those constraints.
Another challenge along those lines. We can't always schedule miracles. By that I mean sometimes we'd like
to do something, but we simply don't know how.
One of my favorite examples actually comes from my own area of research which is high-speed flight,
hypersonics. In 1958, two researchers down at the National Advisory Committee for Aeronautics came up with this
idea for burning fuel in a supersonic stream of air, building something called a scram jet engine, a supersonic
combustion REM jet. It was 1958. The first attempted flight test of that engine was ten years later on the
X-15. The X-15 program was canceled, and it turns out that that engine wouldn't have worked, by the way. We
know now it was a bad design for an engine.
We did actually finally fly the scram jet engine this year—46 years after the concept was first introduced,
and it worked perfectly as expected. So the good news is it worked. The bad news is it took a long time for us
to get it to work, and it's still nowhere near what we'd need to build an operational system, and in fact the Air
Force Research Lab now has a very robust program called the Scram Jet Engine Demo, which is attempting to
demonstrate that operational capability.
Sometimes it's worth waiting. Stealth, for instance, represents an accumulation of research in large part
sponsored by our AFOSR beginning in the 1950s. It took a long time to develop it, but clearly this was worth the
time line required.
And actually, sometimes we can schedule miracles. There are very famous stories of development programs that
occurred in very short order. You heard mention of one just a few moments ago. The Blue-113, developed in 28
days. The Manhattan Project, approximately three years. The A-12 and the SR-71 derivative really developed in
about four years. So there are cases when under stress we can do it.
One of the challenges I think is figuring out what it is that went right in these programs and applying that
across the board.
Another challenge. Science and technology is not a linear process. It's hard to front load S&T and then just
let it go and dump it into the acquisition process.
There are various activities related to the development of the airborne laser. It’s a criss-cross of basic
research topics, applied research topics, development, testing. Back and forth from basic research, coming up
with an idea, testing it, it doesn't work, go back and develop more research, right up to the present day. So in
fact we still have some very basic issues that are being addressed right now by AFRL and AFOSR and their
associated researchers that are going to go right into the airborne laser.
Another challenge. Sometimes science and technology is accused of providing solutions in search of problems.
That doesn't always mesh well with acquisition.
Very often that's not true. Very often, S&T is very much driven by a need and that need is supplied.
Sometimes, however, that is true and I would argue that it isn't always a bad thing to develop solutions and then
look for the problems that they solve.
I've got a partial list of some of the things that were developed first as solutions and then we later figured
out what to do with them. Home computers. No one envisioned that home computers would be used the way they are
back when they were first being developed—lasers, even the airplane. I don't think the Wright Brothers even
envisioned how the airplane would be used 100 years later. Rockets, telecommunications, the Internet, the Global
Positioning System. I don't think anyone would have accurately predicted that in the year 2004 you could walk
into an electronics store, put $100 down on the table and walk out with a GPS unit so you can figure out how to
drive your car. That was a solution that found its application and a very useful application.
If we really knew what we were doing, it probably wouldn't be research, which is a slightly glib way of saying
that research can be very unpredictable. That doesn't always match well with the acquisition process. However,
the payoff for that unpredictability can be well worth it. I'll give you an example here.
There's a lot of interest right now in enhancing our space capabilities. I've got a laundry list of different
technologies, revolutionary technologies that can provide tremendously new capabilities in the space launch area.
Things such as new fuels, advanced fuels for rockets, nano-technologies, very microscopic technologies, all of
which could reduce our cost of launch, improve our efficiency of launch, reduce the size of required payloads and
satellites to make space more operational. Everything on this list is really a risk. We don't know if any one
of these technologies will work, but if any one of them works, then the payoff for the final acquisition process
could be tremendous.
So the moral of that story is sometimes we have to be willing to take risks. We have to be willing to
understand that not everything will pay off, but the ones that do will be the gems that will help us when we
finally acquire systems.
With that I'd like to conclude.
General Martin: First, Dr. Sambur and Dr. Lewis, it's great to be up here with you. I think they've
really set the stage well for some questions that I hope we'll get into as soon as I finish my pitch.
I'd like to take it from a little bit different perspective. What I'd like to do is look at it in terms of
some issues out there in the operational world that we're facing with some needs that the Air Force has, and then
perhaps some areas where the technology can be focused.
We just saw a great pitch that talked about some of the technologies that were moving. Were they ahead of us,
were they satisfying a need that had already been defined, or were they technologies that were looking for a
problem that actually existed but we didn't know that that technology would make a difference?
You'll see at the end of this pitch that there are some things out there we need if we're going to accomplish
what it is our leadership in this nation is pushing us to, and what they're pushing us to is transformation, a
different way of looking at how we operate.
Transformation is about the technology organization and the infrastructure that will enable us to do what
President requires. It's for us as military members to define the battlespace on our own terms.
The problem is that too often in battles, whether we are attacked first or are responding, we find that the
enemy that we're facing is perhaps defining the battlespace on their terms, and that is certainly true in the
global war on terrorism. There's a group of dedicated terrorists out there that are after us. So let's just
talk about some things that will help us in that part of the game. Many of them will apply to high intensity
conflict and much larger warfare, but in the end, let's just look at this problem.
How do we go about defining that battlespace on our terms? Let me give you a war story from an Air Force
Reservist, now a Guardsman, named Master Sergeant Jim Hotaling. Many of you were here for the 12 Outstanding
Airmen of the Year dinner and you also saw other heroes that we had that had been on special tactics teams and
combat control and that sort of business. Hotaling is a special tactics team individual. He was a master
sergeant in the Reserves on the 10th of September 2001; on the 11th he went down to his unit, raised his hand
and said, “I need to go. Sign me up, I'm ready to go in this global war on terrorism.”
By November, he had been deployed into Afghanistan. Many of you will remember that as we started Operation
Enduring Freedom, the only way to get into Afghanistan was by air. They were inserted with a special tactics
group to go to an observation position about 3,000 feet higher than the altitude they were at, which was about
10,000 feet then, up into the snow and to observe areas where we thought Taliban were and report on those and
ultimately begin to help us understand the nature of that enemy in such a way that we could get in to take it
They had 140 pounds of pack each, most of them Hotaling's batteries that his team was having to carry for him
up this hill. It was 3,000 feet. It was the hardest thing he said he's ever done. He was a Washington State
trooper who had been issuing tickets and eating donuts. He says he wished he'd eaten fewer donuts before he
started that trek. He got to the top. They were up there, they were in position, they were reporting, and out
of nowhere came a Bedouin wandering by on top of this 10,000 foot mountain. They didn't know whether to shoot
him, to arrest him, or to let him go. After a couple of hours, they decided to let him go. Two hours later, he
had made it down the hill and he had an attack force that was formed up and coming after Hotaling and the crowd.
They went down the backside, they got onto their all terrain vehicles that they had stashed there, and they
headed out. Now they're being chased by a Toyota with a 50 caliber machine gun on the back and a whole group of
Taliban that are running after him.
He's on the radio asking for help and he got a hold of a Navy P-3 that happened to have an infrared ball and
happened to have the electro-optical ball, who could look down and see what was going on and tell Hotaling and
the crowd which way to turn—go up this valley, turn down that road, nobody's there, stay away from this
population area which is to your one o'clock for about two miles. Go this way. Pretty soon they were able to
get to a position where they could turn around, call in air power, and do away with the Toyota with the 50 cal
on the back and everyone else that was in it. He was beginning to understand about this defining the battlespace
Then, four months later, Operation Anaconda. Same kind of thing. Got into a contact point, had to move along
a ridge line into an area that overlooked Roberts Ridge, you recall, where Seaman Roberts was knocked out of the
bad end of a helicopter that was hit by an RPG and then we sent forces in to rescue him while other forces down
in the valley were beginning to engage. He was in a position where he could see all of that.
In order to get there he had to go through bad guy territory. He was able to get a hold of a Predator that was
above him and the Predator could show him everywhere they needed to go to get to their observation point and then
ultimately be able to direct fire on enemy positions. And he was talking to somebody back in the United States
that was flying the Predator and that was controlling the sensors.
Fast forward one year. He's now in Operation Iraqi Freedom. He's in the western area. His mission was to
stop World War III from happening, i.e., find the Scuds before they launched them against some of our allies.
So he was out there with his team and he had this time a hand-held pointer UAV that could go out seven or
eight miles and report back; could fly for 35 or 40 minutes and show him what was on the other side of that sand
dune. Their objective was not to get into contact. Their objective was to get in position where they could
observe and make a difference.
So Sergeant Hotaling inside of about a year and a half proved the value of having the kind of ISR support at
his control where he was able to define the place that he would choose to engage when necessary.
So I asked him as we went into Fallujah and we went into Baghdad, “what do you think about what we're doing
there?” He said to send a squad of people down an alley in a city and not know what's in there already blows him
away. We have the technology to do it, we need to do it, and we need to transition it as Dr. Sambur talked about,
very, very quickly. That's what we mean by defining the battlespace on your own terms. Get there ahead of the
enemy at least with the knowledge so you can pick and choose where you need to go. That's where technology needs
What do we bring to the fight? We bring range, we bring precision, we can bring knowledge, we can bring the
decision, and we can bring the time factor into effect. So range, precision, knowledge, decision and time.
We're green in the range. We can go anywhere in the world. We're pretty green in the precision. We can hit
coordinates very well. We'd like to have a little matter of terminal guidance so that the coordinates and the
target are exactly the same. We'd like to know more about the battlespace than we do today. We'd like to know
it in such a way we can make a decision quickly and we can act in time. So where are the transformational
vectors for our Air Force and where are the things we really need to work and partner with the Air Force Research
First of all, I think we need to have battlespace knowledge at the touch of a screen. Fused, animated, and
presented in such a way that decision makers can understand it, see it and understand the battlespace without
having to do mental computations and make decisions quickly, so that we can then achieve the kind of effects that
we need in near real time.
So the real watch words for us are the knowledge, the decision, the action, in time. And what we're really
trying to do then is to break the time barrier. That's where we're going in combat operations.
Yesterday, for those of you that were here for the Four Star Panel, you heard the Chief of Staff talk about
something called “time of flight.” “One time of flight.” He used 40,000 feet, 50,000, maybe 60,000 feet, and
to be able to understand the battlespace and to have an effect occur within about a minute.
If you're in a B-2 and you're at 40,000 feet, it's in about 40 seconds. Let's assume that you have a
permissive environment. Let's assume that you have perfect ISR. You know what's going on. Those are the first
two things. Knowledge and decision. You've decided you want to do something. How fast can it happen?
Well, in a permissive battlespace, it really depends on where the system is. Does it have the persistence to
be there? Does it have the persistence to be as close as you want it to be proximity-wise? And then the time of
fall, because the minute you tell the system to activate digitally it's going to plant the coordinates into the
system, it's going to drop it, and you'll have that effect in 40 seconds. So the 40 seconds is what we're after.
What happens when you get into a non-permissive environment, a hostile threat, where we've got surface-to-air
missiles? Now all of a sudden you've got a boundary there that maybe you can't penetrate easily. You need an
aircraft that can get in there and deliver the weapon in that same 40 seconds. Supercruise can help. It gets
you outside the threat area. Perhaps that with a small diameter bomb can make 40 seconds work.
I was surprised, many of you know I'm a geography major, but every now and then I do think about technical
things. Come on now. [Laughter] Those of you that know me know that's not true.
The fact is that I can drop a ball from here and it's going to hit the ground in two-tenths of a second. I
can fire a .22 out there and it's going to hit the ground in two seconds but a thousand yards down range. So
speed matters when it comes to time of flight.
So in this case I can stay outside that threat ring if I have the right system that can deliver the weapon
and meet that 40-second time of flight.
What happens if it grows? Now I'm into supercruise and maybe I need hypersonics or maybe I need some sort
of powered system that is going to take that target out. Again, remember I'm assuming somehow I have perfect
knowledge. I know of something that needs to be done. I'm only talking about the last part of it now which is
actually in time.
What happens when you don't have the ability to be inside that particular radius there that allows you to act
within 40 seconds? What are the things we have to pay attention to? I'm still after 40 seconds. I've got the
knowledge. I'm making the decision. But now maybe I have to go 600 miles. Maybe now I have to go 1,000 miles.
How will I do that in that time of flight? That's where I need help in technology.
The need is for unprecedented survivable endurance. When you have that, whether it be from air and space,
you have it because you've got a sensor network that can give you that battlespace. No matter what happens, you
need unprecedented survival, endurance, so that you can stare and have the knowledge of that battlespace.
Think about that. There's engine technology there, there's stealth technology, there's space technology,
there's interconnectivity that's got to occur, there's bandwidth issues. All of that now is focused on having
knowledge of a significant part of the battlespace where you can watch everything that's happening, know what's
happening, and know what it means to you.
Now you've got to have the next phase which is exactly what we said, the unprecedented fusion, animation and
visualization of that battlespace. Visualization so you can see it, understand, and not have to do computations.
You can make decisions very quickly. Those are the kinds of things you're looking at.
Once again, connectivity not only between the sensors, but the sensors to the decisionmakers. Have enough
horizontal integration so that you know things of meaning across the individual stovepipe systems, not just one
or another. The ability to exploit very fast and then visualize it in such a way that it's friendly to the human
that's using it. Then last, how do you act in time?
Is it going to be by presence? In other words, you don't need anything really new if you're there. But what
if it's a high threat area? How will you get that survivability? How will you get that kind of persistence?
Are we talking about hypersonics, are we talking about directed energy, are we talking about space kinetics?
In the end, if you're trying to be transformational where this Air Force is going from a technology standpoint,
these are very critical things for us to focus our efforts on.
It's clear to me that as we move forward in the Air Force Material Command with the Air Force Research Labs,
with our acquisition support and with our sustainment, we were missing something. We weren't quite tied together.
We now have an organization called XR, integration capabilities or capability integration. Their job is to
know what technology can produce, to know the deficiencies, to know the status of our acquisition programs and
develop roadmaps of opportunity from the bottom up instead of waiting for some of our senior leaders to see it
and then push it from the top. Because what if they miss it? They weren't there for that discovery. We might
miss something important. We need a system that's providing that kind of information.
I look forward to your questions.
Q: Major Sam Lightfoot. I'd just like to know how the President's directive for space exploration has
affected you? What are the requirements for S&T?
General Martin: First of all, I assume you're talking about moon, Mars and beyond, and the commission
that reported out in the June timeframe? Okay.
Frankly, at this point, it has not manifested itself in some series of actions that have been issued to the
Department of Defense. There are some actions that are being reviewed by NASA. So it has not yet, that I know,
to my level, had an effect on where we're going in our S&T area.
Having said that, however, I think it's important for you to understand that during the Space Commission's
report we spent quite a bit of time making sure they had insight into all of the technologies that we were
pursuing in the Air Force Research Lab and in our over a billion dollars worth of work we do with our grant
agencies. Many of those technologies were applicable and necessary for us to pursue some of the President's
objectives, but they were also applicable to the things that we're pursuing in terms of trying to get rapid,
on-orbit situations. Joint Warfighting Space, where we have smaller, more capable, sensor systems that we can
rapidly populate a constellation with; the right propulsion materials; and, of course, other technologies that
will help us reduce the weight and increase its capability with less power.
But I'm not aware of any direction that has come to us to say, “okay, we really want you to accelerate these
because that will make a difference in the moon, Mars and beyond.”
Q: Dr. Sambur, my name is Ted Stokes. I'm from Hill Air Force Base. I had a question about your
acquisition times. I agree with you that many of our systems take too long to acquire, but I was wondering what
methodologies you may be aware of or are considering that can help us reduce those times.
Dr. Sambur: Mark had discussed some of the major programs, for example, the F/A-22, which in actual
fact has taken more than a quarter of a century to develop. Mark’s chart I think had only 20 years so, you were
off by five. [Laughter]
The average duration of a major acquisition program in the Air Force and I think it's true of the Army and
Navy as well, is in the neighborhood of 10 to 12 years. And as Mark showed, not only do you have political
changes that occur in that timeframe, but think of the changes in technology in a 12-year development.
I had a chart that showed, with respect to the F/A-22, that you had a six megahertz processor at the time that
the F/A-22 went into development, and now you have six gigabytes and beyond. So technology is rapidly changing
when developments are very long.
With respect to your question, how do we shorten development times? We did a survey to determine what were
the major causes of developments taking so long, and there were three reasons that stood out as the most
significant with respect to why acquisitions take too long.
The first reason was that requirements keep changing. That we can't keep making up our mind. And we start
down a path and then we say, “we're not quite sure of that, let's do a change.”
The second reason was funding as well changes. When funding changes it has a very disruptive effect on the
The third reason was inadequate systems engineering. Now if you've been following what we're doing in the
Air Force, there is a push right now to make systems engineering a very important aspect of any acquisition
decision and also a very important aspect with respect to how we do programs.
Now I use this example of a FRAM filter. I think many of you have heard of that advertisement in which it
says, “pay me now or pay me a lot more later.” The implication is if you don't do the groundwork at the
beginning you frequently run into problems at the end.
So we're trying to develop a systems engineering philosophy that says at the outset do the systems engineering
right and do it in a way that there is flexibility and robustness in the systems engineering so that we can
accommodate change, the normal change.
I’ll give you an example in presentations that I've made, that when you go out and buy your first house as a
couple and you don't have any children, if you're very clever you go to the person who's building the house for
you and tell him, “we don't have enough money right now to fix and finish the basement, but tell me what the
smart things are to do so that when I do have enough money I can fix my basement in a relatively inexpensive way
without knocking everything off.” And the constructor will come back and tell you put in the piping, do this,
and when we're under construction it's relatively easy to do that.
That's what we're seeking. We're seeking a robust systems engineering approach that allows for changes in
requirements, that allows for changes in technology, and allows us to set a foundation that is robust so we can
meet various requirements.
Now as for changes in funding, what we have right now is a policy that says the reason why we have funding
issues is primarily that when someone bids on a program they usually take the low bid approach. They think the
Air Force or the services really don't understand best value and the only way to do that is by finding out what
the budget of the service happens to be and then bid significantly lower than that. How many people working in
industry have done that? Be honest. Come on, guys. You always bid low.
What we found here is that before we're able to start a major program, an ACAT-1D for example, we have to
have a funding profile that's consistent with the CAG, the Cost Analysis Group. What happens is after the bid
comes in it's usually some fraction of that CAG bid. Usually it's one-half of the CAG bid. What then happens
is that that CAG estimate goes out the window, and then you have to budget based upon the real bid, the bid that
you have in hand, even though our expertise tells you it's going to cost at least a factor of two.
When you look historically and you look at cost overruns, you find out that the cost estimates of the CAG and
also of the services—if you took them as the real funding profile—you would find out at the end of the day the
real cost was within one percent of those estimates. However, if you compared the final cost with what is bid,
what you find out is that there's a 36 percent average overrun. So the bid price is usually a factor of 36% less
than what reality is, but if you take the estimates from the service or the CAG estimate, there's only a
disconnect of one percent.
We are now imposing on our bids a factor that says that if you bid less than what our estimate is, we will
assess that as a risk factor. We will tell you it's a risk factor and if you don't come back and tell us why you
have this incredibly novel way of doing it less than what we estimate, we're going to throw you out because we
cannot take all of this funding perturbation.
We just did that on this last bid with the battle management command and control system on the E-10 and they
were very close to what our estimates were, so we didn't have to throw anybody out. I wrote a letter to all of
the executive vice presidents indicating to them that if you bid low because you think low bidders win, you're
doing it at your own peril. It's working. We have a policy now in place.
So with respect to shortening cycle time what we're trying to do is get rid of the three factors that cause
programs that normally take three to four years end up to be more than a decade. Hopefully, that will achieve
some real results.
General Martin: One thought on that. We haven't, I don't think, transformed outside of the Cold War
mentality which set up a series of reviews and a very strict regimen of documentation to process those reviews,
which then resulted in decisions about going through different milestones and proceeding with the program.
Now, that process is very heavily structured and is reviewed by a whole series of functional folks across the
DoD before someone can get a decision to move forward. It's a Cold War mentality. In other words, we're going
to now produce the next system and we're not in a war, we're not fighting, and we hope there will never be a war,
but this thing will deliver sometime downrange.
Now we're having a war every two years or we're in a constant war. So the concept of spiral and then the
format of how we go through a decision making process and the documentation system has not been streamlined, yet
we're trying to make decisions quickly.
An example, when we came out with the advanced concept technology demonstration concept back in the '94
timeframe, the first system that we move forward on dealt with UAVs. Predator and the Dark Star and the Global
Hawk. The Predator I think delivered very quickly. Within about 18 months they were beginning to fly the
machine and pretty soon we actually wanted it over in Bosnia. So in ‘95, a year and a half later, we're
deploying the Predator into an operational area where it was making a different with respect to battlespace
knowledge. And it has never come home, for all intents and purposes. Yeah, we've brought some of them back,
but in the end the warfighter wants them and they want more of them.
I remember directing ACC to develop the ORD for the medium altitude UAV and back came a machine that was going
to have to go 270 miles an hour, it was going to be up to 48 hours instead of 24 hours, was going to have a
bandwidth thing that was different, it was going to be a different wave form system, and a whole bunch of
I said, “guys, the problem we've got here is that the Predator only goes 70 knots. We might get it to 120.
Do you need 240? Are you going to give up the Predator, stop everything we're doing, and build a whole new
system based on what you've learned?” Advanced concept technology demonstration was to leave behind something
that would be useful to the warfighter and we would spiral to, but the ORD process was going to build a machine
that we would know as un-obtanium. [Laughter]
So what they did is they worked back and they said, “well, we'll have threshold and we'll have objective
requirements,” and we went through it. Of course, then the test and evaluation said it was unsuitable and
operationally ineffective. Dr. Roche talks about that regularly. He's got that assessment in his bottom desk
drawer, yet we're using the Predator every day. In fact, we put weapons on it, we put sensors and lasers spots,
and so for a system that is unsuitable and operationally ineffective, we seem to be doing a lot with it.
The system hasn't transformed yet to understand what spiral is about, what speed is about, and what it means
to have operational capability that can grow and change as Dr. Sambur was saying with his house example.
So we're stuck in an old form versus substance issue and I think we need to get on with it and figure out a
way to move to a new method by which we score the programs, we account for our expenditures, and we move forward
in a way that is operationally effective for that model and version.
Dr. Sambur: It's worth discussing a little bit more on this un-obtainium issue that General Martin
Frequently, the warfighter wants to make sure that when he writes an ord that he enraptures everything he can
conceive of, even though technologically it may be very difficult, the technical maturity is not there. The
reason for that is because there is a lack of trust in the spiral development process. The trust is that if he
doesn't put it in right now, will we ever spiral to the next state? So they write these ORDs, and a lot of times
they not only write ORDs that have great requirements in there, they frequently pad the requirements.
If you speak to a warfighter, and I have countless examples that I can tell you about, he will make a
requirement. Let's say he really wants X but he'll write the ord 1.1 times X. The reason why he does that is
because he knows he'll never get 1.1X, but at least if he makes it 1.1X, 10 percent greater than what he wants,
hopefully he'll get X. So he puts this 10 percent pad in.
What we're doing right now, and we're even doing it with the Chief and the SECAF, is an expectation management.
What we do here is we meet with the requirements people and we tell them about the difficulty of this and if they
want us to go ahead and do that they actually have to sign a document that says that it is our acquisition's
estimate that this is going to take a heck of a lot longer than you anticipate and the funding is going to be a
lot greater than you have budgeted. Do you really want us to do this? They have to sign off for it because
there has to be a realization that if you want un-obtainium, then you better have the funds for it and you better
have the schedule flexibility to wait for un-obtainium to happen.
Q: My name is Patrick Mason with the Air Force Research Laboratory with the Directed Energy
Bio-Effects Division. We have numerous ideas for directed energy. What we really need help on from the lab and
from industry is thinking about how do you make a smaller footprint for a directed energy device? How do you
have out-of-the-box thinking for antenna design? How do you get a smaller footprint? How do you get better
power supplies for the directed energy devices? I think that's where the future is for directed energy.
Dr. Sambur: Thank you for bringing that up.
The real issue, as I tried to point out in my presentation, is that people really don't understand the process
of getting their ideas brought forward to the Air Force or the Army or the Navy or whatever the service is. I
recognize in order to have an effective S&T process it can't only be funded by the Air Force. It has to have the
ability to be synergetic with everything that's out there. The only way to do that is to make available to
industry what our needs are, and also be willing to see what they have and evaluate it.
That's why I had that offer. We're all happy to learn about new ideas. I mean you don't have to hire a
billion dollar consultant to get to the Air Force. We want to know your ideas. The process is there, it's
fairly easy to do, you just have to call and tell us.
General Martin: One other thing that I think you were also alluding to and I think is a very, very
important point, and it gets back to Dr. Sambur's systems engineering.
No technology is an island. The guys that are doing directed energy, that work themselves into a very useful
technology opportunity, have to be married with the people that are going to produce the power for them; have to
be married to the people who are doing the antenna, the focusing of the beam. So it's not just one technology.
One of the areas we have to be careful about is that once we see a value and marry it up to a capability, we then
are going to have to put together a team that actually looks across the different technologies that will actually
operationalize the system that you're talking about.
So that's a great comment and one that I think we can improve on. But that's also the reason we created that
XR organization, that capabilities integration, to help us with that.
Return to AFA Air & Space Conference