[House Hearing, 114 Congress]
[From the U.S. Government Publishing Office]
[H.A.S.C. No. 114-60]
SHORTENING THE DEFENSE ACQUISITION CYCLE
__________
COMMITTEE ON ARMED SERVICES
HOUSE OF REPRESENTATIVES
ONE HUNDRED FOURTEENTH CONGRESS
FIRST SESSION
__________
HEARING HELD
OCTOBER 27, 2015
[GRAPHIC(S) NOT AVAILABLE IN TIFF FORMAT]
______
U.S. GOVERNMENT PUBLISHING OFFICE
97-494 WASHINGTON : 2016
-----------------------------------------------------------------------
For sale by the Superintendent of Documents, U.S. Government Publishing
Office Internet: bookstore.gpo.gov Phone: toll free (866) 512-1800;
DC area (202) 512-1800 Fax: (202) 512-2104 Mail: Stop IDCC,
Washington, DC 20402-0001g
COMMITTEE ON ARMED SERVICES
One Hundred Fourteenth Congress
WILLIAM M. ``MAC'' THORNBERRY, Texas, Chairman
WALTER B. JONES, North Carolina ADAM SMITH, Washington
J. RANDY FORBES, Virginia LORETTA SANCHEZ, California
JEFF MILLER, Florida ROBERT A. BRADY, Pennsylvania
JOE WILSON, South Carolina SUSAN A. DAVIS, California
FRANK A. LoBIONDO, New Jersey JAMES R. LANGEVIN, Rhode Island
ROB BISHOP, Utah RICK LARSEN, Washington
MICHAEL R. TURNER, Ohio JIM COOPER, Tennessee
JOHN KLINE, Minnesota MADELEINE Z. BORDALLO, Guam
MIKE ROGERS, Alabama JOE COURTNEY, Connecticut
TRENT FRANKS, Arizona NIKI TSONGAS, Massachusetts
BILL SHUSTER, Pennsylvania JOHN GARAMENDI, California
K. MICHAEL CONAWAY, Texas HENRY C. ``HANK'' JOHNSON, Jr.,
DOUG LAMBORN, Colorado Georgia
ROBERT J. WITTMAN, Virginia JACKIE SPEIER, California
DUNCAN HUNTER, California JOAQUIN CASTRO, Texas
JOHN FLEMING, Louisiana TAMMY DUCKWORTH, Illinois
MIKE COFFMAN, Colorado SCOTT H. PETERS, California
CHRISTOPHER P. GIBSON, New York MARC A. VEASEY, Texas
VICKY HARTZLER, Missouri TULSI GABBARD, Hawaii
JOSEPH J. HECK, Nevada TIMOTHY J. WALZ, Minnesota
AUSTIN SCOTT, Georgia BETO O'ROURKE, Texas
MO BROOKS, Alabama DONALD NORCROSS, New Jersey
RICHARD B. NUGENT, Florida RUBEN GALLEGO, Arizona
PAUL COOK, California MARK TAKAI, Hawaii
JIM BRIDENSTINE, Oklahoma GWEN GRAHAM, Florida
BRAD R. WENSTRUP, Ohio BRAD ASHFORD, Nebraska
JACKIE WALORSKI, Indiana SETH MOULTON, Massachusetts
BRADLEY BYRNE, Alabama PETE AGUILAR, California
SAM GRAVES, Missouri
RYAN K. ZINKE, Montana
ELISE M. STEFANIK, New York
MARTHA McSALLY, Arizona
STEPHEN KNIGHT, California
THOMAS MacARTHUR, New Jersey
STEVE RUSSELL, Oklahoma
Robert L. Simmons II, Staff Director
Lynn Williams, Professional Staff Member
Spencer Johnson, Counsel
Abigail Gage, Clerk
C O N T E N T S
----------
Page
STATEMENTS PRESENTED BY MEMBERS OF CONGRESS
Davis, Hon. Susan A., a Representative from California, Committee
on Armed Services.............................................. 2
Thornberry, Hon. William M. ``Mac,'' a Representative from Texas,
Chairman, Committee on Armed Services.......................... 1
WITNESSES
Chu, David S.C., President, Institute for Defense Analyses....... 4
Francis, Paul L., Managing Director, Acquisition and Sourcing
Management, U.S. Government Accountability Office.............. 7
Hunter, Andrew, Director, Defense-Industrial Initiatives Group,
Center for Strategic and International Studies................. 2
Pasqua, Joe, Member, Business Executives for National Security... 6
APPENDIX
Prepared Statements:
Chu, David S.C............................................... 60
Francis, Paul L.............................................. 79
Hunter, Andrew............................................... 47
Pasqua, Joe.................................................. 71
Smith, Hon. Adam, a Representative from Washington, Ranking
Member, Committee on Armed Services........................ 45
Documents Submitted for the Record:
``Running the Pentagon Right: How to Get the Troops What They
Need,'' by Ashton B. Carter, Foreign Affairs, December 6,
2013....................................................... 95
Enclosure 13, Rapid Fielding of Capabilities, from DODI
5000.02, January 7, 2015................................... 104
Witness Responses to Questions Asked During the Hearing:
Ms. Speier................................................... 117
Questions Submitted by Members Post Hearing:
Mr. Thornberry............................................... 121
SHORTENING THE DEFENSE ACQUISITION CYCLE
----------
House of Representatives,
Committee on Armed Services,
Washington, DC, Tuesday, October 27, 2015.
The committee met, pursuant to call, at 10:00 a.m., in room
2118, Rayburn House Office Building, Hon. William M. ``Mac''
Thornberry (chairman of the committee) presiding.
OPENING STATEMENT OF HON. WILLIAM M. ``MAC'' THORNBERRY, A
REPRESENTATIVE FROM TEXAS, CHAIRMAN, COMMITTEE ON ARMED
SERVICES
The Chairman. Committee will come to order. Appreciate our
members, witnesses, and guests for joining us today on this
hearing related to acquisition reform. As you can tell, there
are party conferences still going on, and so members will be
coming in as those conclude.
Let me just say that a major priority for this committee
has been, and will continue to be, improving our acquisition
system. Partly, it is to help get more value for the taxpayer
dollars. In my mind even more important, is to have a more
agile system that can better respond to the myriad of complex
national security challenges facing our country.
And it is at least my belief that unless we improve our
acquisition system, we cannot keep up with the many challenges
that we face. At the same time, while we are trying to improve
our acquisition system, the acquisition process has to work
every day. You have got to get that rifle into the hands of
that soldier in Afghanistan and do all the other things that
are required of the system.
And so, I believe we can't have a 2,000-page bill that
fixes acquisition. We have to take it a step at a time. I think
we made some good progress, good first steps in the fiscal year
2016 National Defense Authorization Act, working on some of the
basics when it becomes law.
But there are more steps to go, and that is the reason for
today's hearing, to benefit from the experience and wisdom of
our distinguished witnesses on next steps, and direction for
the acquisition reform efforts undertaken by this committee and
the Senate committee, working with the Pentagon. One thing
everybody agrees on is that we have got to do better, and so
largely this has been a cooperative effort.
Let me yield to the distinguished gentlelady from
California for any comments she would like to make on behalf of
the ranking member.
STATEMENT OF HON. SUSAN A. DAVIS, A REPRESENTATIVE FROM
CALIFORNIA, COMMITTEE ON ARMED SERVICES
Mrs. Davis. Thank you, Mr. Chairman. I am delighted that
you are all here today. We appreciate it very much.
On behalf of the ranking chair, Mr. Smith, I wanted to
submit his statement for the record and also acknowledge how
difficult it is to find that appropriate balance between the
acquisition cycle time, and risk. We know that needs to be
done. And also, how do we nurture innovation and developmental
testing within the acquisition cycle. That is also a big
concern and something that he notes in this particular
statement.
Again, thank you very much for being here. We look forward
to your testimony.
The Chairman. Thank the gentlelady.
[The prepared statement of Mr. Smith can be found in the
Appendix on page 45.]
The Chairman. We will now turn to our witnesses, Mr. Andrew
Hunter, director of Defense-Industrial Initiatives Group at
Center for Strategic and International Studies [CSIS]; Dr.
David Chu, President of the Institute for Defense Analyses
[IDA]; Mr. Joe Pasqua, member of the Business Executives for
National Security [BENS]; and Mr. Paul Francis, managing
director for acquisition and source management from the
Government Accountability Office [GAO].
Without objection, your full written statements will be
made part of the record, and I would ask each of you to
summarize them at this point before we go to questions.
Mr. Hunter, I guess we are starting with you.
STATEMENT OF ANDREW HUNTER, DIRECTOR, DEFENSE-INDUSTRIAL
INITIATIVES GROUP, CENTER FOR STRATEGIC AND INTERNATIONAL
STUDIES
Mr. Hunter. Well, thank you, Mr. Chairman. And it is really
a pleasure to be back here at the committee where I spent so
much time learning from members like yourself, some of whom are
still here and some of whom have departed the Congress. But it
is a pleasure to be back.
And I commend the committee for its focus on acquisition
reform, which is, I know, a focus of longstanding, but remains
a very important focus, and obviously one that brings you into
alignment, as you mentioned, with the leadership of the Senate
Armed Services Committee and with the leadership of the
Department of Defense.
And I do believe there is a golden opportunity here with
this meeting of the minds or alignment of focus to make some
real progress. And certainly at CSIS this has been a focus for
us as well, so it is something that we share.
The focus today is on ``faster.'' I want to briefly mention
the fact that acquisition is about balancing priorities. And so
the old saw, ``Faster, better, cheaper: Pick any two,'' is
something that I just want to start and mention that in picking
``faster'' you have to be willing to sacrifice at least one of
the other two. And when I say ``sacrifice,'' I mean deemphasize
or make a lower priority. And so if you are going ``faster,''
then either ``better'' or ``cheaper'' has to sort of be willing
to give a bit in order to achieve a significantly faster
outcome.
Now, of course, when I say ``sacrifice,'' when I mean
``better,'' better meaning not necessarily the highest end of
capabilities. If you have a really old system, the new system
you are buying is almost certainly going to be better than the
one you are replacing, but it may not be the state of the art
of the most latest technology. And these priorities shift over
time.
And in the Cold War, in most cases, ``better'' was often
the priority. I use in my written testimony the example of the
B-2, which was innovative in almost every way as it was built
and conceived and constructed, and that meant that it was
expensive. And there was a major schedule delay in that program
particularly because they changed the requirements in the
middle of the development. And that was a choice that was made
because ``better'' was what mattered then.
In the most recent time period, with the wars in Iraq and
Afghanistan, ``faster'' has certainly been a priority. In my
time at the Department of Defense, part of which was as
director of the Joint Rapid Acquisition Cell, was really all
about moving faster. And the Department achieved quite a bit of
success in that, and I will get into that, because I think that
is something that there are lessons learned that we need to
take away from that.
But I do want to mention that in the time that we are in
today, ``faster'' is not the only priority. We have an erosion
of U.S. technical superiority that has been taking place over a
number of years for a number of reasons, and we explored what
some of those reasons are, in a CSIS report released over the
summer.
And that is a case where we do need to be fielding at least
some systems that are in the ``better'' category, where
``better'' is a priority so that we can maintain a
technological advantage, which is part of our strategy. And
also with the budget crisis that is currently being dealt with
in the Congress, and maybe there is hopefully some progress
being made there, ``cheaper'' has to be a priority for some
systems. So your system has to be able to focus on different
priorities for different systems at the same time.
Within my time as director of the Joint Rapid Acquisition
Cell, we tried to capture the lessons learned from rapid
acquisition. This was something that really spun up at the
Department of Defense in the 2005 timeframe. And I came in, in
2013 as director of the Joint Rapid Acquisition Cell when we
were trying to capture the lessons learned and trying to
institutionalize those lessons. And Secretary Gates in his
testimony last week made reference to the desire to
institutionalize those lessons.
And I would draw your attention to the article in Foreign
Affairs Magazine that Dr. Carter published in 2013 which goes
into the lessons that he took away and how he tried to
institutionalize those. And I would ask, if you are willing,
that that might be made part of the record for this hearing,
that article.
[The information referred to can be found in the Appendix
on page 95.]
Mr. Hunter. The keys that we identified at that time, the
first is flexible funding. By and large, when you are working
through the Department's regular budget process, it takes 2
years to get money to start. So that is an immediate 2-year
delay in the system. Now, there are certain ways around that,
but they are cumbersome and they are difficult, and they make
it hard to move fast.
And the Congress was very generous during the war period of
establishing flexible funds like the MRAP [Mine-Resistant
Ambush Protected] Transfer Funds and the Joint IED [Improvised
Explosive Device] Defeat Fund, but those funds are really going
away. And so exploring how to extend flexible financial support
for programs that need to move fast is definitely an area of
focus.
Second big area was getting the senior leadership of the
Department involved and shortening the lines of authority, and
that was really what was called the Warfighter Senior
Integration Group that the Department did during the war years,
and I think that is an excellent model for programs that matter
to move fast. And that model is being somewhat echoed in the
Long Range Strike Bomber program the Air Force is about to
initiate with the way that they manage their Rapid Capabilities
Office.
And then the third priority is basically continuous
communication between the acquisition community and the
operational community about requirements, about testing, about
what is acceptable, and about what the art of the possible is
with technology, and whether that is acceptable to the
warfighter. Those three lessons are very much applicable to
rapid acquisition, but they are applicable more broadly.
And the last thing I want to leave you with is the idea of
adaptable systems. If we are always trying to figure everything
out for the next 30 years today and plan that all in, that is a
real challenge. That is slow. That is just an inherently slow
process.
And so focusing on adaptable systems that can evolve over
time where you don't have to have the full answer right when
you start is a good way. And I would use the Predator system as
an example of how that has actually happened in practice. And
that system has evolved in a revolutionary way over time, and
that is, I think, an example to say.
Thank you very much, Mr. Chairman.
[The prepared statement of Mr. Hunter can be found in the
Appendix on page 47.]
The Chairman. Thank you.
Dr. Chu, welcome back to the committee.
STATEMENT OF DAVID S.C. CHU, PRESIDENT, INSTITUTE FOR DEFENSE
ANALYSES
Dr. Chu. Thank you, Mr. Chairman. I appreciate the
opportunity to be part of the panel this morning.
Let me try to summarize my written statement under three
headings: First, what do we know about cycle time defined as
the time from the start of a program until initial operational
capability is achieved; second, to the extent there are issues
with cycle time, what are they; and third, what can we do about
those issues.
Looking at the Selected Acquisition Report data of the last
25, 30 years, I do not see any trend in the cycle time. Cycle
time has been relatively stable, 8 to 9 years over that period.
Our impression is very different, perhaps because some very
large programs have taken a long time: F-22, V-22, and F-35,
which, of course, is not completely finished with its
development program.
To the extent that there is dissatisfaction with cycle
time, I think a good deal of the source lies in the definition
of the program at the start, what people like to call the
requirements process, a term I actually would urge we drop.
Because, in fact, we are always picking a point in the space of
trades among the various features of the system concerned.
And our real interest ought to be how well the system
performs against the mission needs for those in the field who
are going to get it. Too often, from a technical perspective,
looking back at history, we pick a point in the trade space
that is too tough to achieve from a technological perspective
within the timeframe that we might desire.
And that tendency is exacerbated, I would argue, by the
incentives facing those responsible for the system, starting
with the program manager. We reward program managers for
getting programs to production, not for helping the system make
a good decision, which in some cases, is to admit we have made
a mistake and the program ought to end.
The services, likewise eager to have as much content within
the fiscal guidance as they can possibly achieve, tend to plan
for more than can actually be financed. And the companies look
to production for the source of their return on capital have
every incentive to be optimistic about development time and
development needs.
If those are the sources of our dissatisfaction, what can
we do about them? First and foremost, I think at the start we
ought to take what some of my colleagues have called a physics-
based approach to setting the technical parameters. What does
the trade space look like? What point within that trade space
do we want to select?
Second, as Mr. Hunter suggested, as one of my colleagues
has phrased a bit edgily, we should prepare to be wrong. We
should build systems knowing--especially the major platforms--
knowing that we are likely to want to change them to aim at
block upgrades across their lifetime, that means allowing for
extra space, weight, power, et cetera, in the original design.
To be sure that we have picked the parameters thoughtfully,
I think greater emphasis on development testing is essential.
The Department System Acquisition Reform Act of 2009 called for
that, and the fiscal year 2016 National Defense Authorization
Act strengthens those provisions.
And finally, I think we ought to rethink the incentives
that face the program managers and the services as well as the
companies that produce the articles to emphasize, much as Intel
does, as I understand it, that really the rewards are to go to
those who give good advice, and sometimes that advice is the
program is not meritorious, that not every program started
ought to go to a finish.
Let me offer three observations very briefly in conclusion.
First, I think the emphasis I would urge is less on whether or
not we shorten the cycle time and more on understanding how do
we pick the best cycle time for the need that we face. In some
cases, we want an article urgently, we are willing to give up
certain elements of performance in order to get that, or
certain elements of long life that we would otherwise seek.
MRAP is an excellent case in point. MRAP was achieved fairly
rapidly, but was an article we decided not to retain, and we
discarded approximately $40 billion worth of equipment. Some
additional bias for flexibility will be helpful in shortening
cycle times for those articles we want quickly.
Second, I think it is essential to keep our focus on
mission performance as the ultimate standard, not on the
technical parameters per se. It is the mission needs that are
crucial. That includes, of course, deployment deadlines when
those are significant.
And finally, as Secretary Gates' testimony, I would argue,
last week at the Senate Armed Services Committee contended,
perhaps the most important ingredient in success is the human
capital, the quality of the people managing the system and the
technical staff that support them, an issue, I think, that the
National Defense Authorization Act for this fiscal year
recognizes.
Thank you, Mr. Chairman.
[The prepared statement of Dr. Chu can be found in the
Appendix on page 60.]
The Chairman. Thank you.
Mr. Pasqua.
STATEMENT OF JOE PASQUA, MEMBER, BUSINESS EXECUTIVES FOR
NATIONAL SECURITY
Mr. Pasqua. Chairman Thornberry, members of the committee,
my name is Joe Pasqua, and I am honored to be here today as a
private citizen to address you.
Having been asked for ways to address shortening of the
defense acquisition cycle, my statement today will focus on how
the private sector has addressed similar challenges and
increased their ability to adopt innovation quickly.
My testimony is based on over three decades in the
information technology, the IT industry, and also as a member
of BENS, Business Executives for National Security, which is a
nonprofit, nonpartisan organization that for over 30 years has
been a conduit through which private sector leaders can help
build a more secure America. Although the opinions I am going
to express today are largely reflective of BENS's perspectives,
the views I express are my own.
I would first like to commend the committee's efforts at
acquisition reform. I agree with Chairman Thornberry's approach
to make incremental and achievable changes as a path to
comprehensive reform. Because of the central role of IT in
virtually all modern systems, the ability to efficiently
specify, acquire, and adopt IT innovation has become a key
success factor. Smaller, more agile companies are often the
best sources of innovation; however, they can often be the most
difficult to identify and engage with for large organizations.
In the past, barriers for both the Department and these
small companies have impeded building effective relationships.
Traditionally smaller companies haven't viewed DOD [Department
of Defense] as a viable customer because they lack the
specialized knowledge and the time that is required for
operating in this space. It is not that they don't want to
engage with DOD; it is just that it is too high a risk for
these still small businesses.
So the question becomes, how are private sector companies
addressing similar challenges? Over the last 5 to 7 years,
there has been a fundamental change in the way that they
specify and acquire IT. The rapid pace of innovation has made
long, expensive requirements processes untenable. As a result,
we are seeing less of what I referred to as ``big bang''
acquisitions; instead, companies are starting small, conducting
iterative evaluations in real time, and adjusting as needed.
Advances in cloud computing, scale-out architectures, and other
technologies have enabled companies to test concepts quickly
and purchase IT hardware as they need it rather than buying
everything upfront.
This has been a challenge, quite frankly, for large
organizations with high inertia and low risk thresholds. But
even with these larger organizations, we are seeing that they
are becoming more agile as a way to keep pace in a competitive
marketplace. This shift has lessened the bias towards large,
incumbent vendors and has given innovative new players a better
opportunity to compete.
This new approach also helps to remove risk by keeping the
initial investments small. Traditional requirements processes
attempt to mitigate risk by conducting long-term, expensive
studies to ensure all options, every conceivable outcome can be
reviewed in advance of a decision. In contrast, an agile
approach allows companies to start small and scale up as
appropriate, thereby reducing the need for protracted
requirements processes.
In fact, a traditional process has a different sort of
risk, the risk that by the time a long acquisition process is
complete, the solution that is chosen will no longer be
appropriate. Nowhere is this more true in cyberspace where the
threat landscape is changing on a continuous basis. In such a
dynamic space, the requirements process needs to account for an
organization's current needs and be able to adapt to the
inevitable changes that will come. This is one reason why open
architecture is so important. It provides increased
interoperability, modularity, and the ability to incorporate
new technologies without overhauling an entire system.
In summary, I believe that these practices and
understanding and implementing these approaches would help the
Department to become more agile and responsive to innovation,
allow a slightly different, yet still very good risk mitigation
strategy, and encourage participation from a wider segment of
industry.
Thank you for the opportunity to speak today, and I am
prepared to answer any questions you may have.
[The prepared statement of Mr. Pasqua can be found in the
Appendix on page 71.]
The Chairman. Thank you.
Mr. Francis.
STATEMENT OF PAUL L. FRANCIS, MANAGING DIRECTOR, ACQUISITION
AND SOURCING MANAGEMENT, U.S. GOVERNMENT ACCOUNTABILITY OFFICE
Mr. Francis. Good morning, Mr. Chairman, Mrs. Davis,
members of the committee. I appreciate the opportunity to talk
about weapons system acquisition this morning.
I think we know the overall problems being, weapons systems
cost more and they take longer than expected. They perform well
but generally not quite as well as we thought. I believe some
of the top-level consequences are understood, that is the
warfighter is somewhat underserved by systems that come late
and buying power is reduced.
There are other consequences maybe that are less well
understood; that is, when we pay more for a weapons system than
we thought that we would in the beginning, we are making
tradeoffs. We are deciding not to do other things because we
are putting more money into this one system. And those
opportunity costs I don't think are explicit or the tradeoffs
are looked at, so we are not quite sure what we are giving up
to put more money into system A.
The other thing is, weapons systems typically take more to
operate and support than we think. They are a little less
reliable. Again, those costs are kind of hidden once the
acquisition is done. So, some consequences not so clear.
Our position is that the key to getting better acquisition
outcomes, whether they are shorter cycle times, or as Dr. Chu
mentioned, the right cycle times, is a better business case at
milestone B. And I will talk about a business case in two
parts: One is, what happens before milestone B. And that is
when you are sending requirements and you really need your
requirements to be--they need to be clear, flexible, but well
informed by a couple of things: One is technology knowledge.
How much technology is available to meet the requirement; and
your engineering expertise. Do you understand the implications
of the requirement for the design?
So if you come to a milestone B and you are asking for
technologies that aren't mature yet, or you don't quite
understand the implications for the design, you are in trouble.
If you do come to milestone B with a pretty well-informed,
reasonable set of requirements then you are kind of ready for
the second half. And the second half of that is, what is your
game plan going forward.
And we would say the second part of that business case then
is a knowledge-based acquisition strategy that lays out a
logical path for getting the design stable, building
prototypes, testing, maturing the design, maturing production
processes, and laying that out with the schedule and resources
that allow that to be done nonconcurrently.
So, you ask yourself, well, why aren't we getting these
kind of business cases routinely? Which is the David Packard
question. We all know what needs to be done; the question is,
why don't we do it? And I would say what I just described is a
sound business case, but a sound business case isn't the same
as a successful business case. And a successful business case
is one that wins money.
And I still think predominantly in the Department, a
successful business case is one that overstates or overpromises
performance and understates cost and understates schedule. That
is what still wins money today. And I would say the reason for
that is there is still strong incentives, which we refer to as
the acquisition culture in the Department, that put pressure on
these kind of business cases.
And I will give you a couple examples of what is kind of
under the hood. First is the competition for funds in the
Pentagon is pretty intense to start a new program, so that does
create incentives to overpromise performance and understate the
investment cost. Also, weapons systems are highly symbolic.
They are more than just a piece of equipment at the right
price. They involve policies, roles and missions, careers,
jobs, budget shares, so they carry a lot of weight.
If you look at the private sector, when the private sector
does a product development, it is an expense. They are spending
their own money to finance the development and they don't make
any money until they get into production. So that creates real
incentives to get the business case right, because if they are
late, the customer walks. If you are Ford and you build a Ford
Taurus that is 5 years late, it has a $50,000 sticker price and
it gets bad gas mileage, your customer walks, and the
investment is lost.
And the Department of Defense, when you get a program
started, it is a revenue stream. It is not an expense. So you
get a bigger budget share. And those incentives then are quite
different. And at the end, the customer isn't going to walk. So
if it costs too much, it takes longer, it underperforms, the
customer is still going to buy.
So in the private sector, the point of sale is after
development when you are in production. In the Department of
Defense, the point of sale is at milestone B. In fact, I would
say it is before milestone B when you first approve funding. So
it is a completely different psychology.
That is why things like--practices like cost estimating,
everyone wants--or you would say policy says we should have
good cost estimates. We all know how to do a good cost
estimate. But they don't really help your business case. They
are pretty inconvenient if they are high. Same is true for a
fly-before-buy in testing. You would want early test results to
see how good the system is, but they could be inconvenient as
well.
So I think the real kicker is, to the extent business cases
like this win funding approval they are sanctioned, and those
principles then become what policy is, not what is in best
practices or DOD policy.
So what to do, I would just say let's start thinking about
the acquisition process as not something that is broken but
something that is held in equilibrium by a set of incentives
that are stronger than best practices. You know, moving forward
there is a number of things we can do, we will probably talk
about that more this morning, but it is going to take joint
action on the part of the DOD and Congress.
And I will just list a few things. One is we need to
separate technology development from product development; we
need to take risks in the right places, which I would say is
early in programs; and if we have to take a risk on a program
after milestone B, let's declare them and pay for them. Let's
take the risk together and be honest about them.
We have to do something about better aligning funding
decisions with program decisions, because today you are having
to make a funding decision 18 months in advance of a program
decision. So once you put money on the table you can't take it
off. We really need, as Dr. Chu mentioned, a really good
investment in program managers and systems engineering staff.
And then finally, I would say my hope, my appeal is to
Congress to be the game changer in acquisition reform and that
will be manifested by what you do in funding programs. So I
would say for programs that don't measure up to good business
cases, say no. I think a couple of good no's in the process
from the Congress is going to send the right example as to what
you expect.
So that is my hope. I am looking for you to be the game
changers. Thank you, Mr. Chairman.
[The prepared statement of Mr. Francis can be found in the
Appendix on page 79.]
The Chairman. Thank you.
I just want to ask each of you to comment, because each of
you have touched on this in the comments that you have made so
far. And one of the concerns that I have heard over and over
is, especially for complex systems, we invent as we are in
production. Dr. Chu said we reward programs for getting into
production. Mr. Francis was just talking about this. So the
incentive of the current system is to get that program past the
milestone B, that is where you get the dedicated line of
funding, and the incentive is to do that even if the technology
has not been developed that you are going to rely upon.
And so, part of what happens is you are inventing as you
are producing, and that results in delays, cost overruns, and
so forth. So the suggestion has been made to me that if you
separate technology development from production and you don't
take anything to production until the technology is established
and proven, that maybe you could improve that situation with
the adaptability that you all talked about so that as
improvements in technology are developed, then you can plug it
in.
So Mr. Hunter, what is your reaction to that? I mean, part
of what we are trying to do is get below the symptoms, the
surface here, and dig down into deeper root causes that have
caused people concern. Is this a root cause, and is that
something that together we should explore with the Pentagon?
Mr. Hunter. Well, I would say it can be a root cause and
that does happen. And the example that jumped into my mind as
you were laying that out is the example of the F-35 helmet,
which I would say was not mature technology in the early 2000s
when the program went through milestone B and the investment
decision was made and the decision to take that approach was
made, but which is now actually working.
And so some 15 years later it is there but it probably
wasn't there when we made the decision. So it does happen. It
is a cause. It is not the only root cause but it is a cause.
One note I would make about that and something they did there
that actually was a good idea, although I think it was belated,
is they had an off ramp and they had a second helmet that they
could have gone with if the original helmet didn't work. And
that, I think, is a good practice.
There are times when you want to reach a little bit, as I
mentioned when ``better'' is the priority. It is not clear that
on F-35 that was really the intent, but where it is a priority
you may want to reach. But what you can do is have off ramps,
so that if you are not able to invent the thing you were trying
to invent, you still have a workable system and it still meets
your threshold requirements.
And I also think, you know, as I mentioned this idea of
adaptable systems to where you may be trying to invent
something, but again, you don't put it in the baseline design.
It is in a later block. It can be a way when you are trying to
reach for new and innovative technology.
The Chairman. Dr. Chu.
Dr. Chu. I would add two thoughts to the general idea that
you advanced. One, reinforcing, which is there in various
congressional direction the last several years, and that is a
greater emphasis on development testing. We don't do enough
testing early on of the technology ideas to be sure that they
are going to pay off in the way we think.
Second, I think, again, back to deemphasizing the word
``requirements,'' too often we pick a technological point and
we follow these attributes forgetting that in the end what
counts is does it add to mission success or not. And there are
a number of systems where we have picked points that actually
don't have a lot to do with mission success but we keep
pursuing them in the systems development even though they are
not going to have a high payoff, and that often is the cause of
serious difficulty.
The Chairman. Okay. Mr. Pasqua, I want to come back to you
in a second with the private sector.
Mr. Francis, what do you think?
Mr. Francis. I think it is definitely one of the big root
causes. I think part of the solution lies in the fact that we
have to enable those technologies, so we do have to take those
risks to make those gains. So we are not arguing against that,
but the burden needs to be borne more heavily in the science
and technology community. And we typically aren't funding it to
carry technologies that far.
And the mechanisms we have to transition technologies to
programs aren't very good. So programs are a better place, if
you will, to fund the programs, which is not what we want. I
also think when programs are doing their analysis of
alternatives, there is incentives to advertise very high
performance, which means you are counting on technologies that
haven't been invented yet.
So I think we can go forward, bring technologies to higher
level before milestone B. If we still have to take risks, let's
take the risks and pay for them upfront. Or, as I think Mr.
Hunter was suggesting, go forward with the design that is
flexible enough that you can bring in improvements in
technologies during the course of development.
The Chairman. Okay. Now, Mr. Pasqua, how does all this
comport with your experience in the private sector? And is
this, kind of what we have just been talking about, a path
towards a more agile system, in your opinion?
Mr. Pasqua. I think it translates very well to the private
sector, particularly with technology. It is sort of well
understood that the further on you get into a technology
development cycle, the costs of changes and finding and fixing
problems increases close to exponentially.
So you want to make sure that you are doing as much cycle
work as you can upfront to get your technology in place and in
a mode where it is operable, it can be adapted and modularized,
but you don't want to be making changes during the production
cycle, you don't want to be redoing the architecture during the
production cycle. That is the absolutely most expensive time to
deal with those types of issues.
So the idea of being more agile upfront, being able to test
systems before they are in production, before you get into the
most expensive phase for changes, is, I would say, an industry
best practice.
The Chairman. Thank you. Thank you, all.
Mrs. Davis.
Mrs. Davis. Thank you. Thank you, Mr. Chairman.
And I appreciate all of your input here.
I think one of my questions, and perhaps both in business
and in the Department, the extent to which we rely on people
who have done those things before, and with the idea that they
will be able to move forward, you know, do you see in business
particularly putting folks on the project who maybe haven't
been there before? I mean, is there something to be said for we
really do need in management to bring in--talk about different
eyes on the issue. Do you see that happening in the business,
that you are able to do that more than perhaps in the public
sector?
Mr. Pasqua. Well, I can't compare to the public sector
because I can't speak with authority on how often that happens,
but it is definitely the case that having a fresh perspective
is always a good thing. But having people who have--are
experienced with the process and know how the process works and
can operate efficiently in it, I think, is very important.
I think what we are seeing in industry is sort of a bridge
being built between the people and the way processes had
operated for many, many years, and a transition to the way they
are operating now. And I think part of the way that is
happening, particularly in large organizations, is they are
looking at smaller organizations and wondering how these
smaller organizations are so much more effective than they are,
and trying to understand which sorts of processes can be
adapted from those sort of more agile companies into a larger
organization.
It is not easy, frankly. The things that work in a smaller
organization often don't translate directly to a larger
organization. But as Dr. Chu was saying, I think that one of
the critical things is always keeping in mind what the end goal
is.
So as organizations are focusing on not what's written in a
document somewhere about, you know, specific sets of
requirements but actually what they are trying to achieve in
the marketplace or for their customers or for their patients or
clients and being able to adjust based on that north star of
what the actual goals of the projects are rather than
specifically the detailed requirements is one thing that I
think is changing industry in a positive way.
Mrs. Davis. Uh-huh, yeah. I don't know, Dr. Chu or even Mr.
Francis, does sometimes just the culture get in the way of
that?
Dr. Chu. I think there is an issue there with the human
capital, and that is that as the number of new systems has
declined over the last several decades in at least several
platform areas, private aircraft being a principal example, it
is less the case as was earlier true that the design engineers
have prior experience with that design problem.
And so we move from a situation where United States and
let's say the 1950s, 1960s science engineers have frequent
opportunities to try out new design ideas and experience with
the ups and downs of that process to--they may do one or two
designs in an entire career.
And so that base of hard-won lessons from things that
didn't go so well is not as frequently there, and I think that
is one of the issues out there. I think that does lead to a
different kind of technology separation production, which is
perhaps more emphasis on prototypes and prototypes for their
own sake, to try out technologies and to give the design teams
more experience with the tough issue of how you actually make
these trades work.
Mrs. Davis. Thank you.
Mr. Francis, is the reason that we tend to build on
immature technologies that there hasn't been this sort of
change in terms of looking at something, I guess, different
from what they have done in the past? How would you solve that,
I guess?
Mr. Francis. Well, I think that is part of the problem. I
think the issue with program managers is a real one. I think we
are putting really outstanding people in those positions, but
we are often handing them an impossible situation. So we hand
them a business case that no one could execute.
We are not really giving them the training and glide path
to put them in a position and really have the business acumen
to do business with their private sector counterparts. And then
we are not giving them a really good career path. So we don't
put program managers in a good position to succeed, so that is
a remedy that we need.
The other part, touching on your technology, is we are
still short on systems engineers in the Department, and we
particularly need that expertise before a milestone B decision
so that you can work with those requirements and understand the
preliminary design. So I think the work doesn't get done early,
it falls on the shoulders of the PM [program manager], and the
PM is not well equipped to handle it.
Mrs. Davis. Thank you.
Thank you, Mr. Chairman.
The Chairman. Mr. Forbes.
Mr. Forbes. Mr. Chairman, first of all, I want to thank you
for your work in this area. You have done a lot to move this
forward and this hearing is one of those things.
And our committee is always--we spend some time in the
weeds looking at the specificity of what we need to do and then
we move up and get kind of an aerial view. There are times we
need to do the aerial because it shows us the trajectory and
the curve lines that we have.
When Eisenhower left office in 1961, he warned of the
influence of a permanent armaments industry of vast
proportions. But in the years of the Cold War, we have seen the
size of our industrial base shrink along with defense spending.
1961 defense contractors comprised 15 of the top 100 companies
in America and made 30 percent of those top 100 companies'
total revenue. Today, there are only 4 defense contractors in
the top 100 and they make 4 percent of the total revenue.
Now, I know that is in part due to mergers and acquisitions
and also tremendous growth in the commercial sector. But
unfortunately for our national defense many of the commercial
companies are hesitant to work with the DOD due to low profit
margins, huge regulatory burdens, and demands that they turn
over intellectual property.
Could you guys give us your opinion of the curve lines that
you see that worry you about the industrial base and the
acquisition process in terms of the health of our industrial
base and its ability to meet the needs of our military, and
also the flexibility that we might have. You know, we all talk
about often in World War II how we could shift our
manufacturing and produce other things. What worries you today
about those curve lines, and is there anything we can do as a
Congress to impact the curve lines?
Mr. Hunter, do you mind giving us your thoughts.
Mr. Hunter. Yes. I would say if you step back and take the
total aerial view, on the whole I would say the industrial base
is actually still pretty healthy today. Where I get concerned
is where that defense industrial base becomes very much
divorced from the commercial side. And an example I would give
is shipbuilding, right. The commercial shipbuilding industry in
this country is gone essentially. There is a little sliver of
it left out in California, but by and large it is gone.
And so all of the shipbuilding, all of the expense, all of
the overhead of what is an expensive industry is carried by the
Department of Defense. And that is why the Navy works so
assiduously to try and take care of that industrial base which
is certainly an excellent thing that the Navy does. On the
aviation side, it has been much more tightly integrated both in
terms of airframes and engines.
Now, that may also be starting to separate a bit, and so I
do have some concerns that if the aviation side of the
industrial base goes in the direction that shipbuilding has
gone and we get this separation between the commercial aviation
industrial base and the defense aviation industrial base, that
could have real consequences. The decision by the United
Technologies to sell Sikorsky does raise some concerns in that
area. And that was not because the business was going away; it
was more a decision about profitability. But that is a concern.
Mr. Forbes. Anybody else have a thought?
Dr. Chu. Sir, what I would worry about most is the lack of
competition, the consolidation that you mentioned. You look at
history of fighter aircraft since World War II in the early
decades, I think the historians would argue the interesting
innovations came from the firms that lost the last competition
because they realized if they didn't come up with a new idea
they would not be around much longer. That is no longer a
threat to the major suppliers.
Coupled with, as you hinted, and as Mr. Pasqua's testimony
underscored, a set of Federal procurement rules make it very
difficult for a truly commercial firm to do business with a
fellow company. So what you have is basically a firm that
specialized in defense procurement or subsets of firms, such as
the Boeing division between military and civil aviation that
specialize in defense procurement. And again that limits the
degree of competition, most importantly competition about new
ideas.
Mr. Forbes. Okay. Any other thoughts?
Mr. Francis. Mr. Forbes, I think obviously competition is a
big issue. And so as there has been contraction, there has been
less competition, and I think we have come through an era of
really big platforms which made winners and losers out of
industry. So if you didn't get on the next new platform, you
were out of business. I think we are a little past that right
now and so there is not as much big, new platforms coming. So I
have some hope.
I think the other thing is, there are barriers, I think,
that can be reduced for the government to attract more
innovative commercial firms to do business. That may be kind of
limited. I think the government also has to instead adapt to
the fact that the private sector is funding so much more
research and development. So the government has got to learn to
adapt to that.
Mr. Forbes. And my time is up, but thank you, gentlemen.
I would yield back, Mr. Chairman.
The Chairman. Thank you.
Mr. Walz.
Mr. Walz. Thank you, Mr. Chairman. And I too want to echo
my thanks for you grabbing onto this issue we all know is
critically important in working in. Thank you all for being
here.
Just a couple questions, and I am interested amongst the
different services and the performance amongst them because
there appears to be a difference, if there is some lessons
learned or if it is the nature of the service. This idea of
breaches going through the cost ceiling's scheduled performance
or whatever. I started looking at this.
While this pains me to say as a former soldier, the Army
leads in this area with a 38 percent breach. I am kind of
curious, from your perspectives, is this cultural? Is it the
nature of it? Or what is at work there that would set them
apart from the Air Force and the Navy?
Mr. Francis. That is a tough question. I think the Army,
you know, after the Big Five programs of the late 1970s and the
early 1980s, has had a difficult time finding traction with
aviation and its ground combat vehicles and came through a very
difficult phase with the Future Combat Systems, where it was
coming up with a completely different concept for fighting. And
I think that was doomed by relying on technologies that simply
weren't there and they just couldn't execute that.
I don't know that the Army is quite recovered from that.
After the Future Combat Systems, which is predicated on
fielding an array of 19-ton vehicles that could be airlifted,
the next vehicle that the Army developed was the ground combat
vehicle which was a 70-ton vehicle. So I think the Army has had
some difficulty trying to identify just what it needs and how
it wants to fight.
And when it decides on something it has been moving out a
little bit too aggressively, trying to get it fast and
discovers during the process that it is not a good concept. So
I have seen that more with the Army than the other services.
Mr. Walz. Yes, sir.
Mr. Hunter. I would agree with that largely. I think that--
and the essential point that--with the Army not having
recovered from Future Combat System [FCS], I think the issue is
that they don't have a consensus vision within the Army of what
the next Army should look like. You know, the 1980s, they had
the Big Five and that was a pretty consensus vision. And
because there were five of them and within that there were a
number of subsystems, it essentially--every part of the Army
was winning or was getting something out of that approach. And
since the collapse of FCS, there has been no similar vision for
how to move forward.
Mr. Walz. That lack of vision, is that what led to like
Crusader? I always look at that, is that the problem with the
acquisition process, or did we actually see a glimmer of hope
that it was actually killed after a while? I am kind of curious
on that. Is that just part of this culture, they are searching
for the weapon system that didn't fit the battle that was
coming?
Mr. Hunter. I would say Crusader is an example of a case
where there was a vision but it was not an affordable vision.
And so that is another obviously possible failure mode is you
can have a great vision but it tends, you know----
Mr. Walz. So this is a leadership issue then, is the way
you see it amongst the Army or at least vision-wise. It leads
me into my next question about we included the service chiefs
having a say in this in the NDAA. And again, this might be the
chip on my shoulder or whatever, is it important to add those
senior-enlisted people? I would say they are closer to the end
user type of thing. Does that start to straighten this mission
out, or is that a whole different discussion?
Mr. Hunter. Well, where I see the strength of having the
service chiefs more involved is their ability to bring all the
elements of the game together, budgets, requirements, and
acquisition. And I think there is a lot of power in that. And
they are at the top of those pyramids. And so I think that is
the real strength that they have to offer. They by and large
don't bring to the table a lot of technical expertise to
address some of these technical issues because that is not
their training. It is not expected that they should have that.
So I think on terms of the enlisted side, to the extent that
the senior leadership there can, again, help to bring the
aspects of the system together, that is a good thing.
Dr. Chu. If I may add, sir, I think on the end user front,
the end users you may most want to encourage to say more are
the combatant commanders. They are after all the one
responsible forces--at whatever level, enlisted--officer. And
they don't have too large a voice in the present system.
Mr. Walz. Great. I appreciate that. And again thank you for
helping us understand this. I think all of us here do recognize
this is a critical issue. And at some point in time we are
going to have to--and I think the chairman is right again on
this: We would like to fix it all. That is not going to happen
realistically, but these steps forward do make a difference. So
thank you and I yield back.
The Chairman. Mr. Rogers.
Mr. Rogers. Thank you, Mr. Chairman.
I will throw this out for whoever wants to swing at it. Do
you believe there is a bias in DOD goods and services
procurement against public/private partnerships? For example,
does it make sense for OMB [Office of Management and Budget]
and CBO [Congressional Budget Office] to require 1-year scoring
of the entire liability of a public/private partnership but not
the same treatment for traditional goods and services
government to contractor procurement? Dr. Chu.
Dr. Chu. This question of scoring multiyear buys,
particularly if they have a lease-type structure, which is what
you are describing, I think is one that has bedeviled the
Department. Congress, to point to a positive example, offered a
way out for privatized housing and special provisions.
And so on the one hand, I understand the source of the
constraint, which is to avoid signing long-term leases and
dodging the fiscal limits; on the other hand, I think it has
proven injurious to some arrangements that might indeed be
interesting, and I would look to some provision, perhaps
modeled on the privatized housing authority that allows
meritorious multiyear deals to go through.
Mr. Rogers. Mr. Hunter.
Mr. Hunter. I would just say, my point about financial
flexibility is exactly aimed at what you are talking about,
which is the tendency within the system to shut down innovative
financial approach because of scoring issues or legal issues or
other impediments that have been brought in over the years,
usually because of a bad case that happened somewhere in time
and then we foreclosed an entire range of options. And
multiyear funding is certainly an aspect of that, acquiring
material through services is something that can be very
powerful but is very hard to do in the current system.
Mr. Rogers. Yeah, and housing was the example I was going
to go to. It has been a stellar success. And you know, we are
going to own it at the end of this 50-year lease/purchase
agreement. It has been a win-win. But we run into these same
problems with our satellite access. We would like to have
multiyear deals where we could get a lower rate, and we are
locked into the 1-year scoring, which is just a killer for us.
And another area where I would like to see this done is I
would like to take the same model that we use for housing to
re-engine the B-52 bombers. We could pay for that, in my view,
with the fuel savings, but we would get into the scoring issue
again. So I am real interested in your thoughts about how to
get around that.
Let me ask this: There are going to be some areas where it
doesn't make sense to treat goods and services as a commercial
item where DOD can afford to rely on the market to influence
positive private sector decisions. Is the space launch one of
these?
And as you know, we historically have a situation where
Lockheed and Boeing were in the space launch business and
couldn't make a profit, and decided to get out and we went, no,
don't do that. You all get together and put together a
partnership called ULA [United Launch Alliance], and we will
feed you enough business to keep you alive. And now we are
being attacked--or that model is being attacked, as you know.
And I am just wondering, can we rely on commercial
enterprises for essential national security access to space?
Anybody want to take a swing at that?
Mr. Francis. So Mr. Rogers, I know when ULA was formed it
was formed on the basis of they thought there was going to be a
big commercial market. So what they were going to do for the
Department they thought they were going to be able to adapt to
the commercial market and that market did not materialize. So
they became more dependent on DOD.
Now we have opened that up to commercial competition so we
have commercial firms that are competing. We still have ULA, as
one of the competitors now.
Mr. Rogers. Well, we have got ULA for the moment.
Mr. Francis. For the moment, right.
Mr. Rogers. That may not be there by December 1. That is
the whole point.
Mr. Francis. Right. So I think it depends on how good these
commercial offerings are. Can their rockets--right now at this
point we are trying to see whether they can handle those
payloads and be reliable. And will there be a commercial
market. So if the government is the only customer, it is hard
to imagine you can have all of these suppliers. So it is going
to depend on, I think, largely on the commercial market, and
the government is going to have to protect its interest going
forward.
Mr. Rogers. The last statement you made is the key: ``The
government has to protect its interest.'' We have to, from a
national security standpoint, have assured access to space,
which by DOD definition means two sources. We are going to be
doing good to keep one at the rate things are going.
But let me ask you this: Mr. Francis, you have recommended
that, quote, ``stronger and more uniform incentives are needed
to encourage the development of technologies in the right
environment to reduce the cost of later changes, and encourage
the technology and acquisition in communities to work more
closely together to deliver the right technologies at the right
time,'' closed quote.
You point out that there are organizational, budgetary, and
process impediments which make it difficult to bring
technologies from DOD science and technology enterprise into
acquisition programs. What are the impediments and how can we
change this?
Mr. Francis. So one thing is the science and technology
[S&T] budget is relatively fixed. I think if you look over the
past 20 years, it is about 20 percent of the R&D [research and
development] budget. I don't know if that is the right number,
but again, it is a fixed level of funding. Seems to me, if we
are going to get ready for a next generation of something that
maybe that S&T budget needs to be built up.
It is not big enough now to carry technologies far enough
into maturity, so you end up having to hand them over to
weapons system programs too early because they are the big bank
for money. And there aren't really good mechanisms, at least
consistent mechanisms right now, for science and technology
managers to go into a transitional phase where they can work
with program offices and successfully hand off technologies to
those programs.
So I think there is some structural issues. There is
funding, organizational, and then the fact that really it is
the acquisition programs that are more in control of transition
than science and technology organizations, which is different
from the private sector.
Mr. Rogers. Thank you. I yield back.
The Chairman. Mr. O'Rourke.
Mr. O'Rourke. Mr. Francis, I was very interested by your
comments about opportunity costs and that when we have cost and
time overruns, it is not just money and time in an absolute
sense; it is the loss of something else that we could have been
focusing on, spending dollars on, spending time on. I know it
is hard to quantify a negative, but do you have any examples of
what some of those opportunity costs were, specific programs?
Mr. Francis. There are some cases where the tradeoffs were
explicit, so I remember, kind of using a reverse example, when
the Comanche helicopter was cancelled, it freed up money which
this committee, I think, took the lead on making sure that that
money went into other Army aviation investments. So that is
kind of using the negative to illustrate what should happen.
I don't really have good examples for when a program
overruns and you need more money to do it and where did that
money come from? So programs like the F-22, the Joint Strike
Fighter, the Ford-class aircraft carrier that have overrun, we
have made decisions to put more money into those to buy what we
thought we were already buying, but I am not aware of where we
have listed the tradeoffs. What did we give up to provide that
extra money? Now, I have to believe that exists in the
Pentagon, but I don't know if that is a debate that the
Congress is afforded.
Mr. O'Rourke. You also talked about the fact that we
unintentionally incentivized, overpromising on the outcomes and
underestimating on the costs, and you suggested that this
committee or Congress should send a signal by rejecting some of
these programs or projects or systems. Do you have any specific
examples?
Mr. Francis. You know, I was thinking about that before I
came in. I really can't think of examples where Congress said
no. What Congress tends to do----
Mr. O'Rourke. Could you think of some examples where
Congress should have said no?
Mr. Francis. Oh, yes. So Future Combat Systems. I give this
committee a lot of credit for having all the early hearings on
Future Combat Systems. It was simply not possible. It didn't
measure up to any reasonable test of an executable program. It
relied on 50 uninvented technologies, and it was a $200 billion
program, and we were going to do it, I think it was 19 separate
programs, and were going to run all 19 in 5 years, in less time
than it takes to run one program. It just was not executable.
What Congress tends to do and what it did in this case was
it puts strings on the money. It will put a cap. It will put a
condition that you can't go forward unless you report back. But
it never said no to the program, so it took Robert Gates to say
no to it. So Congress is reluctant to give a no. It will give
an angry yes, but that is a yes nonetheless.
Mr. O'Rourke. Let me ask you about another topic that you
brought up, which is the need to take risks earlier in the
cycle or in the process. Can you expand on that a little bit
and talk about our role in doing that or in creating the
incentives for that?
Mr. Francis. Sure. And Dr. Chu, his statement covers that
as well. We need to take risks. So we don't want a situation
where we don't take any risks and we never have any failures.
You have to take risks. We are not going to have perfection
here. Perfection I think would be a bad thing, but we can do
better. We need to take those risks in science and technology.
That provides the environment where failure is okay.
So the purpose of S&T is to discover. But once you get into
product development, the purpose of product development is to
deliver, and you can't invent on a schedule. So I would say we
have to take those risks in science and technology and carry
those risks further and resolve them. And if we can't, then we
make an eyes-open decision that we are not going to take the
risks in product development, and so we are going to take those
out of the requirements; or we are going to take the risks but
we are going to have to put the money up to take them.
So too often we say we are going to take these risks, but
we have a risk mitigation plan in place that is going to make
it okay. But that risk mitigation plan generally lacks two
things, time and money, which are the consequences most likely
to attend risk.
Mr. O'Rourke. Thank you. Mr. Chairman, I yield back.
Mr. Hunter. Can I just add one thing to that? And I agree
with what Paul has said, but there still needs to be path for
that S&T to get into the arsenal. And that is where I would
again mention this idea of adaptable systems, so when you do
prove something out in the S&T or early stage R&D, you need a
way to host it on a platform that the warfighter actually uses,
so you have to make that connection.
Mr. O'Rourke. Thanks.
The Chairman. Mr. Wittman.
Mr. Wittman. Thank you, Mr. Chairman. Gentlemen, thank you
for joining us today. I want to talk about how things happen
within the decisionmaking process on acquisition, and you
pointed out some success models at the smaller scale where you
can take a concept that comes from unit commanders. You develop
that process of developing that idea with systems engineers,
and you have program managers involved and contracting
officers, and you end up with something that works and
resembles what was needed at the very beginning. So that small-
scale process you have shown works.
Let me ask this. How do we take that and graduate it to the
large scale? How do we take emerging technology, compress the
time process, be able to make sure that we have great
communications from the unit commanders, to the systems
engineers, to the program managers, to the contracting
professionals, and getting that done?
It seems like to me today what we have is we have the unit
commanders that are here, and then in a whole separate element
there is the systems engineers that take that concept and
develop it, and then in a separate place, in another area, are
the folks that write the requirements, and then another group
of folks that come up with the proposals and another group of
folks that come up with the contracting process, so all of this
is fragmented, and no wonder decision making goes awry,
especially when it seems to be process-driven. People seem to
be I got to check the box. And if I check the box, then I have
done my job. Instead of saying, you know, let's focus on what
the warfighter needs. Let's focus on getting that technology to
them as quickly as we can. How do we do that?
How do we put authority into the hands of those people at
every point in the process? How do we bring the decision making
together? And how do we make sure that we also have
accountability there so we don't go awry, or if we do go awry,
we can either get things back on track or stop things
immediately? Give me your perspective on how we make those
things happen.
Mr. Hunter. I try to address this in my testimony. I talked
about the importance of senior leadership and shortening the
lines of authority for acquisition. And the model that I would
offer is what we used for rapid acquisition was the Warfighter
Senior Integration Group. And if you could picture it, a giant
room, tons of people around a table. You have got the
acquisition folks there. You have got the logistics and
sustainment folks there. You have got the operators in theatre
coming in through VTC [video teleconferencing] who are actually
setting the requirements, are going to use the equipment.
Everyone around the table, with the Deputy Secretary of
Defense there, and the question is not a debate about should we
or shouldn't we. We are going to do it. That is the bottom
line. And everyone who has a role in the system is there, and a
decision is made. The Deputy Secretary says, here is how we are
going to do it. Everyone go out. These are the marching orders.
Mr. Wittman. Mr. Hunter, I agree. That is a great concept,
but that happens occasionally. How do we make that the rule?
How do we make sure that that is how the acquisition process
takes place, rather than saying here is a great example about
how it works? To me it has to become a culture within the
organization to make that happen. What needs to take place to
make sure that that is the rule?
Mr. Hunter. I would say for programs where faster is your
priority, you can make that similar construct work. I believe
that will be the case for the LRSB [Long Range Strike Bomber]
program because you have this Rapid Capabilities Office in the
Air Force that already works this way. It has a board of
directors, very similar to what we had at the Warfighter Senior
Integration Group.
Where you have instead of 50 layers between the person in
charge, Deputy Secretary of Defense or in this case Secretary
of the Air Force and the Under Secretary for Acquisition, it is
three or four layers or less. To me that is where the real
power of that approach comes. And it can be done. Now can we do
it for every acquisition program in the Department? Probably
not.
Mr. Wittman. Is this a directive that needs to come from
the House Armed Services Committee? Is it something that needs
to come through OSD [Office of the Secretary of Defense]? How
do we operationalize this? Because we have got a lot of great
ideas floating around about how to fix the acquisition process,
but the ideas never seem to make their way to reality. Tell me
where you believe the push needs to come from and the
determination and concrete direction needs to come from to make
this happen?
Mr. Hunter. Well I would say Congress has done one thing in
this most recent NDAA--the one that is still pending, I should
say, after the veto--is to streamline the process. So, you
know, there are a number of documentation, check-the-box kind
of exercises that have been imposed over the years, a number of
them by the Department but a number also by statute and kind of
cleaning the books of a lot of these things can really help.
And then, as I mentioned in my testimony, now that many of
those statutes have been changed, making sure that the
Department follows through to actually change the regulations
because a lot of these things were required by statute. Now
they are in the regs, and so you have to clean that stuff off
the books. And following through to make sure that now that the
statute has been streamlined, that the regulations are also
streamlined is critical.
Mr. Wittman. Thank you, Mr. Chairman. I yield back.
The Chairman. Thank you. Mr. Moulton.
Mr. Moulton. Thank you, Mr. Chairman. Gentlemen, thank you
for being here. I just want to echo the chairman's comments
that he began with, which were that I think that this is partly
about conserving taxpayer dollars, which we know is incredibly
important, especially in this budgetary environment, but even
more importantly it is about responding to the next threats.
And you know, companies used to succeed in the light bulb
industry by trying to be most efficient at making the light
bulb for the least cost, and now companies succeed in that
industry by being the quickest to develop the newest type of
light bulb, to be the quickest into LED [light-emitting diode]
technology. I think in my long tenure of almost 11 months on
the committee, I have been a strong advocate of cutting
programs that we don't need and old systems and legacy systems
so that we can invest that money in the new ones, which I think
is incredibly important.
One of the things that we need to do more of, as you have
said, is be willing to accept the fact that technology
development does not just occur in the Department of Defense
anymore. It occurs outside. And there seems to be a conflict
between the desire to get more commercial-off-the-shelf
technology, and the MIL [military] standard requirements that
this technology then has to meet.
And I think about how much more effective I would have been
as an infantry officer on the battlefield if I could have used
an iPhone. Now, if I were to get killed because my iPhone
didn't meet that 100 percent requirement and failed at some
point, there would be a lot of grief. But on the other hand, if
we don't allow the iPhone on the battlefield for years because
it can't meet that 100 percent requirement, a lot of people are
going to die. And you might not see the news stories about it,
but it will be a loss as well. So how do you think about better
managing that conflict?
Mr. Pasqua. I think that is incredibly important, and to
the earlier point about sort of what worries us in the larger
curves, it worries me that some of the best minds of a
generation in the tech world are focused on how to get you to
click on more ads rather than technology that can be helping
our Armed Forces. And I think one of the big reasons for that,
as I mentioned in my opening statement, is it is just too hard.
Companies don't know the terminology that is being used in this
room. They don't know how to engage in government processes or
DOD processes. They look at what it would take to learn, and
unless they are making a technology that is specifically suited
for that area, they are just not going to do it.
Mr. Moulton. So what can we do to try to fix that? I
visited a company in my district that is developing an iPhone-
based application or mobile-phone based application. And just
out of curiosity I asked them why they had switched from the
iPhone to the Android phone. And they said, well, the problem
is it is harder to access the software on an iPhone. And I said
ironically that sounds like a really good thing if you are in
the Department of Defense, but obviously DOD does not have a
good relationship with Apple; so in this case it might be
harder to hack into an iPhone, but we are going with Android,
and it is nothing against Android, but if that is true based on
what they implied, it seems like a better partnership with
Silicon Valley would help. What can we do to facilitate that?
Mr. Pasqua. I think there is a couple of things. I think
one of the things that would be really helpful, and it has been
happening. We are getting many more visits from different
governmental organizations to the Valley to make companies more
aware of what the opportunities are and how they might get more
involved. But there needs to be, you know, short of changing
all the acquisition processes, there needs to be some
methodology, some help for these organizations to be able to
sell into the Department without having to learn all of the
processes that are involved because they are just not going to
do it.
So whether it is working with larger integrators who
already know the ropes so to speak, or creating conduits by
which some of these technologies can get embedded into other
modular platforms, as we have been discussing, more easily, I
think either of those approaches would do it.
Mr. Moulton. I just have 30 seconds. Dr. Chu, did you have
a comment?
Dr. Chu. I think in terms of getting other firms to be
willing to offer to the Department, what does need to be
thought about is do we need to burden the contractors with as
many special provisions, largely social goals, as current
acquisition statutes require? It makes it very difficult for
the commercial firm to want to offer to DOD. This is a high
wall of expertise and requirements we are going to have to
meet.
Mr. Moulton. Thank you.
Mr. Hunter. I just had one thought there. Other transaction
authority, something that the pending NDAA would make
permanent, is a way to do this to create a special, much
stripped-down agreement with commercial companies. It is
definitely a great tool, and it is something the Department
needs to use more.
Mr. Moulton. Thank you. Thank you, Mr. Chairman.
The Chairman. Thank you. Mr. Scott.
Mr. Scott. Thank you, Mr. Chairman. And I think the
question of is perfection the enemy, is good enough to get the
job done, is something that we all wrestle with. And I also
wrestle with, for example, cancelling the F-22 line before
another plane is ready to go. I don't think that General
Electric or any manufacturer in the world would stop one line
of refrigerators before they had another line that had proven
that it was capable of doing so.
I want to talk with you a little bit about the rapid
acquisition process and the JSTARS [Joint Surveillance Target
Attack Radar System], that E-8 platform flies out of Robbins
Air Force Base. We have discussed for a long time the rapid
acquisition process. And this is effectively a platform where
the technology is ready to go. It is just a matter of getting
the go-ahead if you will and the decision from the Department
of Defense with regard to which platform they want to go with,
and we are going to end up with an operational capability gap
because of the depot maintenance that is going to be required
on that platform, and there is not a battle management platform
that can take its place going forward.
So, just if you could speak to why is there the delay when
the Air Force knows what it wants, when they know the rapid
acquisition process would save money, they know they have to
field a new fleet, and the delay is actually going to result in
an operational capability gap of a couple of years before full
operational capability comes back, what factors contribute to
programs being lengthened, even when they have those high
levels of technological readiness for the major systems and
subsystems?
Mr. Hunter. My perspective on the JSTAR's recap program is,
it gets back to this question of vision. I think the Air
Force's vision for that program as a battle management control
asset is something that the broader Department is still waiting
to embrace, maybe would be the way to put it. Other services
have thought traditionally--I think the Army has thought
traditionally--of JSTARs as more of an ISR-type [intelligence,
surveillance and reconnaissance] asset, so these are slightly
different missions that the services see coming from this
platform.
And if you think of it as an ISR asset, you say, well, we
have got lots of ISR assets, and we have got unmanned ISR
assets, and so why is a manned platform the right way to go?
The Air Force sees JSTARs fundamentally as something bigger
than or more than an ISR asset. And I think that is kind of the
dialogue that has been going on. And they are making kind of
that first step, and I mentioned in my testimony that Paul
mentioned about the milestone B decision, which is a threshold
decision of do we want to invest in this or not?
And you want to be cautious making that decision, but once
you have made it, and you have made a good decision hopefully,
then I think the priority is to proceed with all speed into
program execution. And that is where I think JSTAR's recap
would be a good example of something that ought to be one of
these adaptable programs that I mentioned, particularly given
the pace of evolution of electronics technology, that you want
to have a design for that system that you can upgrade
continuously throughout its life cycle because you know the
technology is going to be light years away in 30 years in ways
that we could never imagine today.
Mr. Scott. Well, we already have some of that. We have the
ability to plug and play, if you will, with certain camera
systems and other things, and with respect as far as the Army,
you said waiting to embrace, that is not my understanding from
the Army. That platform flies continuously and has for
approximately 20 years. We have got a problem with corrosion
now. I mean, these planes are old. These are 707s, I think,
from the 1970s that we have updated and updated and updated;
and to send these units through another major round of depot
maintenance, it would make much more sense to spend the money
on totally new systems.
Mr. Hunter. And I agree with that, sir. I am not in any way
suggesting we shouldn't move forward on the program. I think
this dialogue that has been going on as they have gone to this
investment decision has been more about the vision than it has
been about the specifics of the program.
Mr. Scott. What about just the example of shutting down,
for example, the C-17 line without another lift capability
ready to go? I am out of time. I apologize. I yield.
The Chairman. Interesting questions. Mr. Castro.
Mr. Castro. Thanks, Chairman. I know that we talked about
the competitiveness of the process in contracting and how it
has become dominated over the years by a smaller group of
companies. Some of that is due to consolidation. We also talked
about how hard it is for commercial enterprises to break into
defense work.
So I guess let me ask you, we also have in front of us
information on the cost overruns and breaches for each of the
divisions. Is there any penalty for a contractor who
experiences a cost overrun on a contract? Or I should ask what
is the penalty?
Dr. Chu. Depends on the contract. If you have a cost-
showing provision or if it is a fixed-price situation of some
kind, the contractor will obviously earn a lower rate of
return. I think the ultimate penalty for the contractor is
something Mr. Francis touched on, which is, if the system
proves more expensive, the Department may decide to buy fewer
of those systems. And so the length of the production run or
the volume of business the contractor enjoys is thereby
diminished.
In the worst case, the Department will decide, has
occasionally done so, this is too much. We are going to stop
and thereby lose the opportunity to further production. So any
contractors have an incentive to try to keep, not necessarily
to meet the guidelines that were pledged in the acquisition
process--that is a whole different issue--but to keep the
production price of the article still competitive with the
mission need.
The real issue in all these cases is, is this worth
investing in to perform the mission we have in mind, or has it
become too expensive relative to the return that it will yield?
Mr. Castro. And is it fair to say that over the years
particular contractors have again and again gone over on cost?
Dr. Chu. Since we now have a small number, I don't think
any of them has been exempt from that problem. I think it is
important to keep the cost overrun issue in perspective. The
large cost overruns are largely percentagewise on development
contracts, not production contracts. Once we get to production
beyond the first few lots, we generally have a fairly good idea
what it is going to cost, and people stay within those
parameters.
The typical program doesn't actually overrun. That is not
always true. That is the legend out there that they all
overrun. That is not fair----
Mr. Castro. Let me ask you this. We have been speaking
about each contract individually. Is a contractor who
consistently overruns penalized when they bid for a new
contract? In other words, with consumers, for example, many of
us are subject to credit scores; right? So if we demonstrate
bad credit over a period of time, there is a penalty when you
try to get credit next time. Does the same principle apply with
contracting here?
Mr. Hunter. Well, I would say it does to an extent. There
is two ways in which it can apply. One is past performance.
Rarely are major contractors ruled out because of past
performance, so there I would say no. On the other hand----
Mr. Castro. Are they not ruled out because there are simply
not enough options that we have? In other words, is there such
a limited number of these contractors that you just can't go
anywhere else?
Mr. Hunter. It could preclude competition if we ruled out a
major competitor, but what they do often in the evaluation
process is they will evaluate the contractor's price, not
necessarily exactly at what they bid it, but at what the
estimators inside the government think that that would really
translate to. In other words, for example, with the tanker
contract, the last version of that was fixed price, so it was
evaluated at what was bid.
But a previous version, they evaluated the price of the
bidders higher because they thought, we don't really believe
the costs that you are putting forward. And so depending again
on the nature of the contract, they can evaluate a contractor
at a higher price if there is a history that they have
delivered at a higher price.
Mr. Francis. If I may, it can get pretty complicated, so an
overrun, you get into a debate as to whose fault it is. Did the
contractor deliberately underbid and then overrun, or did the
government underestimate and----
Mr. Castro. Well, but is it safe to say, that the
Department of Defense has the most overruns, and the cost is
the highest of any of the agencies of the Federal Government?
You are part of the GAO, so I assume----
Mr. Francis. Right. Actually as much as we talk about the
Department of Defense, they are probably the best in
acquisition. If you go to the civilian agencies, they are much
worse generally.
Mr. Castro. So there are more overruns and more breaches?
Mr. Francis. A much higher percent, yes. This is probably
the subject of another hearing, but the government and the
contractors don't share the same interests. I mean, they are
working together on a program, but where the government may be
thinking it has got a contract to get a product for a certain
price, where the contractor maximizes its profit. The
contractor is also interested in a longer business line,
keeping its facilities amortized and so forth, so they may
sacrifice profit to get a larger volume of business, so two
different incentives here.
Mr. Castro. Thanks. My time is up. I yield back, Mr.
Chairman.
The Chairman. Thank you. Ms. Stefanik.
Ms. Stefanik. Thank you, Mr. Chairman. Over the course of
the war in Afghanistan, we saw ground combat vehicles undergo a
number of upgrades and additions because of new and increasing
threats. But specifically in the Army, protecting the safety of
our soldiers also added additional weight of combat vehicles,
and, therefore, we had subsequent challenges to maneuverability
and rapid deployability.
I represent the Army's 10th Mountain Division, and
lightweight weapons systems, body armor, and operational
mobility are essential capabilities for the 10th Mountain. So
the struggle as I see it is how do you balance the weight
against the protection, against the budget? So for example,
obviously a vehicle built out of titanium would address the
weight challenges, but the cost would be much greater.
So how can we as Congress help the DOD make those needed
upgrades or obtain new and affordable materials which are able
to stop emerging threats?
Dr. Chu. I think your question goes exactly to the thought
I offered, that a more physics-based approach at the early
stage of the program, looking at these tradeoffs explicitly,
would be very helpful. And if Congress were to ask for what the
parameter space looks like, which it doesn't tend to do, in
other words if we want more protection, what are we going to
sacrifice either in carrying capacity of the vehicle, or cost
in order to use a more advanced technology for protection, that
would lead to a more informed debate about why did the
Department pick the particular combination of ingredients it is
recommending in the program that is going forward? That
conversation at the legislative level typically does not happen
now.
Mr. Hunter. I would say, you bring up a point that I think
starts to highlight some of the challenges that the Army has in
moving forward with its acquisition programs and its vision.
Because I know that the Army has looked at what can we do with
combat vehicles.
And one of the reasons why I would suggest and that I have
heard from some in the Army acquisitions system is that they
aren't moving forward on a new ground combat vehicle is because
they don't think they can get one that is significantly lighter
than the systems they have today. And there is some logic or
some mode of thought that says why would I invest billions of
dollars in a system that ultimately isn't going to meet the
objective I want, which is a lighter, more maneuverable
vehicle?
This also relates to the point Dr. Chu made about the MRAPs
and why the government didn't retain most of those. We did
retain some, and actually many of the ones that were retained
are the M-ATVs [MRAP All-Terrain Vehicles] that were maintained
in Afghanistan which were more mobile and able to move around
in more challenging terrain. But the heavier versions that we
used in Iraq have largely been let go because they don't meet
that priority.
Mr. Pasqua. Just a quick comment. I agree with Dr. Chu's
comment about understanding and explicitly choosing the point
in the trade space early on in the process and understanding
what the entire space looks like. I would just add that it is
important also to get a feel for what it will take to make a
move in that space. So we can understand where we can be at a
given point in time, the tradeoffs that we are making to choose
that point, and understand what it will take if we want to move
in other directions, or at least have a feel of the scope of
it.
So that is to say that, you know, we talked about modular
and adaptable systems, but they are not free. They are actually
hard. It is hard to design a system that is adaptable in every
conceivable way; and, in fact, you typically don't want to do
that because it will introduce new limitations or costs. But it
is important to understand, even when making the initial
choice, what the costs will be to make moves to different areas
in the trade space like lighter weight, what the costs would be
associated with that, and in the upfront design decide whether
it is appropriate to enable those moves in the trade space
later. Because as I say, it will take costs to enable that
modularity or flexibility.
Dr. Chu. If I may add one thought, as one looks at the
technical trades, I think it is always important to keep in
mind what mission need are you trying to fulfill? And that may
lead to you conclude that you don't need quite as
technologically ambitious an article as you thought you did.
An example comes to mind on position navigation precision,
one of the technical programs I have had a chance to look at.
When we were aiming at a very high degree of precision, when we
showed operators, back a bit to what Mr. Hunter was
emphasizing, what we could achieve, which was far south of that
objective, they said, no, no, that is good enough. Don't keep
going. We will take what you have already been able to achieve.
Ms. Stefanik. Thank you. Mr. Francis, do you have anything
to add?
Mr. Francis. Yes. I think a good example of what you are
describing is what Secretary Gates brought up, so when the Army
was really putting all of its emphasis on the Future Combat
Systems for the next war, Secretary Gates made the point that
we are not really focusing on the war that we are engaged right
now.
And I think the issue becomes in some cases we are not
anticipating well. So the science and technology community was
not necessarily working on those up-armoring solutions. So,
when the need arises, we have to react, and we have to react
maybe suboptimally, so anticipation is important.
Ms. Stefanik. Thank you, my time is expired. I yield back.
The Chairman. Ms. Speier.
Ms. Speier. Thank you, Mr. Chairman. Thank you to our
experts who are here today. You know it has been said that we
have the best weapons in the world, but our acquisition
pipeline often gets us those weapons late and over budget.
Mr. Francis, you have said that to describe our acquisition
process as broken is an oversimplification because it implies
that it can merely be fixed. Time and time again we have tried
to overhaul this process. We come up with the same challenges,
an ingrained culture, an inadequately trained workforce, an
inexperienced set of program managers, and a dangerous
revolving door to industry.
In March of 2015, the GAO issued a report that 19 of the 38
assessed programs reported they planned to begin production
prior to completing software development needed for baseline
capabilities. A perfect example is the F-35, where software for
even basic capabilities necessary for testing and evaluation
are running months behind. As weapons systems grow in
complexity, this is a problem that will come up more often than
before.
How can we adjust our acquisitions process to better
develop and test the software components of the hardware? And
that question is open to any of you.
Mr. Francis. So, Ms. Speier, I think one of the issues
there is not fully understanding what the requirements require
from the design. And that is something we have had
conversations with the chiefs about. They think they understand
the general requirements, but they don't understand the
thousands of specifications that are necessary to meet those
requirements. And a lot of that translates into software code.
And I am trying to remember on the JSF [Joint Strike
Fighter], and maybe one of of my colleagues here can help me,
but I think it is like 80 percent of JSF's functionality comes
from software. I don't know that that is known in the
beginning. And that is what, when I talked earlier about we
need to know what the design requires and what risks we are
taking upfront and we can make decisions on that.
Rather what tends to happen is we don't know enough when we
start. These risks get played out later on, and we end up with
what I call latent concurrencies, doing things at the same time
that we didn't plan on doing at the same time. So if we are
going to be concurrent, let's agree to it upfront and say we
are taking that risk. If we don't want to be concurrent, then
we have to understand the design better sooner.
Ms. Speier. Mr. Francis, you were here almost exactly 2
years ago on October 29, and I am sure you feel like a broken
record, but on the theme of repeating yourself today, you
mentioned the Ford-class aircraft carrier in your testimony.
Saying that the GAO identified this program as lacking a good
business case back in 2007. That makes the program's current
struggles unfortunate but not at all unsurprising.
What programs are currently in the pipeline that we should
be looking at with greater scrutiny? What aren't we looking at
today that you will be talking about in 2 years?
Mr. Francis. So I will come back for the record with a
list, but I think the Joint Light Tactical Vehicle, the Army
program, is something that needs to be looked at. Long Range
Strike, when that program comes up, that needs to be looked at.
Ohio-class replacement is another big one coming. DDG-51 Flight
III right off the top of my head are big ones that I would
think about.
And I think for Congress, where you really have to weigh in
is, when money is being requested for these, so their milestone
B decisions might be 2017 or so, but you have to ask those hard
questions now when you are putting money on the table. So I
will come back with another list, but perhaps some of my panel
members here know about other new programs coming, or maybe
not.
[The information referred to can be found in the Appendix
on page 117.]
Mr. Hunter. There is a few others. We mentioned JSTAR's
capitalization program which is right on the cusp of entering
the process, and another one that has been a big focus has been
U-class program.
And I assume when Paul says take a look at them, that
doesn't mean cancel all those programs. I am not, certainly,
going to put myself in that position. But I do think the role
of the Congress in examining that investment decision that the
Department has made, to say ``Why is the cost of this worth it
from a warfighting perspective?'' is absolutely the right
question to ask.
And asking it right around the time of milestone B, I would
say ask it around the time of milestone B. Paul is saying ask
it a little earlier than that even. You want to be on the front
end. Once you get deep into the program, there are
constituencies associated with it, and frankly, you are
committed in a way that is just hard to get out of. And that is
why those early milestones are so critical.
Ms. Speier. Thank you. I yield back.
The Chairman. Ranking member.
Mr. Smith. Just following up on that, you talked about all
those programs that have been to be, quote, ``looked at.''
Obviously, they have to be looked at. They are going to cost us
billions of dollars. You are planning on building them. Looked
at, A, from the standpoint of do we even need it? Is that is
what you are saying? I mean, if you could be specific because,
I mean, you rattled off basically all of the major programs
that we are planning on building over the course of the next
decade or so.
And you know, Ohio class, Long Range Strike, JLTV [Joint
Light Tactical Vehicle], are there any of those that you would
say from a warfighting standpoint, why are we building this? We
don't really need it. That is one. And then two, are we making
some of the same mistakes with those programs that we made with
the F-35, basically constructing the plane as it is working its
way down the runway? And those are my two questions.
Mr. Francis. So, what I was suggesting was more on, are
these programs executable? Whether they are needed, and what
are the right solutions? You know, I have to think the
Department really does consider that pretty heavily. You do
have to ask those questions, and I would ask, I think it is
hard to say these programs aren't needed.
The real hard question is, do we need this program at the
expense of this other one? What tradeoff are we making? Those
are fair questions to ask. I don't have evidence to say these
programs aren't needed. But I do think where you can really
weigh in is, we have talked about things like technology
maturity. Is the design understood? Do the requirements reflect
reality, or are they too lost.
Mr. Smith. And as you look at where we are at on those
programs, can you point to a specific red flag? Because, I sort
of get all that. That basically, you know, on these big-ticket
programs now, you know, concurrency is--well, I can't say that
word in a public hearing--but not a good idea.
Basically figure it out, then build it. Not at the same
time. Do you see us making that same--are we counting on that
level of, okay, we will build it and then we will figure it out
as we go? The Ohio-class is an enormously expensive program. I
think it is probably the most expensive one of the bunch. Are
we making that mistake in these early stages in your view?
Mr. Francis. So we haven't yet looked at Ohio class, or
JLTV, or Long Range Strike. We are looking at Long Range
Strike, but that is classified so we can maybe give you some
information on that. So I don't have anything specific to offer
there. On DDG-51 Flight III that is moving very fast. That is
really rapid acquisition. We do have----
Mr. Smith. Is that a good thing?
Mr. Francis. Pardon me?
Mr. Smith. Is that a good thing?
Mr. Francis. I think in this case it is going a little
faster than it should. It has been bundled into the multiyear
for the DDG-51 Flight IIA.
Mr. Smith. But isn't the real issue here just the rapid
pace of technology? You know, I mean, we can all just sort of
logically say, as I facetiously said, don't build the airplane
as it is working its way down the runway. You know, figure out
what you are going to do, and then do it.
But the problem is, while you are building these things,
technology is just boom, boom, boom, boom, boom, moving. I
mean, it is Moore's law. You know, how do you get around the
problem that if you take the opposite approach and you say,
look, we are going understand it, and then we are going to
build it. And we are not going to change it.
Because really what drives a lot of these costs, well, on
the Ford, is change orders; is as we start building it we go,
you know what, now we could do this. Let's do that. Let's add
that to it. I mean, Huntington Ingalls will tell you that if
they were building the same aircraft carrier that they were
told to build when the contract was given to them, it would be
on budget. But there has been so many changes.
But I guess what I am asking is, part of those changes are
driven by just the way the world works these days. Technology
is updated. You can build a better thing. Would you say that we
are better off to say, look, we know the technology is
improved, but we are better off building good enough, than
trying to adjust to that technological improvement that could
make it better? Because isn't that what really traps us on
these things, is the technology, you know, leaps ahead while we
are in the process of building it?
Mr. Francis. Well, two things, Mr. Smith. I think when you
talk about enabling technologies, technologies that make that
platform possible, so on the carriers it is the Electromagnetic
Launch System; it is the Dual Band Radar, and it is the
Advanced Arresting Gear.
If these are enabling technologies, you have to have them
matured before you go forward with the concept because they
make the concept possible. Going forward from that, you want to
have open systems architecture so you can then bring in modular
improvements of that baseline.
Mr. Smith. Upgrade.
Mr. Francis. That is right. So what is going on with the
Ford class is not so much technology refresh as we go, it is
technology discovery as we go.
Mr. Smith. Right.
Mr. Francis. So you have to have a two-pronged approach, I
believe.
Mr. Smith. Yeah. I mean, to a certain extent aren't we--I'm
sorry, go ahead, Mr. Hunter.
Mr. Hunter. Well, I was going to add to that, we have been
having a discussion about adaptable systems. And I think to
your point, an example that I used in my testimony is the
Predator system, you know, which has really evolved. It is
almost unrecognizable now as a system from where it was when it
started. And in incremental changes along the way, it has
adapted to warfighter needs.
And I think it is a classic example of how that can be
done. You start with a relatively simple thing. It is an air
truck. It happens to be an unmanned air truck, but other than
that, it is basically an air truck. But it is a flexible enough
design that as new sensors have come along, new weapons have
come along, many, many changes, I probably shouldn't get into
all of them that have been made----
Mr. Smith. Right. In that case we were able to add it to
the existing Predator. We weren't required to scrap the ones we
had and build a whole bunch of new ones, is that correct? And I
am sure, but----
Mr. Hunter. Well, it is true in part, and untrue in part.
Mr. Smith. Right.
Mr. Hunter. So the design that we have today is
significantly bigger than the original assets that were done.
But it looks roughly the same. So the general design concept
has been fairly constant, but it is a bigger airplane today
than it was before.
Mr. Smith. Yeah, but like on the Ohio class, I suspect that
as we build that thing, there is going to be technological
improvements that we are going to want to add to it. And I
actually would suggest that we are better off not.
We are better off saying, look, we cannot afford to drive
these costs through the ceiling, and yes, maybe it won't be
absolutely perfect or as good as it could be, but particularly
from a competitive standpoint. I think the Ohio class would be
able to serve its function without adding all the new stuff
that is going to be discovered in the next decade.
And I think that is a choice we need to make because it
seems to me, we always make the other choice, which is, you
know, this is my Austin Powers joke: All I want is sharks with
frickin' laser beams attached to them. I use that joke
frequently in acquisition, because, you know, it's like, we can
do this. Let's try it. And we could, but the costs are
prohibitive compared to the gain.
And I think we need to start accepting good enough instead
of, we could put the laser beams on the sharks, so let's go
ahead and give it a shot. But, you know, those are individual
decisions that have to be made program by program by the
program managers and by the Pentagon. I just hope they will
start making the more cautious decision to save us some money.
Mr. Hunter. Ohio class is an interesting example, because
actually, you probably need less than the existing system we
have today. I am not suggesting we scrap them, but----
Mr. Smith. I am sorry, which system?
Mr. Hunter. To the existing system we have today----
Mr. Smith. Right.
Mr. Hunter [continuing]. In the sense that we are going to
fire the same missile----
Mr. Smith. Right.
Mr. Hunter [continuing]. And we don't necessarily need as
many tubes as we have today. Our requirement is not as robust
as it was when the Ohio class was designed. So, of course, we
are 30, 40 years on, and so there is going to be new
technology. There has to be new technology in the system
because you can't go backwards----
Mr. Smith. Right.
Mr. Hunter [continuing]. On many of these things. The one
last thing I would say in terms of submarines, is the Navy
really pioneered this adaptable systems approach with the
combat system on attack submarines, their Acoustic-Rapid COTS
[commercial-off-the-shelf] Insertion Program. And I think that
is exactly what you are describing, is, you know, go into
production with a design that you know works, and then have a
system that allows you to update and upgrade that combat system
as technology proves out.
Mr. Smith. Okay, thank you very much. Thank you, Mr. Chair.
The Chairman. Thank you. Ms. McSally.
Ms. McSally. Thank you, Mr. Chairman. And thank the
gentlemen. I spent 26 years in the Air Force, and I was lucky
that I didn't get any assignments to the Pentagon. So my
perspective comes from the warfighter end of that perspective.
We had a hearing earlier this year and my first question is
about development of aircraft, not Joint Strike Fighter-type
aircraft, but say, follow-on to the EC130, or follow-on to the
A-10 for a light-attack aircraft.
In a hearing earlier this year, I asked the Pentagon
official if we decided today that we wanted to develop a light-
attack aircraft--again, this is not complicated technology. It
is just all of the things that we have learned about what does
permissive CAS [close air support] and does it well to follow
on to the A-10, and we decided, today if we wanted to do it,
how long would it take? And he said about 15 years, I think was
his answer.
I look at the EC130. We know the guts of the EC130 is
working. It has got a great mission set. We know it needs a new
platform, but, you know, we struggle to take forever in order
to figure out how to adapt what we have and put it in a new
platform.
You know, what is it we can do specifically in like these
types of things? We are not developing new stealth technology,
fifth-generation fighter, but we are just learning from
everything we have had, and we have just got to refresh and put
it into maybe a different package. Like, why can't we do this
faster? And what can we do to help especially in those two
examples, you know, to be adaptive, to put these smart brains
together, and develop something in 3 to 5 years that could be
follow-ons to these type of platforms. Anybody want to jump in?
Mr. Hunter. Well, I can't resist, having spent time doing
rapid acquisition. We do do it faster. We have done it faster.
This is something that I guess amazes me after my time in the
Department that, you know, go to Iraq, go to Afghanistan----
Ms. McSally. For an aircraft. Specifically for an aircraft.
Mr. Hunter. Yeah. There are systems that are flying. Hosts
that just simply didn't exist even 3 years ago.
Ms. McSally. So why do they still say 15 years then? What
are we missing?
Mr. Hunter. So what that is an answer to is a fleet of
aircraft that we are going to sustain for 50 years. It takes 15
years, essentially to--if you are lucky, to have a program that
is going to be a 50-year, large aircraft fleet type of a
system.
And there may be opportunities to accelerate that, but on
the other hand if you think you are going to sustain something
for 50 years, it probably makes sense to take a little extra
time to get it exactly right. But we don't need to do that in
all cases, and I think that is kind of the key.
Ms. McSally. Yeah.
Dr. Chu. I am not sure I want to defend the 15-year
estimate as being meritorious.
Ms. McSally. Yeah.
Dr. Chu. Certainly, if you insist on starting every element
of the new design over, you are going to add to the time scale.
I think part of the genius, and it is hinted in the way you
phrased the question, is can we take some existing designs--
which might be foreign, by the way, not necessarily in the
United States--and adapt those to whatever purpose we have
mind. And I think a more--the approach that builds more on what
we already know would allow you to field capabilities faster.
Mr. Francis. So a couple thoughts. I have been around long
enough to remember when the A-10 was being developed and the
Air Force wasn't particularly in love with it either.
Ms. McSally. Still isn't.
Mr. Francis. So you have to want to do it. I think what Mr.
Smith was saying is important. The 80 percent solution has to
be okay.
Ms. McSally. Right.
Mr. Francis. And that is hard to sell because you have to
show you can crush all the alternatives. So you need a 200
percent solution. Eighty percent has to be okay. The other
thing we haven't talked about is there are cases where you want
to put a time constraint on the development. So if you put time
in there as a constraining factor and say, I want to get
through the development phase in 4 years, what can I do then?
That has a way of affecting the requirements of the design.
Ms. McSally. Yeah. Great. Thanks. You know, my other
experience with this is, I have spent a lot of time in air
operation centers, and joint operation centers, and spiral
development is something that we worked on in JEFX [Joint
Expeditionary Force Experiment] programs and time-sensitive
targeting.
And, boy, that seemed to work great, but as was mentioned
earlier, that is not the norm. That is kind of a one-off where
you have got the warfighter and the people who are developing
the technology for command and control. Which is basically
about collaboration, real-time decision quality information.
This is not rocket science. I mean, this is just allowing the
information to be collaborated for the exact type of mission
that you are looking for and adapted, and that worked really
well. But that is really not the norm.
You know, just basic geospatial information we were trying
to develop in the Joint Operations Center at AFRICOM [U.S.
Africa Command] and had that vision, but it just seemed like we
were dealing at the speed of bureaucracy instead of broadband.
And when it comes to some of these other issues with command
and control, collaboration, information sharing, there is some
great stuff that is, obviously, way out in front of us in the
civilian world, in the private sector.
What do we need to do in order to very quickly bring that
in to make sure that, you know, we are allowing our command and
control system to not be bogged down? Because it was a quite
painful experience that I went through in both the Air
Operations Center, and the Joint Operations Center, just not
being able to adapt quickly enough. And anybody want to jump
in?
Mr. Hunter. Well, I would say one thing is that there are
real impediments in the system that make it very hard to do
that. And we have talked a bit here about the agile approach to
acquisition. And there is a real challenge to utilizing that
approach, and I mentioned, I have talked about it as adaptable
systems, which is the system is designed to say, give me a
clear baseline, everything you are going to do, and then I will
grade you as to whether you have met that baseline or not, or
whether you have gone over. And if you have gone over, I am
calling you for cost growth or schedule growth.
And there is, now, and I have said a lot of things that I
like that are in the pending NDAA [National Defense
Authorization Act]. There is one that I don't like, which is
the provision that says, we are going to penalize the services
year after year after year, if they experience any growth above
baseline. Well, if you have got a fragile system, or adaptable
system, of course, you are going to grow the baseline. That is
the whole idea. That is the point.
But effectively, you know, this provision is going to make
it so that the services, in order to do that, it is going to be
like going over the salary cap for an NFL [National Football
League] team or an NBA [National Basketball Association] team.
They are going to have to pay a penalty every year because they
are trying to do something to make the system better. And I
think that there are many other barriers because it becomes
very hard to baseline these programs where you know you are
going to evolve them, but you don't know exactly how yet. So
that is a real issue that we need to work through, and I talk
in my testimony about we need to come up with a new paradigm,
not for everything, but for some of these systems that we think
we need to be highly adaptable.
Ms. McSally. Yeah.
Mr. Pasqua. There is an approach in industry called MVP,
for minimum viable product, and the whole idea there is, don't
build the be-all, end-all. Don't boil the ocean. Build what is
actually needed to accomplish whatever it is you are trying to
accomplish. And build the minimum thing that is needed.
Because, in fact, you are not going to know all the details of
how it is going to grow and how you are going to want to adapt
to use it.
So instead of trying to build the be-all, end-all, the goal
is much more to build the smallest thing that meets the
requirements with the adaptability to go in different
directions that you don't necessarily know today. And that has
two sort of beneficial outcomes.
One is, it happens fast. It is small. It tends to focus you
on what is really important rather than on contingencies that
may be important some day. And it gets you to focus on the
adaptability of the architecture that you are building, so that
as you actually use it and find what is important, or your
needs change, it is easier to actually take the system in the
direction you want to go in a much more cost-effective way.
Ms. McSally. Great. Thank you. Thank you, Mr. Chairman, I
yield back.
The Chairman. Thank you. Mr. Langevin.
Mr. Langevin. Thank you, Mr. Chairman. I want to thank our
witnesses for your testimony and appearance here today. It has
been a very interesting discussion, obviously.
So I serve as the ranking member on the Subcommittee on
Emerging Threats and we oversee all of our R&D efforts, DARPA
[Defense Advanced Research Projects Agency], and ONR [Office of
Naval Research], and others. And just in S&T directly, you
know, the basic nature of S&T development means that several
failures often proceed as successful technology. DOD culture,
and acquisition processes, and congressional oversight are
often risk-averse at best, and punish failures at worst. How
can this culture be changed or mitigated, or should it?
Mr. Francis. Well, Mr. Langevin, I would say definitely
making the investments in the science and technology community,
and giving them the budgets to take things further, so I think
you have to have a wide funnel in the front. I think you want
to have a lot of failures early, and then as you are paying for
more mature technologies and higher levels of demonstration,
obviously, you have to be more discriminating there.
But the S&T community, I think, should have the
organization and the resources to take those technologies
further up and be okay with having those failures early. Right
now, I think it is not so much that we are afraid of failures,
but we put things that haven't reached the point of failure
yet, and we put them in an acquisition program, and then we
discover what the failures are.
So I don't have really a problem. We were talking earlier
about the carrier. I don't have a problem with what those key
technologies are going through. The problem I would have is,
where they are going through them. It is right during
construction of the ship. So those are the risks that we have
to take earlier.
Dr. Chu. Indeed. I would agree that it is not clear to me
it is a risk-averse culture. In fact, we take the wrong set of
risks, I think is what Mr. Francis is arguing. And we
underinvest in technology development, without--or let me put
it the other way around--that too often we see technology
development as always needing to lead to a new system, and that
is not always going to be true.
I think more willingness to sort through the technological
choices in an organized way and to reward people for giving
good advice about which are the promising paths versus the ones
that should be shut off. In the current incentive system,
managers talk about the value. Everybody thinks he has to get
his or her technology into production. That is the sign of
success.
Now, I would argue differently. Success is having a broad
portfolio of choices to start with and narrowing down to the
most promising ones. That does include, as Mr. Francis I think
has emphasized to you, much more emphasis on developmental
testing than has been true in the recent past.
Mr. Hunter. One thing I would add in terms of risk is that
we can do better at managing risk. So I would agree. I don't
know that our system is unwilling to take risk, but it does
struggle to manage risk. And in many cases, you know, you will
see these risk charts, you know, and there is always one item
that is either high yellow or red, and everything else is kind
of green or in the mid-yellows, and they all look roughly the
same. Because there, really, again, it is in some cases,
unfortunately, more about selling the system than it is about
managing the risk.
And that requires real discipline, and this is where the
quality of the workforce comes in. So that the government
workforce really understands what the risk is, and what is the
plan to manage it, to burn it down over time. And I think the
biggest key there is leadership and then the human capital
issues that Dr. Chu has referred to.
Mr. Langevin. A follow-up question. What changes are needed
to allow for a rapidly changing investment area such as
cybersecurity, which I spend a lot of time on, where
generations of technologies can pass within a single budget
cycle, and to what extent do current budgeting processes impede
the deployment or development of technologies?
Mr. Pasqua. This was a particular frustration area for me.
I ran the global research organization at Symantec, which is
the largest cybersecurity firm in the industry. And one of the
challenges that we had, given the rapid pace of change of the
landscape, was that we develop new technologies in our research
organization and want to get those out and into the hands of
our government counterparts, but oftentimes the cycle of doing
that, just being able to discuss it and go through the process,
was so long that the window of opportunity for dealing with a
threat had passed by the time that we were through it.
And I always wished that there was a way for us to build a
relationship that didn't start and stop; that provided a way
for us on an opportunistic basis to say, hey, we have got an
interesting technology for you that we think is of interest for
you to get into service today or very soon. How can we make
that happen quickly and not have to start, you know, a whole
cycle of discussions to make that process happen that then made
the technology irrelevant 9 months later.
Mr. Langevin. I know my time is expired, but that was an
interesting question and response. I appreciate your thoughts.
I yield back, Mr. Chairman.
The Chairman. I thank the gentleman. Mr. Cook.
Mr. Cook. Thank you, Mr. Chairman. I apologize. I had to
meet some constituents, so I probably missed some of the
responses to maybe some of the questions that I am going to ask
right now. And one of the problems I have is just the, and I am
speaking as somebody that has been on the receiving end of some
of these systems where they don't work, or they break down, or
they just don't fit into the mission that you are supposed to
have.
And I don't have the complex rocket ships and everything
like that. I am talking about the M-16 when it first came out
in Vietnam where a lot of Marines died because you had a lot of
things wrong with it. And I actually had an opportunity to
change it by talking to a guy by the name of Omar Bradley who
had to be about 86 at the time where they dragged certain
people from the field. That, the Gama Goat, the M-203, which
was 5 years that I knew it was on deadline, never even saw it
fired. I can go on, and on, and on.
And I don't know if we--and then I was just at a CODEL
[congressional delegation] where somebody was asked a question
about cybersecurity, and they said, we are just starting now.
And in about 5 years, we are going to show it. We don't have 5
years. No one knows how long we are going to have with what is
going on in this world right now.
So I think from somebody that is on the frontline, the
troops, they want something right away, and something that fits
their mission. And you can have a lot of different--but
basically, it is to close with and destroy the enemy. And I am
wondering if we launched that philosophy in World War II where
we had certain systems that came out that didn't work, where
almost within weeks, we had changed it. Unbelievable that we
could do that.
And you look at what happens with the Sherman tanks in
Normandy, where it was a field expedient by a sergeant that
changed the whole thing, turned it into a Rhinoceros tank,
changed the whole battlefield almost overnight. What did it
cost, $15? And then they did it to all the tanks.
I can go on and on and on. But I think Congressman Walz had
a good point. Sometimes I think the individual troops or what
have you, the customer, the end recipient, the ones who have to
live and die with this system--Iraq, when we had to upgrade our
Humvees and the MRAP came down, great success, but in the
interim X amount of people died or were wounded.
And I am just saying, to me, I think we have got to
expedite that, and make it cost efficient and we have to put
certain deadlines. And if it doesn't work, there has to be
consequences. That is the bottom line. And some of these
systems we can do it.
When we changed from the old bazooka to the 3.5 rocket
launcher, sounds simple, but the bazooka was not able to
penetrate the Russian tanks that were made; the 3.5 was. Now,
3.5 is long gone. They replaced it with the LAW [Light Anti-
Tank Weapon], which was another piece of crap because it didn't
function in humid conditions, so we had all these things come
out to the system and what happens. All that stuff then went on
and on and on.
So I think we almost need to incorporate that philosophy,
what has happened in the past. And the best example I can give
or hope that you would look at are the Israelis. The Israelis
don't have time. You look at what they have done with their
missile programs. You look at what they did in the Yom Kippur
War that they almost lost and they changed certain things.
You look at the battleground, 2006 against Hezbollah along
the Lebanese border, where they changed their MPCs [military
personnel carriers] and tanks because the threat was there. And
they didn't have 5, 10, 15, 20 years. They had to do it or they
were going to not exist as a country.
And if you could comment very briefly, and I yak too much.
Dr. Chu. You have named some of the notable failures over
time. I do think part of the----
Mr. Cook. I am old, so I named all the old systems.
Dr. Chu. We did get better, actually, over time, I would
argue.
I think part of the solution is what others have advocated,
which is often from the field perspective the 80 percent
solution is good enough. And so one of the reasons for longer
times to solve the problem than is meritorious is we aim too
high. We ought to aim at, as I think your comments emphasize,
what is most essential for accomplishing that mission. What
does the troop really need in order to do a good job. If we get
that done, we could then add to that success in a more
evolutionary approach.
Mr. Cook. Thank you.
I yield back. Thank you.
The Chairman. Well, actually, I think that is a good way to
end, because DOD may be better than government agencies, but we
have also got more at stake. And so that is part of the reason
that I very much appreciate you all's input today.
I know we will continue to engage with CSIS and IDA and
BENS and GAO, but I want to encourage you all to continue to
offer us your input. Don't wait for us to ask. This, as you
know, is a complex subject with a lot riding on it, as we are
going to be in an iterative process to try to improve it. And
we need the assistance of people with valuable expertise and
insights to help us do that.
So I appreciate today, and I appreciate your contributions
in the future.
With that, the hearing stands adjourned.
[Whereupon, at 12:03 p.m., the committee was adjourned.]
=======================================================================
A P P E N D I X
October 27, 2015
=======================================================================
PREPARED STATEMENTS SUBMITTED FOR THE RECORD
October 27, 2015
=======================================================================
[GRAPHIC(S) NOT AVAILABLE IN TIFF FORMAT]
=======================================================================
DOCUMENTS SUBMITTED FOR THE RECORD
October 27, 2015
=======================================================================
WITNESS RESPONSES TO QUESTIONS ASKED DURING
THE HEARING
October 27, 2015
=======================================================================
RESPONSE TO QUESTION SUBMITTED BY MS. SPEIER
Mr. Francis. The decision to start a new program is the most highly
leveraged point in the product development process. Establishing a
sound business case for individual programs depends on disciplined
requirements and funding processes. Key enablers of a good business
case include: firm, feasible requirements, mature technology,
incremental, knowledge-based acquisition Strategy, and realistic cost
estimates.
Every year, there is what one could consider a ``freshman'' class
of new acquisitions. This is where DOD and Congress must focus to
ensure that programs embody the right principles and practices and make
funding decisions accordingly. Congress will need to focus on oversight
of programs in the President's Budget projected to begin Engineering
and Manufacturing Development phase by holding a Milestone B decision.
What that means, is, for example, for a program with a projected June
2017 Milestone B, the funding for that program will be in the
President's Budget presented in Feb 2016. Congress will need to have
started its oversight of this investment before that budget comes in
and must conclude before markup. This criterion would provide a list of
programs that Congress can most influence. Based on that approach, as a
starting point, congressional oversight could focus on ensuring sound
acquisition strategies using knowledge-based acquisition principles are
established for the following three programs projecting Milestone B's
in fiscal year 2017: (1) Presidential Aircraft Recapitalization; (2)
Unmanned Carrier-Launched Airborne Surveillance and Strike System; and
(3) Joint Surveillance Target Attack Radar System Recapitalization.
In addition, our March 2015 assessed ongoing programs against
acquisition best practice criteria to identify specific acquisition
risks. The risks we reported provide an opportunity for targeted
congressional oversight of programs already underway. We will be
publishing our next annual assessment of selected weapons programs
later this spring which will include updated assessments. [See page
29.]
?
=======================================================================
QUESTIONS SUBMITTED BY MEMBERS POST HEARING
October 27, 2015
=======================================================================
QUESTIONS SUBMITTED BY MR. THORNBERRY
Mr. Thornberry. What can be done to encourage more incremental
development and deployment of weapon systems?
Mr. Hunter. Some of the concepts required for incremental
development, such as planning and developing ``Blocks'' of capability
that are separately or sequentially developed over time and programs to
prototype new systems and subsystems, already exist. An example is in
the F-35 program, in which the Department has been planning and
developing Block 4 capabilities for F-35 even as it works to develop
and deliver Block 2B capabilities that are being used by the United
States Marine Corps and Block 3F capabilities that are being used by
the United States Air Force in their initial operation of the F-35.
Another example is the separate development of mission modules for the
Littoral Combat Ship (LCS). However, the approach to blocks of
capability that prevail in today's programs remains focused on
developing and delivering full system sets of capability that are
tightly integrated, and that as a result, take many years to plan,
program and deliver. While this approach is faster and probably cheaper
than having entirely separate acquisition programs, it still does not
match up well with the pace of technology innovation. It is necessary
to supplement the Block approach by decomposing system level
requirements into smaller sub-increments or ``packages'' of capability
that correspond to the subsystem design level. This would allow the
incremental development of these packages of capability that can be
designed, prototyped, tested, and relatively rapidly deployed, in
months rather than years, allowing for the incremental development and
deployment of technology at the subsystem level. A potential barrier to
this approach is in the challenge it presents to ``baselining'' systems
for purposes of budgeting, scheduling, testing, and tracking program
execution. Congress should work with the Department to establish a more
dynamic approach to baselining acquisition programs where incremental
development and rapid deployment are identified as priorities.
Mr. Thornberry. What must change in the DOD acquisition process and
culture to better transition technology from the ``lab'' to a program
of record?
Mr. Hunter. The key gap in the current system for transitioning
technology from the lab to programs of record and fielded system is the
fact that while experts within the DOD enterprise have tremendous
awareness and knowledge of technology developments (developments in the
``lab'') across the full spectrum of the in-house and industry-led
technology sectors, this knowledge is not spread broadly across the
enterprise, and in particular, it is frequently not the case that
decision makers in the requirements and resourcing communities have
knowledge of the latest technology developments. So there is a gap in
turning expert knowledge into enterprise knowledge that can be acted on
by DOD leadership. Similarly, the number of opportunities for
technology insertion in programs of record and fielded systems are too
far apart. In the worst case scenario, a program could go from
Milestone B all the way to full rate production and Full Operational
Capability (FOC), usually a period of roughly ten years, without
significant consideration of technology insertion beyond the minimum
needed to address issues of obsolescence. Many programs work to create
some additional technology insertion points in between these major
program phases, but they are essentially working against the system in
doing so. In contrast, the pace of technology change suggests that
technology insertion points need to occur no less frequently than every
six months. The Department traditionally addresses this need to create
additional opportunities for technology insertion by creating
``Blocks'' of capability that deliver incrementally over time. However,
in most cases these blocks are themselves separated by several years of
time. Creating additional opportunities for technology insertion
requires allowing the requirements, budgeting, and acquisition
processes to decompose system-level requirements into sub-increments,
``packages,'' such that these sub-increments could be swapped out or
updated independently on a timeline of months rather than years.
Mr. Thornberry. What is needed to achieve integration of the
requirements and acquisition processes in the DOD to facilitate
requirements tradeoffs prior to acquisition programs being initiated?
Mr. Hunter. The requirements and acquisition communities need to be
in continuous dialogue on trade-offs relating to acquisition programs
before program initiation and during program execution. While this
dialogue exists in many ways today, it can be hampered by the fact that
both sides are not always speaking the same language (in a figurative
sense). That is, the requirements community is talking in terms of
capability gaps and key performance parameters (KPPs), concepts that
are fairly absolute and unqualified, while the acquisition community is
talking in terms of costs, risks, and timelines for development, things
that before program initiation are estimates that are inherently
uncertain and imprecise. For the dialogue to be productive, a bridge
between these languages is need. In the world of rapid acquisition,
this dialogue was bridged by the concreteness of short fielding
timelines. Both sides could work backwards from an expected fielding
date as a basis for understanding how to characterize bottom-line
needs, in the case of the warfighter, and the art of the possible, in
the case of the acquisition community. An interesting example of how
this was done on a major defense acquisition program was with the
Combat Rescue Helicopter program, where Under Secretary of Defense for
Acquisition, Technology, and Logistics, Frank Kendall, asked the
requirements community to put a dollar value on how much more they
would be willing to pay to get a capability that met their objective
requirement, rather than just the threshold requirement. The
acquisition community was surprised that the requirements community
indicated that they would willing to pay only about 10% more to obtain
the increased capability of the objective requirement. When this value
criterion was incorporated into the Request for Proposals for the
program, it had a decisive effect on industry's evaluation of how to
compare cost with capability. In the CRH example, the bridge in the
conversation came from developing a concrete measure of value in the
difference between the threshold and objective requirement.
Mr. Thornberry. To what extent is government systems engineering
expertise available and applied in the requirements-setting process? To
what extent should it be available and applied?
Mr. Hunter. Systems engineering is fundamental to the process of
flowing down requirements in systems design. In this sense, it informs
the ``requirements'' that flow down in the design process from the
system level to the subsystem level and on down through the design
process. However, systems engineering does not play a large role in the
setting of KPPs in the systems-level requirements process as far as I
am aware. I can think of one example in my direct experience, on an
Army truck program, where the maintainability requirements were
modified after initial review in the Defense Acquisition Board (DAB)
when it was brought forward by the test and systems engineering
communities that the maintainability requirements set for the program
exceeded any previously demonstrated standard for similar Army
equipment by an order of magnitude. In this way, the systems
engineering work that is done in support of the DAB process can, and
has on occasion, been used as an impetus to revisit the requirement,
and in the case of the Army truck program, the Army requirements
community revisited that requirement and established KPPs more in line
with previous Army experience. If the requirements process is modified
as suggested in my answers above to decompose requirements that allow
for more incremental development at the subsystem level, systems
engineering would have to play an increased role in the requirements
process to ensure that system integrity is not compromised.
Mr. Thornberry. What capabilities or efforts may be needed to
streamline the DOD's requirements and acquisition review and approval
processes to increase ``speed to market'' or getting the right weapon
systems from concept to fielding?
Mr. Hunter. In addition to turning expert knowledge about
technology developments into enterprise knowledge that can drive
action, and decomposing system-level requirements and adding additional
technology insertion points to enable incremental development, review
and approval processes should move from processes that focus on
examining documents in sequential fashion for procedural compliance to
processes with greater direct interaction and involvement of the
various stakeholders focused on developing a common understanding of
the strategy being pursued, the risks being taken, the plans for risk
mitigation, and the benchmarks of success. The goal should be for the
acquisition community to leave these sessions prepared to support the
program manager in pursuing the program, rather than a situation where
the larger acquisition community observes from the sidelines looking
for stumbles.
Mr. Thornberry. Are there lessons learned from the Department's
rapid acquisition programs that can be applied to accelerate other DOD
acquisition programs?
Mr. Hunter. Absolutely. Chief among these is the need for
continuous dialogue between the acquisition and requirements
communities that updates and informs those setting requirements on the
likely costs, timelines, and results of development efforts and that
allows the requirements community to rapidly inform the acquisition
community about emerging threats and to refine requirements as
additional information about technology developments emerges. This
ability for the two sides to meet regularly and exchange information
regarding urgent operational needs was a powerful mechanism for
accelerating action and it is applicable in many ways to other kinds of
acquisition programs. Another key for rapid acquisition was the support
of senior leadership to overcome obstacles by identifying and
transferring funding ahead of the normal funding cycle, approving
waivers or taking extraordinary action to acquire long-lead items when
necessary, and alerting other offices throughout the Department to move
rapid acquisition programs to the front of the queue whenever approvals
were required. Because it is difficult for senior leadership to play
this sort of role in a large number of cases, it was critical that the
rapid acquisition process was associated with a discrete, definable
universe of urgent operational needs with a proven and credible
requirements approval process. Lastly, the availability of flexible
funding was essentially to avoiding the long delays associated with
obtaining funding for new start programs, a process that generally
delays even the initiation of action on new efforts by at least two
years.
Mr. Thornberry. What can be done to encourage more incremental
development and deployment of weapon systems?
Dr. Chu. Incremental development and deployment, as the Committee
appreciates, is a design and acquisition strategy. The decision to
employ it must weigh its benefits, and their likely realization,
against the costs, and the realism of those cost calculations. One
benefit may be more rapid fielding, depending on the circumstances
involved; another may be the ability to adjust the article's features
in response to early field experience or evolving threats; yet another
could be capitalizing promptly on research progress. One downside may
be the additional complexity (and cost, e.g., for training) that
deploying a variety of models could entail.
Accepting the premise that benefits will at least sometimes
outweigh costs (certainly true of several major system upgrades over
the last two decades, which constitute one version of incremental
development), it may be sufficient to ensure that this alternative
strategy is one of the options considered by the Analysis of
Alternatives that should precede any major investment decision. The
evidence on benefits and costs should be sufficiently persuasive to
make the case for selecting the strategy.
Acknowledging that systems will change in response to early field
experience will reinforce that case. Indeed, anticipating the need for
changing configurations will reinforce the case for an incremental
approach. But it does require that the design effort facilitate such
changes (for example, by providing larger margins for weight growth, or
space and power for additional features). Two of my colleagues make the
case for just such an approach in their aptly titled paper, ``Prepare
to Be Wrong''.\1\
---------------------------------------------------------------------------
\1\ Patel, Prashant R. and Michael P. Fischerkeller, ``Prepare to
Be Wrong: Assessing and Designing for Adaptability, Flexibility, and
Responsiveness'', ISA P-5005, April 2013.
---------------------------------------------------------------------------
As Patel and Fischerkeller argue, this approach advantages the
adaptability of equipment as circumstances change. In its 2010 Summer
Study, the Defense Science Board recommended that we tie program
objectives to planned deployment dates, to buttress just such
adaptability. Doing so also places a premium on constraining
development objectives, in order to enhance the likelihood that the
needed schedule will be honored.\2\
---------------------------------------------------------------------------
\2\ See Report of the Defense Science Board 2010 Summer Study on
Enhancing Adaptability of U.S. Military Forces: Part A: Main Report.
Washington, DC: Under Secretary of Defense, Acquisition, Technology,
and Logistics (USD[AT&L]), January 2011, p. viii-x.
---------------------------------------------------------------------------
In short, three managerial tools may produce a greater use of
incremental development: explicit consideration in AoAs, generous
margins for changes, and using deployment dates as a disciplinary
instrument.
Mr. Thornberry. What must change in the DOD acquisition process and
culture to better transition technology from the ``lab'' to a program
of record?
Dr. Chu. In my judgment, the fundamental problem lies in the
current incentives facing the potential acquisition partners--
government labs, the Military Services, the Defense Advanced Research
Projects Agency, and private industry. While all generalizations are
risky, the current incentives for government labs focus their efforts
on improving existing equipment, supporting existing Service concepts
of operation, not on generating new ideas per se. The Services, which
control the ``requirements'' process, tend to focus on their views of
operational needs, sometimes only vaguely informed by what might be
technologically feasible. To the extent the Services consult the
technical community, beyond their immediate staffs and those of their
OSD overseers, the exchange centers on the dialogue with private
industry via development contracts. It's well established that industry
sees the development contract as an economic ``prize'', leading to the
source of most profit in the American system--the production of
finished articles.\3\ Provided the proposed development contract
responds reasonably to ``requirements'', there is no particular
incentive for private industry to work with the labs.
---------------------------------------------------------------------------
\3\ See William P. Rogerson, ``Profit Regulation of Defense
Contractors and Prizes for Innovation'', Journal of Political Economy,
Vol. 97, No. 6, December 1989, pp. 1284-1305.
---------------------------------------------------------------------------
Accepting these generalizations as broadly applicable, if there is
a desire for government labs to play a larger role in the development
process, incentives must be adjusted. A mechanism to reward the labs
for generating successful ideas is required, as a necessary condition.
In the private sector for civilian products, that mechanism is the
return that accrues from intellectual property. A simplistic analogue
for the government lab would be some financial return for
``successful'' ideas.
But that alone would be insufficient, without a mechanism that
encourages a dialogue between the government labs and the Military
Services, and among the labs, the Military Services, and the producing
contractors. Could some early development work be awarded the labs,
presumably based on the potential excellence of their early ideas
(which by itself might provide a needed financial incentive)? Would it
be feasible to create partnerships between government labs and
production firms that did not generate undesirable conflicts of
interest? Might one form of partnership be development of operational
prototypes embodying new technological approaches? Could that enhance a
culture of experimentation, using experiments with prototypes to
sharpen the appreciation of both real-world limits and the tradeoffs
that must be confronted? Would that also help DARPA test its best
ideas?
If successful, such changes in incentives would change the routine
behavior of the acquisition process participants. But it is also
possible to over-ride current routine behavior using the Secretary's
(or Congress') authority, nurturing promising technologies that the
labs might develop until they take root. It might be argued that is how
cruise missile technology became such an important part of the defense
portfolio. In the best of all possible worlds such nurturing might
bridge the transition to a better set of incentives, worked out through
trials of the sorts of ideas sketched above, both to explore their
feasibility and yield insights into the unintended consequences against
which the Department must protect itself and the public interest.
Mr. Thornberry. What is needed to achieve integration of the
requirements and acquisition processes in the DOD to facilitate
requirements tradeoffs prior to acquisition programs being initiated?
Dr. Chu. ``Requirements'' cannot be separated from the ``physics''
of the problem--that is, the trade space among potential attributes for
a system that technology provides. The technology constraint (the
``frontier'') may not be a bright line, but rather more likely a fuzzy
zone. As you approach it, and perhaps try to move toward its outer
boundary (or beyond), costs and risks increase, arguably
substantially--even in a nonlinear fashion.
With a healthy respect for where that fuzzy zone begins, it's
typically feasible to depict the tradeoff space among attributes of a
potential system quantitatively. Approaches to do so are available.\4\
They are not extensively utilized, but should be. Effective utilization
will require appropriate ``human capital'' (perhaps capitalizing on
what some of the Federally Funded Research and Development Centers can
provide)--and a bureaucratic process sympathetic to their intent. The
last will require significant leadership from the Department's seniors.
---------------------------------------------------------------------------
\4\ See, for example, Patel, Prashant, et al, ``Defining
Acquisition Trade Space Through `DERIVE' '', IDA Research Notes,
Acquisition, Part 1: Starting Viable Programs, Fall 2013.
---------------------------------------------------------------------------
A complementary approach is to re-invigorate the use of Analyses of
Alternatives (which will also require leadership from the Department's
seniors). The alternatives considered could include alternative bundles
of the desired system attributes, thus illustrating the benefits, costs
and risks of the tradeoffs involved. AoAs should respect the
uncertainties of future operational environments, and of the fiscal
limits under which the proposed system solution must be pursued,
constraining the tradeoff choice to one that is realistic in the
context of likely future budgets.\5\
---------------------------------------------------------------------------
\5\ The Army's critical self-examination of its acquisition
system's performance noted the ``weak'' use of AoAs. See Army Strong:
Equipped, Trained and Ready: Final Report of the 2010 Army Acquisition
Review, January 2011, p. x (sometimes known as the Decker-Wagner study,
after its two co-chairmen, Gilbert F. Decker and Louis C. Wagner).
---------------------------------------------------------------------------
A final promising ingredient is to encourage a series of
``feedback'' exchanges with those who will use the proposed system,
requesting their vision(s) of how the system might be employed,
recognizing the limitations of their abilities to foresee how that
might actually develop. (The challenge is reflected in the allegation
that had we asked cargo users what they wanted at the dawn of the
automotive age, they might have replied ``a faster, stronger mule''!)
Mr. Thornberry. To what extent is government systems engineering
expertise available and applied in the requirements-setting process? To
what extent should it be available and applied?
Dr. Chu. The Weapons System Acquisition Reform Act of 2009
appropriately mandated a resurrection of DOD developmental testing
capacity, and improvements in systems engineering capacity, but the
full realization of WSARA's vision will require yet more effort. These
two capacities should be foundational elements in the early DOD
Milestone deliberations--they should have proverbial ``seats at the
table'', perhaps more elevated seats than they currently enjoy.
How much capacity is needed will differ by warfare area. Some of
the needed capacity may be provided by the Federally Funded Research
and Development Centers. There is unlikely to be an easy generalization
about overall needs.
Applying these capacities energetically will require the commitment
of the acquisition community leadership, accepting the tensions that
competing perspectives can generate. All concerned must be willing to
accept some of the ``hard truths'' that these communities so often
provide--especially about competing or incompatible ``requirements''.
Mr. Thornberry. What capabilities or efforts may be needed to
streamline the DOD's requirements and acquisition review and approval
processes to increase ``speed to market'' or getting the right weapon
systems from concept to fielding?
Dr. Chu. As I testified, cycle time (presumably the motivation for
streamlining) may not be quite the issue we imagine. But as I also
testified, many of the oversight and reporting burdens are symptoms of
the problem we face, not the cause. Government decisions, especially
the high-profile decisions associated with major weapon systems, are
inherently political. Given the decided lack of sympathy for error, we
should not be surprised that managers at every level demand
significant, careful review before each step is taken. In my judgment,
a greater political tolerance for ``mistakes''--an ``error budget'', so
to speak--would eventually allow us to streamline the process.
But we can also speed delivery by more frequently considering
alternatives that update existing systems instead starting afresh.
Modifying what we already have to aim at the desired performance (or
cost) improvement should be a consistent option in Analyses of
Alternatives. The relative success of the F/A-18 E/F program provides
an example worth considering. It may also illustrate the limits of this
approach.
We can likewise start with articles developed outside the United
States. The rapidity with which we could deploy the Mine Resistant
Ambush Protected Vehicles illustrates the potential of this
alternative--as well as its limitations.
And perhaps most important, if we honor the limits of what's
technically realistic in the design tradeoff selected, as I argue in
response to Question 3, and if we are faithful to what's really needed
in systems engineering and developmental testing (Question 4), we are
much more likely to field articles expeditiously--articles that shine
in their operational tests, and that live within the cost limits we
need to impose. Indeed, as I speculate in response to Question 1, if we
start with the desired fielding date as a key parameter driving program
design, allowing adequate margin for incremental improvements, we may
reach the result to which so many aspire--acknowledging that there are
some technical developments that may require long periods of
investment, notwithstanding the preference for quick results, as the
history of the pure power geared turbofan aircraft engine
demonstrates.\6\
---------------------------------------------------------------------------
\6\ See Coy, Peter, ``The Little Engine that Could Reshape the Jet
Engine,'' Bloomberg Businessweek, October 19-25 edition.
---------------------------------------------------------------------------
Mr. Thornberry. What can be done to encourage more incremental
development and deployment of weapon systems?
Mr. Pasqua. Massive improvements in incremental development and
deployment practices have been a major contributor to rapid growth in
the Tech Industry over the past decade. There are several contributors
to these improvements. First, there has been a move to the notion of
``minimum viable product'' or MVP. A MVP contains only those features
which are required to meet key requirements while providing a basis on
which new features may be added over time without re-engineering the
entire platform. This allows products to get to market sooner, provide
tangible customer value, garner feedback, and be enhanced to provide
enhanced functionality that may not have been anticipated or
prioritized before real usage. To make the MVP approach work, one must
embrace a continuous deployment process. That is, it must be possible
to enhance the deployed product rapidly without requiring major
infrastructure changes or end-user retraining. Because this can be more
difficult to do with hardware systems, more and more systems are
``software defined''. This allows the hardware to provide a flexible
substrate that can be molded to meet new requirements with software.
Incremental development can't work if the specs/requirements are
overblown to begin with or if there is no way to practically deploy
incremental improvements.
Mr. Thornberry. What must change in the DOD acquisition process and
culture to better transition technology from the ``lab'' to a program
of record?
Mr. Pasqua. Navigating a large bureaucracy can be difficult, but it
can be achieved by finding a customer that demands a product. This
takes the acquisition process out of the realm of the theoretical, and
is achieved by strengthening the bond between the customer and the
``lab'' or the developer. There are two primary ways to achieve this.
First, an organization can allow the lab to have greater access and
contact to the end-users in the field. This may not be feasible,
however, for a variety of reasons, including the risk of fielding
unproven technology. Therefore, the second way is to bring the end-
users into the lab. Often an end-user doesn't actually know what he
requires until a prototype is developed. Building a stronger connection
to the lab allows the customer to provide rapid feedback on prototypes
as they are developed. One of the benefits is not only the real time
test and evaluation, but also the dialogue on what is possible. The
operational work force may not have the technical knowledge to
articulate challenges to consider for technical solutions because they
are not informed on what are the current outer limits of technical
capability. Bringing the lab to the field reduces that knowledge gap by
bringing the science and technology closer to the problem, educating
the operational work force on the realistic bounds of current
capabilities, and illuminating operational challenges for the science
and technology work force to provide vectors for innovation. Thus,
better transitioning technology from the lab to the customer can be
achieved by strengthening the bond between the two; in essence:
bringing the lab to the field, or vice versa.
Mr. Thornberry. What is needed to achieve integration of the
requirements and acquisition processes in the DOD to facilitate
requirements tradeoffs prior to acquisition programs being initiated?
Mr. Pasqua. In larger organizations, often requirements are set and
given to the developers without the latter being included in the
requirements process. A stronger and more responsive feedback loop
between the requirement setter and the developer is necessary to
identify and reconcile any tradeoffs before a product is engineered and
acquired. Often, organizations conduct market research to inform their
requirements for a product. However, while all of those requirements
may be true, they may not all be essential. Yet, because the developers
are left out of the requirements process, they will engineer according
to the requirements they are given, regardless of need (i.e. perhaps a
70% solution is sufficient) or cost (i.e. perhaps a 100% solution would
double the cost of the program). Absent a strong link between need and
cost, there is a risk of over-engineering a solution or spending too
much. To avoid this challenge, there should be a responsive feedback
loop between the requirements setter and the developer. It should be
part of the developer's (or Research & Development team's)
responsibility to meet with the customer and agree on the need and cost
before embarking on the acquisition process. Often, the larger the
organization is the more specialization there is. This specialization
only creates more layers between the developer and the customer, and
risks weakening the link between need and cost.
Mr. Thornberry. To what extent is government systems engineering
expertise available and applied in the requirements-setting process? To
what extent should it be available and applied?
Mr. Pasqua. Although I am not in a position to comment on the
availability of systems engineering expertise available within the
Department, I can say that in the private sector the importance of
having the expertise available is growing. The rapid deployment and
iterative development of products and systems requires this expertise.
In industry, we struggle with the question of how to build more modular
and adaptable platforms that allow us to incorporate innovation over
time. This is a systems engineering issue. For example, open
architecture is now part of the lingua franca of software development
and implementation. The same can and should be said for major
Acquisition Category (ACAT) I programs. My sense is that should systems
engineering expertise been available and applied during the acquisition
cycle for the now cancelled USMC Advanced Amphibious Assault Vehicle
(AAAV) program, outcomes may have been very different. As requirements
grew in the development phase to ultimately unsustainable levels, a
system engineering perspective could have identified the technological
and fiscal impracticality of proposed adjustments before they were
articulated to the contracted developer.
Mr. Thornberry. What capabilities or efforts may be needed to
streamline the DOD's requirements and acquisition review and approval
processes to increase ``speed to market'' or getting the right weapon
systems from concept to fielding?
Mr. Pasqua. Industry looks like it is moving quickly because we are
seeing the aggregate progress made by an entire ecosystem of
organizations. Individually, many projects within these organizations
fail, but the overall effect is fast forward progress spurred on by
competitive pressure. The DOD may benefit from an approach that allows
more small scale experiments to occur quickly to ensure the feasibility
of new approaches and their suitability to the requirements. There will
be more small scale failures, but the ultimate result will be faster
time to fielding of capabilities that are best suited to the needs. In
some cases a ``minimum viable product'' approach should be employed
that allows a capability to be fielded sooner and improved later.
Mr. Thornberry. What additional steps are needed to ensure industry
understands what the Department's capability needs are for the future?
Mr. Pasqua. Beyond Defense Contractors and specialized commercial
organizations, there is very little understanding of the Department's
capability needs, how to learn them, and how to navigate the
procurement process. Unless the last item is addressed, industry won't
feel a strong need to overcome the first two items. Having said that,
outreach events including both academia and industry can be quite
helpful. Establishing a network of individuals and organizations from
industry who are co-sponsors of these events can help to attract the
right attendees. For example, the Venture Capital community can be
leveraged to bring new innovative companies from their portfolios to
sessions to learn more about the Department's needs. Organizations like
BENS also have members with broad networks who could also help attract
key participants. In addition to understanding the Department's
capability needs, I believe it would be valuable for industry to
understand more about the missions to which these capabilities are in
service. I find that once people understand the importance of the
mission, they are more motivated to find a way that their organizations
can contribute.
Mr. Thornberry. What can be done to encourage more incremental
development and deployment of weapon systems?
Mr. Francis. Greater discipline by DOD when setting requirements
and when establishing business cases, as well as reinforcement through
congressional oversight will be needed to encourage more incremental
development and deployment of weapon systems.
DOD will need to better integrate the requirements development and
acquisition processes so that trade-offs informed by systems
engineering take place before programs start. This will require a
recognition by DOD officials that requirements cannot be truly set and
a sound, incremental business case established until the requirements
technical feasibility and affordability can be fully determined. Our
recent work shows that DOD officials appear to recognize the need to
take additional steps. We reported in June 2015 that ``Several service
chiefs noted that more integration, collaboration, and communication
during the requirements and acquisition processes needs to take place
to ensure that trade-offs between desired capabilities and expected
costs are made and that requirements are essential, technically
feasible, and affordable before programs get underway''.
In addition, every year, there is what one could consider a
``freshman'' class of new acquisitions. DOD and Congress must ensure
that these programs embody the right principles and practices, such as
incremental acquisition strategies, and make funding decisions
accordingly. Through our reports and testimonies we have determined
that a key enabler to getting better acquisition outcomes is
establishing an incremental, knowledge-based acquisition strategy.
However, there are strong incentives within the acquisition culture to
overpromise a prospective weapon's performance while understating its
likely cost and schedule demands. Encouraging more incremental
development and deployment of weapon systems will take the joint
efforts of Congress and DOD. As I recently testified, the principles
and practices programs embrace are determined not by policy, but by
decisions. These decisions involve more than the program at hand: they
send signals on what is acceptable. Programs that present well-informed
acquisition strategies with reasonable and incremental requirements and
reasonable assumptions about available funds should be given credit for
a good business case and funded. Similarly, a few healthy ``No's'' by
DOD decision makers and Congress to programs that request to begin
without a sound foundation, including incremental approaches, would go
a long way toward shaping the expectations of programs and contractors
as to what is acceptable.
Mr. Thornberry. What must change in the DOD acquisition process and
culture to better transition technology from the ``lab'' to a program
of record?
Mr. Francis. DOD has long noted the existence of a chasm between
its science and technology community and its acquisition community that
impedes technology transition from consistently occurring. This chasm,
often referred to by department insiders as ``the valley of death,''
exists because the acquisition community often requires a higher level
of technology maturity than the science and technology community is
willing to fund and develop. We have reported extensively on shortfalls
across DOD's technology management enterprise in transitioning
technologies from development to acquisition and fielding. In June
2005, we found that DOD technology transition programs faced challenges
selecting, managing, and overseeing projects, and assessing outcomes.
In September 2006, we found that DOD lacked the key planning,
processes, and metrics used by leading commercial companies to
successfully develop and transition technologies. In March 2013, we
found that the vast majority of DOD technology transition programs
provide technologies to military users, but tracking of project
outcomes and other benefits derived after transition remained limited.
More recently in November 2015, we found that programs progress through
Defense Advanced Research Projects Agency (DARPA) without the agency
head fully assessing whether transition strategies make sense. Such
assessments, if measured against key transition factors, could improve
a program's potential for transition success. Transition
responsibilities then fall almost exclusively on individual program
managers, who are often not sufficiently trained to achieve the
favorable transition outcomes they seek. Further, when the program
manager's tenure expires, the primary advocate for transitioning the
program's technology is also lost. This turnover increases the need for
technical gains to be appropriately documented and disseminated so that
user communities have visibility into potential solutions available to
meet their emerging needs. An important part of this process is the
tracking of transition outcomes, as we recommended DOD undertake for
its technology transition programs in March 2013, and which we have
also found lacking at DARPA. Our recent November 2015 review of
technology transitions at DARPA offers a place to start. While there
are a number of factors that determine whether a technology effort
successfully transitions to a program of record, we found that science
and technology development organizations should regularly assess
technology transition strategies, improve transition training for
Science and Technology program managers, and increase sharing of
technical data on completed programs. Among the most significant
factors that contribute to transition success are whether there is
military or commercial demand for the planned technology, linkage to a
research area of sustained interest by DARPA, active collaboration with
potential transition partners and achievement of clearly defined
technical goals. Finally, there is the issue of money. Technologies and
concepts that are taken to a higher level of demonstration are more
likely to transition to programs successfully. This is key to success
in the private sector. But taking technologies to higher levels of
demonstration is expensive. It is often difficult for DOD labs to
afford such demonstrations. Conversely, programs of record have much
higher levels of funding available, which creates incentives to
transition technologies sooner than in the commercial world. This can
have negative consequences for transition. For example, a program may
be less willing to accept a technology that a lab has not been able to
fund to higher levels of demonstration. Also, technologies that do
transition early may cause problems for programs of record because they
will still be going through the discovery process associated with
higher levels of demonstration, with attendant discovery of problems
and complications. While normal for technology demonstration, this is
disruptive for a program of record that is operating within a formal
cost and schedule baseline.
Mr. Thornberry. What is needed to achieve integration of the
requirements and acquisition processes in the DOD to facilitate
requirements tradeoffs prior to acquisition programs being initiated?
Mr. Francis. This year's NDAA took several steps to increase the
role and formal authority of the service chiefs and ensure they are
consistently involved in program decisions. In addition, as I testified
to in 2013, DOD's better buying power initiatives are also having a
positive effect including making early trade-offs among cost, schedule
and technical performance requirements. However, more can be done. Most
current and former military service chiefs and vice chiefs GAO
interviewed from the Army, Air Force, Navy, and Marine Corps in our
June 2015 report collectively expressed dissatisfaction with
acquisition program outcomes and believed that the Department of
Defense's (DOD) requirements development and acquisition processes need
to be better integrated. Several service chiefs noted that trade-offs
informed by systems engineering must take place before programs start
so that requirements are better defined and more realistic cost,
schedule, and performance commitments can be made. GAO recommended that
DOD ensure sufficient systems engineering is conducted to better define
requirements and assess resource trade-offs before a program starts.
DOD concurred with the recommendations, citing recent policy changes.
To the extent that service chiefs will be taking an increased role in
program decisions, additional consideration should be given to whether
they have received sufficient training and experience in acquisition as
well as whether they have appropriate staff support to successfully
execute these additional duties. Finally, DOD will need to ensure that
it has sufficient workforce expertise to provide support during the
requirements setting phase. In 2015 we reported that ``. . . in areas
such as cost estimating and systems engineering, our work found that
DOD may not have adequate resources to fully implement recent weapon
system reform initiatives.''
Mr. Thornberry. To what extent is government systems engineering
expertise available and applied in the requirements-setting process? To
what extent should it be available and applied?
Mr. Francis. DOD often does not perform sufficient up-front
requirements analysis via systems engineering on programs to determine
whether the requirements are feasible and whether there is a sound
business case to move forward. Programs continue to be proposed with
unachievable requirements and overly optimistic cost and schedule
estimates and, usually, participants on both the requirements side and
the acquisition side are loathe to trade away performance. Almost all
of the service chiefs we interviewed in June 2015 stated that there is
a need to further enhance expertise within the government, and several
specified expertise in systems engineering. Several service chiefs
indicated that systems engineering capabilities are generally lacking
in the requirements development process, and do not become available
until after requirements are validated and an expensive and risky
system development program is underway. Some service chiefs advocated
that having systems engineering capabilities available to the military
services during requirements development could help to ensure earlier
assessment of requirements feasibility. The service chiefs' views on
the importance of systems engineering is consistent with our prior
acquisition work. We recommended that DOD ensure sufficient systems
engineering is conducted to better define requirements and assess
resource trade-offs before a program starts. DOD concurred with the
recommendations.
Mr. Thornberry. What capabilities or efforts may be needed to
streamline the DOD's requirements and acquisition review and approval
processes to increase ``speed to market'' or getting the right weapon
systems from concept to fielding?
Mr. Francis. The process used to manage the acquisition of weapon
systems is inefficient, cumbersome, and bureaucratic. A contributing
factor to this inefficient process is the significant time and effort
required to complete information requirements before an acquisition
program can proceed through a milestone to the next phase in the weapon
system acquisition process. DOD leadership has acknowledged that too
much time is invested in preparing for key milestones, including the
documentation and oversight of information required by statutes and
policy, which takes time away from conducting day-to-day core program
management tasks such as contractor oversight, engineering, and risk
management.
We surveyed 24 programs in February 2015 and found that it took DOD
over 2 years on average to complete the entire set of documents
required for review and approval at key decision points. In the end,
program officials felt almost half of these information requirements
were not of high value. The challenge is to find the right balance
between having an effective oversight process and the competing demands
such a process places on program management. If information
requirements and reviews are not clearly linked with the elements of a
sound business case and/or the key issues facing acquisitions today,
then they can be streamlined or even eliminated. If they are linked,
but are not working well, then they warrant re-thinking. These
requirements, as well as ones that take a year or more to complete,
could serve as a starting point for discussions on what documentation
is really needed for weapon acquisition programs and how to streamline
the review process. If DOD does not eliminate levels of review, but
only makes the existing process more automated, inefficiencies are
likely to continue.
We recommended that, in the near term, DOD identify and potentially
eliminate (1) reviews associated with information requirements, with a
specific focus on reducing review levels that do not add value, and (2)
information requirements that do not add value and are no longer
needed. We also recommended that, as a longer-term effort, select
several current or new major defense acquisition programs to pilot, on
a broader scale, different approaches for streamlining the entire
milestone decision process, with the results evaluated and reported for
potential wider use. DOD concurred with both recommendations. A place
for the committee to start would be to monitor DOD's progress
implementing these recommendations.
Users may well be willing to live with a less ambitious set of
technical outcomes than those at which the technical community is
aiming. Or they may want something more, reflecting changes in the
operating environment since designs were originally considered, sending
us back to the trade space drawing board.
[all]