DoD News Briefing, Dr. Gansler, April 15, 1998
Mr. Bacon: Good afternoon. We have Under Secretary of Defense for Acquisition and Technology Jacques Gansler here to talk about the F-22 and answer some of your questions. He has to leave by 12:25 at the latest, so you might have to ask your best questions first.
Secretary Gansler: I apologize for the schedule problem, but Ken had suggested that since a number of questions have come up over the last few days about the F-22 that it would be a good idea to clear the air. It may sound like I'm crying wolf, but it's not that. It's that we think it would be appropriate to give you what the story is and where it stands.
It's actually an "in-process" story. We've been reviewing it, and I'll explain why.
I have one chart which we have copies of for you, it will make it a little easier to explain the status of the program and why we're having some reviews of it. In fact some of you have probably seen the GAO report written about the program.
As you all know, I've been in this job for about five months so it's a new program from some perspectives for me, although I've been in this business for 40 years so it's not that new. I've been following it for some time.
I have with me today General George Meullner from the Air Force Acquisition community, and George Schneiter who is on my staff and who has been following the program as well. If you have any historic questions or background questions, the two of them may want to answer them for me.
The two issues that have come up, that were raised by the GAO that caused me to go back and look into the program and to gain more confidence myself in where the program stands, have to do with the fact that the test program was delayed. The basis for the delay in the test program was primarily that associated with manufacturing process issues. This is not a design deficiency, but it was a manufacturing process. Some of the structural elements they found in the manufacturing process had some cracks in them. George can go into the details of that. As a result, because they were major structural elements, that caused a significant delay in the test program.
We then had a very successful flight test of the system. The plan was always to instrument that first aircraft with detailed strain gauges and so forth, so the plan was followed; the aircraft was taken apart; all the strain gauges were added to it. While it was taken apart it was flown out to California for the flight test program, and it's going back into flight testing.
The problem that arose was that when you would like to have release for production, there's not sufficient confidence in it in order to be able to have enough flight test data to say yes, this is really the right time for release for production. So in terms of exercising my stewardship as the acquisition executive, I was reviewing it from the viewpoint, do we have, will we have at the time of production release, adequate flight test data to demonstrate high confidence in it.
The second issue, and I'll cover both of these when we look at the schedule, that the GAO raised was the question of what will be the cost of this program when it in fact is in production, in quantity production. There the issue is can we really demonstrate with this aircraft some of the new modern manufacturing techniques or not? Most of you have read the book about the machine that changed the world where the automobile industry was dramatically streamlined. The Air Force has been going through for some time now a major program on the lean aircraft initiative, which you probably have heard about. The idea being basically to try to modernize aircraft manufacturing and therefore get significant cost reductions on the program.
The difference between the independent cost analyses, done based upon historic information, and the projections that the Air Force and the contractor have been making for the out years, is based upon the ability to introduce these lean manufacturing methods into the aircraft to really move this aircraft into the next generation of manufacturing, and therefore lower the cost quite significantly. The challenge is whether that will be realized or not. So those are the two issues -- the question of having delayed the flight test and therefore how much flight test data was accumulated, and the second one was what would be the likely future cost of this program, and will it realize the projected costs or not.
The plan for the program is shown here on this chart. This is the design development program and the initial operational testing. Part of this design, development and test is a much more extensive simulation and testing associated with literally thousands of hours that have already been accumulated on testing of parts, of wind tunnel testing, of simulation and modeling, things that we're becoming much better at doing than we used to, and which in many cases is a good complement to flight testing. Therefore, in a sense, the objective is we want to end up with less flight testing, more simulation and modeling because of the cost savings, but you still need a sufficient amount of flight testing to gain high confidence.
The plan was to have the first two lots, and this was originally called lot one and this was called lot two, released for long lead parts here in June, for production. This is fiscal years, by the way, let me emphasize that so you understand this chart. But this represents June of '98, this represents December of '98 when one would release the high cost portion of building these two aircraft. That was the piece that the GAO suggested should be moved one year out.
The Air Force had negotiated with the contractor a combined buy of this quantity of two and quantity of six so they could have a sufficient quantity to gain some confidence in the cost. Also, the contractors in dealing with their subcontractors could have some high confidence that the quantities were going to be sufficient to be able to get lower cost. It's a very attractive package that's been negotiated with the contractors, one that we would like to try to hold on to.
In addition, if you were to slip this, then the vendors would be impacted by having a discontinuity in their program, which you wouldn't like to have. There was a big gap between the building of these development contracts and the production.
It's required by law, when you begin the operational testing out in here, that you have production versions of the aircraft. That's a requirement. It's a sensible one. These two, unlike these development models which need not be production representative, these production representative test vehicles would be matched with two more that came out of the initial development program. But it's primarily these that are the production representative birds that would be used to demonstrate the operational testing on the program.
So what we have chosen to do here, and as I say, we're staffing this now so it's not a firm decision, but what seems to me to make a lot of sense at this point would be to maintain the program with these two and then six birds that are planned for demonstrating first of all the production design; and secondly, the process design of this lean production--trying to demonstrate that, in fact, you can get the lower cost that has been projected on this quantity of eight systems.
I should point out that the buildup as shown here for these early birds on this program has all along been a very slow one. It's not like we have done in some of the earlier programs like the F-15, for example. By the time we were at the point of having a few hundred hours of tests, the earlier lots were much larger quantities. This is a very slow buildup (for the F-22).
By the time we get to the point where we have to make the first decision on quantity build for these two, we will have around 200 hours of flight testing. That should give us reasonable confidence at that point with this quantity of two. I will hold a detailed review in November of this year prior to that December release for those quantity of two.
We will obviously have a lot more flight test data a year from then -- December of '99 -- when we will make the decision on the quantity of six, which is a much bigger release, and I'll hold another detailed review in November of '99, by which time we will have a lot more flight test data. Also, we'll have a better indication of costs because we'll have gotten through this earlier phase and subcontracts associated with those first two. So we'll have a lot more confidence.
So in terms of stewardship, if you will, of the acquisition process, I'm comfortable with the idea of these small quantities now -- first for the advanced buy, but particularly this release in December of '98, first for the big dollars associated with these two, and then secondly, the advanced buy, the long lead parts for these six. By this point, as I said, we'll have a couple of hundred hours worth of flight test data and we'll have much more confidence by the time we get out to here where we'll make the first production.
The distinction here is the importance of what is called the first production release for the long lead items, which would now be made in December of '99. There's no reason for concern relative to the program based upon all the data that we've seen to date. In fact it's been quite spectacular in terms of the first flight being very successful. But on the other hand, just in terms of what makes sense relative to having the confidence to have a lot more flight test data here and a lot more here. I'm more comfortable saying that this is the first production long lead, low level production releases.
Of course the full production release is out here in '03 and that's when you get into the real quantity production. That was the reason I made the observation about the slow buildup. This is where the decision for major quantity release is.
I think one other point that I should make is what would be the impact if we followed the GAO recommendation? In other words, if we took and moved them out a year, which is the recommendation they made, and which obviously I therefore had to consider.
There are a couple of problems with that. One, as I mentioned earlier, that would cause a break at the subcontractor level because when they finish the DDT&E birds, there would be a break in the supplier's manufacturing operation. But I think is more important is there would be a very significant cost impact to the overall program. The estimate made by the Air Force was something like $500 million for the EMD phase and about $3.5 billion for the production phase as a result of trying to slip the program out or try to accelerate by not having this gradual buildup that we have.
So the fact is that we have a commitment by the contractors to a very challenging, but we think achievable, program -- we had the independent cost estimating people go through in detail the assumptions that were made to realize the more optimistic cost numbers. In other words, what had to happen in the manufacturing process in order to achieve this lean production.
We think it's very desirable to hold them to the commitments that they made for these two (aircraft). This is a fixed price commitment that's been made. The overall program makes sense because there's no reason not to continue on with it, given the fact that we haven't had any flight problems or anything of that sort.
I think frankly the concern that people had, and I think that may have caused Ken to suggest this conference, is that people might think there's something wrong with the flight test data, or that we're withholding cost information. I think it's important to recognize what is happening. We have had a successful flight test, but we haven't had enough (flight test) hours; and we have a cost estimate which is based upon traditional ways of doing things that was higher than what the contractor and the Air Force had. Based on those two, we wanted to make sure we had sufficient confidence to go ahead with these two lots, these two quantities. That's what we did. We had a detailed review that was reported in the paper last week. We then had another meeting to go over some of the questions that came up at that time relative to the cost, the schedule, and so forth. Now we're in the process of staffing a decision paper that will probably come out pretty close to what I've shown you here.
Q: Can I just ask one thing in lay person's terms? What you're doing is expressing confidence in the program.
Q: And yet you're saying you're delaying for a year essentially a decision on low rate production. You're going to call these first two planes, you're not going to call those production models, is that...
A: No. I should point out they are production models, and it's very important that you recognize that. A, it's important that they be production models because of operational testing; and B, it's important that they be production models because we want to demonstrate the lower cost production process. So they are production models. In fact, the plan would be to phase those back into the fighter wings at some point in the future. But you're right in the statement that you made, that I want to exercise good stewardship in delaying the production decision in order to gain more confidence as a result of more flight testing which doesn't exist today.
Q: Can I get you to elaborate on that? Come December 1998, if you stay on your current path you will make a decision to commit to production of these two planes, correct?
A: In December of '98 I will make a decision to commit to building these two production representative test vehicles.
Q: Based on 200 hours of flight testing?
A: Well, it depends on how much you get. The conservative estimate is that we'll get 180 to 200 hours. What we've actually found is once you get into flight testing you end up getting a lot more. It could be up to 400 hours, but that's the kind of numbers it will be.
Q: So in December of '98 you're going to commit to Lockheed some additional money, correct? For production.
A: That's correct.
Q: So that is a production decision, correct?
A: In fact at that point we'll also release some long lead parts for that first production. This is the time when we are going to make the firm production decision.
Q: How much money are we talking about releasing in December of '98? About?
A: Do you know the numbers?
A2: I don't. We'll have to get that...
Q: What are you delaying for a year? If you're going to make a production decision this December, what are you delaying until December?
A: The way that our acquisition process is set up we have three major milestones. One which was made early on which was the decision to begin major development.
A second major milestone is the time at which you say you have enough confidence based on flight test and other analysis that's been done to say that you're ready to begin low level production. Because of the fact that that is such an important decision, I feel more comfortable making that decision when we have more than just the initial flights that we've had to date. So I was reluctant to say we're really releasing this to production.
On the other hand I have no reason for saying I shouldn't do it. It's just a matter of exercising my responsibilities in terms of that major decision milestone. It would be the same when we get out to this milestone, which is the full rate production. It's another major decision point in the program.
Q: Is that on the two or on the next six, the second milestone you talked about?
A: The first one is when we make a decision to go to production. That's December '99 for the initial production. There's a major milestone for high rate production which takes place in '03.
Q: You said December '98 was your production...
A: No. That's the point I'm trying to make. In December of '98 what we'll be doing is releasing the quantity of two, the one's you've just asked about, for test vehicles.
Q: But they're production models, right?
A: They are production representative models.
Q: And it's production money, correct?
A: It is production money.
Q: ...the production line?
A: They are the first two that will be coming off that line.
Q: Are you going to do anything different than you were going to do with the first two production planes? Anything at all in the way of testing or as part of this?
A: One of the things we are doing is looking at getting more testing in the avionics area with these vehicles. As a result of the test plan to date, we're going back now and looking in detail at what would make the most sense for the test plan with these vehicles in addition to these. But in essence, they are representative. They have to be representative of the production process.
Q: I guess what I don't understand is, the only thing you're doing in this exercise is you're inserting a second DAB review, essentially -- the one you had now, and now you're inserting another one, which essentially... What I don't understand, in your position you could have gone to milestone this year, and if you had to solve problems after the first two years you could have just said...
A: ...call for another meeting...
Q: Then the same thing we just went through with the E/F essentially.
A: I could decide that I was going to release this as though it was a production program, and not hold another review unless there were some issues that came up. I would rather see the results of the testing and see the results of the costs for another year, and then make the decision. I'm delaying the decision a year. I'm not changing the manufacturing. That's a correct interpretation.
Q: How much do you expect to get in that year, and how much confidence can you get in terms of cost?
A: We get quite a bit in both.
A2: General Muellner: The prediction was 180 to 200 originally, and by delaying that a year you'll have another 400-500 hours of tests. So you'll have anywhere from two to three times as much test as a result of that one year delay in the decision process. That's going to be a lot more confidence from the standpoint of understanding the flight envelope.
Q: You had a couple of numbers in terms of what the contractors and the Air Force said it would cost. Are those cost growths accurate?
A: Secretary Gansler: No. The cost growth was the potential cost growth if we delayed the program a year. Those numbers that I said that would be the impact if we decided to accept the GAO recommendation of delaying the program a year, there would be significant cost impacts. By doing what we're talking about here, there is no cost impact. That's one of the main reasons for doing this. There is no...
Q: You stay inside the fixed price contract...
A: We stay inside of the fixed price contract that has been negotiated. It's not any higher risk to us. But we have delayed that management decision point by a year at which point we will then have significantly more flight test data and additional cost visibility.
Q: I'd like to ask you a question about how these kinds of stories are generally perceived by the public. When they hear of any kind of delay, whether it's a decision or any kind of problems in the production of parts that have delayed, it raises the question about whether or not this already expensive airplane is going to end up being more expensive. Can you just address that perception?
A: I don't have any reason for perceiving that it will be more expensive. I'm trying to control the cost by this action. I'm trying to get them to demonstrate, in fact, the lower costs that have been projected for this aircraft by introducing these modern manufacturing approaches.
As I'm sure you've read in lots of the analyses and studies that have been done in the past, delaying programs in fact often significantly raises their cost. So we're not delaying the program, we're delaying the decision point, and we're trying to live to the costs.
Q: This is not a troubled program.
A: This is not a troubled program, and there is no cost growth on this program. What we're trying to do is to gain higher confidence in the product before releasing it to production.
Q: There were four pre-production vehicles in the program originally which were cut last year as part of the most recent restructuring. Was that a mistake?
A: No. I think at the time it wasn't. I think today, in a sense, these two vehicles are representing what you would have had at that time with those extra ones. We're gaining it back, in a sense.
A2: General Muellner: Exactly. Those airplanes would have been doing IOT&E; those airplanes would have gone, after IOT&E, to Nellis just like these airplanes would. So two of them, in reality, would have done about the same thing.
Q: But we did give Boeing... I think it wound up with Boeing, a lot of money for additional hardware-in-the-loop facilities, so we wouldn't need pre-production vehicles.
A2: That's a mistake. The additional hardware-in-the-loop facilities were done to mitigate avionics risk, and not associated with manufacturing or these production vehicles at all. Those were ways of driving down the avionics risk. The Joint Estimating Team said that was probably the higher risk element of the program and we needed to deal with it. That's the reason we added that.
Q: Is it safe to say that the action you're taking here. while not really affecting the program, per se, it's giving you an extra year before you're actually committing for good to building these planes. Is that a good way to explain it?
A: Dr. Gansler: I think as a summary statement, what it does is it allows us another year to make a firm decision for the production low level release; gives us much more flight test data. But it is important to emphasize what has been repeated a number of times. A, there are no problems in the program that we know of at this time. B, that by sticking to this program the costs don't grow. The costs are essentially the ones that we've been living with and stay within the intended total program costs. And that this is basically a decision being made in terms of postponing the decision because of the stewardship question, not because of any issues associated with the performance of the aircraft or the performance of the contractor.
Q: The issue... Are you going to obligate $595 million in December and another $1.5 billion in December of '99? Is that still going to happen under this profile?
A: Those are the planned numbers. We will continue the program in terms of the planned program commitments in terms of the long lead items in June of this year; then here in December of this year for the quantity of two, and the long lead for the next six. That was planned, and it's going to stay.
Q: December of '98 is still the first release of production money, and if Tony's right, it's $595 million. Is that...
Q: That's a big decision, isn't it?
A: It's a big decision, certainly. It's the choice of do you delay it for a year and spend the additional half billion dollars or so on the R&D and additional maybe $3 billion, or do you try to maintain this program because you have no reason for not doing it?
Q: Why are you calling that $595 million production money and you're making the decision on that? That doesn't mean you've made the decision to go ahead with low rate production.
A: With subsequent low rate production. We have not made that decision.
Q: ...the first two production models, right?
A: These will be the production representative test vehicles. (Laughter)
Press: Thank you very much.