The New Iron Triangle: Achieving Adaptability and Scale in Defense Acquisition
United States Representative, First District of Virginia
Deputy Director, Strategy, Policy, and National Security Partnerships, Defense Innovation Unit
Principal Military Deputy to the Assistant Secretary of the Army for Acquisition, Logistics, and Technology
Senior Vice President, Army Programs, Palantir Technologies
Executive Vice President and President of Mission Technologies Division, HII
Chief Executive Officer, Ursa Major
Chief Strategy Officer, Saab
Chief Engineer, Epirus
Vice President for Government Relations, Varda Industries
Senior Fellow
Nadia Schadlow is a senior fellow at Hudson Institute and a co-chair of the Hamilton Commission on Securing America’s National Security Innovation Base.
Senior Fellow and Director, Center for Defense Concepts and Technology
Bryan Clark is a senior fellow at Hudson Institute. He is an expert in naval operations, electronic warfare, autonomous systems, military competitions, and wargaming.
This event is part of the Apex Conference Series.
The war in Ukraine offers numerous lessons regarding the future of military operations. One of the most important—and most underreported—is the value of adaptation. Ukrainian troops, previously on the offensive thanks to Western precision weapons, are now on the defensive as their rockets and bombs miss targets due to Russian electronic warfare. In the Black Sea, Ukraine’s early naval losses suggested Russian dominance. But lethal new naval drones have restored Ukraine’s access to the open ocean and constrained Russia’s fleet to its own coastline.
The United States and other North Atlantic Treaty Organization militaries will likely face a similar challenge in future confrontations against Russia, China, or their proxies. Merely stockpiling today’s weapons or expanding their production capacity could lock in obsolescence against technologically sophisticated sophisticated opponents. US and allied militaries will need an industrial base that can both modify today’s weapons or combat systems and produce them in volume—then be prepared to repeat the cycle in response to enemy countermeasures.
Join Hudson and the Apex Conference Series for a three-part event discussing the challenges and opportunities facing Western militaries and defense industries as they attempt to achieve relevant capability at scale.
Agenda
12:00 p.m. | Remarks and Fireside Chat
- Rob Wittman, United States Representative, First District of Virginia
Moderators
- Bryan Clark, Senior Fellow and Director, Center for Defense Concepts and Technology, Hudson Institute
- Dan Patt, Senior Fellow, Center for Defense Concepts and Technology, Hudson Institute
12:45 p.m. | Lunch
1:15 p.m. | Panel 1: The DoD’s Efforts to Achieve Relevant Capability at Scale
- Aditi Kumar, Deputy Director, Strategy, Policy, National Security Partnerships, Defense Innovation Unit
- Lt. Gen. Robert M. Collins, Principal Military Deputy to the Assistant Secretary of the Army for Acquisition, Logistics, and Technology
- Mitch Skiles, Senior Vice President, Army Programs, Palantir Technologies
- Andy Green, Executive Vice President and President of Mission Technologies Division, HII
Moderator
- Nadia Schadlow, Senior Fellow, Hudson Institute
2:15 p.m. | Panel 2: Industry’s Efforts to Develop New Approaches to Adapt and Scale
- Joe Laurenti, Chief Executive Officer, Ursa Major
- Michael Brasseur, Chief Strategy Officer, Saab
- Michael Hiatt, Chief Engineer, Epirus
- Josh Martin, Vice President for Government Relations, Varda Industries
Moderators
- Bryan Clark, Senior Fellow and Director, Center for Defense Concepts and Technology, Hudson Institute
- Dan Patt, Senior Fellow, Center for Defense Concepts and Technology, Hudson Institute
3:00 p.m. | Reception
Event Transcript
This transcription is automatically generated and edited lightly for accuracy. Please excuse any errors.
Bryan Clark:
All right, we’re going to go ahead and get started if everybody wants to grab a seat. Excellent. Well, thank you all for being here, and thank you to those online who are watching. Welcome to the Hudson Institute. I’m Bryan Clark. I’m a Senior Fellow here at the Hudson Institute. I’m Director of the Center for Defense Concepts and Technology here. We’re here to have an event today that we call The New Iron Triangle, looking at ways to expand adaptability and scale in the defense industrial base.
And to explain a little bit about the premise here. So program managers in the defense department often talk about an iron triangle of cost, schedule, and performance that govern how they go about building a new program, an airplane, a ship, whatever. What we’re finding today is that we’re experiencing a need for increased defense production. We’ve heard lots of arguments both here and abroad about how defense production is has kept up with the needs both to fight current wars like in Ukraine, but also to deter future wars such as in the Indo-Pacific.
But program managers are challenged to achieve that scale. At the same time, we’re seeing in places like Ukraine, the need to adapt. Ukraine received a lot of western equipment, Javelin missiles, JDAM bombs, HIMARS rockets that are now being countered by Russian GPS jamming and other electronic countermeasures.
So they’re having to adapt despite the fact that they got the highest end military equipment the west could provide. So adaptability and scale are now becoming the buzzwords or the watchwords for future defense acquisition. We can’t just adapt through innovation. We can’t just produce scale through enhanced production. We’ve got to somehow figure out a way to adapt and then scale the production of those things that we developed through adaptation.
So that’s we’re going to try and focus on today is this need for a combination of adaptability and scale and thus far that the DoD in particular has been challenged to achieve that we have a lot of great innovation efforts going on through places such as the Defense Innovation Unit, which we’ll talk a little bit about today with Aditi Kumar, and also on the services side so General Collins from the Army will talk about that.
And industry has been trying to support them, but those innovation efforts have not necessarily yielded the kind of scale we need to be able to prepare for a future fight potentially against China and to certainly support the current fight against Russia and Ukraine. So we need to come up with a way to innovate and then scale those things. There’s a need for new models of defense production. There’s new models for acquisition and we’ll talk about those over the course of today.
So thank you very much for being here for that. I also want to shout out our partners in this, the Clarion Defense, which puts on such great conferences like the DSEI Conference in London and in Japan. They’re partnering with us on this and they are putting on a conference next year called APEX in January of 2025.
And we’re working with them because one of the ideas behind APEX is this idea of achieving adaptability and scale. How can we produce other kinds of capabilities we need and then evolve them over time in response to a changing security environment? So thank you very much to them and we’ll hear a little bit more from them later on today.
But to start, our first discussion is going to be a panel discussion or rather a small panel discussion between me and Congressman Rob Wittman. Rob Wittman is a representative in the US House from the first district of Virginia. He is the Vice Chair of the House Armed Services Committee and the Chair of the Tac Air/Land Subcommittee, which has actually had the largest portfolio within the House Armed Services Committee. He’s also on the Natural Resources Committee where he does a lot of great work at protecting Chesapeake Bay and has been diligently working for a decade or more trying to improve the ability of the Defense Department to both produce at scale as well as adapt to a future environment.
So join me in welcoming Rob Wittman to the stage and we’re going to have a discussion.
Rob Wittman:
Thanks, Bryan. Appreciate it.
Bryan Clark:
Well, thank you very much, Congressman Wittman.
Rob Wittman:
Thanks, Bryan. Appreciate it. Thank you.
Bryan Clark:
Thank you for being here. So to start with, obviously it’s a very busy time. Over on the hill we’re doing final markups and approval of the Defense Authorization Act for 2025 and obviously a lot of other activity up there as well. Can you lay out for us where you see one, the challenges that we’re facing in the US military and our allied militaries, and then what some of the opportunities are to address those?
Rob Wittman:
Sure. Well, we’re in the process now. The Rules Committee has met, they’ve made their final determinations of the almost 1400 amendments that have been put in. There are 350 that have been accepted in what’s called In Order. Many of those will be on block. It looks like about 58 of those amendments will be debated. So we’ll go through that process today and tomorrow and then ultimately final passage vote on Friday.
So our desire is to make sure we get that passed. The last thing I want is for it to not pass and then us have to go through a suspension process. I think that sends a horrible message to not only our members of the military but also the nation.
And the challenge for us is to do several things. First of all is there is an issue with recruitment and retention. So there’s a lot of focus on this year’s NDAA with quality of life, and that includes raising the pay of our junior enlisted, making sure we’re taking care of families, understanding the nuts and bolts about what is necessary to make sure we retain those members of the military and what they do, but also create the incentive for us to recruit the best and brightest.
And you can’t do that by saying, “Well, the starting pay for a junior enlisted is $23,000 a year.” Which is $11 and 50 cents an hour, and we’re asking people to raise their right hand and swear to uphold and defend the constitution and put their lives on the line for $11 and 50 cents an hour. They look at it and go, “I can go down to Chick-fil-A and serve chicken sandwiches in heating and air conditioning and get healthcare benefits and make $16 an hour. Where’s the motivation there?” Now listen, many people have a greater sense of contributing, but that’s a big issue for us to be able to take care of.
The other is how do we modernize at the speed of relevance? And there’s been a lot of great concepts that have been projected by the service branches and by the Pentagon. Things like Replicator, things like how do we produce expendable systems, attritable systems? So the concepts are there and even some of the elements of RDT&E, research development, technology and engineering, and looking at how we bring those concepts to light.
But what’s lacking is the execution. And that’s been, unfortunately, the problem with the Pentagon through the years is great concepts, lots of talk, lots of back and forth and some great science experiments. But really the key is how do we get these things to scale quickly and how do we not let perfect be the enemy of the good? How do we make sure that we say “We’d rather have the 70 percent solution in the next 30 days than the 100 percent solution five years from now?”
And the problem is we have to overcome that paradigm. In closing with this, the Pentagon, since its very inception, has been a hardware-centric organization. Great at going through years of writing requirements and acquiring big exquisite platforms. And we’ve been used to living in a world where the exquisite platforms have been the overwhelming superiority that we’ve enjoyed. But if nothing else, we’ve seen what’s happened here in recent history, especially in Ukraine, and know that it will not be expendable platforms by themselves that allow us to make sure we counter the threats.
It will be how do we quickly and effectively feel expendable systems that can be literally tuned on a daily basis based on the nature of the threat. And then how do we produce attritable platforms? Because the most valuable asset in the United States military is not the hardware, it’s not really even the software, it’s our men and women in the military.
How do we find ways to put them in situations where they don’t have to be in harm’s way? I can think of many situations where there’s no need to put a human being out there to collect ISR. There’s no need to put a human being out there to do jamming and in some instances not even need a human being to go out there and launch weapons. Now they need to be in the loop. So the question is how do we do those things? And again, the Pentagon needs to go from a hardware-driven organization and now be software-driven. Software needs to inform everything that the Pentagon does. Anything in the hardware realm needs to be designed around what the software can do and then we’ll actually be able to operate the speed of relevance.
Bryan Clark:
Right? Absolutely. Those are excellent points, which we’ll dig into in the course of this conversation. So when I mentioned the idea of the Iron Triangle, the old Iron Triangle cost-schedule performance is all based on a set of requirements that came over from the operational community and they were treated like Moses tablets. We don’t change the requirement. We just work on trying to satisfy that need.
And obviously as you’re talking about with a 70 percent solution, it kind of implies a New Iron Triangle where the operator’s got to be part of that discussion. So the operational concept, the tactic is one corner of the triangle and then maybe relevant capability and relevant capacity are the other two, as those are the things you’ve got to meet as a program manager. You’ve got to have enough of them, they’ve got to be good enough and they have to satisfy an operational concept of the war fighter.
So for that new triangle, how do we get the operators to be more involved to be able to define what is that 70 percent solution?
Rob Wittman:
Bryan, that’s a great question. If you look historically when acquisition was done, you had the operator was in the same room as the requirements writer who was in the same room as the acquisition professional. So they all talked to each other and they knew basically what the platform was about, what was the challenge that the operator was facing so there was this conversation.
Today, the operator’s over here, the requirement writer’s over here, and they’re reading what they believe is the mission objective and they go, “Well, this is what I think it is.” And then in the acquisition professional’s over here reading the requirement and go, “Well, I think this is what they mean with the requirement.” So the operator says, “I need an apple.” And then the requirement writer writes it up and it’s really an orange. And then the acquisition professional gets it and goes, “We need to acquire a pineapple.”
And then the operator gets the pineapple and goes, “What the. . .? This isn’t what I asked for.” And then they look at it and go, “It’s not what we need. It’s not going to be operational.” So what do they end up doing? They’re very innovative. So they say, “Well, let’s take the pineapple, let’s cut the top off of it. Let’s cut the rind off of it and we can get it to somewhat look like an apple and maybe function like an apple.” That’s really where we are today.
You have to make sure you flatten that and you also have to make sure that time is of essence. You have to look at it and go, “As much as we strive to have these perfect systems, the world is operating at a very fast pace, especially as these threats morph and develop.” And as we see adversaries are pretty innovative. It used to be we looked at it and said, “We have the intellectual capacity here, the monetary capacity, all the requisites to make sure that we can predominate there.”
And we see today that those assumptions really aren’t as critical as timeliness is. And we see that folks that don’t have the amount of resources that we have or don’t necessarily have the massive amount of intellectual capacity can be pretty creative, can be pretty imaginative, and they come up with things quickly that can counter what we do.
Listen, there are lessons learned even prior to what’s happening there in Ukraine. There should be lots of lessons learned from what happened in Iraq and Afghanistan. We saw very rudimentary systems defeat massively expensive and complicated systems. And then we finally said, “Well, maybe we ought to go back to simplicity and speed and cost to be able to do that.”
So the New Iron Triangle really is going to be how do we create a system? How do we make sure to that timeliness is part of that? Cost is going to be part of it too. The competition today is not just with what China does, but the way we won the Cold War is we just out-resourced Russia. We had a bigger checkbook and Russia wasn’t interested in the economic side of things. China’s very different. China says, “We’re going to defeat you at every level. We’re going to go after you economically, we’re going to go after you strategically, we’re going to do things faster and better than you.” Because they start out with a blank sheet of paper.
When you go to the Pentagon and try to do acquisition, your paper starts out with a bunch of nos. “No, you can’t do this. No, you can’t do that. No, you can’t do this.” And you have to find pathways through that massive paper with a bunch of nos on it. So we have to find ways to be able to innovate and create.
So the new triangle needs to be timeliness. Not just cost, but how do we get value? How do we get more for our dollar than the Chinese get for their yuan? And then how do we make sure to that we change the paradigm that is based upon operational capability through execution. So execution has to be part of that New Iron Triangle.
Bryan Clark:
So you brought up software centricity, which is something we’ve talked about a lot here. The role that it plays in making for more adaptable force, and that’s not just both within the skin of an aircraft or a weapon, but it’s also between weapons and trying to create new force compositions that allow you to be effective even in a contest environment where an adversary is trying to defeat you.
So JADC-II or CJADC-II, the Joint All Domain Command and Control Initiative is the mechanism by which the DoD is trying to create this more recomposable force. Obviously it hasn’t gone swimmingly, they haven’t yielded a lot of output. So what do you see as some of the key elements though that could enable that more recomposable force? Obviously software is a part of it. Are there other elements there that we need to achieve the objective of JADC-II, even if it’s not accomplishing it?
Rob Wittman:
Well, listen, the concept of JADC-II is great, CJADC-II and we need that, but again, it’s an area where the concept is ahead of where we need to be to enable that. So you look at what we do right now to communicate lots of information, ISR information, targeting information, the system’s called Link 16. Now Link 16 is a pipe that’s about this and the amount of data that we need for JADC-II, and the coordination of all these systems, the pipe necessity is like this.
So the question is if you’re going to do CJADC-II, maybe you to build the pipe bigger first before you look at all this enabling software and things that need to be done. So first of all, we have to make sure that there’s bandwidth and duplicity there and bandwidth because you know the first thing our adversaries are going to do is to take down the ability for us to link those things. So how do we create layering there?
That’s why this year under the Constellation Security Program that Space Force has, which is called CASR, what we’re doing is we want to put that on steroids. We want to make sure that we have layering there. We want to get lots of LEO satellites, lower earth orbit, mid-earth orbit GEO satellites we want to get many of those in the private sector to have bandwidth reserved for the United States military. So that way we have layering and that way too, we have enough pipe to be able to move information around.
Second of all is how do you make sure that you have systems that can talk to each other? And that’s a key right now because a lot of legacy systems aren’t even designed anywhere close to, whether it’s the frequency itself or whether it’s how they exchange information. And I think too, what the Pentagon needs to do is they need to narrow the scope of what they want to try to accomplish initially with CJSC-II and CJADC-II, listen, they want to do all these things and it is great. Conceptually, it’s great. We want to coordinate, we want to make sure you can use different weapons systems, those kinds of things.
Let’s start here. Let’s start here. Let’s make sure we have a pipeline to just do a good job in gathering, targeting information, and then we can figure out how to get targeting information to different places. Let’s just make sure that it does a basic component, really good first. And if it does that, that’s the building block to say, “Okay, now let’s take the next step.” Instead of, again, the 100 percent solution that becomes very, very difficult and very difficult to achieve. Instead of saying, “Maybe let’s start with it being an excellent targeting data platform.” That’s how they need to go about it. So I think they need to draw in the scope of where they’re trying to go with CJADC-II.
Bryan Clark:
And it seems like they’re starting to do a little bit of that in terms of a bottom-up approach to building out different kill chains as opposed to trying to make the entire force recomposable and every shooter in every sensor can be working together instead say, “Well, which kill chains do we need to do first and building those up?”
Rob Wittman:
Exactly. And how do we make sure too, how do we understand the sensors? How do we modernize them? How do we get them to talk to the combat systems out there out there that can deploy that. The human being’s going to be in the loop there somewhere, so don’t expect too much performance out of it to begin with. Where it can really be a value is with all the sensors we have, all the data we have, how do you make sure you can synthetically annotate that data, put it in a form that’s targetable or targeting data and then get that to the war fighter? That would be a massive improvement just by itself.
Bryan Clark:
And I think people don’t. . . they lose sight of that because I think we think with all the sensing data that’s out there, all the kind of ubiquitous commercial sensing, you’d think, “Well, I can see everything on the surface of the water or the earth.” Which may be true, but that’s not target quality data.
Rob Wittman:
That’s right, exactly.
Bryan Clark:
The difference between those two things is enormous, and that ends up being a big roadblock to a lot of the operations that DoD wants to do or US military wants to do.
Rob Wittman:
And it’s essential with our kill web, especially as you look at where we are today with magazine depth, our magazine depth on long-range, precision weapons is horribly shallow. So you cannot afford to be firing any of those weapons without really high-grade targeting data. So what you don’t want to do is to go, “Well, this data is good enough,” because guess what? If you miss, then you create a double effect on what your challenge is on the next shot, and the next shot.
Bryan Clark:
And going back to the idea of adaptability and scale, what that highlights is if you’ve got better space-generated targeting data, that might give you a lot of flexibility in terms of what your weapons might look like and I employ weapons that maybe aren’t quite as sophisticated as maybe the preferred ones of today? And I can use unmanned vehicles that don’t necessarily have the autonomy that you’d want in a fully autonomous vehicle, but they can get the job done because they’ve got this great access to targeting data coming from outside.
Rob Wittman:
That’s it. Bryan, you talk about mass, when you have that ability to get that targeting data, all of a sudden mass and expendable platforms means something. And listen, the Chinese have looked at that and said, “Listen, we’re going to generate mass and capability.” Our challenge is to look at capability and capacity and say, “How do we generate that in mass?” Because that has an effect and it doesn’t have to be an exquisite platform.
I would argue the best way for us to generate mass is through expendables and attritables. I think the greatest opportunity for us as we look at challenges in the tactical aircraft space is CCAs. That to me is the gap closer that we can do quickly. I am excited that the Department of Defense has decided to award the initial contracts to two smaller companies, to innovative companies. So they’re getting away from the paradigm that everything has to be large-scale companies that. . .
Listen, I don’t want to in any way, shape or form criticize them. We need them, but we also need other smaller companies that are innovators and creators. So the second tranche of awards on CCAs, the remaining competitors will be part of that. So we need everybody to be involved in this. The gap closer there is how quickly and effectively can we do CCAs? I would argue that what they do to enable an F35 if we can ever get the improvements to F35 done, we’ve done significant things to close the gap.
Bryan Clark:
Right. So CCA, that’s a great point on that. And the fact that they’ve gone to two private companies that are putting a lot of their own money to work in that I think is an important point. Do you think that that’s a model we need to look at more going forward when we start to try to field more unmanned systems that are going to be 70 percent solutions? Are we going to depend on industry doing a lot of that legwork to develop that system, anticipate what the war fighter might consider to be good enough, as opposed to getting this into a long R&D cycle where the department’s managing this centrally, this idea of industry independently coming up with potential solutions?
Rob Wittman:
Listen. Bryan, there’s already an incredible amount of work that’s been done by these small companies out there and they have very, very innovative and creative ideas about how to do this, and they’ve actually fielded those systems and tested them. Why do we want to recreate the wheel? Why don’t we want to say, “Let’s compete with those companies that already have those platforms there.” And the Navy’s starting to do some of that with unmanned surface platforms. Let’s do that to scale. And if you’re going to do that, take some risk and say, “Listen, it’s not just one company, but maybe there’s a half a dozen companies out there that all do really good things. Let’s feel those systems. Let’s do it masse.”
Here are the two case examples that we ought to look at. When we saw the need for unmanned aerial systems, it was Congress that said, “Let’s do it.” And guess what? We did multiple company acquisition. When Congress said “We got to have something that protects members on the ground from IEDs in Iraq and Afghanistan.” We did acquisition on MRAPs. And instead of saying, “This one company does it.” Guess what? We had a half a dozen companies do it and they manufactured a thousand MRAPs in six months.
Same paradigm needs to be used for these unmanned systems. We’ve got a lot of great companies out there. Let’s take some risks. Let’s put a bunch of them out there and let’s figure out, “Well, this one works really well, this one not as much or maybe we can make a change to that one.” If we’re going to generate mass quickly, it cannot be with a single company because remember, a single company is also a single point of failure. So let’s take some risks, let’s feel those platforms. Let’s get a bunch of them out there. If we’re really going to expand the industrial base to have the capacity we need, you’ve got to be able to acquire expendables and attritables and even for that matter, exquisite platforms from multiple companies.
Bryan Clark:
Right. And so one of the challenges that comes up with that is transitioning those into use and actually fielding them and sustaining them. And we’re going to hear from DIU here in a little bit. And one of the challenges they have is they’re going to buy a bunch of systems under replicator, but then how do they get those in the hands of the operators somewhere out in the Indo-Pacific for example.
So how should the department be thinking about that challenge of fielding these systems? Especially when you think about fielding a joint system of systems, which you can’t just give it to the services because then it’s got to be brought back together by some poor operator in the field. Do we need to have a different construct for how do we get these systems out there in a joint unit to start with?
Rob Wittman:
I think what you need to do is you need to be able to engage the operator essentially at the unit level. So let’s take a soldier that’s operating at the squad, platoon and company level and say, “Okay, how do you and your mission objective for your unit, what are the things that you would need in a UAS?” And then let them with whatever platforms are out there, let them experiment with it.
And in general, George I think is this concept in mind to be able to do that, his transformational effort there to actually do it at a tactical level, and that’s great. And then you get a lot of feedback. So the operator tells you, “This is what we need, this is good, this is not so good. Let’s add those things in.” You have to be able to do that on a pretty fast pace and on a quick trajectory to be able to get that to the point where now you can feel that in mass.
And what we did in this year’s NDA and it got through is to create a drone corps in the Army. And listen, you’ll hear from the Army folks a little bit later on. Now listen, they’re resistant to that. They go, “Listen, we’re already doing that. It’s at the operational level. Our soldiers are going to develop this and we don’t want to this structure within the army, we think we can do it here.”
The problem is this is, I don’t disagree with the concept of saying “We have to develop these operational concepts and what needs to be in those platforms.” That needs to happen at the unit level. But the key is is that if it happens at the unit level and every unit’s going to have their own little idea about what needs to be in that UAS or that counter UAS system. So the question is then how do you acquire it in a timely way?
There needs to be some entity there that says, “Okay, we’re taking all this data in and we believe the best combination of that is to do this combination of things and we can get it out there and it can do. . .” Again, 80 percent of what the operators want it to do instead of starting with the apple and giving them a pineapple. That’s where I think things need to be. So what we want to do is to make sure that there is that level of what we call the drone core, where there’s a centralized effort to bring all this information in from the operators and go, “Okay, now we’re going to operationalize this. Now this is how we can bring this to scale quickly.”
And what I get concerned about is we’ve watched the service branches and all of them have done that through the years, and that is they do a great job in doing experimentation with a lot of things, but the challenge is to get across that valley to actually get it acquired.
And the example is SOCOMs figured this out. Special operators have said, “Hey, listen, we get it. We know what we want and we know how to acquire it.” Why? Because they come up with a paradigm that says, “We do fast acquisition. It’s part of who we are, it’s part of the nature.” And that’s how they operate. The question for us should become, why is that the exception? Why shouldn’t it be the rule? How do we design a system to where all of a sudden we do that just as a matter of fact and how we operate? So you have to be able to merge both of those.
So you have to have the drone core concept and still have the transformational activities that General George talks about that are bottom up, that are operator up. We can do that. We can do both. You can’t just say, “Oh no, we don’t want drone core. We’re going to do all this organically.” Because the concept sounds good, but operationalizing it in mass is always lacking.
And listen, I love the army. They do a great job, but they’ve also had a pretty challenging history with aviation acquisition. You go back to Comanche, you go back to all the things that they’ve tried. You look at FARA this year, discontinued FARA. Listen, by the way, I think that was the right thing to do. I wish they’d have done it a year earlier, but they are saying “We’re going to do UAS and counter UAS.” So there are good things happening there. The question is how do you integrate all of those things, both at the operational level and at the acquisition level?
Bryan Clark:
So it seems to argue for the idea that the program manager needs to be part of that discussion. So you can’t have the program manager off, “Well, tell me when you’re done and I’ll get to you.” They’ve got to be in the middle of this conversation between the operators and the industry people.
Rob Wittman:
They all have to be in the same room. Listen, the requirements writer needs to be there. In fact, I would argue in some instances, you don’t even need a requirement to be written. You need to make sure the acquisition person is there. So the acquisition person needs to be embedded at the unit level. So if you’re going to do that in timely ways, the acquisition person needs to be there. They need to be out in the field talking to the operators as they’re experimenting with these things. If you want to do things at the speed of relevance, that’s the way to do it.
Bryan Clark:
Right. And the SOCOM model was sort of based on the idea that, “Well, SOCOM’s worried about near-term problems.” And so the technology of today is all that matters to them and that the services are worried about the longer-term challenges that the DoD faces. But technology has probably changed that, right? So technology now means that I can both worry about today’s problem and evolve that to address tomorrow’s problem.
Rob Wittman:
Software centric and give me the 70 percent solution tomorrow. I don’t want the 100 percent solution a year from now.
Bryan Clark:
Right. And so you’re bringing up software, and I wanted to ask you about that. So the DoD is trying to do a better job of buying and paying for software through new software acquisition pathway, new software appropriation, but it seems like they’ve been slow to adopt it in any meaningful way.
How does the DoD need to be able to embrace the kinds of software-centric approaches that we’ve been advocating, that you’ve been advocating? Do we just need them to embrace these new models? Does it need a different relationship with industry, which might need to be continuously building systems or software for them?
Rob Wittman:
Well, I think you start there. You have to understand what are the capabilities that software can even bring to the tables. You have to understand the scope of what’s there, and then you have to be able to say, “Okay, now how do we write an acquisition strategy around that?” And that’s difficult because software tomorrow is different than what it is today, especially if you’re going to adapt and it’ll be different the day after.
So how do you create enough latitude in that contract where you allow the operator to talk to the acquisition person and then the customer can then go to the company that delivers the product, the software, and say, “This is how we want it to change.” That’s a very difficult paradigm. And where it’s also difficult too is that as authorizers get more and more of that in mind and what we see needs to be enabled through the military, listen, we get it. And we’re trying to push in that direction.
The challenge is is appropriators don’t like any flexibility. “We want to tell you how to spend the money.” And listen, I understand because they feel they’re responsible for doing that, and that’s part of their reach and part of their power. But also on the congressional side, this has to be a team play. This has to be the Pentagon, it has to be the executive branch. It has to be the legislative branch that say, “Okay, we’re going to give you some flexibility because we know what you’re acquiring is a very dynamic product.”
And listen the private sector does it all the time. A great example and Doug back there at DIU is the perfect way to be able to transcend this outside of DIU to the rest of the Pentagon, is Apple had a perfect strategy to be able to do that. Apple looked at it initially, Stephen Jobs said, “We’re going to make the iPhone and we’re going to do everything internally. We’re going to do it ourselves. We’re going to develop the software and the hardware.” And they figured out that there’s some really good stuff out there that they themselves didn’t develop.
And when they tried to solve these problems, a great example is when you dropped your iPhone and the screen cracked, they got all these returns because the screen cracked. They go, “Well, let’s come up with a glass that’s resistant to fracture.” And they found out that they could do all this experimentation, but there’re actually companies out there that are doing it now where they could acquire that much faster and much less expensively. So look at Apple’s model and what they do, they’re looking at it and go, “We’re not going to try to be the monopoly on every technology in our phone, but you know what we’re going to do? We’re going to find the best technology and we’re going to be the systems integrator and we’re going to integrate that into our phone. And the iPhone is still going to be the best phone out there.” Now listen, Samsung customers, Samsung thinks the same thing. So I’m not making a judgment on those phones, but that’s how those companies operate.
Doug gets it, and what Doug I think needs to be able to do from DIU is to expand that concept about how you take the best of what’s out there and actually integrate that into systems. The government, instead of being the acquisition central place that determines everything that says, “We have to make it, DOD has to make it. And if we don’t make it, then it’s not what we need and we’re going to drive all the requirements.” And instead of that, the government needs to say, “You know what? We need to do a better job in acquiring software and we need to be the systems integrator.” So we need to think, where’s the best software? And it may be a combination of software. Maybe this software does this really well, this does that very well. How do we combine those? How do we team some of those companies that do that? And then how do we team to the hardware that’s out there and how do we make sure the software enables the hardware?
If we do that and when we do that, that’s how we get to the place we need to be to operate at the pace of relevance. And that’s a tough place to go because the Pentagon is very structured, there’s a very static model about how they do acquisition. And it all begins with, again, that requirement writing. And I look at MTAs and OTAs, again, appropriators get really mad when you use OTAs because it’s like we don’t like to give all this flexibility in the acquisition. But OTAs and MTAs need to become more the practice than the exception. Changing the acquisition paradigm is necessary.
Bryan Clark:
So when you’re buying software, I think one of the things too is the idea that you might be able to have multiple companies competing to deliver the same software. Because one of the challenges that the appropriators are worried about is, “Well, once I get locked in with a software provider for this capability, I’m stuck with them forever. I’m not going to have this ability to pivot to another provider because somebody’s already so entrenched.” But if you could have multiple providers delivering the same kind of software-
Rob Wittman:
Yeah. They ought to encourage teaming. Listen, we do it for Virginia class and you bring the best of those two companies. Why don’t you say, “Hey, listen, we’re going to put an RFP out there and we’re going to score those based on the highest level of innovation and creation that you bring to the table. And if you want to partner with another company that does this well and you do this well, and another company does this well, we’re going to score that even higher.” So you’re going to encourage teaming. And then all of a sudden, how does innovation and creation happen? A number of different minds working on it. And this company may say, we got this viewpoint. And you combine that with another viewpoint, all of a sudden you go, “Wow, I didn’t even know that was possible.”
Bryan Clark:
Right. And you can recompete it so that you can say, “Well, we can change the teaming arrangements as you go.”
Rob Wittman:
Exactly.
Bryan Clark:
And so the last thing I want to bring up before we go to audience questions is this idea applying that model to other parts of the supply chain, if you will. So one of the obviously strengths of Apple is its supply chain understanding and strength of providers.
Rob Wittman:
Yes. Right.
Bryan Clark:
So for DOD, we tend to build very kind of narrow supply chains for each of our systems. Do program managers need to be given an incentive to basically create a more resilient supply chain, even if that might mean the system costs more, maybe the performance is going to be less than what optimally could be provided if you had the exquisite set of providers.
Rob Wittman:
Well, Bryan, I think the most fundamental thing that needs to happen is certainty, certainty in the acquisition realm. And Congress is at the very forefront of creating uncertainty. Why? Because every year we adopt CRs, there’s no way that you can stay on a known path if you do CRs every year. It wastes money. It doesn’t allow you to advance at the speed of relevance. So Congress needs to get its act together. So if you’re going to have those innovative and creative opportunities, there has to be certainty for folks.
Listen, people are willing to invest. There’s a ton of private equity and venture capital dollars on the sidelines there that said, “If you send us the demand signal, we are going to do that.” But you can’t be up and down. And then two, you have to make sure that there’s continuity in what happens. And a great example is this year in the Virginia class submarine program. We’re now in AUKUS and AUKUS is going to require United States sell some submarines to Australia. And we’ve already said we have to be producing on average 2.3 submarines a year, which means you really have to do on a cadence of three, two, three, two. So this year we came back in and the President’s budget comes over with one submarine and the Australians go, “What the. . .” And we said, “You’re right. We have to stay at least two.” And the industry is trying to recover from COVID, so they’re adding new people in the workforce, so they’re ramping up to be able to get to that too. And so we’re doing that, but then it looks like the appropriators are going to go back to funding one.
So how do you go to the Australians and say, “We’re serious about that.”? How do you make sure that the industry gets the demand signal and say, we are going to ramp up and hiring people and developing that skill set? And I don’t care what you say. Well, some of this will go to AP. Listen, there was $3 billion in the Taiwan supplemental to go to submarine industrial base. That is the necessary step to be able to build that portion of it. But the only way that you’re going to get to that production rate is you can’t say, “Well, we’re going to fund some AP and some ship sets and that’ll be the substitute for another boat.” No, it’s not.
I mean, these yards are not going to hire people on for the promise of another boat next year. And by the way, when you lose a year of building another boat, you’re never going to make that up. So even if you say, “We’re going to go back to two next year.” So the industry goes, “Well, now we got to bring people back,” and it’s not going to happen. We also have to send a constant demand signal and we have to stay on track. Because what happens now is that industry isn’t only trying to battle the uncertainty of a CR, but they’re also trying to battle the uncertainty of spotty acquisition strategies where the Navy says, “Let’s build two. We need 2.3, but let’s only go to one next year.”
And then it becomes a self-fulfilling prophecy. And the reason they say is, “Well, we don’t think the industry can do it.” Well, duh, if you don’t send the demand signal, the industry’s not going to do it. So if you send the demand signal, guess what, and demand that from the industry, I think the industry can perform. So again, it becomes a self-fulfilling prophecy. So it sets my hair on fire when I hear the Navy say those things. But that’s where we need to be. And Congress needs to hold the Pentagon’s feet to the fire to say, “No, you are going to do this.” So we’re kind of working through with the appropriators how to make sure that happens.
Bryan Clark:
A component of that too is are you designing your program and are you managing your program in a way that tries to increase the number of suppliers that are possibly able to deliver components for your program, and are we willing to deliver the kind of money that is necessary to incentivize those suppliers to come on board and stay on board?
Rob Wittman:
Right.
Constant demand signal and making sure those suppliers can look out in their windshield five years, 10 years and go, “Yeah, I’m going to continue to do business there.” And not say, “Well, I’m going to have to manufacture one ship set. This ship is going to be out the door in two years. What happens after that? I don’t know.” So those companies most of the time say, “You know what? I’m not in that game. I’ll go somewhere else. Do something else.”
Bryan Clark:
So I want to open up to questions from the audience. So if you have a question, please raise your hand and Morgan will bring the microphone around to you. And if you can state your name and affiliation and keep it to a question, not a comment. Oh, and I’ll start with you, Vago Muradian.
Rob Wittman:
Yes.
Bryan Clark:
That wasn’t intended just for you.
Rob Wittman:
Vago.
Vago Muradian:
Vago Muradian with Defense & Aerospace Report. Terrific event, off to a great start. I have two questions, but they’re interlinked. Sir, great comments, US Air Force is looking at a more consolidated requirements approach to both be able to siphon up good ideas from the bottom, but also enforce discipline and figure out crossover as opposed to each of the individual branches of the Air Force doing this. Do you like that overall approach as maybe a model for the other services, given that nobody has a consolidated requirements? We’ve tried this with Futures Command, is that a template? And I have a follow-up.
Rob Wittman:
Listen, I do like that because it’s driving the idea of writing requirements or needs from the bottom up. So you use it at the operational level, you look at the tactical needs there, you bring that up. But then you have a unified strategy to say, how do we then acquire in a timely way in its scale and with value? And those are the three metrics that have to be done. And that’s why our idea about the Drone Corps in the Army is it can’t just be an operational idea of up flow. It has to also be, now how do you combine those? How do you synthesize all those things to do acquisition in time and in scale? And that has to happen.
Listen, the Air Force has got a pretty good record of doing that. You look at our Air Force RCO and B-21 acquisition and things that they did there, they were great at systems integration there, and they were embedded with Northrop as Northrop started. And listen, when I chaired CPOWER, I was somewhat skeptical. I went out to Palmdale there with Northrop, with Air Force RCO, and I said, I don’t know if you guys can pull this off. They were just hiring engineers there. And I thought, this is a big enterprise, most advanced aircraft ever in the world. Can you pull this off? And to their credit, their credit, because they were embedded with Northrop, everything that happened happened at the speed of relevance, all of a sudden they were able to do this, and then acquisition followed what they knew operationally would work.
So Air Force I think is trying to copy that paradigm is to make sure that now they can scale that. So I think that’s the way to go, is how do you get the operational level impact? How do you make that scalable? How do you make sure too then that becomes how the service branch operates. And again, it’s value, it’s timeliness and it’s innovation.
Vago Muradian:
The follow-up I was going to ask was about the cyber corp. There is a sense, for example, I mean, had the Army been using air assets, no criticism of the Army, we wouldn’t have created an independent air force to be able to use air assets differently as opposed to tying them to units. Now we’re saying, okay, there’s a recognition of these unmanned assets. But we are looking at individual purposed organizations, whether it’s a drone corp, whether it’s space force, what are your thinking on whether or not we need a cyber corp? Because cyber is more granular and interspread through each of the fabric of the services and indeed conjoined increasingly with electronic warfare through General Schmidle’s innovations. What’s your sense on whether we need a cyber corp or better integration actually across the fabric of the force? Thank you very much.
Rob Wittman:
Well Vago, that’s a great point because where we are right now is we have really changed. It used to be the service branches everybody had their own hardware, their own software, even using floppy disks in the day. And now what’s happening is everybody’s going to the cloud. And listen, the cloud is a great asset, but to think that somehow there’s no threats within the cloud is also pretty myopic. And the key is if everybody’s operating in the cloud and we know how that platform operates, are there commonalities that we can put forward to say, how do we protect that exchange of data? How do we protect what we do in the cloud? How do we make sure that cyber protection is there?
So yes, there is application to do a Pentagon-wide effort to look at how do we put those basic protections in? And I would argue the way the military operates today, which used to be in buckets, this was your operating platform, this was your way that you used and kept data is very different than today. We do data on the cloud.
And then another thing too is that how we are changing today is as data started out in being converted to a usable form, it was human annotation. So you take a bunch of photographs of intelligence that you’d gather and you say, “Well, this is a Yuan class destroyer China.” And so you’d have to put all those together. So human beings would’ve to look at that. Where we are today is being able to use synthetic data annotation, but the question then becomes is then how do you protect that? There are all kinds of aspects of what we do digitally that if you don’t protect that and somehow that gets corrupt, and your adversaries say, “I don’t even have to build new ships. All I have to do is to trick the system to think that this is a Chinese ship that has capability.”
So there is a place. . . And that goes across all the aspects because every service branch will be using that data that comes out of that. So how do we make sure we do those things? And we have to be able to think in front of where the threats would go. And I would argue too that the cyber realm is not just an effort in what we do on the defense side, but we have to be more thoughtful, and I would argue more aggressive on what we do with offensive capability. And I know that we have it, most of it’s classified, but the question is how do we make sure that that is an integral part of what we do and how do we make sure too that we are more thoughtful about how we could apply that? And I would argue that is a strategic asset, and it is in many instances, much more effective than a kinetic effort.
And if we do that in a smart way and understand what the threats are, we can get out in front of the threats. And you talk about deterrence, if you want the biggest deterrence for how our adversaries are developing capability, offensive cyber capability is right there. So I think that has to go into how do you do that in a uniform sense? Because all the service branches are going to experience the same sorts of threats and the same capability that they need to build.
Bryan Clark:
And that’s doing a project for DOD right now on the non-kinetic capability supply chain, basically how do those capabilities get provided to the war fighter. Which there’s a lot of improvement that could be achieved there in terms of getting scale and adaptability in terms of what we can afford there. But yeah, that’s where a lot of the fight’s going to happen, especially if you look at dissuading a fight rather than fighting the fight.
Rob Wittman:
It is going to be in that realm, and I would argue major conflicts could be avoided if we get that right.
Bryan Clark:
Other questions? Yes. Yes. Right there.
Michael Hiatt:
Hi, Mike Hiatt from Epirus.
Rob Wittman:
Hi Mike.
Michael Hiatt:
Question for you on adaptive acquisition framework. Do you feel that that is doing the job of getting that timeliness that you’re talking about? And do you think, are there more ways that we can help smaller companies sort of bridge that infamous valley of death to get through that so that private equity money sitting on the sidelines see that it’s possible? Thank you.
Rob Wittman:
Well, listen, the adaptive process I think has started, but it is not happening fast enough. And what I want to make sure is that those small companies out there that don’t even know that they might have a product that is a great solution to the challenges that we face, how do they make sure that they can at least see into what the Pentagon needs? And we’ve been advocating for years that the Pentagon needs to have a windshield for those folks that aren’t intensely familiar with the acquisition process there. To where somebody can put up whatever their product is, software, hardware and say, “Hey, this is what we have,” and have some place there where the military can say, “Well, listen, we have these situations where maybe this could be applied.” And to have sort of a trial place where someone can says, “Well, this is what it can do,” to be able to search out and understand what are the possibilities out there right now.
It’s a go to the mountains. So they’re expecting people to come to the mountain instead of the mountain going to the people there. DIU is trying to say, let’s be the search component. Listen, and Doug and his group do a great job. They’re searching every day to try to find those innovators and creators. And listen, they spread a pretty big net to try to bring those in, but they can’t cover everybody and they can’t really be out there in front of what’s developing at the speed of relevance. So I think we have to have those things in addition to DIU in being able to discover where is this capability and capacity today and where may it even be at a developmental stage to say, “Well, maybe we need to get out in front and really understand what’s going on with that.”
So I think that those are things that we can do, and the acquisition strategy needs to be multifold, it can’t just be DIU or even OSC with Jason Maranci. It has to be how do we create a windshield in the Pentagon for somebody to say, “Well, I think I’ve got a solution. Let me try it.”
Bryan Clark:
Thanks. So do we have one time for one more question Hallie, or do we have to go? Oh, okay. Well then we definitely should let you go.
Rob Wittman:
Yes, thank you.
Bryan Clark:
We’ll leave it there. But join me in thanking Congressman Rob Wittman for being with us today.
Rob Wittman:
Thank you folks. Thank you. Thanks, Bryan. Thank you so much. Thanks for the opportunities.
Bryan Clark:
Thank you very much. So for everybody else here, we’re going to be taking a break until 1:15, so everybody can go get lunch and bring it back in here. We can get started at 1:15. We’ll have a panel with Aditi Kumar from Defense Innovation Unit, General Collins, assistant secretary of the Army for Acquisition and Logistics and technology, as well as Andy Green from HII and Mitch Skiles from Palantir in conversation with our own Nadia Schadlow. So we’ll see you all at 1:15. Thank you.
Rob Wittman:
Thanks.
Nadia Schadlow:
Thanks for taking time out of your busy day to be here, we really appreciate that and appreciate everyone in the audience as well. I thought I would start with General Collins and overall ask a few questions. We’ll try to have a conversation and then I’ll turn to the audience for the last 15 minutes or so for questions.
So General Collins, tell us a little bit about what the DOD is doing vis-a-vis Ukraine, Ukraine is an example that keeps coming up today but clearly it’s one of the most important ones in terms of the subject of today, modularity, adaptability, the ability to produce at scale, and what are we learning from Ukraine in terms of also addressing some of these problems of adaptability and scale?
Lt. Gen. Robert M. Collins:
Yeah, great question. And first, thanks for hosting this forum, allow us to have this dialogue and being up here with these participants. And I think it’s a great question. I think the strategic importance, certainly with Ukraine, I think obviously it has also allowed us in the department specifically in the Army to learn a number of lessons. We’ve got a number of modernization focuses. I’d say two that come to mind that have allowed us to really hone in on the munitions of fires and also when you think about defensive posture, air defense, those are probably two areas in particular.
When it comes to, part of the theme I think, is doing things at scale, producing velocity, doing things quickly. There’s the saying that production at scale is also a form of deterrence in itself. And so I certainly think that we’ve learned from a number of these areas that whether it’s government, munitions facilities, the defense industrial, there are probably some areas where we need to look at things that we can become more sophisticated and move into the digital age. Probably heard a lot, Javelin, Stinger, our guided multiple launch rocket system, Patriot. I’d probably hone in maybe a little bit on some 155 artillery, it’s probably been one particular area. Probably about a year ago. I wasn’t necessarily an expert on how to assemble a 155 artillery round. There’s a lot of things that go into it from the projectile to the casings to the energetics and all the raw material.
And I think that what we’ve learned is that there is a degree of sophistication that we need to look at and how we can move faster at scale, more multifunctional. I’ll specifically comment down in Mesquite, Texas, some of you may have seen the article down there. We just opened up a universal artillery production line. And one of those things does is it adds in things like autonomous vehicles to be able to move steel, robotic arms to be able to do things for the safety of folks that used to have to touch labor. When you start doing the forming of a projectile using software to be able to tune. So things like that allow us to move more quickly. And then I’d also say gives us a little bit of flexibility so that we don’t necessarily have to do one type of munition. We’ve got some versatility at how we do that. So those are some areas that I think that this has allowed us to learn and to be able to invest in our industrial base so that we’re ready to do things at scale.
Nadia Schadlow:
Was that factory stood up relatively quickly? It seems like it was. And a theme of this morning very much in the previous conversation was time, right? So it seems like that’s a big, actually, that’s a positive point in terms of DOD acting quickly.
Lt. Gen. Robert M. Collins:
I think one of the things that we’ve learned with Ukraine, and we’ve kind of put ourselves, I would say on a war footing, that we’ve been able to learn things not only about requirements, but acquisition contracting. We’ve been given some very special authorities to be able to move quickly, but that one in particular with the help of Congress, some of the supplemental dollars, we were able to take that facility to bring that capability and stand that up. In fact, we were all just down in Mesquite, Texas about two weeks ago, and they’re going through their first article testing. So that ramp and we’re route to a tenfold increase in our 155 artillery of what we had been producing about a year or two ago. So I think as you look at the speed of what we’ve done in past, I would put that as moving very rapidly.
Nadia Schadlow:
Amazing. Aditi, I’d like to ask a little bit about what DIU is doing in terms of allies and partners. We spoke about that a little bit offline and you were describing some interesting things to me.
Aditi Kumar:
Yes, thank you. And one point, just to extend what Lieutenant General Collins was just saying, we have done a lot of work with Ukraine and in that space on autonomous systems as well, in addition to traditional munitions, which you talked about. A lot of lessons learned on uncrewed systems and the use of those systems by both parties. And I believe the last time that I had the privilege to be here, I had just come back from Warsaw, Poland where we hosted a UAS conference with Ukrainian warfighters telling us what lessons they were learning on the battlefield. And the primary lesson was that we need to be able to not only upgrade the hardware, but definitely upgrade the software on those systems and that cycle that they described, which was shocking to me at the time, was every 90 days. Every 90 days is when we need refreshes.
And so actually just last week, the team was back in Poland in Kraków this time, and we were doing the second iteration of that UAS conference. And in addition to of course, dialogues, which are many very, very meaningful, we actually did a hackathon focused on, okay, how could you deploy something like this to the systems that we have in theater? And we actually had companies working with us to do, for example, last mile navigation for FPV drones that we can then use to integrate into the hardware solutions. And so these are all lessons learned about hardware, software integration and modularity that we have definitely picked up from Ukraine and in partnership with the Army and the other services are now incorporating into how we view autonomous systems going forward.
And of course, in addition to Ukraine, DIU in its current growth stage is focused a lot on deepening relationships with allies and partners more broadly and becoming the Department’s liaison when it comes to commercial tech adoption not just at home, but with our allies and partners. And that is taken on a few flavors, and we’re really executing that by embedding our folks in the DIU equivalents of our closest allies and partners. And so we have very strong partnerships, for example, with AUKUS nations, with India, Japan, Singapore, others, where we are working towards personnel exchanges so that we can really share information in real time, putting in place challenges where two nations together put out a requirement for something that both of us, or in the case of AUKUS, all three of us, have a need for.
We just did a challenge with AUKUS partners focused on electronic warfare, all three nations put out the same statement. We on the US side had about 50 companies apply, we’re in the down select process. And we’re starting to build those bridges internationally when it comes to commercial tech sector and creating sort of those ecosystems of founders, funders getting to know each other and us working across borders.
Nadia Schadlow:
So in a way, you’re setting a foundation for the systems integration across with allies and partners similar to what Bryan and Representative Wittman were talking about earlier.
Aditi Kumar:
Yeah.
Nadia Schadlow:
On one hand, it could be very complex, though. You could be adding complexity. We’re having our own problems within the DOD, but.
Aditi Kumar:
I think there is a broad recognition that we need to leverage the collective industrial base to tackle these problems. And so there is a lot of work to do in creating the international agreements to be able to do that, but we’re paving the way so that we develop the familiarity with our tech ecosystems, as I said, and we work towards these contracting opportunities that really materialize and scale across borders.
Nadia Schadlow:
And so speaking of AUKUS, General Collins, and then I’ll turn to Andy and then Mitch.
Lt. Gen. Robert M. Collins:
Well, I would just underscore we will always fight as a coalition or with allies and partners, one. Two, the more that we can share technically interoperability, training, all those procedural type things make us a more effective force. I think too, there are things that we can share. We’ve looked at co-production agreements. As you start to look at the luxury of a contested logistical environment, you can’t always, getting things to the point it need or having those things that are in place. Certainly when you start looking at particular types of regions.
I’d also say too, as we look at things for foreign military sales, it offers us an opportunity as you start looking at production and scale, as we ramp up some of these production lines, the more that we can have and other partners that are looking at US capabilities, I think those are things that can help us even from our industrial perspective. And so I think there’s a number of opportunities that we’ve learned with Ukraine and a whole host of other things that we can do across the FMS arena, and then taking early looks at what are those things from an exportability area that we just need to take into consideration and making sure that we’re doing our due diligence. And so it’s just kind of getting after that. And I think we’ve certainly seen increases in our foreign military sales over the last 12 to 24 months. So I think it’s a great point.
Nadia Schadlow:
And Andy, you’re right in the middle of FMS and Australia.
Andy Green:
Yes. By nature, Huntington Eagles is in the middle of AUKUS. But when you think about Pillar 1, that’s one thing with the submarines and the infrastructure and so forth. But when you think about Pillar 2, that’s when a lot of Aditi and General Collins, what y’all were talking about, we’ve got to smooth the pathway for technology transfer and do it in an orderly way, but we’ve got to be able to do it in an expedited way. So every last change or discussion or whatever, it doesn’t take six or seven months of an approval process. So I think y’all hit it right on the head.
I’m glad you brought up a contested logistics because it kind of extends into that contested logistics environment where you’ve got to have that ability to share information with our coalition partners very quickly and safely and securely, and that will kind of build off of these relationships that we’re creating through AUKUS.
Nadia Schadlow:
How is that linked to adaptability and modularity and scale? How are some of our partners thinking about that and how are you thinking about it as you AUKUS? And then, Mitch, I’ll ask you to comment on the software side.
Andy Green:
Well, we’re thinking about it in terms of adaptability and modularity kind of go hand in hand. And when you think about some of our coalition partners, they have the same challenges that we have. They’re kind of burdened by the old style as Mr. Wittman and as Bryan said, the old way of acquiring big platforms over a very long term. So we’re going to have to be more flexible, they’re going to have to be more flexible, and we’re going to have to be able to do that together and in sync and very quickly in theater. And so I think when you think about modularity and adaptability, I think that designing programs and designing platforms that facilitate that, and we’ll talk about it here in a little bit, Aditi and I talked about it. The Lionfish program is a great example of the type of program that is good for the US. It’s very modular, very adaptable, and would be just a fantastic program to also get our partners engaged in.
Nadia Schadlow:
Okay. Want to give Mitch a chance to jump in, but then maybe we can go back and you can tell the audience a little bit more about Lionfish. So Mitch, from the software side of things.
Mitch Skiles:
Yeah, I mean we’ve had the unique privilege of sitting at the nexus of where General Collins and his team are really increasing the production capacity on the, we think of hardware platforms and energetic side and then what Aditi touched on as far as actually delivering that capability to allies and partners and even our own forces that are supporting around the world. There’s this question when it comes to production capacity as a mechanism for deterrence, I’d certainly agree with that, but then ultimately, I think what drives the deterrence piece of that is what are we using that production capacity for and now can we actually demonstrate that we’re exploiting that to the maximum extent possible?
And that’s really right now what we’re finding and conflict in Ukraine has been an example of this, even comment on some recent examples in the Sankham region as well, where we can actually use software to go and understand the full availability of capacity, whether that’s 155 rounds or the particular types of effectors that we have distributed. All of the intelligence information that we have available to us and all of the objectives that commanders have, we can bring all that information together and then actually start to, whether through processes of automation or applications of AI or other types of capabilities, we can actually improve at dramatic rates the ability for the units downrange and our partners to actually go and deliver the effects on the battlefield that actually exploit that capacity to the point where I think our software and the partnership is actually driving the demands that General Collins was commenting on.
And it’s through that kind of full visibility in the supply chain and then actually seeing the results on the field where we’re delivering effects downrange at anywhere from 10 to a hundred times faster and more voluminous than has ever happened before, is really driving a level of fear and really effectiveness in the context of these partners that hasn’t been seen before. I think it’s starting to actually change an understanding of what software can actually do in terms of really bending the curve on I think the American deterrence priorities writ large.
Nadia Schadlow:
And it sounds like you’re working well with DoD. DoD is enabling or at least not creating obstacles to the speed of adaptation that you need. Has that been worked out over the past few years?
Mitch Skiles:
Yeah, I think something that we’ve seen change probably in the last decade or so is this recognition by the department. I think the Army led the way on this and now we’ve started to see this adopted more broadly. DIU, I think is a shining example of goodness right now where there’s this idea recognition that commercial companies and particularly commercial companies in the technology sector can and are innovating through their own capital and developing capabilities that can be utilized inside the department. And so historically the way this would’ve typically worked is you’d take RDT&E funding and then you’d identify problems inside the department. You’d use groups like DARPA or in the Army, groups like DEVCOM to actually go and start to develop new capabilities against those requirements, mature those, and then try to transition them over into larger programs, programs of record and get them downrange.
And what we found over the last 10 years or so is a great acceptance by the acquisition community to go and say, Hey, we can actually just buy capability directly from commercial industry and the amount of time, energy, costs to modify that into actually facilitating outcomes for DoD actually is a significant advantage to the department. And I think that realization by the department and then us working in close partnership to actually get some early results on that has really started to pave the way, not only for Palantir, but I think new entrants across the board. Anduril being, I think a great example of a quick follower in that. And now you have hundreds of startups emerging right now and tons of venture capital being poured into companies that can follow suit.
Nadia Schadlow:
So you sort of opened the door for my next question to Aditi about DIU pushing the department as a whole toward a more modular adaptable industrial base. And Lionfish is one example. Maybe you and Andy could talk about that a little bit, but could you comment on that in general also?
Aditi Kumar:
Yeah, I mean more generally I think there is a broad recognition that we need to move towards this space, and my boss is going to love that I use this analogy, but it would be like Apple putting out an iPhone and saying only Apple can develop all the apps. We don’t get the best product that way, nor does it make good business sense for Apple to do that. And so we need to bring that sort of model to defense capabilities and create the incentives for some of these primes to see themselves as platform providers that are then open for other firms to provide the best of breed, software, subcomponents, et cetera. And in some cases the prime may be the best place to provide those pieces just like Apple in some cases is best place to provide the app, but that doesn’t need to be the case. And so we’re really pushing that model because I think it makes sense for the warfighter, it makes sense for the taxpayer and it should make sense for the industrial base too.
Lionfish is a great example of that where we’ve partnered with HII. It’s an undersea vehicle with countermine measure capabilities and it has an open architecture as Andy mentioned, and one project that DIU undertook called Project AMMO is to again enable faster, more rapid software integration into this capability. And so it used to be six to 12 months and not floppy disks, but certainly hardware was involved in updating the software. It felt very archaic and through this Project AMMO, we’re able to develop an ecosystem and a pipeline and we have tested vendors updating the software in a matter of days. It’s 97 percent decrease in the time that it took.
So Lionfish is one example and Andy can extend that. There are others. We just announced the enterprise test vehicle that we’re doing with the Air Force, which again is a modular open architecture system. The vehicle is designed to test various subcomponents, so inherently needs to be open and it’s another great example of us pushing that capability forward and driving towards lower costs open capabilities.
Nadia Schadlow:
Andy?
Andy Green:
So just to add to what Aditi said about Lionfish, one of the things that I think made that program maybe not completely unique, but I think special and a special success, an example of success, is how it took, you had essentially the end user, the operator, you had DIU, you had NIWC Pac and you had us as industry all working together so that the customer was involved every step of the way, like we were talking about earlier with Bryan and Mr. Wittman, but the customer was involved every step of the way. DIU was involved every step of the way. We were constantly inserting new technologies.
Now, it started out broader and then they kind of down selected as they went through and then ended up, this happened over a handful of years and ended up in a situation where you now have this product and we’re talking about a UUV that’s about six, seven feet long depending on the configuration that two people can carry around. So very, very portable, very modular. We ended up with this UUV that the customer really, really liked, that our coalition partners really like and they really want because they want to be interoperable with the US and they like the modularity.
This particular UUV, I won’t go into technical details of it, but it was designed to be very modular so you could put whatever sensors you want on it, whatever payloads you want on it. Something in the Black Sea could be configured very, very differently from something out in the western Pacific and you could put on whatever, take commercial off-the-shelf sensors and put those on there, whatever you want to do. So there’s a lot of flexibility built into that. I think it’s just a great example. It’s small, it’s not a huge program, but I think it’s a great example of how DIU, the government, industry, the customer can all work together to really very quickly hone in on a good solution.
Nadia Schadlow:
General Collins, can you comment a little bit or build on that in terms of how the army has been thinking about open architecture systems and modularity and is there resistance to what Aditi and Andy have been talking about?
Lt. Gen. Robert M. Collins:
Yeah, and I do want to comment some more on the software piece because I do think it has becoming a national security imperative, but I do think within the open architecture arena we’ve started to promote several activities, whether it be within the communications perspective, whether it be in the autonomy perspective. And really as you start to look at open architectures, it really enables that modularity. So certainly within the unmanned aerial systems, we’ve been looking at a specific type of architecture as you look at whether that unmanned aerial system may have a lethal capability, may have a kinetic lethal capability, a non-kinetic capability, it could be a comms relay, could be an ISR, is something that we’re taking a look at. I would say one of the areas too that we’re also continuing to promote is within our communications suite we’ve got an open systems architecture for position navigation timing, communications ISR, and it’s almost kind of an analogy of an open chassis, which kind of has an open standard and then you can slot cards.
And so I think not only does it allow operational flexibility for the mission, it allows the flexibility as the threat continues to evolve to be able to do that. But also too, back to the point of now you have increased opportunities for competition to be able to create that level of modularity. I think where we need to continue to do is make sure that we promote those standards not just as a government directive, but I think when we put those out with industry consortiums and we get feedback, we make sure that we’re doing those in partnerships and then we also make sure that we’ve got investment dollars so that we know that there are opportunities to be able to compete. That’s where we’ve seen a lot of the industry from that perspective.
And so I think across the board, US loiter effects, lethal effects is another opportunity that we have. And so I think there’s just a growing opportunity within the open systems architecture to be able to move that forward. And I think it’s very good for scale and for the modularity construct.
Nadia Schadlow:
And that leads us, we haven’t yet touched on the concept of operational concepts and how all of this adaptability and modularity in the field also then provides opportunities for new operational concepts. And I don’t know, Mitch, do you want to comment on that or General Collins?
Mitch Skiles:
Yeah, I mean I would say, I think sometimes we kind of get caught up in this. . . We talk about open systems architecture and then we talk about modularity and we start to treat modularity as kind of the end state that we’re achieving. But we actually look at the reality of how these systems are used, why do we build platforms or systems inside of the department? We’re trying to build things that are more lethal, more survivable, more maneuverable. And then really we want to ensure that those systems are sufficiently adaptable so that you can preserve those first three attributes throughout the lifecycle of when that capability is deployed downrange and the adversary gets evoked. And we have to make sure that our systems are sufficiently adaptable to that.
And so I think the way that we’re really thinking about developing systems in partnership with the department is that we look at operational effectiveness as kind of the principle driving measure really. And if we’re not able to deliver capability that’s one, able to actually slot into current operational forms such that they can achieve their aims more effectively, and then two, not enabling those operational concepts to be adapted to take advantage of new and emerging capabilities in new and interesting ways, then we haven’t really done our job.
One thing that I would just comment on the modularity piece and in reference to a comment that the representative made a little bit earlier, so I think we sometimes use the iPhone analogy as what goodness looks like as far as a system that is highly operationally effective, but oftentimes we’ll use that as a counterexample to modular open systems architecture. But I think the actual lesson to learn from the case of the iPhone is that concepts of modularity are being applied, as the representative noted, are being applied to develop that system, but that a company like Apple is developing the iPhone with the end user in mind as its sort of principle function and the consequence of using modularity as an engineering tactic to deliver capability that is measured in terms of usage and user effectiveness and simplicity is the magic of that type of capability. And I think finding and striking the right balance there is one of the key challenges the department has right now.
And I think what I would encourage us to think of as we have these kind of open systems architecture discussions is really how can we look to identify companies and engineering practices that the department can partner with rather than try to adopt and transfer those techniques themselves. So in the context of an Apple, it’s like instead of looking at Apple and saying, how can we be more like Apple, I think the answer should be how can we partner with those types of companies that are practicing these things in new and innovative ways with an emphasis on the operational effectiveness as the kind of key result to measure.
Nadia Schadlow:
So it’s a broad definition of talent. You’re thinking of talent not just within the department, but externally, the commercial sector and the department needs to think about talent more broadly too. Talent is in itself sort of a modular part of the modular concept. General Collins.
Lt. Gen. Robert M. Collins:
One of the things I think too, just kind of bringing us up into the digital age too, as we design systems, we’ve started to promote the digital engineering aspects of some of our platforms. And so that’s something that’s kind of a relatively new emphasis that we put out there. And when you start thinking about the ability to make decisions at speed, that truly is what the digital engineering process is. I think back to my career in the acquisition flowing requirements down through the old processes, I think those type of digital engineering processes allow us to decompose those requirements, allow us to make sure that we’ve got good solid designs, allow us to look at where we’ve got points of interoperability, allow us to look at where we may have obsolescence or supply chain. These allowed us to opportunities to look at operational outcomes and things that we may not necessarily have thought of. So we’ve not only identified some focus areas, some pilots, where are the areas that we can focus in on.
And then two, true the digital upskilling of the workforce to be able to take a look at that. And I think all of that feeds that modularity aspect and is just another digital tool that allows us to do that a little bit faster. So I think that’s a great point and something that we’re starting to take a look at to see where are there opportunities that we can continue to push on and promote.
Nadia Schadlow:
Okay. I’ll ask one more question and then turn to the audience, but Andy, did you have a comment? No. Okay. So it seems in a way a challenge for the department will be to figure out how to ensure that soldiers and operators now understand that these new opportunities for innovation and creativity and operational concepts exist. It’s a PME challenge, almost professional military education challenge to get soldiers to really understand that these are opportunities that they can take advantage of and be more creative with. Do you think that that’s starting to happen?
Lt. Gen. Robert M. Collins:
Yeah, I’ll comment. I think, and probably on the software side too, I think that is one that’s very pervasive on software. It’s increasingly part of every weapon system and there’s a degree of proficiency of how we do that. I think it was mentioned, you can’t just do that on an annual scale. We’re getting into sprints where we need to be 90 days, 30 days, even less than that.
So we’ve started to look at everything from a soldier perspective as we start the requirements process, how do we do that with kind of more broad requirements where we’ve got users that are involved throughout the process, kind of the processes that we go by through where you’ve got developers, you’ve got users, you’ve got testers, you’ve got everything that can do things in parallel, as we collect the process reform, how do we do that, test reciprocity, looking at all those types of things even from a funding and then just the educational part too of the up-skilling across the entire army, whether it’s the user community, whether it’s the developers, the testers, all of that. And so that’s been all part of what we’ve looked at from our software aspects, our software directive.
And then really the last part, the acquisition piece where I think the congressman was talking about the various adaptive framework of pathways, software pathway is kind of a newly-minted pathway. How do we look at things we can do more quickly, more open? We do it more rapidly. And so that’s part of the area of the education campaign that we’ve started to put out in the army and really start to promote that.
Nadia Schadlow:
Great.
Aditi Kumar:
And I would just add, this is a change in the way that DIU is approaching how we think about delivering the capability to the warfighter as well because we talk about the valley of death and bridging the valley of death, but there are actually a lot of valleys and they happen all along. So if you don’t have the right concept of operations and it probably won’t scale, if you don’t have the right experimentation schedule that makes sense for the capability in the right environment, it probably won’t scale. If you don’t have the right training in place for our warfighters at the right time and they don’t know how to deploy it on day one, it probably won’t scale.
And so taking all of those processes along the life cycle and shepherding a capability through those all the way to the end is something that DIU is really focused on with now our partners in the building, and I work very closely with General Collins on a couple of Army where we’re doing this now, where it’s the combat and command joint staff, service all around the table saying, okay, well this is how I’m going to use the capability. Can I do that? If you can’t do that, let’s go back and update the capability or update the CONOPS and make sure that this works. Let’s give it to warfighters, see how they use it, let’s get that feedback back into the system. And that’s obviously really relevant to Ukraine as well as we put capabilities out there to field, getting the feedback loop back into our acquisition system is really, really important.
Nadia Schadlow:
Mitch, and then we have some questions.
Mitch Skiles:
Just to kind of follow on with that. I think a change that I’m starting to see is even a couple years ago when it comes to questions of talent, there’d be these conversations of, well, how can the department retain, let’s say, software talent or develop software talent and then how is the department’s talent compared to industry? And I think your point on how we think about soldier education as an example in terms of how can we operationally utilize capability that broadly exists and that’s being provided by the broader industrial base is kind of critical because the change that we’re starting to see now is not an assessment of how does defense software talent compare to industry software talent, but rather how does American software talent, the combination of effective operators utilizing capability that’s provided by the millions of software engineers, data engineers that support the entire tech economy and then of course all the capital and talent and resource and everything else that goes into that is now being measured against the adversaries.
And then just to kind of bring this back to the point on partners, what we’ve really started to see in the context of Ukraine or again, some of these other cases is that we’re even broadening that definition to be the entire capacity of this technology innovation base across all of the US and its partners in a way that is all oriented towards the operational effectiveness. That is a pretty significant change even over the past couple of years. And I think it’s starting to produce pretty significant results as a consequence.
Nadia Schadlow:
Thank you. Now, even though this is quite a, it’s a very nice crowd, Bryan is going to help me manage some questions from. . . So any questions?
Bryan Clark:
So make sure, name. . .
Paul Tennant:
So Paul Tennant, I work for HII, but I’m going to ask a question channeling my recently departed 32 years as an ally and partner in the British military, if I may. So as an ally and partner and a very close one, I have become slightly resigned to a commitment to the strategic advantage, so frequently stated of alliances and partnerships in a rhetorical form and not much more. And I’m really encouraged by the coverage of it on this panel. And I’m really encouraged by AUKUS. To my mind it comes down to a risk calculus that which you can take in risk today and as soon as you possibly can, not only because it will accelerate your pathway towards being better together, but also because it has an immediate deterrent effect and the risk that you increase if you don’t take that opportunity as soon as 2027 by many people’s estimates, when we might have a serious punch up, I’d be really interested in the panel’s views on whether we are doing enough fast enough and taking enough risk.
Nadia Schadlow:
Does anyone want to comment?
Andy Green:
I’ll take a first shot, especially since we work for the same company. So I would say no, no, and no. I don’t think we’re moving fast enough. I mean we’re getting better and we’re moving faster and we’re taking more risk, but I think we could go a lot faster. I think it was representative Wittman that was talking about special forces and how fast they move this morning or may have been Bryan that brought it up. I see that in my own organization and we’ve seen it, SOCOM is the customer of ours and we see it there. They are very willing to go fast, get what they need, deploy it, and are there risk? Yes. But that’s what we get paid to do. And so I personally think that we could do better, but I do think that a lot of people recognize that and we are starting to move that direction. That’s a personal opinion.
Nadia Schadlow:
DoD?
Aditi Kumar:
Yeah, I think we’re moving in the right direction. I don’t think I could ever say that we’re moving fast enough just because of the imperative to work closely with our allies and partners and how much we have learned. I mean, in supporting Ukraine, the fact that the Ukraine defense contact group has come into existence and held together for the past three years, 50 plus nations coming together almost on a monthly basis supporting this nation and the co-production gains and the sustainment gains that have also come out of that I think have been an important lesson that we will carry forward.
When DIU is looking at this problem, we’re thinking about, okay, practical solutions. We don’t need to boil the ocean. We don’t need to solve the most extensive international cooperative agreement in every case. We just need to show partnership on capabilities that matter and let’s pick what those are in each context and let’s go after it. And so we’ve done that in a number of instances with the UK, as I mentioned in the electronic warfare space and the cyber space and others, we’ve done that with Australia in the autonomy and other spaces. And so that’s what we’re trying to do is find that practical solution that we can move forward, understanding that there are constraints on both ends there, national policies and things that we need to work through. But there is a realization that one, there’s redundancy in the types of capabilities that we’re all going after. And so we can actually maximize our gains and maybe even get economies of scale if we join forces. Then there’s just the knowledge share that has been super invaluable as well. We’re just looking for those applications.
Nadia Schadlow:
Another question?
Chip Walter:
Chip Walter, Marlinspike Partners. Spent some time in Northrop Grumman and CIA before that, but really appreciate the comment about the partnering with HII from DIU. That is incredible and it needs to happen more. Can you comment on other stuff that you’re doing with other larger companies because that is the focal point in terms of starting to beat down the transition of a company?
Aditi Kumar:
Yeah, I mean maybe I’ll comment more broadly on our process, which is that we at DIU are using the commercial solutions opening process to bring competition into a variety of areas. And this is where it’s really important that we have two partners. One is the service that’s going to be the engine of scale. At the other end is the service, they’re catching the things that we’re bringing into the department. And the other is the warfighter. So working with the combatant commands to say what is the operational imperative? What is the problem that we need to go after? And so we’re using those two stakeholder groups in the department to figure out where do we want to place our CSOs and where do we want to bring companies in? And so we’ve had a number of CSOs out there.
Recently, I mentioned ETV Franklin is a big one where we’re able to pull in non-traditional players to contribute to this capability. We recently posted one which is called PRIME, which is a unmanned surface vehicle. And in writing those CSOs, we have been very intentional to say, hardware providers, welcome software providers, welcome, teaming arrangements, welcome, incorporate advanced manufacturing techniques. We’re very open to those types of partnering relationships. And I think part of that, and maybe I’ll throw this to General Collins as the true acquisition contracting expert here, is the department creating those incentives both for the primes to have those partnerships, but also for the smalls to have the protections that they need of their IP to make that arrangement work for everybody.
Lt. Gen. Robert M. Collins:
Yeah, no, and I think this has provided us the avenue to be able to demonstrate to industry that we’re serious about moving out quickly and moving out at scale. And so I think we’ve been able to demonstrate that. I do think the intellectual property is an area that we’ve had to make sure that we kind of protect that process. I even think even small companies coming in, there’s simple things they can do of cooperative agreements that kind of no cost, that benefit one another. We certainly have found that. But I could not underscore enough though as we bring that in and we get the operational concepts, we understand how we’re going to employ that, that’s where it’s been an advantage here from an army perspective. We’ve got early in requirements process we think we know how we’re going to employ, but now we start looking at what are the different mechanisms and how we’re going to use that, what units, what structure, what doctoring, what organization, those whole concepts. While at the same time demonstrating out to those companies that may look at the department and say, “Hey, there is a barrier to entry here, and I’m a little bit concerned now opening it down, breaking down those barriers, and opening up that communications,” I think has all been beneficial to start to move that forward.
Mitch Skiles:
I’d also encourage you to take a look at the CDO’s recent initiative under Dr. Plum called Open Dagger. I think it was just announced about a week or two ago. But the whole intent there is can the department actually look to identify novel mechanisms to break down the barriers to entry for new and emerging technologies.
I would say there’s really three kind of broad pillars of what makes it challenging for startups and smalls to participate. One is the acquisition pathways, were just covered. The second one, interestingly I think, is actual technology barrier to entry for new participants. Sometimes for software we think just contain a resume thing and throw it in a container environment, we’re good enough. But the integration in cybersecurity overhead is pretty challenging.
Something that Palantir’s been doing is actually taking a look at all the infrastructure that we’ve developed over the last 20 years, saying, “Can we actually just provide that to the rest of industry to actually facilitate rapid onboarding?” Where acquisition pathways like these, like Open Dagger, exist.
The third one, which I think is the one that is most in need of some reform, is the amount of funding that’s actually then reserved to resource these new and novel entrants. I think small, $100,000, million dollar CIBRs is not sufficient to attract and incentivize them to participate. But certainly on the first two fronts, a significant amount of progress was made recently, and now the third one, I think, is the next domino to fall.
Nadia Schadlow:
Thank you. One more.
John Troops:
Thank you. John Troops from the Atlantic Council. I’ll be a bit provocative following up on the first question. We think we are in a great power competition with two revanchist powers, at least, China and Russia. There’s talk about China maybe considering a strike across the Straits of Taiwan in a few years. There’s a chance that if Russia were to win Ukraine, we might see a strike . . . Certainly much greater provocations against our Baltic allies. We’ve learned from the war in Ukraine that great power war is greatly costly in terms of equipments. I feel like I’m in the land of the Munchkins talking about our improvement in defense production, when what we need is a top-down call for greatly ramping up our arms industry. Any comments would be welcome.
Lt. Gen. Robert M. Collins:
Well, I think starting to scale up some of the things that we’ve started to do with our munitions, I think, is certainly an area. We are also looking at long-range. As you start to think about different types of regions, extended range capability is another area that we’re starting to take a look at. I would say within the army too, from a combat platform perspective, where are the areas that we need to look at and not necessarily manned but maybe un-crewed systems that we could take a look at. There are certainly areas that when you start looking at competition, there are things that we can do that have exponential impact and very basic things of how we move logistics or how we conduct breaching operations or other activities. And so I think we’re starting to take a look at how do we better invest in our industrial base to make sure that it’s prepared to be able to scale?
And I think that also kind of goes back to, as we work with our international partners, we can’t necessarily do this alone. Having these co-production agreements to be able to do things at scale and be able to reach those areas and the levels that we need to be able to produce at to be able to sustain a long-term conflict. So, I think this is something that certainly highlighted that.
Aditi Kumar:
I would say that ramping up production of capabilities, and General Collins has been at the forefront of this, is certainly a key lesson learned and one that we’re going after aggressively. That needs to go hand in hand with actually diversifying the industrial base also, right? Yes, we need to ramp up existing production and support these capabilities, but then we also need to bring new players in and diversify the base so that we have more options.
And I think we also, as part of that, need to think about how do we change the cost curve. And some of that has to do with some of these capabilities that we’ve talked about today, which are moving us towards lower cost, although not low enough and we need to keep working on that, especially in a defensive space.
And then another area of focus where cost has come to the fore is on the sustainment side. Because I think it’s 70 percent of the cost is actually in sustainment. And so how do we do that more cost effectively, cheaply, closer to the edge, et cetera? And so there’s been a lot of work done on the sustainment side as well. Which, again, incorporates some of these agile production methods, advanced manufacturing methods, so that we can leverage that for spare parts and things, and get those to theater. That has been a huge part of what we’re doing in Ukraine as well.
Nadia Schadlow:
So, Andy, and then I’ll close.
Andy Green:
Yeah, one quick comment. I would say that we use some examples today, and I think we all talked about some specific examples that aren’t in an of themselves going to move the needle, right? But I think what I tried to look for, when we were thinking about this panel, was demonstrators and how do we show that we are, as a country, we have the ability to move in the right direction and we are going that way, and we can prove it by these examples, right?
At the same time, you got to remember, my company’s building virtually every capital ship in the Navy right now, right? And so they’re investing billions, 4 billion over the past several years, in infrastructure. The Navy is investing alongside us. So, I would say that they haven’t lost sight of the big problem or the pink elephant in the room, right? I would say these smaller examples are just that, examples of how we can migrate the broader system. That’s kind of the way I look at it.
Nadia Schadlow:
And ultimately, John, by the way, it’s nice to see you. Ultimately your question is a bigger political one. Is the country ready? Should it be on a war footing? Should our- . . . Right. Right. Thank you and thank you everyone. Thank you to my great panelists.
Bryan Clark:
All right, well thank you, everybody, for sticking with us. I’m Bryan Clark, again, from the Hudson Institute. And I am joined by my fellow senior fellow Dan Pat from the Hudson Institute. And we’re going to have a panel discussion now about how does industry, and in particular how do some of these new defense industry players, provide that scale and adaptability that we’re talking about today? So, how do they provide the new capabilities that the DOD needs or that the allies need? How do they scale those capabilities as we identify opportunities to gain an advantage? And then how do we adapt over time to continue evolving the force to remain relevant against opponents that are thinking and are technologically advanced on their own?
So, for that discussion we brought in several players from key industries across each domain. So, we’ll start, Mike Hiatt here is the Chief Engineer from Epirus, and they build high-powered microwave systems and they could be used for a lot of missions. We’ll talk about that as we go. Joe Laurenti is here from Ursa Major. We’re going to talk a little bit about how modular weapons and what Ursa Major’s doing in propulsion and new approaches to weapon design. We also have Josh Martin here from Varda, which is a commercial space company that is doing great work in terms of on-demand orbit and de-orbiting capabilities, but we’ll talk about commercial space more broadly. And over there on the end, Michael Brasseur from Saab, where he is the Chief Strategy Officer and also Director of Skapa, which is their technology innovation organization. We’ll talk about that a little bit more as we go.
So, thank you gentlemen for being here. And thank you, Dan, for co-moderating this panel with me. So, to start out with, what I wanted to ask you each, this challenge that we talked about this morning of scale and adaptability. The DOD seems to be putting a lot of effort into improving our ability to produce more of the stuff we have today, and we just talked about that on the panel with General Collins and others. Or we’re doing a great job at adapting and creating new innovative ideas and new capabilities, but not necessarily scaling them. So, we can do either of those things, but how is it that we’re going to get the ability to do both? How do we get to being able to both innovate and then turn around and scale and then innovate again? So, Michael, I’ll start with you.
Michael Hiatt:
Great, thank you. I think there’s a lot of industrial base here in the United States, and a lot of desire to solve some of these problems. But when we look at sort of this new paradigm in warfare that we’re seeing, where it’s heavily leveraging consumer items, it’s heavily distributed, and the threat is heavily networked. Our existing solutions are just really not designed for that. We’re designed for that Cold War era fight of exquisite solutions going against other exquisite solutions.
And so we have to be, I think, very thoughtful given the resource constrained environment that we’re in on what we choose to scale. And then look for new innovative ideas that say, “All right, if my adversary has a significantly stronger industrial base on consumer electronics,” which is the case with China, “we won’t be able to compete one-for-one on small consumer electronics based things there.” So, how do we take then a different approach? How do we look instead of solving a 1-v-1 type solution? How do we turn that on to a 1-v-many solution, as an example? So, I think we have to be very thoughtful about what we choose to invest our dollars in, and make sure that that’s countering that asymmetric threat.
Bryan Clark:
That’s a really good point. We’ll talk a little bit more about this idea of how do we lean on US advantages when it comes to trying to get adaptability and scale. Because scale may not just mean mass, like you’re saying, just don’t fight mass with mass, we got to come up with something that’s going to fight mass with something that will work against it.
Michael Hiatt:
Absolutely.
Bryan Clark:
So, Joe, CEO of Ursa Major, what do you think?
Joe Laurenti:
We tend to think about fielding products as the end state in a series of flywheels. In the early phase of development we have this really small fast-moving flywheel where we can iterate really quickly, we can learn a lot from our hardware. Somewhere in the middle there you start to engulf requirements. You have heavy qualification, your flywheel starts to slow down inherently a bit, and in the end, in production, you should have this really stable big slow-moving flywheel.
Scale is really on the far end of that. Adaptability is really on the front end of it. So, we tend to think about design for interoperability across all three. How can we keep all three of those flywheels moving kind of in their intended states? And a lot of that comes down to early product fielding decisions. How are we thinking about acquisition? Unfortunately, while a company like us may think about those series of flywheels, acquisition doesn’t work the same way. So, we tend to think through how do we field systems where they’re intended to be fielded?
We’ve heard a lot of talk today about DIU, some of these fast moving organizations. It’s actually really easy for us to adapt to requirements if we have all of the stakeholders bought in from the PEO where they combat and command up through a fast moving organization, like DIU.
So, I think it comes down to a few things. It comes down to how do we address listening to the end state customer, starting to plan out that larger flywheel of production, while acknowledging that the early phases are going to be products that we will not field, products that we will not field. And I think what that results in is capital expenditures that are really flexible. We think about manufacturing mixed model manufacturing, producing many products out of one factory. We think about the design and iteration process being very vertically integrated, learn as much as you can. That probably goes back to my SpaceX days of blow up a lot of rockets so that you can fly very reliably.
And then we tend to think of production being highly automated, highly digitized so that when you do have to make changes to that flywheel, it is a much more adaptable process than a traditional production line.
Bryan Clark:
So, we’ll talk more about this idea of to what degree integrating design and production makes sense, to what degree decoupling it makes sense. So, definitely great, thanks. Josh?
Josh Martin:
I think where a company like Varda thinks about how we can assist the US government in rapidly iterating on technology, developing technology is mostly in the test and evaluation vertical. Being able to access commercial platforms, like Varda’s, allows the US government to test in a manner that they have not been able to until now. And do it in a manner that de-risks by testing a little a lot, right? And so I think commercial solutions like that, being able to get technology to good enough, and then iterate and improve that through T&E over time, I think is something that we are working very closely with our government partners on. And I think that it’s a solution that just really hasn’t been wholly available in our particular sector.
Bryan Clark:
That’s a great point is we often, in the course of this discussion, about innovation and then scale, we don’t talk about, okay, how does somebody actually do the T&E that makes the program able to move forward into full rate production.
Josh Martin:
Right.
Bryan Clark:
It tends to be sort of an afterthought. And the T&E infrastructure of the country is probably not designed for this sort of adaptable rapid innovation. But maybe Michael has a solution to that.
Michael Brasseur:
Yeah, and so for me the solution is all about getting very, very close to the problem, and understanding it in a way that is so intimate that you almost feel like it’s your problem. And that means going to the edge, going to where the problems are.
I work for a Swedish company. You can feel when you go to Sweden that this war is very, very real, and I’m talking about the war in Ukraine. It’s literally a distance from here to Jacksonville is the distance from Stockholm to Kiev. So, you can feel the sense of urgency. There’s a mindset piece of this.
And then in terms of looking at novel production models, additive manufacturing for me is a game changer in this. Not only just in terms of production but in terms of resilience. Getting it closer to the edge so you can literally print replacements to the stuff and what Joe was saying about the digital design you can iterate at the edge. And you may have different capabilities emerge in Ukraine vis in the Pacific because that development and that production is closer to the edge, and that also includes, in many cases, local sourcing. So, you can build resilience into this whole thing and then hopefully that will enable the scale that we’re talking about.
Dan Patt:
Yeah, so it seems that one of the underlying themes across all of this is this tension between two things. A tension between prediction and a tension between learning. I want to get your comments on this. On the one hand, when we talk about when there’s a clear requirement, that sounds like there’s somebody really smart who made a prediction. You know what? If they’re smart and made a prediction, I can check the boxes and I can deliver that. But to do some things, like set up production of physical items, I have to know some of this stuff. I have to have some predictions that I can bet down on. I have to make orders for materials and for suppliers, right? I do require some prediction to get these production volumes.
On the other hand, there’s this tension with learning. We talked about this in a few different aspects. We talked about getting design engineers to see the results of their design. Maybe blowing something up or something not working out. That’s really important. We have to have this coupling between engineers who do things and the products of their work.
But it’s not just about engineers and technology that they develop. There’s also the learning and how we use the equipment, right? When we talk about a requirement, that’s not just implying that we know what the materiel that we want is, it’s implying that we know how to use it. And that may also be open. And it may be used in new combinations.
I’d love to get some of your feedback on these tensions that are bubbling under the surface here. How should we think about this? What’s the right role for prediction and what’s the right role for learning? And how do we break these apart as we really try to innovate in how we fight, innovate in what we fight with, and be able to scale? That’s what we’re trying to do. How do these things play out? Mike, start with you.
Michael Hiatt:
Thank you. I think a big part of it has to do with . . . I love that you brought up materiel. Not forgetting about the rest of DOTMLPF, and making sure that we consider that end user in mind at the outset, at the basic architecting. The traditional RDT&E focus has been heavy on performance. Get performance to where it needs to be, prove that technology can hit the performance metrics required, and then, oh, by the way, let’s add utility. And you’re sort of seeing that bear out in how the laser solutions that are going to the field now, very low operational availability. They break a lot, you need a clean room to service them, things like that. Now those are solvable problems, but they’ve spent decades on performance, and now try to add utility.
So, if you, from the outset, focus on utility as you ramp up performance with the rapid iteration and the rapid prototyping, then you’re able to, once you actually start to think about, all right, how is a very young maintainer going to keep my product operational? How are we going to handle sustainment and logistics? How are we going to train folks on how to operate the system and make sure that they can be trained very efficiently and quickly on how to use it? And then how do we make sure that the manning requirements . . . I can’t have five people to operate a widget. It needs to be one person operating a couple widgets, and so good use of things like machine learning, decision aids, sprinkling AI through that to help on those decision aids so that operators can make use of all of that, the abundance of data that’s coming out so they make the right decision. It’s considering that whole DOTMLPF at the outset.
Joe Laurenti:
I think that’s dead on. It’s a lot of what we think about for interoperability. Going all the way up to acquisition, sustainment, and training, but even in that early phase of development and qualification. If we are working on a propulsion system for an air launched effect, and we know a small design addition or a small qualification addition can lead to certification for ground-based capabilities, then we are more likely to design that in in the early phase versus waiting for that requirement, waiting for that feedback.
I’ll really touch on something Michael said. I think predictability is a hard word to use right now. With the last six months in Ukraine not much has been predictable. So, I think we need to plan to parallel path acquisition, planning to the fullest extent that we can, while acknowledging particular systems. I won’t shout out any systems because I don’t know who’s in the room here, but some systems are less effective than they were three weeks ago and we are actively modifying those systems. That’s the right approach.
The right approach is not to say, “Those systems are no longer effective with the W environments in the Western front, so we are going to get rid of them and plan for the next five years of acquisition toward another system.” So, I think it’s a parallel path answer.
Michael Brasseur:
Yeah, I have some comments from this from two perspectives. One, as a former founder and Commodore of Task Force 59, where a business model innovation capability as a service really unlocked and unleashed the flexibility that we needed to iterate and get basically the last technology now vis legacy stuff and two to three years from now. And so that feedback loop, because it was capability as a service, the companies were operating their own equipment and we were giving them very raw feedback in the operational environment versus real problems. And for me that’s the special sauce in terms of rapid capability iteration.
So, from the business side of the perspective, we at Saab launched a new initiative called Skapa, which is Swedish for to create, to make, to shape, and we’re also looking at business model innovations. How can we deliver capability in a way that’s adaptable to the world that we live in? And a big core of that is much of the same premise that Task Force 59 was. Focus on real problems in the real world alongside operators. And that loop, the smaller you can get that loop the quicker we can solve those problems. So, just two different perspectives there.
Bryan Clark:
So, building off that idea, there’s a lot of commercial companies clearly moving in to the defense space. Some of you are coming from those companies. They talk about capabilities as a service, so we’re using capabilities as a service as a way to, in some cases, supplement or complement what we already have in terms of government-owned capabilities. And in some cases we’re using it as a way to experiment with new ideas.
But those are often not well transitioned into scaled employment. The good ideas that come out of that don’t really reach scale inside the traditional force. So, we’ve had a difficult time making that fielding construct work.
So, how should we be thinking about transitioning these capabilities that are in capabilities as a service into something that the government can employ either by owning it or having a better arrangement to be able to tap into it when the time comes? Maybe start, Michael, with your experience with that.
Michael Brasseur:
And it doesn’t work for everything. Obviously when you go kinetic, and then there’s also a case where you would prefer actually a tradable, a disposable system. I think the key is to follow the money. I think one of the issues with scaling that model was a matter of resource. With more resource, we built the model to scale at Task Force 59, and with more resource it would have scaled.
But I think if we, if the US government could design a system that clutches in with that at the front end and then a program of record or capability of record on the back end, I think that would be the ideal situation.
Joe Laurenti:
Yeah, I’m glad that you called out that this might not work for kinetics because we are 100 percent kinetics.
Michael Brasseur:
Yeah, yeah.
Joe Laurenti:
Doesn’t apply everywhere. But I’ll actually hark back to the last panel. I heard General Collins mention collaborative research agreements. In an environment like ours, where an enormous part of the enormous cost to the ability to scale and maintain flexibility is capital expenditures. Something like a collaborative research agreement with different services can become really useful. We’ve had success in the past with the Navy in that regard specifically developing energetics, that just would have taken a long time for us to bring in the cap X capability. So, I do think that there are bridges there even on the kinetics side.
Dan Patt:
Yeah, so there’s a related question that I would like to focus in on and, Josh, focus on you. When we talk about industrial mobilization, we talk about scale, and look at the historical lessons. A lot of this involves taking the existing commercial base and trying to harness that for defense purposes. And there’s a role for new defense entrants with new technologies and new business models and new concepts, but there’s also a strong role for crossover areas. And space is one of the places where we’ve seen the most vibrant crossovers, where there are both viable commercial business models and viable defense models and there’s mutual benefits to industry players there.
So, I would love to just hear a little bit, Josh, about how you think about how can there be mutually beneficial alignment between commercial and defense models? And we don’t always just have to strictly segregate these things.
Josh Martin:
Yeah, no, I think that that’s a good question. I think sometimes it’s an understanding on the government’s part that we can co-mingle these products and services in a manner that they shouldn’t strive to standardize. This is probably an area where standardization and modularity maybe aren’t always desirable. For instance, for the Department of Defense to sort of force us into a more standard payload integration on our reentry capsule, that may be great, you may be able to squeeze out some integration time for those government payloads as we’re trying to mix and match commercial payloads on our reentry vehicle as well. But that plate is also probably adding weight. That means you’re probably reducing the number of experiments and payloads that we can fly for the US government when we can solve that problem on our own. So, I think under having that government understand that is really important.
I think another piece, because it’s related to space operations, is that-
Dan Patt:
Just for the benefit of our audience, just go one level deeper one. So, one, you’re suggesting there’s this inherent tension between performance and modularity, which is interesting. But just help give us some context. When we talk about payloads for reentry, for people who aren’t tracking exactly what Varda’s doing, can you just contextualize that?
Josh Martin:
You bet. We’re talking about heat shield material. We’re talking about guidance systems. We’re talking about communication systems. We’re talking about scientific equipment to measure the reentry environment, to understand exactly how that vehicle is performing in that high stress environment.
Dan Patt:
All right.
Josh Martin:
The other piece of that though real quick is infrastructure. For space operations, infrastructure is really important. SpaceX and other commercial launch operators could not do what they do without government investment in launch infrastructure. I think that’s another piece of this that the government needs to understand, that infrastructure related to novel space is also important. It would be sort of like, hey, we want the truck to do the thing, the commercial truck to deliver the product, but we don’t want it to drive on the interstate. I think commercial space companies need to have access to a diverse set of government infrastructure from launch ranges and reentry ranges too.
Michael Brasseur:
And I just had an additional thought on what I was talking about is you’re really seeing some of these organizations in the DOD grab on to this and lead in this space. For example, the disruptive capabilities office in the US Navy. Obviously Aditi and the team there at DIU. And then you’re seeing that matched with the resource. And so momentum is building. I just think we just got to pour gas on this thing because we really need to hustle in terms of deterring a war in the Pacific and winning the war in Ukraine.
Bryan Clark:
So when we come up with a good idea, I think the problem has been transitioning that into the hands of war fighters and then scaling the production of that good idea into being something that’s at relevant capacity. Joe, this is something you guys are working specifically on, but obviously this is something you all work with. How do we get production scales up to where they need to be while also retaining the ability to be adaptable and not get to the point where we’re basically stockpiling things that are going to become obsolescent in the near future? So for weapons, how do we get that ability to scale production kind of on demand and then turn around and have to adapt to scale something else on demand?
Joe Laurenti:
I mean, we had the major advantage of, we’re a relatively young company, we are nine years old now, so we were not building this new effort to stockpile or to replenish munitions and stay really flexible on the heels of five decades of capital expenditures and factories and training. So our approach was starting largely with additive manufacturing, which Michael brought up, but really a design for manufacturing approach across all of our systems, whether these are next generation hypersonics or rebuilding forty-year-old Stinger missiles, we can do it all under one roof because we have a design for manufacturing approach that allows novel materials to be created very quickly.
We have metallic alloys that we’ve developed in-house really for their 3D printability and their consistency and reliability in mass production environments. So one way to take out 10 machines is replace them with one machine with the right capability set and the right design backing. Another one is to think through processes maybe that weren’t around in initial system fielding. So for us, a lot of automation around composite capabilities, we think with a lot of our propulsion solid rock motor systems about building these systems inside out, we think about what were the design constraints? What were the tooling constraints of the last 50 years that might not be constraints if we’re starting from scratch.
So it’s not always doable in the case of when we need to come in and backfill a system that’s qualified, there are really strict requirements. But what we’re finding is again that parallel path of, I keep using that term interoperability, we really think of it as the flexibility that is being afforded us by the state of the battlefield today. A couple particular systems that we are working on are US systems, US heritage, but they have to be flexible enough to potentially be used by former Soviet ground systems equipment. And so that’s not going to be done in a factory that was built in 1975. It’s going to be built in the factory or it’s going to be built in factories that are more modern than that.
Bryan Clark:
But this gets to this relationship between design and manufacture. And so what you’re talking about when you guys vertically integrate your design and manufacturing-
Joe Laurenti:
Absolutely.
Bryan Clark:
. . . you’re talking in a lot of cases about the detailed design for the propulsion system of a missile that somebody else designed the overall missile, so an SM-II or whatever, somebody designed that back whenever, a long time ago. And you guys are looking at ways to improve the detailed design associated with a component of it, which is an interesting relationship, is that as opposed to being vertically integrated from the weapon all the way down to its subcomponents, you’re talking about being able to do detailed design of a particular piece of something to allow you to manufacture it in the most efficient way possible?
Joe Laurenti:
That’s right. And it doesn’t matter if I’m in the Pentagon or under the roof at a prime, just about everyone that works with these systems and has worked with Ursa Major has heard me say, we work far better designed to spec than designed to print. We don’t want to take and translate old drawings and tweak them for modern manufacturing methods. We’d rather take a propulsion system designed for SM-II and enable higher range, more flexibility and lower cost because we have designed it around modern manufacturing methods.
Bryan Clark:
And Mike, I was going to ask you a similar question. I mean, you guys obviously work in a software defined environment, right? Building solid state software-defined radios essentially that do high power microwave operations plus other things. So does that approach of using solid state software-defined radios, does that give you the ability to scale because you can tap into a supply chain that’s highly distributed, or do you end up having to have a very exquisite supply chain that you’re forced to work within?
Michael Hiatt:
Just by nature of how we approach the problem of that software-defined radio and leveraging a phased array technology, where you have a lot of elements and you get them all talking together to make a very large aperture, that lends itself to scale inherently. So we’re sort of scalable out of the gate by design. Now we also are leveraging heavily on the commercial market. Look at the investments that DARPA made 20 years ago on the gallium nitride industry in the United States. That ended up being a fantastic investment. Now the 5G cell phone industry has kind of taken that, all the 5G base stations are using gallium nitride. So now you see a commercial side and then we’re able to leverage that sort of commercial process and the commercial viability of those companies is now able to provide for a defense industry use. And so we’ve really tried hard to make sure that we don’t design in, I’ll say, issues with a supply chain.
Dan Patt:
So let me dive into this. To what extent do you view design of a system that you create as a bottoms-up process? Which is what are the components out there? What are the leading power electronic components, these GAN components that I’m going to pull off the shelf from some other application? Or alternatively, what are the signal processing components that I’m going to pull from a 5G application? How do you think about what is bottoms up? And then where do you think about, no, there’s a component I need that is in some corner case of performance where that doesn’t exist and I’m going to ask somebody to design it or build it for me or I’ll do it myself? How do you think about that relationship between what is top down and what is bottoms up sourcing?
Michael Hiatt:
Yeah, great question. We like to do both, and I’ll explain on that is leveraging what Representative Whitman said this morning, is we can’t let perfection be the enemy of good enough. Because by the time we get a product to market, it will not be viable or it’ll be obsolete or something else will have taken that 70 percent solution while we’re chasing Nirvana. And what I think that we’ve been able to do is, all right, I can start the learning process with that 70 percent, 80 percent solution. At the same time through that while building that I can identify, all right, what does that next solution need to be? Where do I need to be in 3, 4, 5 years and what’s that technology roadmap?
And then how do I then take that and influence, whether it’s partners in the supply base or whether it’s something we do internally and work on? Where do I need to be? And then I can start to work towards that sort of more boutique-style component. But it can’t be, we’d never get to the 100 percent solution if we don’t do the 70 percent solution, meaning our company just won’t exist if we waited five years to put a product out. When you look at, there’s a lot of desire to have more of the VC community involved in this process. Well, the VC folks have a certain amount of a time horizon that they look at. And so you start to say, well, my investors are going to get impatient if I don’t do anything for five years.
Dan Patt:
So let me try to paraphrase that. I mean, you view this on the basis of time horizons, right? You’re going to time bound the problem. And the reality is, if I have this time horizon in the near term, I’m sourcing from something that exists. I know because I have engineers working the problem coupled to operators, I know where I might like to be and maybe I have levers of influence. Maybe I tell my supply base that if I’m a big enough customer. Or maybe it’s so specialized that that’s something that I take on, but I’m using time to decide whether I’m sourcing, bringing things bottom up, right? I’m time bounding the problem or whether I’m trying to push with technology. Is that a fair summary?
Michael Hiatt:
Oh, absolutely. And this kind of goes back to another thing Representative Whitman said, which is being able to have demand signals is so important because I have a lot more leverage going out to supply base if I’ve got the army saying, all right, we’re going to go out and buy X of these systems. That’s a tremendous demand signal that gives us a lot of credibility going into the supply base. And so we’ve had a great partnership so far with the army and really want to leverage that. But the stronger that demand signal, the more influence we can have in those far out road maps.
Bryan Clark:
So Michael, to take this to the system of system level, a lot of the work you did at Task Force 59 in sort of bringing unmanned systems into the Navy at scale, which was really revolutionary, was looking at the system of system level and being able to combine things in new ways that enable different operational concepts. I know that you’re continuing that kind of work at Saab, and so what are the challenges in terms of you’re working with a bunch of other people’s equipment now and trying to make it all work together into an operational concept and you’re leveraging, I assume, commercial communication technology maybe? So how do you put together those battle networks, if you will?
Michael Brasseur:
Well, so thanks Bryan. Brilliant question and a lot of lessons learned from Task Force 59 where the vision was sort of this convergence of all these emerging technologies to give a new capability. And we were looking at sort of next level maritime domain awareness, almost boarding on maritime domain cognition and understanding so rich of the environment that you can predict. At Skapa we’re sort of designed from the ground up to partner and move very, very quickly. So I’ll give you a very specific example. We have an autonomous boat with an operating system, much like the previous conversation about modularity. The operating system controls the core autonomy but enables applications for third-party capabilities. And so that is really the personification of the way of working that I think enables the systems of systems approach. So for example, we can work with a lot of the fast-moving startups in this space, and that helps you create capability much, much quicker. And then hopefully that will scale as well.
I just want to double-click on one of the comments that Joe mentioned. For me, as I said at the beginning, it’s all about understanding the problem and I think one of the most valuable pieces of technology that we have is the whiteboard. And the whiteboard with a world-class operator right beside you and just going vertical on the problem and then bunches of different angles focusing on the problem. So you can have an engineer that says, “No, that’s nice, but you can’t do that in the physical world.” And you can ask, “Well, why not? Or can’t you invent something that can solve this?” So I think that sort of collaboration is key to enabling the system of systems. It has to be at the very beginning at the whiteboard along with your industry partners and your government partners.
Bryan Clark:
And that’s a very bottom-up sort of approach to the system of systems problem as opposed to, I mean, there’s a top-down component to it, I guess. Fundamentally some senior commander has to define what are we trying to accomplish in this mission? But there’s a lot of bottom-up in terms of how do we actually cobble together the systems that’ll do it?
Michael Brasseur:
I think from the top down, it’s been very, very clear. When you look at the combatant commanders, you look at Admiral Poparo has been very, very clear what he needs and when he needs it. General Kurilla obviously has a ongoing conflict in the Middle East. He’s been very clear. And then in Europe I think they’re very clear on, we need this and we need it now, and it’s over to us to sort of deliver it at the pace that’s required to make a difference. I don’t know if that’s answering your question.
Bryan Clark:
Yeah, yeah, yeah. Because Dan brought up the idea of within a system, there’s this relationship between top-down and bottom-up as you try to figure out the most resilient supply chain. Because if you’re going to scale, you can’t have some very narrow supply chain. You’re going to have to have multiple potential suppliers you can draw upon. And at the system-of-system level, I’ve got to have a way for the mission to get done with multiple different combinations of things because stuff’s going to break, stuff’s going to get lost. You need different ways to adapt also to respond to the adversary. So I think you see at the system level and the system-of-system level, this need for modularity.
Michael Brasseur:
And I think the diversification of the industrial base is a very good thing to enable that. And it’s a really exciting time in the defense tech space right now.
Joe Laurenti:
I was just going to add on to that. I completely agree, and I think this notion of modularity feeds directly into that. If you’re only used to selling a Patriot missile, every problem looks like it can be solved by a Patriot missile. But if you’re actually listening to the customer, if you’re actually listening to what is needed in the field, it might be we have seekers, we have command and control, we’re just missing this one piece. And the more modular your systems are, the more you can solve that problem.
Dan Patt:
Name one problem that can’t be solved by Patriot missiles.
Joe Laurenti:
They’re great, they’re great.
Dan Patt:
A question for you there, help us contrast if there’s a version of you in aerospace basin of Southern California in 1955 versus the version of you today, tell us how industrial capacity, industrial base is different. What are the new opportunities that you take advantage of? And then what are the things that you wish were there and that you have to work around and innovate around? How has the US economy, the US industrial base, where there is talent and capacity changed, and how does that drive how you think as a company?
Joe Laurenti:
We’re really top-heavy. There are far fewer companies at every level than there were back then. So I completely agree with what Michael said. It’s exciting but we are in the early days of that excitement. We’re seeing a lot of early phase companies. And one of the big fears I have is that the embracing of defense that we’ve seen out of Silicon Valley and out of private equity in the last two, three years could fall apart pretty quickly without the right steps in progress. If we look at the companies that have made it out of Silicon Valley into selling directly to the government, I think all but one were founded by billionaires. And that’s a really small circle. We’re seeing more and more companies broach that, but it’s at a very different timescale at a very different pace. So we need to see more companies like Epirus, like Varda really jump to that next level and we’re in the early days of that phase.
Dan Patt:
Or more billionaires, one way or the other, that might also. . .
Joe Laurenti:
As a non-billionaire founder, it doesn’t put us in a good spot.
Michael Brasseur:
I thought you were talking about yourself. It’s not you?
Joe Laurenti:
It’s not us.
Bryan Clark:
So Josh, we’re going to go to audience questions here in just a second, but one question I have for Josh is commercial space has obviously been a huge contributor in Ukraine. It’s becoming a huge contributor in the US military’s operations, and not just Starlink and Starshield, but much more broadly. So how, as opposed to just being as a service provider, how are you working with the government to enable it to do things that go beyond just feeding it information on the service basis, but instead providing it capabilities that go at different parts of the kill chain?
Josh Martin:
That’s a good question, and I think that the department has made some really good steps recently with its commercial integration plan. I think to that point of how can we ingest what that commercial service is without trying to sprinkle some sort of additional government requirement on top of that, how can the commercial sector meet that need? And I think certainly there’s always going to be an insatiable need for sensing and remote sensing and communications that are available to you because of space-based operations.
But I think then looking past that next layer, how can you fully leverage the space domain I think is something of interest. Certainly we are trying to meet that demand with the test community right now with both strategic systems and hypersonics organizations that are looking to leverage that space-based ability to test and improve technologies. I think DIU is looking at some interesting things related to space-based applications for delivery of product anywhere around the globe at any point. The Air Force is looking at that as well as with Rocket Cargo. And so I think the application growth is evident right now, and there seemed to be some commercial players that can fulfill that need.
Bryan Clark:
It seems like if you’re looking at doing things like space electronic warfare or space logistics, you’re going to have to tap into commercial capacity.
Josh Martin:
Yeah, you’re not going to build that on your own right now.
Bryan Clark:
Right. Depending on how dispersed you need it, how distributed you need it. So questions from the audience? We can, oh, the young man there in the. . .
Chris Servello:
Hi, Chris Servello from Provision Advisors. Can you talk a little bit about your access to your customer in terms of helping them help you with the requirement definition? A lot of times they may not know the technology that you have and how that fits into their requirement, and then that as part of a larger time horizon, whether it’s for you as a company, but also for the larger security time horizon. I mean, how do we make that process a little bit faster?
Bryan Clark:
Mike, you want to start? We can just go down.
Michael Hiatt:
Sure. Thank you. Yeah, I mean that’s a great question because near and dear to my heart is know Epirus is right now focusing on the counter drone mission, the counter UAS mission. And so we had the good fortune that I could buy a lot of those. I didn’t need to wait for a government person to tell me that drones were a problem. That was on the news, and it’s getting only worse. I spend far too much of my time watching things that are happening in Eastern Europe and how effective and disruptive drones can be, and now you’re seeing that in the Middle East. But as we sort of start and show, okay, there’s a minimum viable product, there’s a capability that works, now I need to start honing in on what does the Army actually want in the case of the Army or the Navy or the Air Force.
And that’s been, that is a challenge. I will say the Army has done some great work in standing up with Army Futures Command, the office that we work with, the Rapid Critical Capabilities and Technologies Office. They’ve been very forward leaning in terms of buying prototypes and getting out and testing them. As we start to look at towards the next step though, well, big Army now needs to decide, all right, what are we going to do about the drone problem as an example, and how does that fit into existing doctrine? That is a problem that a Patriot could solve, but very much not cost effectively. I know what everyone focused on with the Houthis and the Red Sea is as the USS Carney’s lobbing SM-II’s at stuff, the cost asymmetry. But magazine depth is a huge problem as well. There’s only, I think four Patriots per battery in their little interceptor boxes. And the USS Carney only had so many missile tubes before it has to go and reload.
And so you end up with a lot of logistics challenges there and things like that. And so I think the DOD as a whole acknowledges that, but then they’re trying to figure out, all right, with this broad array of technologies to solve these problems, how do we pick that? How do we pick some winners? And I think they’re definitely struggling with that as you start to look at how do they, through a traditional acquisition process of having a capabilities and description and go through analysis of alternatives and really hone that into a doctrine, the threat is rapidly changing. Two years ago, nobody was thinking that armor was as vulnerable as it was, as what you’re seeing in Ukraine. We’re learning rapidly as we go. And so I think that’s definitely been the problem. But I think the big way to do that is like what Task Force 59 did and what General Carrillo is doing out in CENTCOM, is really just cut out the middle folks in some way and really get industry as close to the front as possible and really understand that threat for that rapid iteration.
Bryan Clark:
Mike, you want to weigh in?
Michael Brasseur:
Yeah, I had two sort of perspectives on this one. We put, for example, Skapa in San Diego, right near Naval Special Warfare, right near the Pacific Fleet, right near Coast Guard, US Marine Corps. So go to where the operators are, and then two, go to the edge. Our deputy CEO, Anders Carp has been to Ukraine multiple times. And just go to the operators and go to the problem space. And it just so happens that know Skapa is located 50 meters from a brewery, which is very convenient for eliciting raw, unfiltered discussions about problems. And so focus on the problems, go to the operators and go to the edge, I think are some of the keys.
Bryan Clark:
Yeah, Joe?
Joe Laurenti:
Couldn’t agree more. We use a fantastic tool called whiteboards. So agree with Michael on that. Maybe back to your question earlier, Dan, something that is a commonality probably, or maybe it’s trending in the right direction between now and say 1955, is we’ve never met a closed door. We may not be able to make thousands of Patriot missiles a year yet, but we’ve never had someone in the Pentagon say, “We’ll take that meeting in five years.” Everyone is open to ideas, everyone is willing to hear what companies like us are doing today and the trajectory we are on. So getting boots on the ground, meeting with the end users, getting to the edge is really important.
Bryan Clark:
Josh?
Josh Martin:
No, I would agree with that. I think the defense ecosystem and the ability to interface with decision-makers and people that are interested in your technology is better than it definitely was 10, 15 years ago. And I think our experience is that the interchange between our government customers is very high quality. I think that they’re very interested in learning about our capabilities and how we can help them push boundaries. I do think that one area where our system or Department of Defense almost fails though is with the regular budget and planning process in that it is, I think, very difficult for program offices to ingest new technologies in a meaningful way because of the very complicated and lengthy process that the department undergoes on an annual and future year looking budget planning process. And I know that there are people that are a lot smarter than me that have tried to tackle that over the years, but to me that is one that is, I think, consistently almost broken.
Bryan Clark:
Well, thanks Josh. Yeah. We have time for one question real quick, and we’ll need to wrap it up. Morgan can sort it out.
Speaker 2:
We talked a lot today about modularity as an enabler for scale. Josh, on your point earlier you mentioned that there was a performance trade off on the basis of a standardization initiative. To be really, I suppose, precise with that, what would’ve needed to have been different for that to have not been a constraint? Because I suppose it flies in the face of this modularity opportunity.
Josh Martin:
I think it comes from good intentions that the government perspective is if we can standardize this particular interface, that means that’ll squeeze money in time out of integration activities. And that is not always the case. And in some cases it may add an element, and particularly with our architecture, it actually could reduce the amount of experiments that we could fly on a given flight. And so I think understanding that, again, the intentions I think are good, but sometimes counter-cultural, I think, or counter to the intentions and that needs to be discussed. Yeah.
Bryan Clark:
Well, thank you very much. Thanks Josh for that answer. So I want to thank you all for being here today. We appreciate the time that you took and the great questions. I want to thank our panelists. So Michael Brasseur, the Chief Strategy Officer of SAAB Inc. Josh Martin, the VP for government relations at Varda. Joe Laurenti, the Chief Executive Officer of Ursa Major, and Mike Hiatt, who’s the chief engineer for Epirus. And then Dan Patt from here at the Hudson Institute. I want to thank you all for being here. Let’s give our panelists a round of applause for putting up with us.
And with that, I want to also thank our partners in this effort, Clarion Defense, who is going to be putting on the APEX Conference coming this next January. And we’ll be doing some more events over the course of this year to explore this idea of how do you solve problems that the Defense Department and other militaries are facing, taking advantage of new technologies. So look forward to seeing you at APEX in January of next year. And with that, I want to thank you all and thank you all for being here online. We’re going to have a reception here to follow. If you’re at home, feel free to have your own reception. Good night.
Join Hudson Senior Fellow Bryan Clark for an event with Chief Digital and Artificial Intelligence Officer Dr. Radha Plumb.
Congressman Mark Green will join Hudson’s Dr. Jonathan Ward to discuss the importance of cybersecurity, critical infrastructure defense, maritime and border security, the fentanyl crisis, and more.
Join Hudson for a discussion on the trade and technology relationship between Washington and Taipei with US-Taiwan Business Council President Rupert Hammond-Chambers and Taiwan Semiconductor Manufacturing Company Senior Vice President Peter Cleveland.
Join Hudson for a discussion with renowned intellectual property experts including former United States Patent and Trademark Office Director Andrei Iancu, Dinsmore’s Brian O’Shaughnessy, and the Special Competitive Studies Project’s Rama Elluru. They will explore potential policy changes, challenges, and opportunities for the innovation and creative sectors in the new administration.