[House Hearing, 118 Congress]
[From the U.S. Government Publishing Office]


                                   
                          [H.A.S.C. No. 118-44]

                   CAN IT WORK? OUTSIDE PERSPECTIVES 
                       ON DOD'S REPLICATOR PROGRAM

                               __________

                                HEARING

                               BEFORE THE

                     SUBCOMMITTEE ON CYBER, INFORMATION 
                         TECHNOLOGIES, AND INNOVATION

                                 OF THE

                      COMMITTEE ON ARMED SERVICES

                        HOUSE OF REPRESENTATIVES

                    ONE HUNDRED EIGHTEENTH CONGRESS

                             FIRST SESSION

                               __________

                              HEARING HELD

                            OCTOBER 19, 2023

                                     
[GRAPHIC NOT AVAILABLE IN TIFF FORMAT]

                              __________

                   U.S. GOVERNMENT PUBLISHING OFFICE                    
55-527                     WASHINGTON : 2024                    
              
-----------------------------------------------------------------------------------   
  


    SUBCOMMITTEE ON CYBER, INFORMATION TECHNOLOGIES, AND INNOVATION

                  MIKE GALLAGHER, Wisconsin, Chairman

MATT GAETZ, Florida                  RO KHANNA, California
LISA C. McCLAIN, Michigan            SETH MOULTON, Massachusetts
PAT FALLON, Texas                    WILLIAM R. KEATING, Massachusetts
DALE W. STRONG, Alabama              ANDY KIM, New Jersey
MORGAN LUTTRELL, Texas               ELISSA SLOTKIN, Michigan
JENNIFER A. KIGGANS, Virginia        JARED F. GOLDEN, Maine
NICK LaLOTA, New York                PATRICK RYAN, New York
RICHARD McCORMICK, Georgia           CHRISTOPHER R. DELUZIO, 
                                         Pennsylvania

               Caroline Kehrli, Professional Staff Member
               Michael Hermann, Professional Staff Member
                    Brooke Alred, Research Assistant
                            
                            
                            C O N T E N T S

                              ----------                              
                                                                   Page

              STATEMENTS PRESENTED BY MEMBERS OF CONGRESS

Gallagher, Hon. Mike, a Representative from Wisconsin, Chairman, 
  Subcommittee on Cyber, Information Technologies, and Innovation     1
Khanna, Hon. Ro, a Representative from California, Ranking 
  Member, Subcommittee on Cyber, Information Technologies, and 
  Innovation.....................................................     2

                               WITNESSES

Clark, Bryan, Senior Fellow, Hudson Institute....................     3
Greenwalt, William C., Nonresident Senior Fellow, American 
  Enterprise Institute...........................................     5
Scharre, Paul, Executive Vice President and Director of Studies, 
  Center for a New American Security.............................     7

                                APPENDIX

Prepared Statements:

    Clark, Bryan.................................................    29
    Greenwalt, William C.........................................    51
    Scharre, Paul................................................    67

Documents Submitted for the Record:

    Statement of the Association for Uncrewed Vehicle Systems 
      International..............................................    81

Witness Responses to Questions Asked During the Hearing:

    [There were no Questions submitted during the hearing.]

Questions Submitted by Members Post Hearing:

    Dr. McCormick................................................   103
    
.    
     CAN IT WORK? OUTSIDE PERSPECTIVES ON DOD'S REPLICATOR PROGRAM

                              ----------                              

                  House of Representatives,
                       Committee on Armed Services,
      Subcommittee on Cyber, Information Technologies, and 
                                                Innovation,
                        Washington, DC, Thursday, October 19, 2023.
    The subcommittee met, pursuant to call, at 9:00 a.m., in 
room 2118, Rayburn House Office Building, Hon. Mike Gallagher 
(chairman of the subcommittee) presiding.

OPENING STATEMENT OF HON. MIKE GALLAGHER, A REPRESENTATIVE FROM 
    WISCONSIN, CHAIRMAN, SUBCOMMITEE ON CYBER, INFORMATION 
                  TECHNOLOGIES, AND INNOVATION

    Mr. Gallagher. The subcommittee will come to order.
    Good morning, and welcome to today's CITI [Cyber, 
Information Technologies, and Innovation] Subcommittee hearing. 
The hearing really seeks to answer two fundamental questions: 
What is Replicator and can it work, how can it work?
    The Deputy Secretary of Defense announced Replicator's 
expansive vision in late August, which is to field thousands of 
all-domain attritable autonomous systems.
    So what is that A, A2, A--this is what my biggest problem 
with A2. I finally got A2/AD [anti-access/area denial] in my 
head and now I have to think of a new acronym here.
    But to field all-domain attritable autonomous systems 
within 18 to 24 months to combat the Chinese Communist Party's 
greatest asymmetric advantage, which is sheer mass.
    Now it's 2 months later--that would be 10 percent of the 
way through the 18-month timeline--and this subcommittee, the 
Armed Services Committee more broadly, and industry, I would 
argue, including small and medium-sized innovative companies 
seeking to supply our troops who would presumably be part of 
Replicator, and the American people, are still left without any 
details on Replicator, what is necessary to make it successful, 
whether it is feasible, and what counter-effects it seeks to 
offer to counter Chinese military capabilities.
    Now, I agree with the Deputy Secretary that the threats we 
face warrant a revolution in the defense industrial base, 
expanding the number of supplies, reducing fielding timelines, 
scaling production, and modernizing the culture of acquisition 
across the armed services.
    And though a small part of my soul dies every time we enter 
a new acronym into the DOD [Department of Defense] lexicon, I 
want to be clear. I want this to work. I think we need to be 
moving heaven and earth to enhance near-term deterrence across 
the Taiwan Strait.
    Otherwise, we will be--we may stumble into a war that would 
make the current conflicts in the Middle East and Eastern 
Europe look small in comparison.
    So I want this to work. And, in fact, someone gave me this 
book yesterday which I'd never read and I haven't finished yet. 
It's a small book, though. So even I as a Marine can read it. 
It's called ``Men, Machines, and Modern Times,'' by Elting 
Morison. Dr. Greenwalt seems to know this book.
    It's the 50th anniversary edition I hold in my hands. And 
thus far, what I can deduce is that he kind of talks about this 
process of innovation and resistance to change and kind of 
identifies a cycle of resistance that involves, really, three 
stages.
    The first is ignoring, the second is rational resistance, 
and the third is name calling. I want to be conscious of not 
falling into this cycle. Obviously, we're holding the hearing 
so we're not ignoring Replicator, though I worry about our 
ability to deliver on the timeline.
    I'm not trying to resist just for the sake of resisting. 
And as for name calling, well, we already have the name for the 
initiative so I'm not going to denigrate it. Again, I want it 
to work. I want to work with the building and we're lucky to 
have a group of experts from outside of DOD to get their candid 
advice and perspectives on the challenges and opportunities 
presented by the Replicator initiative and the general theme 
which we can all rally behind, which is procuring more hard 
power faster.
    We're joined by Mr. Bryan Clark, the senior fellow at the 
Hudson Institute, Dr. Bill Greenwalt, the nonresident senior 
fellow at the American Enterprise Institute, and Dr. Paul 
Scharre. Is that--am I saying that right? Yeah. Did I get that 
right, Scharre?
    Dr. Scharre. [Off mic.]
    Mr. Gallagher. Okay. Okay. We have had any number of 
discussions in person and I still have trouble with your last 
name. Executive vice president, director of studies, Center 
for--Center for a New American Security.
    This is an amazing lineup of people. I'm really looking 
forward to this discussion. And with that, I will turn it over 
to the ranking member.

STATEMENT OF HON. RO KHANNA, A REPRESENTATIVE FROM CALIFORNIA, 
      RANKING MEMBER, SUBCOMMITTEE ON CYBER, INFORMATION 
                  TECHNOLOGIES, AND INNOVATION

    Mr. Khanna. Thank you, Mr. Chairman. Thank you to our 
witnesses.
    Like each of you I've followed the Department's Replicator 
program and I commend the chairman for holding this hearing and 
the briefings that resulted because of the hearings. I think 
it's important for the Department, really, to brief Congress 
about what this program is and why we need it, what the 
obstacles are.
    As I've often said, briefing Congress is like briefing the 
American public. So the American public needs to know in simple 
terms what is exactly the Replicator technology, what are our 
goals, and how are we going to achieve them and by when.
    As you know, advanced technologies alone aren't--don't 
ensure effective capability. A new system hinges also on 
contracting, management, a committed industrial base, and being 
backed by training units and robust support functions.
    I'm particularly interested in our industrial base's 
ability to support the Replicator program and where there may 
be vulnerabilities in our supply chains and want to hear what 
continued challenges you face and how this committee can help 
in ensuring that you hit your goals.
    To each of our witnesses, thank you for lending your 
expertise and, Mr. Chairman, I look forward to this discussion.
    Mr. Gallagher. Thank you.
    Mr. Clark, you are recognized for 5 minutes for an opening 
statement.

   STATEMENT OF BRYAN CLARK, SENIOR FELLOW, HUDSON INSTITUTE

    Mr. Clark. Thank you, Mr. Chairman.
    Chairman Gallagher, Ranking Member Khanna, distinguished 
members of the committee, thank you for inviting us here to 
talk about Replicator and by extension DOD's other unmanned 
system initiatives.
    Congratulations also on getting the DOD to actually start 
talking about Replicator and what it might be, because this 
hearing has led to a series of conversations that we have had 
both with the DOD and with others about the initiative.
    So Replicator, you know, is pursuing a worthy goal. The 
initiative is trying to solve operational problems being faced 
by today's operational commanders, so combatant commanders, and 
trying to do that in a way that leverages new technologies.
    It has a lot of the elements of a successful quick turn 
technology development program. We just did a study for the 
U.S. Navy that looked at this, and Replicator has some of those 
elements built into it.
    But I'd say that there's a few things that it's missing, 
that it really will need to address in order for the program or 
the initiative to be successful.
    First of all, it's aiming--it's kind of aiming at the wrong 
problem. You know, so the idea of using mass alone to be able 
to counter China's mass is probably not the right solution for 
the U.S.
    You know, China as a resident major power we're going to be 
operating in their backyard. They can generate targets at a 
lower cost and more easily than we can generate successful 
shots on targets.
    So if we are going to try to match mass for mass we have 
probably picked the wrong adversary, the wrong scenario, if 
we're thinking about an invasion of Taiwan scenario.
    So what we need to do is think about not just using mass 
but using our innovative, you know, more creative operational 
capabilities. So just like Secretary Hicks mentioned in the 
announcement for Replicator, we need to enable operational 
innovation on the part of our warfighters.
    Their creativity, their ability to generate new concepts 
and force compositions, will allow them to undermine China's 
planning ability, create uncertainty for China, degrade the 
ability of concepts like system destruction warfare to be 
successful.
    So we can't just rely on mass alone. Replicator, especially 
in its first instantiation, which they call Replicator 1, 
should be aimed at something other than mass. It should be 
aimed at enabling operational innovation.
    The--then that kind of suggests, well, what was the process 
by which they analyzed the solution set that Replicator 1 may 
pursue and that kind of brings to question what is the 
Replicator initiative's design in terms of how do they analyze 
potential solutions, come up with them, and I think that's a 
weakness of the effort thus far is they don't have a strong 
analytic background for the initial solution sets they're 
pursuing.
    We're doing a study right now with Australia's DOD and this 
is one of the areas where they actually seem to have an 
advantage on the U.S. is they've got a very robust way of 
analyzing potential solutions, both looking at gaps and 
potential opportunities for asymmetric advantage. So not just 
trying to fill holes but also looking for ways they can create 
problems for an enemy.
    This is an area where Replicator could improve, you know, 
by looking at a better way of doing the analysis. And if we 
want to be able to gain warfighting advantage the biggest 
challenge that Replicator faces and that they aren't really 
addressing yet, I don't think, is integrating unmanned systems 
together in ways that are going to enable them to communicate 
with one another, be managed by a command and control system, 
and then operate in a way that creates creative operational 
concepts for us to pursue and dilemmas for enemies to deal 
with.
    So Replicator is going to have to create a process that 
allows the software of the vehicles, the software of a command 
and control system, and the networks that are connecting them 
all together to interact. The government's going to have to 
manage that. This is not an area that DOD has particularly been 
good at and the current way that Replicator envisions doing 
this is by having DIU [Defense Innovation Unit] buy vehicles, 
hand them over to the services, and have the services pursue 
this integration.
    As we know, DOD doesn't do a great job of integrating 
between services or between domains and so that's going to 
create the challenge of integration and then maintaining that 
integration through successive changes to operational concepts 
and tactics that we're going to want to pursue if we want to 
create continued dilemmas for China to have to deal with and 
deter their aggression against an ally like Taiwan.
    So, you know, overall I would say Replicator has good 
ambitions. It has a good goal of solving near-term operational 
problems faced by commanders. But in its implementation it's 
going to have to make some changes that start aiming at the 
right types of objectives, not just mass but looking at ways of 
enabling innovation that creates dilemmas for an opponent and 
then allowing for the integration that's necessary for unmanned 
systems and the existing force to work together to create those 
force packages, those concepts, those tactics, that are going 
to undermine China's ability to be successful.
    If we don't do those things then we're simply going to 
create more of the same and make our own force bigger but not 
make it any more difficult for the Chinese to defeat.
    Thank you very much.
    [The prepared statement of Mr. Clark can be found in the 
Appendix on page 29.]
    Mr. Gallagher. Thank you, Mr. Clark.
    Dr. Greenwalt, you are recognized for 5 minutes.

 STATEMENT OF WILLIAM C. GREENWALT, NONRESIDENT SENIOR FELLOW, 
                 AMERICAN ENTERPRISE INSTITUTE

    Dr. Greenwalt. Thank you, Chairman Gallagher, Ranking 
Member Khanna, and other distinguished members of the 
subcommittee.
    I'd like to thank you for the opportunity to testify this 
morning on DOD's Replicator initiative. I ask that my written 
statement be part of the record.
    And so to answer the subcommittee's question about whether 
the Replicator program can work, I think one first needs to 
figure out specifically what it is and then establish criteria 
for success.
    I'm not really sure yet that any of us or even the 
Department can articulate how and with certainty how this 
concept will evolve. Still, I think there's enough information 
out there to say that deploying thousands of cheap autonomous 
sensors, weapons, communications nodes, and targets that serve 
as a deterrent and can complicate Chinese and other potential 
adversaries' calculations in a future conflict is probably a 
pretty good idea.
    In that context, the Replicator-type drone program has the 
potential to be a significant game changer and may well be 
worth pursuing but only if it's done correctly and does not 
crowd out funding for near-term munitions and other critical 
requirements, given the current rising threats around the 
globe.
    An even better idea that's coming out of the Replicator 
initiative and embedded in this initiative and the one that is 
probably really most radical is to eventually replicate the 
business processes necessary to eventually achieve rapid time-
based innovation in operational deployment over a much wider 
portfolio of technologies.
    Replicator proposes to be a vanguard for a massive and I 
would say much needed change in how the Department innovates at 
scale. Each of the initial stated Replicator goals, the short-
term drone deployment and then the medium-term business 
development process change, face significant challenges.
    The issue for Congress to consider is whether any of this 
is realistically achievable. I think it could be but the odds 
at the moment are pretty much stacked against success.
    The biggest challenges revolve around DOD's culture and 
processes. DOD's culture and the management systems derived 
from that culture are stuck in a 1960s paradigm that has 
consistently rejected new approaches and commercial technology 
incorporation that does not conform to DOD's ingrained 
thinking.
    The greater defense innovation problem is multifaceted and 
one that manifests itself in the extensive time it takes to 
deliver capability. The system is based on a planned step-by-
step predictive process that takes decades to deploy anything 
of substance.
    This is diametrically opposed to how DOD used to develop 
capabilities in the 1940s and 1950s and how things are done 
today in the commercial sector. It should come as no surprise 
to this subcommittee that the vast majority of technologies 
that the Department of Defense needs in the future reside in 
the commercial marketplace, which has adopted a time-based 
innovation system, and not in the Department of Defense or in 
the traditional defense industry.
    DOD in the 1950s essentially created the time-based 
innovation playbook that Silicon Valley--that was the basis for 
Silicon Valley's subsequent success, and Replicator is perhaps 
one last chance for Silicon Valley to return the favor and 
reinvent DOD.
    I say all this because the lesson from past DOD innovation 
efforts is that if Replicator is not taken outside of the 
acquisition budget bureaucracy and rules it will fail. The 
Pentagon's acquisition system is simply not capable of acting 
on the proposed timelines contemplated in the Replicator 
program except in very limited circumstances and then only when 
conducted outside the normal rules of acquisition and 
budgeting.
    This committee and its sister committee in the Senate gave 
the Department 8 years ago a new playbook to create a time-
based innovation and acquisition system.
    It had created production other transactions, improved 
commercial item procurement, enhanced rapid acquisition 
authority--a new middle tier acquisition authority--all 
designed to bypass the cumbersome acquisition and budgeting 
processes.
    The results to date have not necessarily been what we would 
hope to--hope to see. Older bureaucratic mandates particularly 
within OTAs [other transaction authority] and MTA [middle tier 
of acquisition] authorities have restored the ability to 
achieve the successes that I think Congress was looking for at 
the time.
    I expect Replicator will face many of the same problems in 
trying to implement these authorities as it tries to move 
forward. I don't think it's hopeless. It's a fight worth 
fighting. I think it can be successful.
    But it's something that DOD needs a lot of work and a lot 
of top-down direction to do so. I think the--this model is much 
needed and the stakes are really high to implement it and so I 
hope they can achieve some success.
    Thank you again for the opportunity to testify on this 
topic and I welcome any questions.
    [The prepared statement of Dr. Greenwalt can be found in 
the Appendix on page 51.]
    Mr. Gallagher. Thank you, Dr. Greenwalt.
    Dr. Scharre. Did I nail that?
    Dr. Scharre. Perfect.
    Mr. Gallagher. Okay.

    STATEMENT OF PAUL SCHARRE, EXECUTIVE VICE PRESIDENT AND 
    DIRECTOR OF STUDIES, CENTER FOR A NEW AMERICAN SECURITY

    Dr. Scharre. Chairman Gallagher, Ranking Member Khanna, 
distinguished committee members, thank you for the opportunity 
to testify today.
    Replicator has a bold agenda to field thousands of 
attritable autonomous systems in 18 to 24 months. This effort 
is not only valuable, it is necessary if the U.S. military is 
to reverse its decades-long death spiral of rising costs and 
shrinking quantities for major weapons platforms.
    The United States will need large numbers of low-cost 
attritable systems to generate the mass necessary to prevail in 
a conflict with China. We have seen in Ukraine the scale of 
violence that is possible in an interstate war and the 
possibility that such a war could drag on for years. The level 
of casualties seen in Ukraine would render the U.S. Army combat 
ineffective in a matter of months.
    The U.S. military will need expensive capital assets such 
as aircraft carriers and stealth bombers but it will also need 
low-cost uncrewed systems to bring greater mass. We have also 
seen in Ukraine the value of such an approach.
    Both Russia and Ukraine have flooded the skies with low-
cost drones to find and track ground targets. Ukraine is 
reportedly losing 10,000 drones a month, yet it is able to 
reconstitute those forces using cheap commercially available 
drones.
    This is a way of fighting with mass that the U.S. military 
does not have in its current force. But it is a part of U.S. 
history. In World War II at the height of production allied 
factories were producing over three and a half times as many 
tanks and airplanes as the Axis powers.
    Today, the war in Ukraine has strained the defense 
industrial base's capacity for munitions production. For the 
United States to be ready for a conflict with China it must be 
able to produce forces at greater speed and scale than it has 
done to date.
    Replicator is a test of the Defense Department's ability to 
generate the types of systems it needs, in the quantity it 
needs them, and at the speed required to deter China. The 
Pentagon's bureaucracy is likely to face challenges along all 
three of these dimensions.
    Low-cost attritable systems and autonomous systems are both 
paradigm-busting concepts inside the Defense Department. They 
have supporters, including the Department's current leadership, 
but they challenge the Pentagon's traditional ways of doing 
business.
    DOD will also be challenged to move at the speed that is 
required. Even if DOD succeeds in its goal of fielding 
thousands of systems in 18 to 24 months, by then at best it 
will be 2025. Replicator is an important step but it is one the 
Department should have taken a decade ago. Speed is now 
essential.
    We saw in Iraq and Afghanistan that the bureaucracy can 
move quickly when needed. DOD rapidly fielded ISR 
[intelligence, surveillance, and reconnaissance] aircraft, 
counter-IED [improvised explosive device] technologies, and 
MRAP [mine-resistant ambush protected] vehicles. Yet, each case 
required direct intervention by the Secretary of Defense to 
circumvent traditional processes that were too slow and were 
not focused on urgent needs.
    Similar senior level attention by Department leaders will 
be required for Replicator to succeed, along with congressional 
support. The DOD has had recent successes in breaking the mold 
and moving quickly. The Defense Innovation Unit has brought in 
commercial technologies in a matter of weeks and months, not 
years and decades.
    But too often these innovation success stories have been 
small scale. To field thousands of systems DOD will need to 
operate quickly at scale, something it has often struggled to 
do.
    Replicator is essential not just because of the 
capabilities it aims to field but because of the path it is 
blazing through institutional red tape. Speed and scale of 
acquisition and new warfighting paradigms will be needed not 
just for autonomous systems but across the force if DOD is to 
transform itself rapidly to counter a rising China.
    If successful, Replicator can train the institutional 
muscle memory inside the Department--not just in the Pentagon 
but in industry as well--to move quickly, experiment with new 
concepts, and scale production.
    Is it possible? The average time from program start to 
initial operational capability for major defense acquisition 
programs today is 11 years, yet the United States mobilized for 
and fought the entirety of World War II in only 6 years.
    The United States can move faster when the moment demands 
it. Nor are these episodes confined to our distant past. Just 
recently, during COVID with government support U.S. industry 
rapidly scaled vaccine production to produce over 100 million 
vaccines per month by the end of 2021.
    Despite repeated warnings, the U.S. defense establishment 
is not on a war footing today. Replicator is a chance to change 
that dynamic and congressional support will be essential for 
this vital effort to succeed.
    Thank you.
    [The prepared statement of Dr. Scharre can be found in the 
Appendix on page 67.]
    Mr. Gallagher. Thank you very much.
    First I want to ask unanimous consent to enter into the 
record a statement from the Association for Uncrewed Vehicle 
Systems International.
    Without objection, so ordered.
    [The information referred to can be found in the Appendix 
on page 81.]
    Mr. Gallagher. Also I want to note how it feels a little 
bit scandalous to be holding a hearing in the absence of a 
Speaker but I kind of like it. So I'm glad--I'm glad we're 
doing this.
    Okay. So let me confess that when I first read the speech 
announcing Replicator my rational resistance kicked in to the 
effect of, well, if we're going to do a crash program in the 
next 18 months let's replicate the existing systems we know we 
need in the Indo-Pacific: Long Range Anti-Ship Missile, SM-6 
[Standard Missile 6], JASSM [Joint Air-to-Surface Standoff 
Missile], JDAM [Joint Direct Attack Munition], NSM [Naval 
Strike Missile], right. Like, the critical munitions, the long-
range precision fires we know we need to be moving to maximum 
production rates of but we're still at minimum sustaining rates 
even after the collapse of deterrence in Ukraine.
    So now my concern is that--is how do we accomplish 
Replicator without crowding out other programs if they say they 
don't need new money for Replicator. Put differently, my 
biggest concern is that, particularly when we know that the 
administration tends to prefer RDT&E [research, development, 
test, and evaluation] over procurement, is that hard power 
programs like munitions are going to end up being bill payers 
for this and I worry about that tradeoff.
    So maybe I'll start with Mr. Clark. Can you comment on 
that? Is that concern valid?
    Mr. Clark. Absolutely. I mean, you think about where 
they're going to find money in the near term to be able to 
reallocate towards, you know, procurement of vehicles.
    It's going to have to come from something that's relatively 
fungible and munitions are--you know, any procurement account 
is a place you can go to get that money. So something that's 
not got fully obligated funds will be a source of money for 
Replicator.
    So that's a concern and I think, to your point, we need a 
combination of both the continued maximum production of long-
range precision munitions but we also need what Replicator can 
bring in terms of creating a more recomposable and more 
flexible force structure that's able to create some dilemmas 
for the Chinese that are separate from what the long-range 
fires complex can do, because they've planned for and are 
dealing with, you know, in their systems destruction warfare 
our approach to long-range fires.
    So more long-range fires is just going to make us better 
able to continue doing just business the way we do now. So we 
have got to also have what Replicator provides in terms of more 
flexible and creative operational constructs that are going to 
create dilemmas for the Chinese.
    So if we're going to have to--we have to do both and I 
agree with you that that money cannot come from Peter to pay 
Paul.
    Mr. Gallagher. Yeah. And I--the spirit of the question is 
that this shouldn't be an either/or choice. But if you were 
going to say to the SECDEF [Secretary of Defense], okay, this 
only--something can only happen in 18 months with SECDEF-level 
involvement and pushing the bureaucracy every single day and 
you can only focus on one thing, I would focus on long-range 
precision fires, right.
    I mean, that would be--that is something you could make 
meaningful progress with SECDEF-level involvement and we're 
just not moving fast enough.
    Maybe, Dr. Greenwalt, comment on this point but also 
perhaps tie it to the need for flexible funding that--in order 
to achieve Replicator's goals. You've offered extensive 
commentary on the use of expired unobligated balances at DOD in 
the defense modernization account.
    Dr. Greenwalt. Yes. No, I mean, the potential of getting--
crowding out other critical needs is one of my biggest concerns 
about Replicator right now and I think we need to basically 
take existing systems that we need in the INDOPACOM [U.S. Indo-
Pacific Command] and ramp them up and I think your recent 
proposal legislation in the FIRES [Funding Pacific Readiness 
and Enhancing Stockpiles] Act is exactly what's needed to 
basically harvest money at the end of the year and put it into 
production-ready assets in the munitions industrial base.
    Funding flexibility and innovation is absolutely critical 
and right now there are very few pots of money that the 
Department can actually use and move around in the year of 
execution to essentially focus on things that look like they're 
going to be something that we can scale up.
    In other words, what happens now--and Paul has talked about 
11 years to achieve capability. That's when programs start. 
Right now it takes almost 8 years to get to the decision and 
get money to start that program. So it's 8 plus 11 so it's 
really 20 years that we're talking about.
    One needs to be able to start a program, get on contract, 
and get money within year of execution. That's the way we did 
it in the 1950s but that's not the way we do it today.
    So these various funds, whether it's the defense 
modernization account, which Congress tried to create a decade 
ago and has not been implemented by the Department to basically 
harvest about a billion dollars to focus on anticipated 
requirements in MDAPs [Major Defense Acquisition Programs], 
it's very difficult for the Department to want to execute and 
use flexible funding, for whatever reason.
    Mr. Gallagher. I'm going to have to come back in a second 
round. But I want to foot stomp this point and highlight a lot 
of Dr. Greenwalt's analysis, which suggests that over the last 
decade we have lost hundreds of billions of dollars--$127 
billion--in buying power, money that was appropriated to DOD 
and not spent and then it goes into Cinderella land, abeyance 
for 5 years in the Treasury and then it evaporates.
    For example, we lost $11 billion dollars a few weeks ago in 
part not just because of intransigence here in Congress among 
certain committees but also a lack of enthusiasm from the 
building and the Comptroller's office. And I think it would be 
wise for us to take advantage of that money, which is, again, 
going to disappear and be used for nondefense purposes, in 
order to provide sustained, predictable funding for both long-
range precision fires and potentially Replicator.
    So help me help you, DOD. We want to give you this money.
    With that, Mr. Khanna is recognized.
    Mr. Khanna. One of the challenges and vulnerabilities of 
democracies is that we're openly critical of our own systems. 
You would never have the Chinese discussing whether their 
systems work or are strong enough.
    I guess my first question is how much of a deterrent effect 
is just the announcement of a Replicator program and that we're 
doing this, putting aside, obviously, that we want to do it 
successfully?
    Mr. Clark. Well, so I would say that there is a dissuasion 
effect. You know, maybe if you're--you know, if you think about 
this from the Chinese perspective they're looking at, you know, 
what are ways that the United States is going to present new 
challenges to us that we have not prepared for and this would 
be an example of that kind of challenge, this along with what 
Admiral Paparo talked about recently in terms of wanting to 
create a hellscape, you know, in the Taiwan Strait to help 
thwart an invasion of Taiwan.
    If you put those together and say the U.S. could be 
fielding large numbers of unmanned systems with the intent of 
orchestrating them in ways that are going to defeat or at least 
slow down our ability to rapidly invade Taiwan, that starts to, 
you know, maybe undermine their confidence in their ability to 
mount that invasion.
    So, yeah, there's an effect that starts to chip away at the 
confidence that Chinese leaders might have about their 
likelihood of success. But you got to follow it up, right, 
because then it'll abate.
    You know, so you got to follow it up with some actual 
tangible changes that are going to continue the--to drive that 
uncertainty and then you're going to have to evolve it over 
time. So it can't be just we send out a bunch of these unmanned 
boats and now they don't--you know, they do the same thing year 
after year.
    You got to continue to evolve them. I think that's one of 
the challenges for Replicator is to present an evolving set of 
capabilities that are going to be--force dilemmas on the 
Chinese.
    Dr. Scharre. I mean, certainly, there's nothing like seeing 
for believing and so, obviously, the Chinese military is going 
to be tracking what the DOD is saying.
    But I do think that DOD needs to think strategically as 
they move forward about doing demonstrations that can 
demonstrate capability in ways that are going to have a 
deterrent effect.
    We have certainly seen this in the past. Several years ago 
DOD did a big swarm demonstration out at China Lake with a 
hundred drones. That was followed not long after by China doing 
their own swarm demonstration.
    We notice the things that they're watching, and so there's 
going to be a lot of capabilities we need to conceal but some 
that we want to think about strategically showing for deterrent 
effect.
    Mr. Khanna. What then, if we convene a year from now, as 
I'm sure the chairman will, do you think are the, let's say, 
top three things we want to have seen by then to demonstrate 
that this is a success both internally and in what we're 
projecting to the Chinese?
    Dr. Greenwalt. I think deployment of capability. In other 
words, to actually have had programs with capability that are 
out in the hands of the warfighter who then are suggesting new 
improvements and we're making the next level of changes that 
are necessary. I think you have to show that you can actually 
deploy capability fast and in such scale that it makes a 
difference.
    Mr. Clark. I think another thing is having a rapid tempo of 
prototyping. You know, the DOD does prototyping now and that's 
some of the things they would do with Replicator. But if that 
tempo is, like, one or two exercises a year, that's not going 
to cut it.
    The DOD is going to have to have a more rapid pace of 
prototyping in order to both assess a variety of potential 
solutions but also change those solutions over time, and if we 
don't see that in a year, that's an indication this is kind of 
going the way that other programs have.
    Dr. Scharre. I think one of the things that we have seen in 
the past with new technologies like this is finding out how you 
use it effectively is one of the most difficult things.
    So integrating this rapid prototyping, getting it in the 
hands of warfighters, with a tight loop between warfighters and 
developers so that they're figuring out how to use this most 
effectively is going to be key.
    Mr. Khanna. And, Dr. Greenwalt, I read in your testimony 
about the defense industrial base and I've been concerned about 
that.
    I was surprised that the DOD didn't track exactly where all 
of our component parts are made, and we had an amendment in the 
NDAA [National Defense Authorization Act] this time just to say 
where are our vulnerabilities, what are we getting from China, 
what are we getting from other countries.
    Do we know that this program--what our vulnerabilities are 
and is there an effort to make sure that we have a robust 
defense industrial base?
    Dr. Greenwalt. I think that's a question that the Congress 
needs to continually ask. I've talked to a number of drone 
manufacturers who would like to scale up and they're looking at 
their supply chain and a lot of that supply chain is not 
American or even allied. And that's going to be a real problem 
when you try to scale up and the time to actually get those 
parts into the United States is going to be a delaying factor 
in achieving their objectives.
    Mr. Gallagher. First up is an Alabaman, an EMT [emergency 
medical technician], and a ``Price is Right'' contestant, Mr. 
Strong.
    Mr. Strong. Thank you, Chairman Gallagher, and witnesses, 
thank you for being here today. And my questions are for each 
of you just whoever wants to take it.
    Through Replicator the Pentagon is committed to fielding 
thousands of attritable autonomous systems within 18 to 24 
months. We're not talking about just research and development 
or a small-scale pilot program. This is a large undertaking and 
the speed of it all within 2 years.
    Don't get me wrong. I agree there is an urgent need for 
this type of a program. But I'll take Huntsville, Alabama, for 
instance. We waited on Space Command basing decision for more 
than 2 years when we came in first throughout the entire 
process and the Biden administration still chose the fifth most 
qualified location.
    Now we're working to fix that but there are countless 
examples of programs and procurements being held up by protest 
and general delays.
    My point here is that the DOD is not exactly known for 
being speedy as they are claiming they can implement Replicator 
with no additional funding or authorities. So my first question 
is do you see challenges with the DOD implementing the 
Replicator program based on current funding lines and existing 
authorities?
    Dr. Greenwalt. I would say absolutely. But this committee 
has given in the past the Department ways to achieve those 
objectives if they so use them.
    So the first one on funding is the rapid acquisition 
authority account which would allow for rapid reprogramming and 
then congressional notification after the fact. This has been a 
very useful authority ever since 9/11 to get capability in the 
hands of the warfighter faster.
    Other transactions, which is a contracting authority, 
allows the government to get on contract quickly. Could be 30 
days versus--and without protests. This was very successful in 
many, many DIU efforts and, frankly, was how the Falcon 9 was 
created at NASA [National Aeronautics and Space 
Administration]. So you have that.
    The middle tier acquisition authority is a way to move 
around the cumbersome requirements and an acquisition process 
which adds years to the process. So there are tools they could 
use but if they use the traditional ways of doing business at 
the Department of Defense it will take decades to achieve this 
capability.
    Mr. Strong. Dr. Greenwalt, do you think that the Pentagon 
has identified and begun addressing these challenges?
    Dr. Greenwalt. I think the Pentagon--parts of the Pentagon 
have. There are a number of innovation hubs that are taking 
advantage of this authority.
    But as Paul has pointed out, it's not at scale. And so the 
Pentagon needs to focus these authorities laser like on 
programs of scale to bring these capabilities as fast as they 
possibly can to the warfighter.
    Mr. Strong. Thank you. The DOD can't do this all on their 
own. It's been made clear in the last year that our defense 
industrial base has atrophied. Is industry ready to scale up to 
meet the needs of the Replicator program?
    Mr. Clark. I would say not yet. They need the investment in 
order to be able to, you know, make the kinds of investments on 
their own that they're going to need to be able to build up the 
facilities because right now DOD's investment in unmanned 
systems has been pretty anemic, which has led industry to sort 
of back off and, you know, keep their powder dry, waiting to 
see if that demand is going to start to manifest.
    And so as a result we don't have an industrial base right 
now to draw upon to scale at the type of pace that Replicator 
is envisioning. So it's going to take some time to make that 
happen even if we take advantage of emerging, you know, 
technologies for manufacturing.
    Mr. Strong. Thank you. Are there any previous autonomous 
programs of similar scale that the Replicator program should 
take lessons learned from?
    Dr. Scharre. Well, we have certainly seen other attempts by 
the Department to scale quickly with commercial technologies.
    You mentioned protests and delays, and I think it's worth 
pointing out that some of the challenges in addition to those 
inside the Department exist outside of the Department. The JEDI 
[Joint Enterprise Defense Infrastructure] cloud computing 
contract is an example of the Department trying to do this. And 
when the Department is using sort of small-scale innovation 
solutions, small money, they can do that.
    When you got to a $10 billion contract with real money on 
the table then we saw protests and lawsuits that mired the 
entire process in 3\1/2\ years of delays. So I think that is a 
big risk when we start to see scale.
    Mr. Strong. Thank you again.
    Mr. Chairman, I yield back.
    Mr. Gallagher. Thank you. Next up a product of Pennsylvania 
and the United States Naval Academy, Mr. Deluzio.
    Mr. Deluzio. Thank you, Mr. Chairman. And good morning, 
everyone. Thanks for being here.
    I want to ask about several nondefense AI [artificial 
intelligence] autonomy organizations, Google and many others, 
who have now adopted codes of ethics saying they're unwilling 
to support the use of their technology in defense, for defense 
purposes.
    So I'm curious maybe from each of you, whoever wants to 
weigh in, to what extent are you thinking these principles 
impact DOD's ability to accomplish goals like those set forth 
in Replicator. And these are ostensibly American companies, I 
should note.
    Dr. Scharre. Yes, thank you. We saw, I think, a big concern 
by those in Washington, in Congress, and in the Department 
several years ago when Google did not continue to work on 
Project Maven.
    I think one of the things that we have seen since then is 
that while there are some companies that have said they don't 
want to work with the military, in fact we have seen a number 
of major tech companies stand up and say that they are American 
companies and they are going to work with the government, with 
the U.S. military, despite in some cases protests from their 
employees.
    We saw the leadership at Microsoft and Amazon both say 
that. In fact, Google is now working with the Defense 
Department again. So, you know, and we have seen since then a 
huge growth in startups that are specifically oriented at the 
Defense Department as their customer--Shield AI, Anduril, many 
others.
    So I think in practice this has been maybe more noise than 
substance and in fact there's just a huge segment of the tech 
community that's eager to work with the U.S. military.
    Dr. Greenwalt. I think even the biggest problem that some 
of these companies face are not necessarily related to AI 
ethics but are related to how the government does business, how 
the government treats intellectual property, how the government 
contracts, the nonunique commercial processes that each of 
these companies has to create to be--to comply with having a 
government contract.
    That's why other transactions and other means of having a 
commercial-type contract is important to these type companies 
but something that the Department of Defense is not exactly 
stellar at using.
    Mr. Clark. I'll just add one small corollary to this is 
that the new rules for the use of autonomous weapons that the 
DOD has put out are going to force Replicator to think through 
some really complicated mechanisms for command and control of 
these unmanned systems that allow them to meet rules of 
engagement requirements and a lot of commanders and operators 
to have a chance to make a decision whether a system is going 
to be used in a lethal manner or not.
    That's not trivial and it's going to require Replicator to 
have, you know, some longer term effort to do the software 
integration and the updating of these systems even after 
they're fielded.
    Mr. Deluzio. Thank you.
    Dr. Greenwalt, I want to come back to something you said 
about intellectual property. That was my next question so 
thanks for predicting it.
    Look, several of these systems that are in DOD's inventory 
the government does not have intellectual property rights. So 
I'm curious if you want to start and if others want to weigh 
in, what should be DOD's strategies here around intellectual 
property for systems like the ones we're discussing where 
because DOD may not have those rights they can't grow and learn 
and gain insights from the systems that otherwise might be able 
to?
    Dr. Greenwalt. Intellectual property in contracting is an 
incredibly hard subject and the Department needs to ahead of 
time understand which IP [intellectual property] it needs to 
maintain and operate systems and which IP it doesn't need so it 
can incentivize not just a monopoly provider but an entire 
ecosystem of competition and then systems can be replaced based 
on that. That is a very difficult process.
    But at the same time it's something that the Department 
needs to become the expert at because it is so important, 
because this incentivizes industry but it also is a barrier to 
the Department using its own things that it's bought. So I 
think it's definitely one of these things that the committee is 
going to have to continue to shepherd and look at.
    Mr. Deluzio. Any other thoughts from the panel?
    Dr. Scharre. I think there's just on IP and maybe more 
broadly a genuine tension for the Department in moving fast 
versus having the right procedures in place, having ideal 
contracting structures or requirements for systems. And to some 
extent if you're controlling for time, as the Department is 
trying to do here, there may be situations where you're 
purchasing something that you're not going to sustain over the 
long term and that's okay. And I think that we should 
acknowledge that that's fine because that goal of we're going 
to buy something that we're going to keep for 20, 30 years is 
the enemy of moving quickly.
    Mr. Clark. So--and one other thing is the government can 
exert its rights to own the interfaces or to manage the 
interfaces between software--and a lot of this intellectual 
property has to do with software--and the government can allow 
companies to provide their own software and develop that on 
their own but then, you know, force them to show that they can 
integrate through the interfaces that the government owns, in a 
way is similar to what iOS does with, you know, new apps trying 
to get into the app store.
    Mr. Deluzio. Thank you. Mr. Chairman, I yield back.
    Mr. Gallagher. Next, a salty Marine infantry officer from 
Salem, Massachusetts, Seth Moulton.
    Mr. Moulton. Thank you very much, Mr. Chairman.
    So Ukraine has done remarkably well at holding off the 
Russian bear in large part thanks to our help and yet all their 
most innovative warfighting is not from us because we didn't 
help them, you know, innovate with drones.
    We didn't help them innovate with cyber. We're just trying 
to teach them tactics that America learned 50 years ago.
    So what lessons are you taking from Ukraine's drone program 
to influence how you get Replicator up to speed so quickly, as 
it needs--clearly needs to be done?
    Mr. Clark. Well, I'd say one thing right off the bat is 
Ukraine has been able to do what they're doing not through 
necessarily just mass and throwing a lot of unmanned systems at 
the Russians but instead it's how they orchestrate, how they, 
you know, organize themselves, how they use the tactics that 
the unmanned systems enable, how they employ those tactics in 
the field, how they, you know, basically sequence their 
operations.
    So it's a lot of the kind of operational art is what 
they're bringing to bear that's actually yielding success 
rather than just throwing a bunch of, you know, mass at the 
wall and hoping that the Russians get overwhelmed.
    So I think the lesson there is we've got to figure out ways 
to enable our unmanned systems to be employed in a very 
flexible way that tactical operators can then adjust in the 
field.
    Mr. Moulton. But what about from an acquisition 
perspective?
    Dr. Greenwalt. I think that what we're seeing there is that 
the proliferation of commercial technology is influencing how 
an agile acquirer, and the Ukrainians are being very agile in 
this, and pulling together various commercial technologies to 
achieve military significant means, and I think the lessons 
learned that we should be taking from there is that the 
commercial marketplace is maybe light-years ahead of where the 
Department of Defense is and we need to figure out how to be 
agile and bring that in and then bring the unique military 
knowledge that we have to bear on that commercial tech. And 
that is going to be a process that moves against the Department 
of Defense's culture.
    Mr. Moulton. Well, I would just suggest you also consider 
that at least my understanding of how much of this is working 
in Ukraine is they're taking their DOD personnel--their 
military personnel--and actually embedding them in the 
companies as opposed to what you just said, which is taking the 
companies in and bringing them into the bureaucracy of the DOD.
    I think we can all intuitively know which one would 
probably produce quicker results. So I think--I think we should 
look very, very carefully about that and, honestly--look, Kath 
Hicks came out in, I think, the end of August and announced the 
start of the Replicator program.
    Just in terms of timeline, because speed is so important 
here and we all agree on that, at what point will Replicator 
have the capabilities of the Ukrainians?
    Dr. Scharre. I mean, certainly not on any timeline that 
we're currently on. I mean, even if they achieve all of their 
goals we're looking at Ukraine is fielding tens of thousands of 
drones on a regular basis and Replicator doesn't even have that 
goal. So that would be years away.
    One of the things that we have seen from an acquisition 
standpoint Ukraine do very successfully, they have a very 
decentralized approach. They have civilians, you know, sort of 
like spontaneously working drone operators working with 
industry, working with the military.
    There's downsides to that. They have a very heterogeneous 
fleet so, you know, things like maintenance is hard. But if 
you're losing them at high volumes and you are focused on 
replenishing, that's fine. And one of the advantages of what 
they're doing is not only can they then scale through that 
direction but they do a lot of experimentation in the 
technology and then how it's being used because they're 
allowing a lot of freedom among those on the front lines and 
the developers to try things and then figure out what works.
    Mr. Moulton. Well, does anyone know what percentage GDP 
[gross domestic product] Ukraine's is compared to ours at this 
very moment? I don't know either.
    But what I--what I intend to do is write a letter to the 
Department requesting a basic chart that simply compares 
Ukraine's speed of fielding, acquiring these technologies--
acquiring and fielding and getting them in the hands of 
warfighters compared to Replicator.
    And we can come up with some basic metrics of how to--how 
to gauge that, but that should be a basic yardstick. I mean, I 
would say that if Ukraine's GDP is one one-hundredth of ours 
then we should be 100 times faster.
    But be that as it may, let's at least just compare one to 
one how we're doing in terms of speed. I think that would be a 
great metric for how successful this program is.
    Mr. Chairman, I yield back.
    Mr. Gallagher. Next, from Norwood, Norfolk, and Nantucket, 
Mr. Keating.
    Mr. Keating. Thank you, Mr. Chairman, and your efforts at 
alliteration are wonderful this morning. The sailor from Salem 
and whatever.
    Thank you very much for holding this hearing. I just have 
one kind of overall question that's--and maybe you can shed 
some light on for me.
    Number one, the idea of a swarm of drones or low-cost 
decoys overwhelming an opponent, an enemy, it's--that's a 
pretty old concept. It's been around for a long time.
    So my question is why now? Has something changed so this is 
now something we're going to implement, and how is the 
Replicator different than these old concepts?
    Dr. Scharre. Why now, you know, I think I'm not entirely 
sure what's driving the Department. Certainly, the China threat 
is one that they are focused on and trying to move quickly.
    We also can see quite clearly in Ukraine that a lot of 
these concepts are working and this idea of low-cost attritable 
autonomous systems has been a very contentious one in the 
defense community. It's not new. It's been around for at least 
the last 10, 15 years. People have been talking about it.
    But right now we're seeing that it works. Like, you can 
build large numbers of low-cost things that are not 
individually survivable and, yet, you can replenish them 
quickly and that is effective in delivering combat power. And I 
think that that kind of real-world demonstration doesn't hurt.
    Mr. Keating. Has the technology changed that much?
    Mr. Clark. Yeah. I mean, what we have also seen, and 
Ukraine has exhibited that, is that the commercially available 
vehicles have been, you know, very effective. They've reached a 
point where they can be operationally useful in a military 
sense. So you can get large numbers of them very quickly into 
your forces' hands.
    On the software side we have been able to manage them and 
control them in ways that are going to be more effective than 
just having a swarm go and try to overwhelm the enemy.
    So I think on both the kind of hardware and software side 
we have seen technology evolve in a way that allows this to be 
an effective military capability.
    And then to Paul's point, you know, the military threat 
right now--the potential for invasion of Taiwan by China--has 
gotten to the point where our traditional approach to dealing 
with that problem is not necessarily going to be as solidly 
effective as it has been trusted to be in the past. And so I 
think that's the other thing driving us is that the operational 
urgency is there.
    Dr. Greenwalt. I also think it's a desire for the 
Department to create a new innovation model and drone 
technology is a segment in which they could actually do that.
    They're essentially, you can argue, already doing that with 
space assets and the commercialization of space and so that's a 
successful model that they can replicate in Replicator and 
autonomous drones are kind of that next level. But there 
probably are a couple other technologies that they could pursue 
as well.
    Mr. Keating. Thank you very much, and I yield back.
    Mr. Gallagher. Okay. I'm going to entertain a second round.
    In my opinion, the only thing better than producing mass 
quantities of long-range precision fires ideally with advanced 
energetics inside of them is doing so in concert with our 
closest allies, particularly the Aussies and the Brits under 
AUKUS.
    As you know, autonomous attritable weapon systems and 
technologies are a key part not only of AUKUS but the national 
technology industrial base and many of our bilateral 
agreements.
    What needs to happen with our allies and partners to build 
out our magazine depth in this regard? And I'll just go--I 
don't know who--does someone want to volunteer?
    Dr. Greenwalt. I'll start with that we need to incentivize 
the working together of the scientists and engineers in not 
just the government but in industry in the three AUKUS 
countries. And in that regard there is a need to reform our 
export control and our ITAR [International Traffic in Arms 
Regulations] process.
    Right now that process designed to protect technologies 
that unfortunately now are many generations old are keeping our 
best scientists and engineers from working together to solve 
these problems and even to the point where Boeing Australia 
created a new loyal wingman drone that's ITAR-free.
    Why? Because Australia itself wanted to have sovereign 
capability and the ability of not having to go every time they 
use that product to the State Department for approval on how to 
use it.
    So we have to reform our export controls to incentivize 
that structure.
    Mr. Clark. That's a starting point, clearly, and I--and to 
build on Bill's point, Australia is concerned that if they 
provide the Ghost Bat to the U.S. military and then the 
military--U.S. military modifies it, then the Australians will 
not be able to get it back because now it'll be ITAR 
controlled.
    So there would be a version of Ghost Bat that we use that's 
maybe more sophisticated and the version Australia has is 
still, you know, whatever--you know, baseline. So that's a 
concern that they had that they raised to us when we were down 
there for the study we completed.
    I'd say the other big thing is, you know, the scientists 
are all working together. We talked with these guys and they're 
very interactive right now. But there's not a demand signal for 
buying the systems and fielding them and doing the 
experimentation because right now there's no experimentation 
venues that are across all three allies where they're actually 
looking at prototypes of these systems being employed in real 
operational settings.
    We need to create that and have a pretty rapid tempo of 
that. And then we have got to create a demand signal in terms 
of investment, whether it's the U.S. making the investment or 
all three countries making the investment, that's going to draw 
these capabilities, you know, across the finish line into some 
useful operational form or else industry is never going to 
start ramping up the production.
    Mr. Gallagher. Dr. Scharre, any thoughts?
    Dr. Scharre. I mean, these export control barriers are 
killing us and the Ghost Bat example is a painful one and we 
can't--the problem is we're not going to be able to get to the 
things we need to be doing working jointly with them developing 
concept of operation if we can't do the basics in terms of 
sharing the technology. Those are things at the starting line 
that are really holding us up.
    Mr. Gallagher. And also, like, at the most basic logical 
level if the problem is, like, PLA [People's Liberation Army] 
mass. Like, they can produce a lot of stuff and we can't sort 
of like counter that one for one. Then the logical thing to do 
would be, like, to assume risk and work with our partners and 
allies, which is our asymmetric advantage, right. They don't 
have allies and partners. They have vassal states and----
    Dr. Scharre. Right. Great complexity for them, great 
uncertainty. Multilateralize the problem. You know, make it 
more difficult for China to feel like they're going to get a 
free shot at a place like Taiwan.
    Mr. Gallagher. Yeah.
    Dr. Greenwalt. But our allies are moving out. So, for 
example, not just Ghost Bat but Ghost Shark which is 
essentially a commercial unmanned submersible that's going to 
be--take commercial technology from actually the U.S. and the 
Australians are going to militarize it and try to, you know, 
flood the zone with a lot of autonomous undersea drones.
    That's huge, but we--the U.S. Navy should be working 
together with the Australian navy on that concept and on the--
on those cheap, you know, undersea attritable drones, so to 
speak. But we're not.
    Mr. Clark. So in the Navy's defense they are working 
together on programs like that and I think, you know, programs 
like Ghost Shark are a great example of how commercially 
derived technology can sort of circumvent these ITAR regulation 
limitations.
    But I think, again, there's no prototyping environment for 
them to go and actually go play with these things and there's 
no demand signal to buy them.
    Dr. Greenwalt. But I'd argue the Navy is also cognizant of 
the fact is that if it does go in on Ghost Shark it ruins it 
for the Australians and ruins it for them because they will be 
ITAR'd.
    Mr. Gallagher. Clark is contractually obligated to defend 
the Navy. He's a homer. So, okay. I probably can't ask this 
next one in 25 seconds. Okay. Well, I'm going to do a third 
round anyway. So yeah. But do you want--do you want to ask a 
question?
    Mr. Khanna. I just have one question.
    Mr. Gallagher. Okay. I'll yield to you.
    Mr. Khanna. My question actually builds off what the 
chairman started out with, which is that we obviously need NSM 
and we need SM-6. We need these anti--long-range anti-air anti-
ship missiles. We also need the drones.
    Obviously, we need everything in some way. But how would 
you prioritize in terms of what is more urgent when you have 
limited resources in the allocation?
    Mr. Clark. Well, I'd say that we probably need to take some 
money, even if it means cutting into some of the munitions 
programs, and get these drone programs up and running because 
this provides a hedge against that kind of high-risk maybe low-
probability potential of an invasion of Taiwan.
    So if we can give Admiral Paparo what he's talking about in 
terms of a hellscape, even if that means we maybe don't buy 
three or four or five or six additional missiles, that may be 
worth it in order to get this hedge, you know, going and be 
able to create the potential for China of a more complex 
situation in the strait instead of adding kind of more of the 
same in terms of the fires network that we're already thinking 
of employing against that kind of force.
    Dr. Scharre. Yeah, I mean, I certainly--I don't think that 
the kinds of strike munitions that we need in this conflict 
should be the bill payer for this kind of effort and certainly 
we have seen historically that because of the flexibility in 
production that munitions are often a bill payer when you are 
looking for near-term money.
    My colleagues, Dr. Stacie Pettyjohn and Hannah Dennis, at 
CNAS [Center for a New American Security] have analyzed, you 
know, sort of by year the rapid volatility in munitions funding 
that then has strained the industrial base and has, you know, 
directly led to the situation we are in today when we try to 
scale munitions production for Ukraine or for deterring China.
    We're in a difficult place. But there are lots of other 
places that are bill payers in the Department, in ground 
forces, in legacy short-range TACAIR [tactical air] and lots of 
other things that are not going to be as relevant in this kind 
of conflict.
    Mr. Gallagher. Have you--by the way, have you seen any 
innovative efforts on containerized fires from DOD, recognizing 
we're in an unclassified?
    Mr. Clark. Yeah. Yes, I have seen them. I don't know if we 
can go in detail on that. Yeah.
    Mr. Gallagher. Okay. Interesting. Okay.
    In the testimony that I entered into the record from the 
Association for Uncrewed Vehicle Systems International it lays 
out that the U.S. currently lacks the industrial base needed to 
meet the goals of Replicator in part due to Chinese drone 
dumping, an illegal trade practice that's been flooding the 
global market with subsidized Chinese manufactured drones.
    It's estimated that Chinese drones account for more than 90 
percent of the consumer market, 70 percent of the enterprise 
market, and 92 percent of the first responder market.
    What steps can the United States take to revitalize our 
domestic drone manufacturing industrial base in order to help 
Replicator? I'll go reverse here and start with you, Dr. 
Scharre.
    Dr. Scharre. Well, certainly, this is a problem DOD has 
been working on for a while with the blue UAS [unmanned aerial 
system] program, trying to find ways to build up some kind of 
domestic commercial base.
    Part of the problem is even if you have drones that are 
made here all of the supply chain comes through China because 
the drone is, in essence, in many ways a flying smart phone. So 
I think that this particular problem is embedded into a much 
bigger problem of the U.S. tech relationship with China, deep 
degrees of entanglement, and finding ways to put in place 
incentives for industry to be moving production elsewhere, to 
partner or allied countries, is going to be really important.
    Dr. Greenwalt. I think we missed an opportunity in this 
technology, not from the Department of Defense standpoint but 
from the regulation of the FAA [Federal Aviation 
Administration] as far as encouraging the type of testing and 
development of technologies and to open the airspace to do 
this, from the Department of State, whose interpretation on 
missile technology control regime that led to the--not 
companies investing and, frankly, the Turks taking certain 
marketplaces and the Chinese taking the market.
    So we had--you know, we developed this technology. We were 
ahead 30 years ago and now we're not and it's going to take a 
whole lot of effort to try to overturn that. But it's a multi-
governmental agency effort. It's not just the Department of 
Defense. The Department of Defense can't drive this technology 
and drive the industrial base. It's going to have to come from 
other agencies' help as well.
    Mr. Clark. Yeah. I mean, it requires supply chain 
transparency. We need to do the analysis to figure out, well, 
what are the components that are coming from China and what's 
the risk associated with that because, you know, batteries, for 
example, are going to be probably--inevitably are going to come 
from China and we probably have to accept that. But other 
things maybe we don't want to have come from China.
    And then also, you know, creating that demand signal on the 
part of DOD for, you know, drones that are in this, you know, 
like Bill said the group two, group three size range where it's 
not your little DJI hobby drone but it's like a, you know, TB2 
or something like that that's militarily relevant.
    Creating a demand signal for those drones from the DOD will 
kind of force some of these controls to be put in place and 
help the emergence of a more stable U.S. domestic drone making 
market.
    Dr. Greenwalt. They're also going to have to export to our 
closest allies and because we--our demand alone is not going to 
drive this market. So we need to--we need to share and that's 
going to be a--goes back to the State Department again.
    Mr. Gallagher. Yeah. And ITAR specifically. Yeah.
    Well, I'm cautiously optimistic we can fix ITAR in this 
Congress. We're making progress for the first time in a while 
and I want to commend the efforts of Chairman McCaul on the 
Foreign Relations--Foreign Affairs Committee. It's a very 
fraught issue.
    A point of clarification. Okay. So Replicator is going to 
be coordinated through the Defense Innovation Steering Group--
DISG, another acronym--with the support of DIU and we have 
great new leadership at DIU with Doug Beck. Just tell me how do 
the combatant commands fit in and the services, or how should 
they, rather?
    Mr. Clark. So the way that innovation is being managed 
right now at the Department is the combatant commanders, in 
particular INDOPACOM, is driving the definition of operational 
problems to be solved.
    So that's the entering point is they provide the 
operational problem they want to address. That goes into the 
Defense Innovation Working Group and Steering Group and then 
they make decisions about prioritizing which operational 
problem to solve in which order.
    Mr. Gallagher. Okay.
    Mr. Clark. And then that goes over to then the kind of 
vague solution identification process and that's what I was 
referring to earlier. It's not well defined, not well supported 
necessarily.
    But then DIU is part of that and then will generate 
solution ideas and then DIU is responsible for going and 
procuring those solutions to get them to the warfighter.
    Mr. Gallagher. Interesting. Any other comments on that 
basic management structure?
    Dr. Scharre. I mean, I guess just we have seen this problem 
of this gap in the ability of the Department to respond to 
COCOM [combatant command] needs over the last 20 years. It's 
not a new one. There's been a lot of organizational solutions, 
JUONs [joint urgent operational needs] and JEONs [joint 
emerging operational needs] and other things. We'll see if this 
works.
    But I think that there are just huge structural barriers 
inside the Department towards, you know, orienting the, you 
know, RDT&E, the production processes, or driving material 
solutions to the immediate needs of the warfighter or the near-
term needs.
    I'm encouraged by the level of attention by senior 
leadership but they do have a lot of work to do.
    Dr. Greenwalt. I think from a managerial perspective you 
need to bring the operators in consistently as you do serial 
operational prototyping.
    In other words, minimal viable product goes out to the--to 
the warfighter, they test it, look at it, see if it's 
operationally useful or not, suggest changes. Goes back into 
the process.
    You prototype again, repeat, and this will be repeating 
over and over again and, frankly, should be repeating over and 
over again with software changes, with hardware changes. And 
the key thing that we have right now is this wall that we like 
to throw things over. And going to Congressman Moulton's idea 
of embedding, you know, the operator in certain areas is 
probably a real, real good one. The question is how do you do 
that.
    Mr. Gallagher. Go ahead. Did you have a comment?
    Mr. Clark. Yes. I would say the--this gets to the need for 
rapid prototyping to be done on a tempo that's much faster than 
what the DOD is currently planning, which is like two or three 
of these events per year.
    You know, that's not going to get you to where you need to 
be. And then the problem with the DIU idea of they're going to 
go procure these vehicles is who's going to actually do the 
prototyping and the evolution of these vehicles because DIU is 
more of a procurement agency, not a integration and R&D 
[research and development] type of agency. So I think those are 
challenges.
    Mr. Gallagher. Yeah. Ro, I got one more. I'm sorry, but 
you're going to like this one.
    Mr. Khanna. You're educating me.
    Mr. Gallagher. Well----
    Mr. Khanna. I'm, like, I'm listening to your terms and, 
like, I'm going to--I'm going to put those in my speeches. You 
know, make me sound good on defense.
    Mr. Gallagher. The speeches will put everyone to sleep if 
you do that. But you'll like this one. So I'll do--are we in a 
fourth round now? I don't know what. Whatever. I'll just keep 
asking questions.
    Okay. So let's say it's January 2025 and the newly elected 
President Ro Khanna comes to you and says, you know what, I've 
reviewed the intel and I think Xi Jinping was serious about 
this whole I'll take Taiwan by force if necessary and 2027 is 
the date. I just think that's--we got to plan against that. And 
I'm betting the farm--and you are each the Secretary of Defense 
in this fictional universe, which actually, I mean, there's--
you know, there's a probability this could happen.
    I'm betting the farm on Replicator to prevent this from 
happening. Like, you as Secretary of Defense, what do you need 
to make this successful in simple terms that the American 
people can understand?
    Mr. Clark, we'll start with you.
    Mr. Clark. Well, I'd tell President Khanna that I think 
we----
    Mr. Gallagher. Has a nice ring to it.
    Mr. Clark. It does. It does.
    So I think we need to, you know, quickly get the operators 
to define how they intend to solve this problem, how would they 
go about tackling the--you know, stopping the invasion. And I 
think it'd be a combination of slowing down the invasion with 
unmanned systems that are going to create the hellscape that 
Admiral Paparo talks about and then the joint fires network 
that's going to be used to then eliminate those ships before 
they're able to get combat power ashore.
    So that's the basic construct. What I need is the money to 
be able to go rapidly buy a bunch of unmanned systems that are 
going to enable me to gum up the Strait of Taiwan and I need to 
sustain the investment in the--in the munitions production 
that's going to get me those missiles to be able to eliminate 
the problem once they've been slowed down.
    Mr. Gallagher. Thank you.
    Dr. Greenwalt. Three things: money, production lines at 
full capacity, and then the third areas that we should already 
be working on is how to deliver these and so the issue is 
platforms. Are we going to do these out of P-8s? Out of 
sonobuoy? Are we going to do it out of C-130s? Are we going to 
create a new arsenal plane on a commercial aircraft that's able 
to deliver these type of things?
    Same thing undersea and on the surface. We should be 
thinking about those now and have these platforms ready to go 
and in 2025 we should just be cranking up full production lines 
on all of these things.
    Mr. Gallagher. Sorry. Dumb question to interject. Are the 
ranges such that you could even have them prepositioned in, 
like, southern Japan, northern Philippines, or does that not 
work in the same----
    Mr. Clark. Right. Oh, no, it's very feasible.
    Mr. Gallagher. Yeah.
    Mr. Clark. You could position these in Luzon, in the 
Sakishima Islands of Japan, or on Taiwan itself.
    Mr. Gallagher. Interesting.
    Secretary Scharre.
    Dr. Scharre. Yeah. So flexible money, lots of it, and the 
Department is going to need to have a very tightly integrated 
system of experimenting with these systems to figure out how do 
you use them effectively and then what are the upgrades you 
need. We see in wartime that that cycle is measured in weeks 
and months. We have these very quick innovation cycles.
    We see it in Ukraine. We saw it effectively for the U.S. in 
Iraq and Afghanistan with things like counter-IED [improvised 
explosive device] technologies where we saw this, like, very 
rapid innovation where we'd see the enemy do something, our 
folks would be innovating in our tactics and our materiel 
solutions, and the Department is going to need to have 
structures to do that that are ready to go in--ahead of a 
conflict to be effective.
    Mr. Gallagher. Any other questions?
    Mr. Khanna. That's good. It was very helpful and 
educational.
    Mr. Gallagher. Yeah. Thank you very much. This was a very 
productive discussion. Thanks for your analysis, your 
expertise, and your continued work on this important issue.
    And I would like to reiterate our desire and request for 
the Department to come testify on Replicator, ideally the 
Deputy Secretary, perhaps, with the head of DIU.
    Mr. Khanna. And I just echo that. You know, on the 
Democratic side we support that.
    Mr. Gallagher. Great. See, we're very bipartisan and we all 
look forward to the Khanna administration fixing all of these 
problems.
    Mr. Khanna. We just need a Speaker. Gallagher [inaudible] 
Speaker first.
    [Laughter.]
    Mr. Gallagher. No comment. The subcommittee hearing stands 
adjourned.
    [Whereupon, at 10:08 a.m., the subcommittee was adjourned.]     
=======================================================================

                            A P P E N D I X

                            October 19, 2023
      
=======================================================================


              PREPARED STATEMENTS SUBMITTED FOR THE RECORD

                            October 19, 2023

=======================================================================

      
   [GRAPHICS NOT AVAILABLE IN TIFF FORMAT]
      
=======================================================================


                   DOCUMENTS SUBMITTED FOR THE RECORD

                            October 19, 2023

=======================================================================

      

   [GRAPHICS NOT AVAILABLE IN TIFF FORMAT]

      
=======================================================================


              QUESTIONS SUBMITTED BY MEMBERS POST HEARING

                            October 19, 2023

=======================================================================

      

                  QUESTIONS SUBMITTED BY DR. McCORMICK

    Dr. McCormick. In order to make Replicator a success, Dr. Hicks has 
said that DOD must rapidly scale and field thousands of attritable 
autonomous systems by leveraging existing resources. One effective way 
to do that is to maximize existing DOD programs that have already been 
tested and are ready to scale up. Do you agree that DOD should leverage 
existing successful programs, like the Army's Short Range 
Reconnaissance Program and the Marine Corps' Short Range, Short 
Endurance Program to rapidly get small, smart, survivable systems to 
the warfighter?
    Mr. Clark. Yes. The DOD should take advantage of existing systems 
to achieve the scale necessary for Replicator to succeed. However, the 
systems chosen will need to be relevant to the operational problem 
being addressed. If the operational problem is, as DOD officials 
suggest, slowing or disrupting an amphibious invasion of Taiwan, 
Replicator should pursue a combination of undersea, surface, and air 
vehicles. That kind of multi-domain approach would create dilemmas for 
the invasion force that would require multiple different 
countermeasures to defeat. For example, by themselves, undersea systems 
could be swept like mines, surface systems could be defeated with guns, 
and small airborne systems are unlikely to cause substantial damage to 
large troop transports. Together, though, these systems can slow and 
disrupt the invasion and force the invaders to activate defenses that 
will make them more easily targeted by U.S. and allied long-range anti-
ship missiles.
    Existing defense programs like the Army's Short Range 
Reconnaissance Program and the Marine Corps' Short Range, Short 
Endurance Program could contribute to the counter-invasion effort. 
Although they are small, these systems would stimulate the enemy's 
defensive systems such as radars, which will be detectable and can help 
target U.S. and allied missiles. These small airborne systems can also 
confuse enemy operational pictures and decision-making.
    Most important, because they are already in the DOD inventory, 
these systems can avoid one of the likely challenges of Replicator, 
which is establishing training, maintenance, and logistics. In its 
current plans for Replicator, the Defense Innovation Unit has not 
addressed how it will support the new vehicles. DOD officials have 
implied Replicator vehicles would likely be sent to the military 
services, which will provide for their support. However, this approach 
risks creating a set of disconnected units that will not be able to 
integrate into a coherent system of systems, which will be necessary to 
address the likely operational problems faced by INDOPACOM.
    Dr. McCormick. As you know, since the Cold War, the Department has 
leaned toward developing and fielding expensive, exquisite weapons 
systems. The ever-increasing cost of these programs, coupled with 
breakthroughs in technology, have finally forced DOD to instead focus 
on small, smart, mass-produced platforms, like drones, that offer 
greater operational flexibility. How do we ensure that Replicator 
doesn't turn into yet another program for acquisition of expensive, 
large, long-range systems?
    Mr. Clark. Much of the reason the DOD has gravitated toward 
expensive and large systems is the way requirements are established for 
new weapons. The DOD generally defined requirements for new systems by 
analyzing the needs for U.S. forces to succeed in projected future 
confrontations. These analyses assume U.S. forces will use a planned 
set of operational concepts and tactics that are similar to how U.S. 
forces would operate today. As a result, when threats improve the 
capability needed to continue current U.S. operational approaches also 
must improve. To break this cycle, U.S. forces will need to consider 
alternative operational approaches that can employ less-sophisticated 
systems.
    Replicator is attempting to take this path to enable the use of 
less-expensive systems that can be produced at greater scale. Instead 
of defining requirements by analyzing future operational scenarios, 
Replicator is starting with contemporary operational problems and 
available systems. Instead of keeping the operational concept fixed, 
Replicator would instead vary the way U.S. forces operate to enable 
available systems to be effective in addressing the operational 
problem. By starting with available systems, Replicator can avoid the 
``requirements creep'' of previous defense programs.
    Dr. McCormick. Mr. Clark and Dr, Greenwalt, given Replicator's all-
domain requirement, it is clear that there will be multiple vendors 
with different autonomy algorithms, also known as stacks. For 
Replicator to be successful, the DOD will need to integrate various 
autonomy software stacks into a single platform backbone to enable 
continuous integration and continuous development cycles.
    1. How is the Department of Defense planning to integrate the 
autonomy software of various vendors across domains and ensure success 
of the program?
    2. How does the Department plan on ensuring communication and 
interoperability between these platforms?
    Mr. Clark. The software stack for Replicator will likely end up 
similar to the figure below:
[GRAPHIC NOT AVAILABLE IN TIFF FORMAT]

    .epsAt the top of the software stack is the software ecosystem, 
like the iOS system for iPhones or the Windows operating system on 
personal computers. This ecosystem is shown in gray in the figure above 
and would incorporate the autonomy and command and control programs for 
a group of Replicator vehicles. Like iOS or Windows, the ecosystem 
would establish interfaces between autonomy software for individual 
vehicles and the ecosystem as well as for software routines in the 
ecosystem, such as mission planning programs that establish vehicle 
behaviors. The ecosystem software would also provide a way to build and 
assess operational concepts through live/virtual/constructive 
simulations and evaluate the ability of new Replicator units to 
integrate into the force.
    Each Replicator vehicle will have its own vehicle control software 
and software associated with onboard mission systems such as radios and 
sensors, including algorithms for automated target recognition. This 
software is shown in yellow or green in the figure above. To minimize 
the integration challenges associated with combining Replicator 
vehicles into a system of systems, the autonomy software within 
vehicles will likely be vehicle-specific and separate from the autonomy 
software of the overall ecosystem. Vehicle autonomy software will 
interact with ecosystem command and control and mission planning 
software via the ecosystem's government-owned interfaces, but the 
software for vehicles and the ecosystem will likely be built and 
maintained by individual vendors.\1\
---------------------------------------------------------------------------
    \1\ This approach is described in Bryan Clark and Dan Patt, 
``Hedging Bets: Rethinking Force Design for a Post-Dominance Era,'' 
(Washington, DC: Hudson Institute, 2024), https://www.
hudson.org/defense-strategy/hedging-bets-rethinking-force-design-post-
dominance-era-bryan- clark-dan-patt.
---------------------------------------------------------------------------
    To address the two questions above:
    1) How is the Department of Defense planning to integrate the 
autonomy software of various vendors across domains and ensure success 
of the program? Replicator will likely use government-owned software 
interfaces, like that between the Apple iOS operating system and iPhone 
apps, to allow software from vehicle providers and ecosystem developers 
to be independently built and shipped. To enable this approach, the 
government will also need to establish a development environment that 
enables automated testing of the executability of new software and its 
integration with the Replicator ecosystem and command and control 
software.
    2) How does the Department plan on ensuring communication and 
interoperability between these platforms? As shown in the figure above, 
the different vehicles involved in Replicator will interact with each 
other through the overarching software ecosystem and its government-
owned interfaces, similar to how a ride-sharing app on a smart phone 
can access the contacts in the user's contact list or saved locations 
in a mapping program.
    Dr. McCormick. Mr. Clark and Dr. Greenwalt, the Department of 
Defense's announcement of Replicator caught many in government and 
industry by surprise. While Congress commends the Department of 
Defense's initiative to deploy autonomous vehicles to offset China's 
advantages, there are questions about the programs ability to validate 
the safety and trust thousands of drones within two years.
    1. Do you believe the Department of Defense can fully test and 
validate the safety and trust of thousands of drones within two years? 
If not, how can the testing be accelerated?
    2. Should the Department of Defense work with industry to 
accelerate the testing and evaluation of these drone at scale?
    3. Do you believe commercially proven modeling and simulation 
capabilities could help the Department of Defense stay on timeline?
    Mr. Clark. The Replicator initiative is an ambitious effort that is 
designed to stress the DOD's ability to deliver new capabilities at 
scale in an accelerated timeline. However, the Pentagon has succeeded 
in similar efforts in the past, including the production and delivery 
of Mine-Resistant Ambush-Protected (MRAP) vehicles to U.S. and allied 
troops during Operation Iraqi Freedom. Like that effort, fielding new 
systems at scale will depend on a willingness to use operational 
innovation to employ available technology, flexible requirements, and 
abbreviated testing.
    1) Do you believe the Department of Defense can fully test and 
validate the safety and trust of thousands of drones within two years? 
If not, how can the testing be accelerated? Yes. The most significant 
challenge associated with fielding Replicator could be testing. In 
large part, this is because the initiative tacitly assumes that tactics 
will change in response to available vehicle technology and changing 
operational problems. As a result, comprehensive testing of Replicator 
systems in representative operational settings will be difficult.
    The DOD is trying to address this challenge by implementing more 
virtualized testing, in which a simulation using digital models 
assesses if a new system will be operationally effective and 
sustainable. Under this approach, live testing is used in part to 
validate digital models, which are then used to evaluate the viability 
of various operational concepts and associated systems of systems in 
which the Replicator vehicles could be employed. This approach could be 
used to accelerate the testing associated with vehicles purchased under 
the Replicator initiative.\2\
---------------------------------------------------------------------------
    \2\ Director of Operational Test and Evaluation, ``DOT&E Strategy 
Implementation Plan - 2023,'' (Washington, DC, U.S. DOD, 2023), https:/
/www.dote.osd.mil/Portals/97/pub/reports/DOTE _Strategy_Imp_Plan-
Apr2023.pdf?ver=jQHyC5uHXsvM25sYurv5Zw%3D%3D
---------------------------------------------------------------------------
    2) Should the Department of Defense work with industry to 
accelerate the testing and evaluation of these drone at scale? Yes. To 
support the automation of testing as described above, the DOD could 
work with industry to build digital models or ``twins'' of uncrewed 
vehicles as they are developed that can be used in simulations.
    3) Do you believe commercially proven modeling and simulation 
capabilities could help the Department of Defense stay on timeline? 
Yes. Using model-based system engineering, the DOD could automate some 
aspects of the testing process.\3\ For example, program offices and 
operational units could use simulation to assess different operational 
concepts and system of system configurations that address an emerging 
operational problem. With digital model or twins, testing of the new 
system of system could be conducted automatically. The simulated test 
could also highlight areas of uncertainty in the digital model that 
need to be checked through live testing. By live testing the system 
against the limited points needed to check the model, the program 
office could validate the system's utility in much less time compared 
to the comprehensive testing usually conducted today.
---------------------------------------------------------------------------
    \3\ Undersecretary of Defense for Research and Engineering, ``DOD 
INSTRUCTION 5000.97 DIGITAL ENGINEERING,'' (Washington, DC, U.S. DOD, 
2023), https://www.esd.whs.mil/Portals /54/Documents/DD/issuances/dodi/
500097p.PDF?ver=bePIqKXaLUTK_Iu5iTNREw%3D%3D
---------------------------------------------------------------------------
    Dr. McCormick. In order to make Replicator a success, Dr. Hicks has 
said that DOD must rapidly scale and field thousands of attritable 
autonomous systems by leveraging existing resources. One effective way 
to do that is to maximize existing DOD programs that have already been 
tested and are ready to scale up. Do you agree that DOD should leverage 
existing successful programs, like the Army's Short Range 
Reconnaissance Program and the Marine Corps' Short Range, Short 
Endurance Program to rapidly get small, smart, survivable systems to 
the warfighter?
    Dr. Greenwalt. The Department should first begin consideration for 
initial Replicator funding those programs that have already been tested 
and fielded with the operators. For the initial tranche of Replicator 
programs there is a need to begin with those capabilities that already 
have a level of maturity and producibility. For a second tranche, DOD 
should look at the creation of new capability that can be delivered to 
the warfighter in less than 18 months for testing and evaluation so a 
subsequent production decision could be made. Finally, there should be 
plans for additional tranches and additional technologies that could be 
developed and used in the 3+ year timeframe.
    What will be significant to consider for these programs is not just 
the ability to operate today but to have the flexibility to mix and 
move in and out different types of sensors and weapons packages over 
time, so as we are not buying a static technology. For some programs, 
there will be a need for these platforms to be modular and have open 
systems interfaces to allow for medium- and longer-term adaptability. 
For others there may not need to be any upgrades except perhaps for 
software.
    Dr. McCormick. As you know, since the Cold War, the Department has 
leaned toward developing and fielding expensive, exquisite weapons 
systems. The ever-increasing cost of these programs, coupled with 
breakthroughs in technology, have finally forced DOD to instead focus 
on small, smart, mass-produced platforms, like drones, that offer 
greater operational flexibility. How do we ensure that Replicator 
doesn't turn into yet another program for acquisition of expensive, 
large, long-range systems?
    Dr. Greenwalt. To keep that from happening will need effective 
congressional oversight. Unfortunately, since the end of the Cold War, 
the type of oversight that Congress has relied on has created 
incentives for the creation of those expensive exquisite weapon systems 
in the first place. This is primarily driven by how Congress 
appropriates funding for large weapon systems or MDAPS and how DOD 
budgets under the PPBE process.
    There is a need for a new paradigm for the types of programs 
Congress funds that should be time-based. The development process and 
the evaluation of the success of programs should be limited by time and 
a fixed budget. There should be a commercial or existing technology 
preference to fielding systems. Finally, there should be a maximum use 
of rapid acquisition (RAA), middle tier authority (MTA) and tailored 
Other Transaction Authority (OTA) projects that are limited by time to 
development. Production at scale should only be funded when a program 
is mature, but limited production can happen under RAA, MTA, and OTA 
authorities. The key thing is not to spend time and money solving 
problems in production. Rather, a program should either meet its time-
based development objectives or be cancelled. Better to write off money 
spend in development than waste excessive amounts of money in 
production on capabilities that will never work as advertised.
    Dr. McCormick. Mr. Clark and Dr, Greenwalt, given Replicator's all-
domain requirement, it is clear that there will be multiple vendors 
with different autonomy algorithms, also known as stacks. For 
Replicator to be successful, the DOD will need to integrate various 
autonomy software stacks into a single platform backbone to enable 
continuous integration and continuous development cycles.
    1. How is the Department of Defense planning to integrate the 
autonomy software of various vendors across domains and ensure success 
of the program?
    2. How does the Department plan on ensuring communication and 
interoperability between these platforms?
    Dr. Greenwalt. I am not sure how DOD will be doing this. It has 
many options--establish standards, outsource the problem to a large 
prime, be its own systems integrator, or rely on its own trusted 
private integrator as the National Reconnaissance Office (NRO) once 
did. Regardless, the management and integration of this software will 
be a significant challenge for DOD.
    The problem with each of these options is it requires DOD to be 
more effective in incorporating not only commercial software and 
technologies but understanding the best practices of commercial 
software and information technology trends. The role of a 
knowledgeable, neutral, and objective systems integrator who does not 
have any hardware or software skin in the game may be critical and DOD 
should take a look at the old NRO systems integration model to achieve 
this objective.
    DOD has been lagging the commercial market in the information 
technology space for at least three decades. It is a key weakness in 
the defense acquisition process and is one for Congress to spend time 
overseeing and being a forcing mechanism to get DOD to incorporate the 
best practices that exist in the commercial market place.
    Dr. McCormick. Mr. Clark and Dr. Greenwalt, the Department of 
Defense's announcement of Replicator caught many in government and 
industry by surprise. While Congress commends the Department of 
Defense's initiative to deploy autonomous vehicles to offset China's 
advantages, there are questions about the programs ability to validate 
the safety and trust thousands of drones within two years.
    1. Do you believe the Department of Defense can fully test and 
validate the safety and trust of thousands of drones within two years? 
If not, how can the testing be accelerated?
    2. Should the Department of Defense work with industry to 
accelerate the testing and evaluation of these drone at scale?
    3. Do you believe commercially proven modeling and simulation 
capabilities could help the Department of Defense stay on timeline?
    Dr. Greenwalt. 1. Plain and simple the answer is no. The Department 
will not be able to do the type of testing that it is used to doing if 
it wants to move systems faster to deployment. Currently the testing 
approach revolves around providing a system that has been in 
development for decades and has been produced at low-rate initial 
production numbers for several years while the testers provide their 
inputs unconstrained by time or practicality. That will have to change 
and the Department and particularly DOTE is not ready for that type of 
change in testing frameworks. In any new time-based developmental 
environment, minimally viable products will be the first to go to the 
warfighter in operational situations. Different testing criteria will 
be needed for feedback to improve those products over time while DOD 
benefits from the near-term operational usefulness of these products.
    2. Given the lack of internal resources and capabilities in the 
organic DOD testing community, it may be appropriate to consider 
contracting with industry for neutral, unconflicted testing services.
    3. Yes, the Department needs a new testing paradigm and modeling 
and simulation can offer a tool to accelerate this testing. It also may 
need different testing standards. For example, instead of its current 
operationally suitable standard it may need to substitute an 
operationally useful standard as defined by the warfighter not the DOTE 
bureaucracy. In addition, if we test to FAA manned flight safety 
standards for drones rather than something more realistic and 
appropriate, we will fall behind our adversaries. M&S and other 
commercial tools and approaches will likely need to be used at scale if 
DOD is going to innovate at speed.
    Dr. McCormick. In order to make Replicator a success, Dr. Hicks has 
said that DOD must rapidly scale and field thousands of attritable 
autonomous systems by leveraging existing resources. One effective way 
to do that is to maximize existing DOD programs that have already been 
tested and are ready to scale up. Do you agree that DOD should leverage 
existing successful programs, like the Army's Short Range 
Reconnaissance Program and the Marine Corps' Short Range, Short 
Endurance Program to rapidly get small, smart, survivable systems to 
the warfighter?
    Dr. Scharre. To the extent that DOD has existing programs or 
prototypes that are proven, ready to be scaled, and meet mission needs, 
of course DOD should leverage these to achieve Replicator's goal of 
fielding thousands of attritable, autonomous systems. DOD has said that 
it will prioritize the mission needs of the Indo-Pacific theater, and 
the urgent need for capabilities to deter and, if necessary, defeat 
Chinese aggression is clear.
    The two programs in question, the Army's Short Range Reconnaissance 
Program and the Marine Corps' Short Range, Short Endurance Program, aim 
to provide short-range ``back-packable'' aerial drones to ground troops 
for ``over-the-hill'' surveillance. These are vital capabilities to 
support ground troops and provide them organic, tactical reconnaissance 
and surveillance capabilities. The short range of these systems will 
limit their utility in a China contingency in the Western Pacific, 
which would be a predominantly air and maritime fight. Small, aerial 
drones would likely have some value for Taiwanese troops in the event 
of a Chinese invasion of Taiwan. The war in Ukraine has demonstrated 
the value of small, cheap drones in supporting ground troops, even in a 
major interstate war with high levels of violence. However, other 
capabilities such as mines and anti-ship cruise missiles should be a 
higher priority for Taiwan's military, as they will be most relevant 
for defending Taiwan against a Chinese invasion.
    Dr. McCormick. As you know, since the Cold War, the Department has 
leaned toward developing and fielding expensive, exquisite weapons 
systems. The ever-increasing cost of these programs, coupled with 
breakthroughs in technology, have finally forced DOD to instead focus 
on small, smart, mass-produced platforms, like drones, that offer 
greater operational flexibility. How do we ensure that Replicator 
doesn't turn into yet another program for acquisition of expensive, 
large, long-range systems?
    Dr. Scharre. Indeed, this is a major risk. One of the potential 
failure modes for Replicator is that DOD is unable to be disciplined in 
cost and requirements, and systems become too expensive to be purchased 
in large numbers. There is a natural tension in any program between 
moving quickly and accepting a ``good enough'' solution and taking 
longer to design the ideal system. For major weapons platforms that 
cost billions of dollars and which the DOD will use for decades, such 
as aircraft carriers, it makes sense for DOD to take its time to be 
deliberate in scoping the requirements and design of the system. If DOD 
applies the same methodical approach--which is valuable in some 
settings--to Replicator, it will fail. In order to move quickly, DOD 
will need to accept the technology as it exists today. This will 
necessarily mean that the systems that DOD fields under Replicator, if 
the initiative is successful, will be less than ideal. They may not be 
the ideal solution to meet DOD's needs over the long term. Keeping 
costs low will force difficult tradeoffs in system design. Moving 
quickly will mean making do with the state of technology today, not 
waiting for a future improved version. The end result may be systems 
that have limited shelf-life and are replaced after a few years. This 
would be a success. While this is uncomfortable for DOD--and indeed may 
be uncomfortable for some in Congress as well--these tradeoffs are 
necessary to meet the urgent need to deter Chinese aggression. DOD will 
need the support of members of Congress to move quickly, be disciplined 
in cost and requirements, and be willing to accept the risk of systems 
that are less than ideal in order to field a ``good enough'' solution 
to meet near-term operational needs.
    Thank you for the opportunity to respond to these questions.

                                  [all]