Lieutenant General Ronald T. Kadish is the director of the Missile Defense Agency, Office of the Secretary of Defense, Pentagon, Washington, DC. The MDA is presidentially-chartered and mandated by Congress to acquire highly effective ballistic missile defense (BMD) systems for forward-deployed and expeditionary elements of the U.S. Armed Forces. Additionally, MDA will develop options and, if directed, acquire systems for BMD of the United States. As director, he is the Acquisition Executive for all BMD systems and programs.

General Kadish entered the Air Force in 1970 after graduating from the Reserve Officer Training Corps program at St. Joseph's University. He was the program director for the F-15, F-16 and C-17 System Program offices, as well as director for manufacturing and quality assurance for the B-1B System Program Office. He is a senior pilot with more than 2,500 flying hours, primarily in the C-130. Before assuming his current position, he was commander, Electronic Systems Center, Air Force Materiel Command, Hanscom Air Force Base, MA. He was responsible for the Air Force's Center of Excellence for command and control systems, handling more than $3 billion in programs annually.

Interview by Jeffrey McKaughan, MAT editor

Q: 2003 has obviously been a year of great activity for the Missile Defense Agency (MDA). Could you give our readers an overview of the organization and any changes that have taken place since President Bush's December 2002 announcement?

A: Over the past year we've done a number of things. One is that we have taken the technology that we have proven, and the concept demonstrations of the hit-to-kill technology, and moved from a test bed focus to an operational system that will also serve as a test bed for realistic operational testing. So, as we speak, at Fort Greeley, AK, for example, we have numerous facilities being put in place that will allow us to place on alert a capability to defend against a limited, long-range threat. At the same time we are pursuing aggressively the Aegis SM-3 sea-based defense to counter short, medium and possibly intermediate range ballistic missiles, which will come out a little bit later. Of course we have the Patriot 3 deployed. And we will be augmenting that through the United States Army as we're transferring that effort to them. So it's been a very aggressive set of activities to actually field missile defenses of this type for the country's benefit.

At the same time, we have been running a very aggressive R&D program to support these efforts - improving the kill vehicles as much as we can through the design process and maturing other technologies like airborne laser for the boost phase intercepts. So, from an R&D perspective, we've been very aggressively pursuing the different technologies while we are working to put on alert our basic systems, and then we'll improve them over time. The R&D lays the groundwork not only for the near term, but also for the long-term improvements that will continue to enhance our defensive capabilities.

The third thing we've been doing is putting together the management structure to support those activities for the long term. The Missile Defense Agency, as an agency, is not just doing experiments, but we are into a full range of R&D, testing and deployment activities.

Q: Is the MDA actually directing the focus of the overall R&D effort as far as technology focus and funding priorities?

A: We are directing the entire effort out of this agency. Previously, when we were the BMDO (Ballistic Missile Defense Organization), our strategy was to franchise to the services the development of these activities, and we provided the funds and general direction. Because we have been directed to develop a layered system for protection in all phases of flight against all ranges of missiles - that's the mission we have received as an agency - the ballistic missile defense system has to be integrated whether it's operated by the Army, Navy or Air Force; that's a basing activity. Therefore we need a single design for the research and development activity that we [MDA] control in order to do that. This enables us to make those design decisions and make sure they are integrated from the beginning as much as possible. That's why it's important to have the management activities that support those kinds of decisions, and that's what we've been working to put together.

Some things we've had to invent in terms of being able make progress. This is not only in terms of technology, but the management systems to support the development of the technology, because if you are trying to shoot in all phases of flight and against all ranges of missiles, no one system can do it. So you have a collection of bullets and sensors, if you will, that have to be integrated to make sure those types of defenses work and, as you layer them, they become more effective. That's the whole idea.

The different systems must be able to talk to each other. Otherwise you're kind of shooting separately and you're not providing the maximum effectiveness you can. So we believe if we have a proper integrated activity, small numbers of interceptors can be very effective in that type of approach.

Q: As the clock ticks down to the "on alert" status, how would you characterize the facility construction progress at the basing end of the program, such as Fort Greeley and other locations?

A: There really isn't a "ticking clock." Our direction is to deploy an initial defensive capability beginning in 2004, with the goal of enhancing this capability in 2005 and beyond in the form of two-year blocks, continually improving the effectiveness and military utility of the system. The infrastructure required for an initial capability in 2004 is proceeding on schedule. I think, if I remember the schedules right, by the end of this calendar year, we should have beneficial occupancy of all of the facilities that we need - therefore it's a matter of putting the mission equipment together and then putting it on alert status.

There are some areas of risk: one is the GMD booster program where we still are lagging behind a little bit on the integration and testing, but I expect it to pick up pretty aggressively in the next few months.

Q: Do you mean on the actual system integration of combining the launch vehicle and the kill vehicle to put it into space or when it is ready to separate?

A: Both, because up until this time, we've done repeated testing with a booster that we don't intend to use. Now we're building that combination that we'll actually put in the silos and that's what we need to test. That's lagging a little bit, but I think right now I'm pretty comfortable with where we are. We'll just see how it works and if we don't meet our schedules, we'll have to make some adjustments. But right now I don't see a problem.

Q: Talking more about the specifics of an intercept test, so far, most of the tests have taken place between the United States and Kwajalein. When do you think you'll be in position to test different geometries as far as intercept angles, more head-to-head or trailing?

A: That's the kind of thing that is hard to explain. We get criticized for doing the same geometry but you have to have the facilities in place, like Fort Greeley and Vandenberg, in order to test different geometries.

Future tests will involve different geometries, including target launches from Kodiak Island in Alaska and interceptor launches from Kwajalein and/or Vandenberg. I think that starts the process of filling out the testing in the intercept envelope. This will be relatively near term.

Q: Do you think testing different geometries is as important from MDA's perspective, as it might be from that of the critics of the program? Are there tremendous gains to be made from testing these different geometries?

A: We've always wanted to do that activity. The more, and different, data points you get, the better the engineers like it. It's been a matter of time, money and ability to do this. By ability, I mean we weren't able to do some of these activities because the ABM treaty strictly specified the test range that could be used, and this prohibited us from testing in other locations. Since we no longer have that restriction, we have moved out as rapidly as we can to put those things in place. So the only real change we made with the president's decision to put these systems out in the field over the next couple of years, is we went from having a test bed that could be operational with a few missiles, to an operational system that could be used as a test bed, which is, in my view, the ideal way to accomplish this mission.

We're actually changing the way we're doing business. Over time, we will learn more about the system, and this is not only a technical issue. Part of what I don't think everyone appreciates is that the military services and the combatant commanders, who are charged with the defensive responsibility, have not seen anything other than view graphs for 20 years. So introducing a revolutionary or an unprecedented weapon system into the military structure is a difficult thing, and it changes the way people think. A good example is when we introduced the airplane nearly 100 years ago into the military. People were trying to figure out what to do with it. It started out being a surveillance, reconnaissance activity and then somebody decided maybe we could drop some bombs from this thing or carry a pistol or maybe a machine gun. Well, those ideas came not from the designers but from the people who were trying to use them. Here, we have the idea of spiral development combined with the user community trying to get comfortable with a new technology and thinking about having it available everyday. This is a major issue and an area of uncertainty.

That's why a grand design that people would like us to have is not appropriate. Delivering a 100 percent finished product is not appropriate because I don't know what that finished product will really look like in terms of its operational usefulness. We do know that it will be operationally useful, but it's how we decide to put it together, and I can predict with some confidence that what we think today is probably not what we're going to think three or four years from now.

Q: The technology is that dynamic?

A: The fundamental technology will get better and better and better if we are allowed to continue to invest in it like we have been. The user community, with us, can decide how to make it as useful as we possibly can make it. Overall, I think we have done as much as we can to make this idea operationally representative and fill out the design envelope, and use our models and simulations to predict the performance as good as we know how to do.

And you're talking about test ranges that span thousands of miles. You launch these things and people sometimes don't have the perspective when they see the target going out. They see a launch plume and then they see some kind of a flash on the television screen, but it is difficult and complex to run these kinds of tests. The media, as well as some of our other critics, say that we ought have every test done under every conceivable condition before we move into an alert status. The problem with that is that we're not exactly sure what all the tests need to be in terms of the objectives that need to be accomplished and we're never going to be done. So approaching it the way we're approaching it, I think is a very rational and prudent thing to do.

Q: Critics often focus on whether a hit on a target vehicle occurs to classify a test as a success or failure. In the instances where there was not a target intercept, do you consider those tests failures?

A: Every test has a numerous set of objectives and normally what happens with a test that fails, where we don't get an intercept, is that we accomplish an awful lot of the objectives and so we learn an awful lot out of every test we do, whether it succeeds or fails. In essence, there may be literally hundreds of sub-tests within the overall test and each needs to be measured in terms of performance, and each is important to the overall development effort. However, there is one truth that I learned in this business since I've been in this position and that is that it doesn't matter how much you say you learn. It only matters if there was an intercept. And then, of course, when we have an intercept, we're accused of rigging the test and things of that nature.

The bottom line is that we want to test as we would fly, and fly as we would operate. We're trying to do that to the best of our ability. We are learning an awful lot. We have gathered enough confidence to say that we can put the GMD element on alert and have a practical approach to missile defense for this country, and we're going to improve it as rapidly as possible. We will have failures, we'll have successes and the sum total of that will be the performance of the system under operational conditions.

The other thing I would say about these tests, and it's the same way with most weapons systems, is that we try to get the most out of it for the money. These are very expensive tests, so we don't want to squander any opportunities. We load them up with a lot of things to do, a lot of objectives - not just one objective. It would be nice if all we had to do was say, okay, we're just going to hit that warhead and that's it, but we want to do a lot of things.

In a test, for example, we only shoot one interceptor at one target, but in the operational system, we're not going to shoot one interceptor at one target. We're going to shoot as many times as we need to kill that target - and the system is designed to do that. Now, of course, you could run out of bullets and that's the game we're playing with the countermeasures issue.

One of the things I'd like to do with the testing is to shoot two interceptors at one target. If one misses, the other one has the opportunity to intercept - just like it will be in an operational system. We're planning to do it at some point; we just haven't nailed down exactly when that will be.

If people get the impression that because we failed to have an intercept in any one test, the overall technology will be ineffective or unreliable, and that's the wrong way to think. Things fail for a lot of reasons, some of which have nothing to do with the basic technology involved. One of the things that is pretty clear to us right now is that all of the failures we've had in our program so far have been quality failures with a couple of exceptions. And by quality, I mean that things just broke.

Out of the three failures that we've had in the ground-based system, two have been failures of booster and kill vehicle separation. This is not something that pleases me. We should know how to do that - we do know how to do that. The fact is that we've had failures with pieces of equipment that have worked many times before. So, from a testing stand point we learn from every test, and we gain confidence based on our ability to perform these tests over time. If a component is prone to break, its better that it happens during a test, because then we know that particular piece of equipment needs improvement, and we'll fix it. The end result is that this component installed in an operational system will be very reliable.

We're going to continue to learn about this technology whether we have successes or we have failures, whether we have an intercept or we don't. But each and every test continues to build our confidence in this technology

Q: Going back to the multiple interceptor launches at a target, is the system set up to use a shoot-look-shoot? In that, the system would evaluate that the probably of hit from the first launch and, based on probability, launch a second interceptor or is it based on shoot-shoot-look?

A: No, that's a good point because when we get to the point in the system that we have enough sensors and they're in the right spot, some areas will be shoot-look-shoot, others are shoot-look-shoot-look-shoot and some are coming in too close, so that will require a salvo. The system is designed to handle all of these modes of operation.

Now with the limited inventory we have, we're working through those issues with a firing doctrine and things of that nature. But we look at shoot-look-shoot as being a good thing, and when we're able to get it with our sensor lay downs, then the capability of our fire control systems will be there as the preferred mode of operation.

What we want to do is kill the target. We will do the highest probability of kill that we possibly can within the constraints of the system. If we can do shoot-look-shoot and do it with high confidence, we will.

By the way, that applies to when we get a boost phase stage interceptor as well. We want that warhead to go through as few layers as possible. We want the attackers to know that we are going to be shooting at it all the way, and that their chances of success are minimal.

Q: Most of MDA's funding goes toward RDT&E. Do you think that MDA will ever become a procuring agency, where as they will actually have the ownership of the launch and kill vehicle or the facilities?

A: It depends on what we're talking about. Something like the Patriot 3, where we made it as basically an upgrade to the Patriot system, it's proper that we transfer that as quickly as we can to the Army to procure and manage the effort, and then we maintain an integration responsibility to the larger integrated ballistic missile defense system, or BMDS, of which PAC-3 plays a vital role.

With regard to the ground-based program at Fort Greeley, given the concurrent nature of the development and spiral development activities, it would seem that we are best served by transferring the operation to the services while we retain management of it until it's mature enough to be broken into different pieces for management purposes. One of the things that we have collectively realized over the past year and a half is that when we have the layered system that we call the BMDS, that we are trying to have as our vision, it is truly turned into a joint development effort. When all the services are involved in one way, shape or form, we've had to adjust our thinking as to how we manage that from a development, procurement and operational point of view.

So in the case of GMD and Aegis, to begin with, it's not clear that we're going to transfer the procurement responsibilities as early as we thought. We will transfer operational responsibilities immediately, but we need to retain management responsibility of the program a little bit longer than we probably had originally envisioned. It's a joint development effort where trades are now being made between and among very disparate systems in terms of their basing. That's part of the management activities that we are trying to understand and adapt to.

Q: Are there any key technology hurdles that you want to focus a little more attention on than maybe you thought you would six months or a year ago?

A: I draw great comfort from the fact that I don't think we have to invent things with the possible exception of the ABL. Even those inventions have been done, but we haven't quite put it all in the airplane yet.

I think if there was one area of technology that we need to explore more aggressively is miniaturization. We had it back in the mid-80s when we did the homing overlay experiment; that kill vehicle was 2,000 pounds. The one we are using for GMD is about 120 pounds. We'd like to get down to seven or eight pounds.

And the smaller they are, the more bullets you have. So that's the technology area we're focusing on. Getting the weight down means we can get the quantity up, and that is important. Now that we know how to do hit-to-kill in terms of the mechanics, miniaturization is where I think our major efforts will be over time.

Q: There's always been opposing curves between system and counter system, or intercept and countermeasure. Are you where you want to be as far as measures and countermeasures?

A: The countermeasure, counter-countermeasure issue is always going to be there, so let's stipulate that. The problem that we deal with - and this is where I think our critics kind of have it right and wrong - in a sense, is that we don't have a good way of knowing what countermeasures will be used with any great certainty. It creates this circular argument of: if you can do that, I can beat you with this, and if you can beat me with that, I can do this.

The question is, how do you get intelligence-based information that keeps up with that kind of a cycle? Now if you have it, that's wonderful because then if we know what a countermeasure is then there's almost a 100 percent chance we can figure out how to beat it, at some point in time. However, knowing what it is with certainty when we're trying to build systems like this becomes a real problem.

What we've been trying to do, and I think we're having pretty good success with it - although we won't talk about it outside of classified channels in any great detail - is that we want to be what we're called physics-based in our capability within the system. Normal military systems are built against a threat where the intelligence community does a very good job of trying to find out what the parameters might be of a particular countermeasure, and then we design to that with some degree of uncertainty spread around it.

What we would like to do is not have that situation. Instead, we're taking the physics-based approach where a countermeasure can go from X to Y in any particular parameter and then say how far can we get to that with our system, you have engineering bounds, physical bounds and threat bounds.

The approach we're trying to implement, and I think we are having some success, is that we open up the capability of the system in many different parameters of the threat potential as wide as we can. Then we're looking for the intelligence community to tell us whether they think the threat could be in that space so that we have a capability-based approach as opposed to a threat-based approach. Because the price of being wrong is pretty high and what we don't want to do is be easily fooled. We want to be as tough as we possibly can in opening up those approaches. Now there are practical limits, but we will then try to make these systems as immune as we possibly can to having to know specifically what a threat looks like in order to have success.

And that, combined with our firing doctrine, can provide a very effective defense system in our estimation. Then when you layer it, you further complicate the adversary's problem because it's not only difficult to make ballistic missiles and warheads behave properly - it took many years for us do that - but having countermeasures that are effective are also not as easy as people make it out to be. And even if they are there, we want to make certain that they will be ineffective.

I think we are making great progress along those lines, but I can't say too much in terms of details about how we're meeting this challenge.

Q: As far as international participation in missile defense, do you see the level of international participation gaining speed or other countries coming on board in the program?

A: Well, you have to understand where we started with regard to international cooperation in missile defense. The ABM treaty specifically prohibited us from sharing any substantial data with our allies on missile defense technology. We've only been out of the treaty a year and a quarter, so the time period in which we've been able to very seriously talk about this is relatively short. There are basically a couple of conditions in order for us to collaborate properly with our allies. One is that we have to have the proper industry-to-industry contact. The other is that we have to have government-to-government arrangements that sanction those kinds of practices. The third is we have to have beneficial-type - what we call noble work - to actually work on. It doesn't make sense, for instance, for an ally to spend a lot of money on something that we've already spent a lot of money inventing. So the business cases need to be defined and that takes time, all those things take time. So we are actively pursuing government-to-government agreements with Japan and the United Kingdom, for instance, and others, as well as industry-to-industry agreements that are probably leading the pack.

Over the next year to three years, I expect that we will come to agreement on specific things with the allies and industries involved to do missile defense that make a lot of sense. But right now we're still in the early stages. It's one thing to do cooperative type of development and deep technology activities, but it's quite another to cooperate on weapons systems development.

Q: Will MDA address the cruise threat and Russian missile threats like the Scud and the Iskander systems?

A: The weapons systems of concern to us initially are the ballistic missiles North Korea, Iran and Libya are working on - the proliferation issue. Therefore, our initial activities are oriented to that type of threat. As we spiral develop activities within the missile defense world, who knows what we'll worry about in the future? But at the present we're very focused on the near-term, limited ballistic missile threat to our homeland, our deployed forces, and our friends and allies, and the near-term BMDS we're developing is designed for that type of defense.

Of course, the cruise missile problem for the defense of the 50 states is different than the ballistic missile problem because if they are long-range cruise missiles, the adversary has to be closer to us as opposed to a ballistic missile.

From an overall standpoint, I think it's prudent for us to continue our focus on ballistic missiles, and then as the Defense Department and Congress continues to evaluate the cruise missile threat, we'll see where we go. But we have our hands full right now trying to do what we are doing.