To get lifesaving technologies to the battlefield, BMNT brings together defense officials, warfighters and experts in computing, big data, cybersecurity, energy, robotics and other areas that have military applications--and which are now dominated by tech companies, not traditional defense contractors. Using cutting-edge methodologies developed in Silicon Valley, BMNT gets a diverse roomful of people working together on finding rapidly deployable solutions to battlefield problems.
BMNT was founded in Palo Alto in 2011 by Joe Felter, a retired U.S. Army Special Forces colonel, after he returned from Afghanistan, his final deployment. Felter had earned a doctorate at Stanford University in 2005 and was a U.S. Army War College fellow at Stanford's Hoover Institution from 2008 to 2009. During his time in Palo Alto, Felter developed a strong network of entrepreneurs, investors, engineers and academics interested in national defense. At the fledgling company, named BMNT after the military term meaning "begin morning nautical twilight," the time of day when the sun is still below the horizon but light is just beginning to appear (the best time to attack and the worst to defend), Felter used this network to find consulting and advisory work focused on helping tech companies interested in working with the military.
In 2012, Felter met Army Col. Peter Newell, director of the Army's Rapid Equipping Force (REF), who was visiting Stanford faculty members to get ideas for the REF. The Army had given Newell command of the REF after he served two tours in Iraq, including leading an infantry task force in the second Battle of Fallujah, in 2004, for which he was awarded a Silver Star, and later commanding a brigade combat team in the southern Iraq border region, responsible for halting the inflow of weapons and fighters from Iran.
During Newell's many trips as REF chief to Silicon Valley, Felter organized "garage crawls" to expose Newell to potential tech solutions for REF's problems. BMNT's website notes one particularly important trip, when a senior Google executive was blunt:
" 'We don't want your money; we want your problems.' This reinforced to Pete what he already suspected--that the key to engaging Silicon Valley was not through government money, which is limited and difficult to access, but by challenging these entrepreneurs and visionaries with DOD and other government agencies' problems."
Although many believed Newell was in line to get a star and remain in the Army, instead he retired and later joined Felter and William Treseder, a Marine veteran and Stanford alumnus, to form a new company called BMNT Partners LLC. Today they simply refer to the company as BMNT.
As Newell was driving across the country to take the reins at BMNT, he got a phone call from Jackie Space, a former U.S. Air Force officer who had worked with Newell when he headed REF, coordinating his West Coast operations. "I called him up and said, 'Hey, can I come help you out in whatever it is that you're doing with the startup?' And that's when I became the first employee at BMNT," Space said in an October interview she and Newell did with Army AL&T.
While Newell was getting started with BMNT, he met Steve Blank, a Silicon Valley legend and serial entrepreneur. (See sidebar, Page XX.) After either founding or working for eight startups during a 21-year career, Blank had begun teaching his "Lean LaunchPad" methods at the business schools at Stanford and the University of California, Berkeley. Newell and Blank realized that what Blank was teaching was precisely what Newell had been doing with REF: finding a battlefield problem and assembling a team to validate the problem, test possible answers and develop technologically informed solutions. BMNT dubbed the process Hacking for Defense, or H4D. Later, BMNT and Blank combined to teach a class at Stanford also called Hacking for Defense, in which students formed teams to find solutions to problems provided by government defense and intelligence agencies.
BMNT's work has gotten the Pentagon's attention, and Forbes magazine recently included BMNT in its list of the top 25 veteran startups. One of the company's clients is the Defense Innovation Unit Experimental (DIUx), Secretary of Defense Ash Carter's outreach to the tech community. Other clients include the Joint Improvised-Threat Defeat Organization, the National Geospatial-Intelligence Agency, the Defense Intelligence Agency and the MD5 National Security Technology Accelerator, which is part of the Manufacturing and Industrial Base Policy office in the Pentagon. NEWELL and Space are also visiting senior research fellows at National Defense University. They talked about the evolution of BMNT with Army AL&T magazine.
ARMY AL&T: When you took over REF, and correct me if I'm wrong, you didn't know it existed.
NEWELL: No, I didn't. I got a note from the Colonels Management Office that said, "Congratulations, you're going to take over the Army's Rapid Equipping Force at Fort Belvoir, Virginia." And I literally had to Google it to figure out what it was. Considering that I was one of only six brigade [commanders] in Iraq, that wasn't a good thing.
I moved to D.C. in July, and I think took over REF at the end of the month. I knew nothing about acquisition, nothing about money, nothing about how things work in the Pentagon. I had all this fresh experience from theater, and I was pissed off that the organization had "overlooked" my brigade in Iraq.
As I started my transition, I was given a list of people to talk to by my predecessor and warned ahead of time that there were many people in the Pentagon out to shut down REF, and that I should avoid talking to them until I had done the list of REF fans. I decided to take the opposite track and talk to those who weren't fans of REF so I could get an unbiased feel for what REF was doing, just to get a sense of where the problems were.
Back then, REF was, I would best say, on a targeting board by most of the folks in the acquisition community for their behavior over the past couple of years. Their relationship with the asymmetric warfare group was horrible, and the relationship they had with senior leaders in the acquisition community was horrible. Fortunately, because I hadn't yet formed an opinion, I had a great opportunity just to hear what people had to say. It was good for me.
Shortly after that, I took off and went to Afghanistan for a visit because I had not been there in several years and because that's where the bulk of our work was being done. I think it was October 2010. At that point, REF's headquarters in Afghanistan was in Bagram and there was a satellite office in Kandahar.
One of the stops included the product integration facility that RDECOM [U.S. Army Research, Development and Engineering Command] had just opened there. It was essentially a small manufacturing facility on Bagram that was the size of a warehouse, where they could make just about anything they wanted. It struck me that you could hear a pin drop in the place, because there was nothing happening. After my visit there, I went back to the REF office, where the REF lab guys [were] working night and day doing something in a room the size of a closet. I asked the guy that ran the place, "Why are you guys doing all this stuff internally when you've got that big old facility on the other side of the base?" I never really got a good explanation on why it was, other than, "This is the way we've always done it."
I took a flight down to Kandahar to see the other REF office and to visit units working there. I just happened to bump into a guy I had served with in the Ranger Regiment several years before. He was now in Kandahar commanding the first brigade that had gone into the city as part of the surge in 2010. We were standing at the edge of the airfield talking, and he looks like a cadaver. He'd been in this country, I think, six weeks. And at the time, he's losing, I don't know, 10 to 15 people a day.
At one point I asked him, "What can I do for you?" And his response to me kind of set me back a little. He just looked at me and said, "What I need most is for people to quit asking me what they can do, and just do something." What he meant was that he was so busy trying to fight the daily fight and keep people alive that he didn't have time to do analysis of what his problems were. What he really wanted was for somebody to look over his shoulder and anticipate what was going on, and hand him potential solutions for his guys to try. And he wanted it done in real time.
After that, I went to see the guy who commanded the second surge brigade into Kandahar. When I asked him what I could do to help, he said, "We're having a seriously hard time with IED [improvised explosive device] attacks against our dismounted squads. … Do something about that." I'd had all the briefings by the division staffs and everybody else, and IED attacks against dismounted forces didn't show up on anybody's priority list. Nowhere.
Later, I made one more stop, this time back at Bagram, where I went to see the guy who was the commander of the Joint Special Operations Task Force, another officer I had served with years earlier. When talking about the IED attacks, he became even more direct in what he wanted. He said, "Yeah, my biggest problem is my guys are getting whacked on the way to the objectives as they're walking in." And he said, "I need you to take all of that big, heavy-ass stuff you're building, you know, to go up and down roads, and I need you to shrink it down so it fits in the back of a Chinook and will operate long enough for us to get us to our objectives."
I was frustrated by the time I left Afghanistan because I thought REF was missing the fight. It appeared to me that REF had simply regurgitated in Afghanistan the model that they'd used in Iraq. By camping out in Baghdad in Iraq, REF had access to hundreds of thousands of people. But as we drew down in Iraq, there wasn't anybody in Baghdad anymore. And they weren't getting out at the edge of the battlefield. In Iraq, the model never changed and they missed my brigade in the south. In Afghanistan, they were doing the same thing. … they were camped out on the large bases, but this time well away from where the people who needed them most were operating.
In my mind, they really had become a very passive organization that said, "Yes, when people send us problems, we respond to them." But they weren't looking for anything, which I think was diametrically opposed to the way REF had started. Col. Bruce Jette, the guy who designed REF, did an absolutely beautiful job with it. Unfortunately, between Bruce's departure and my arrival, you got folks who were in the execution mode, who started to celebrate, picking low-hanging fruit off trees and pitching it over the wall at people, and started reading their own press at how great they were and how jacked-up the acquisition community was. But what REF wasn't doing was helping anybody, other than those short, easy successes.
So when I left Afghanistan, I set out to find out why REF was so counter to what I thought they ought to be doing, and also figure out how to re-chart the course for it. Our analysis would show later that between the start of the surge in Afghanistan and when I got there and made defeating IED attacks against dismounted forces the No. 1 priority at REF, there was a six-month gap, and that six-month gap cost us probably 5,000 casualties.
Having left Afghanistan with kind of a mission in my mind, I spent the next year with a host of folks trying to figure out how to enable REF to focus more on finding problems, and then how to rebuild the organization so it constantly pushed itself to the edges of the battlefield.
Part of my time was spent looking for a professional development program that would help REF's workforce keep up with a rapidly changing world. I spent lots of time in the executive education courses at MIT, learning how businesses thought and how they handled problems. As I understood things better, I started sending batches of REF people to MIT executive ed courses to work on things.
We redid the way we spent money and created platforms by which our teams were moved out to the edges of the battlefield. We then spent our money [on] better understanding and validating the problems that they found, and then used that data to build coalitions of people to solve them.
ARMY AL&T: You actually put labs out on the battlefield, didn't you?
NEWELL: We absolutely did. I went so far as to build mobile prototyping labs--took a 40-foot [shipping container] and put in its own HVAC and power system. Added a VSAT [Very Small Aperture Terminal] to it and then loaded it with CNC [computer numeric-controlled] milling machines, 3-D scanners, 3-D printers and other essential prototyping tools. We manned the labs with a scientist, an engineer and a senior NCO who had combat experience in the theater. We then gave the labs to the divisions to move to the units in combat.
The role of the lab team was to reach out to the guys coming into the base from missions and grab problems from them as they came in. They then used the lab's prototyping capability to actually help us better understand things. It wasn't about building parts and things, although we did sometimes. It was about making sure we understood the problem and using that to help generate interest back in Washington, D.C., to actually solve it.
To cut a very long story short, I used that platform that we built, took my $200 million budget, and ended up spending $1.5 billion over the next two years, because we got so good at building these problem sets and building communities around them. By the time I left, we had people bringing us money and saying, "We know we're supposed to be working on X, but we can't get it done. Will you help us get started?"
As I got toward the end of my career there, I really got addicted to what we were doing.
ARMY AL&T: So how did you go from REF to Silicon Valley?
NEWELL: I had already moved my family five times in seven years. I was simply done uprooting their lives for the chance at another promotion. So I decided to retire.
During the process of the buildup for REF's problem-solving mechanism, I ended up in Silicon Valley looking for a solution to a particular robotics problem. And it just happened the day that I showed up at Stanford University to find a particular mechanical engineering professor, the guy who's my partner now, Joe Felter, had just retired from the job as the COIN [counterinsurgency] adviser to Gen. [David H.] Petraeus in Afghanistan.
He got his Ph.D. from Stanford, and when he retired, he went back to Stanford to work for [former Secretary of Defense] Bill Perry. Joe was assigned to escort me across the campus that day. Joe and I were wandering across campus, talking about the problems I had in Afghanistan. Instead of going back to the airport, we jumped in his car and he said, "Why don't you just come with me for a bit?"
And we went on what I would call a pub crawl of startups, where we'd pull into some garage where he knew folks, and I'd sit down with the CEO of the startup. I'd talk about my problems and he'd talk about what they were doing; before long, the napkin sketches came out. It was fascinating to me that I was having an exceptionally meaningful conversation about national security problems with a guy in a garage in Silicon Valley. What I learned over the course of my visits there was that folks in Silicon Valley would respond to big problems. In fact, that was the currency that we actually traded in.
In the Valley, it wasn't so much that I was looking to spend money to buy tech that was valuable as it was that I had really good, hard problems that people loved to play with. Over time, Joe and I worked out a rhythm where I would send him a one-page document that had my top three or four issues on it. Joe would translate the government-speak into something that made sense in the Valley and use that to find the right people to talk to. I'd show up a month later and he'd have rooms packed full of people. Rather than waste my time talking to 1,000 people to get to the two I really needed, at Stanford, we'd knock that down to 40 and I'd get 10 of the right people from the group.
The time savings alone was huge, but more importantly the quality of dialogue we were having was even more improved. I couldn't get to Silicon Valley as often as I wanted to, because of the demands for other things. So REF hired Joe and another guy to actually be our scouts. Essentially, they did for REF then what DIUx is doing for [Secretary of Defense Carter] now.
Not long after that, we found ourselves doing close to $100 million worth of work on the West Coast, and we had a hard time getting our program managers out there to pay close attention to the more challenging programs--some call them high-risk programs. In our case, we called them highly fragile programs. These programs needed somebody's personal presence to give us early warning that something was amiss so we could take remedial action on them, or something was going very well that we could take advantage of.
I shared our problem on the West Coast with an office in D.C. that did similar work and asked them how they were doing it, only to find the No. 2 guy in the office actually lived on the West Coast. They managed an intro to Jackie Space, who turned out to be the ideal candidate for filling a PM role out there.
A few months later, I was at Stanford and let folks know that I planned to retire soon. Joe, being well-trained at influencing peoples' behaviors, grabbed me and said, "Listen, I don't know how it would work, I don't know what it would look like, but why don't you come out here? And we'll figure out how to do this, you know, facing the other way, by pulling government problems into the Valley and building coalitions of folks around them."
It took us almost three years to actually put the puzzle together and do what we first drafted on our own napkin sketch. We ended up forming a new company so that it was a fresh start for everybody.
I was driving across the country on my move to Palo Alto when serendipity struck twice. The first time happened as I was heading into Kansas City, Missouri, when some folks I had briefed on REF's operating model some six months earlier called me and said, "Will you stop in Kansas City? We want you to talk to the director of our facility [the National Secure Manufacturing Center (NSMC)]," which is a DOE [U.S. Department of Energy] facility that does all the classified manufacturing for the government.
Over coffee in a hotel restaurant, the director laid out their challenge, which was increasing pressure to become more responsive to the agencies that used them. But their workforce largely came out of the defense weapons program, which means they had program managers who were accustomed to doing 17-year programs, not seven-month programs. And they were really struggling to understand how they would apply innovation to their business model and how that would work. They asked me to describe what I was doing at REF, and I got halfway through that description, was literally coloring on a napkin in a restaurant. The guy stopped me and said, "We need you to do that for us."
I begged off long enough to finish moving, but I told them that I'd come back and re-engage. We ended up spending two years working with the NSMC team before transitioning to a wider role supporting the DOE Advanced Manufacturing Office--something we still do today.
The second strike came two days later. I was cruising through the plains of Kansas when Jackie called me and basically said, "I'm not doing anything at REF." She asked if I had room for her on the new team we were building in the Valley--knowing full well this was a startup and we had no idea what the hell we're doing, and that we had no idea if it would make money and or be a flop. God love her, she was willing to take the plunge with us.
Those two events were fortuitous, because the DOE work provided us a platform to actually test out a lot of our concepts while pushing us in directions we had not considered before. It was challenging work that also helped us get the company started.
We eventually settled on a couple things. One was that we were valuable as what I would call a disinterested third-party ombudsman to everybody, which means we sat in the middle and translated DOD verbiage and behaviors for Silicon Valley, and at the same time we educated DOD folks on how Silicon Valley functioned. With that common framework, we found we were able to get them to negotiate their way to a common pathway toward deeper involvement together
ARMY AL&T: What was the first year or so of BMNT Partners like?
SPACE: That first year was pretty chaotic. Early on, I felt like when we showed up in the Valley, we got sideways looks from everybody. I don't think anybody really understood what we were doing there from the startup community side, because we weren't raising money around the next one-hour delivery app. Even on the government side, people questioned why we were there.
ARMY AL&T: Did any of you have a tech background?
SPACE: I'm an engineer, but my tech background is in aerospace. Pete, I think, probably was the most well-versed out of all of us, because while at REF he touched so many different technologies and their applications. I think he had the best grasp of what was on the leading edge.
ARMY AL&T: But it's safe to say you were not a typical Silicon Valley startup.
SPACE: No, we were not at all and still aren't. All this preceded the launch of DIUx, so there wasn't anybody out there that was well-known, except for Palantir [Technologies], SRI International or PARC.
That first year was also challenging because our energy was spent trying to establish who we were and what we were doing while also learning about what made Silicon Valley really tick, and then how to get them to work on national security problems. We did lots of small jobs for about a year and a half, I think, before we really transitioned to this Hacking for Defense model. We found that it helped us to be that third-party agnostic voice in the room. So we transitioned to working more on the government side and sourcing problems into the tech community.
When I talked to companies, they sometimes would ask, "What do we get out of this if we show up for an hour in a Hacking for Defense thing?" My response to them was that they were talking about how they all wanted to expand their portfolio in some way--to have more diversification--and that the government problems might provide a means to do so. Up to that point, the only mechanism they had was to basically canvass FedBizOpps and then send people to conferences and trade shows and try to get meetings with government people to try to have a conversation. None of which they were really interested in doing. Silicon Valley's business model just isn't tuned that way.
I've been on both sides of equation; responding to RFIs [requests for information] and RFPs [requests for proposals] takes a lot of time. For a fast-paced company, especially one in the startup stage, there's no way they could spend the time doing this.
ARMY AL&T: They could spend a day looking up acronyms.
SPACE: Oh, it's terrible. We spend a lot of time just translating that aspect of it. It's maddening for a company to get in one government place and get it all worked out, then have to start over again with every new organization they approach. No small company or tech company focused on building a commercial product will waste its assets doing this.
NEWELL: We came to understand most from our work with startups and investors: The minute a startup takes money from an investor, they are on a three- to five-year timeline to sell that company. And their focus is to get a product into the market as fast as they can. There is no wiggle room for them to divert assets to exploring side deals with the government. I've had this conversation more times than I'd like to count. At one point, the CEO of a startup looked at me and said, "Listen, I would love to do this. As an American, I would love to do this. But if I touch that thing, this company will have a new CEO here the next day. Because it has nothing to do with meeting the immediate objectives of the investors of this company."
That's the hard part. If government folks in programs aren't taught to figure out how to get in sync with potential early-stage companies sooner and with a better value proposition, they just won't get much from them. It's unfortunate, because instead of showing up with a solid market analysis of why the company should work on the government project, they simply show up and ask to look at what technology they have to fill the government requirement. Then they are surprised when they get, "Thanks for your interest, but we aren't going to have a $1,500-an-hour engineer write a white paper for you. No thanks."
SPACE: A caveat to what Pete was saying about startups: The talk about DIUx and the dialogue from D.C. tends to focus on "startups." The term startup applies to a wide range of companies: everything from the girl working out of a garage to a company that is fresh off a multimillion-dollar seed raise. There's a big difference among them, and even other small companies that are on solid footing but have never done business with the government before.
When I look at the problems that we're sourcing in Hacking for Defense, what I love about the model is that it's not targeted necessarily to startups. It can bring in any type of solution provider. It could be a big company, it could be an academic, it could be somebody on the startup side. But it's not necessarily geared toward startups. It's geared toward finding the best solution. Almost always, it's a combination of all of them working together.
ARMY AL&T: It seems that what you do isn't so much about technology, but about problem-solving.
NEWELL: So much of it is really about the sociology of building teams around problems.
SPACE: From the culture side, the Defense Department is not used to having to reach across like that, or reach out from the acquisition side. They're used to having the vendors and people come in and pitch them all the time. So this has really been a new sort of cultural way of looking at the behavior of an organization and how they look at their problems.
ARMY AL&T: What kind of feedback have you gotten from DOD? Is there resistance?
NEWELL: If you measure the reaction in terms of the workload, we have more work than we can possibly handle. I still think, though, that we are met with a little bit of--I would call it skepticism. You really have to educate people, and quite honestly it takes a lot of socialization inside some circles of the government to get them to understand how they'll be able to use this in the environment that the systems are currently built in.
SPACE: I've found so far that our initial engagements come from people who would be considered mavericks inside their own organizations. We'll start the conversation with them and over time help them educate others within their organization on how H4D can be used to help them.
People think, well, if you have a general sold or some secretary, then that will make a big difference. But it's really the middle layers that have to adopt and execute on this. That's who we spend the bulk of our time with.
We are getting good responses, I think, not just the hand-waving and gentle pats on the back, but real solid engagements followed by real solid problems to work on. Despite the results, I don't think it's our job necessarily to go out and try to convince everybody to do this.
ARMY AL&T: There has been some criticism of your approach, and for that matter DIUx's approach, and criticism from supporters of some of the big defense contractors, who say that what you're doing is all nice and well, but it's small potatoes and not going to make that much of a difference. How do you respond to that?
NEWELL: I call it shortsighted, but Jackie has a more worthwhile explanation.
SPACE: We've been maybe viewed as being competitive to defense contractors, which I think is silly. We know that the more work that gets done like this, the better off the programs will end up being when the government goes to write the requirements. I don't see how it's not going to help by understanding the requirement better from a technology perspective and from a user perspective.
The large defense programs--there is a place for that, when it comes to building ships, tanks and fighter jets. But when you're looking at all of the R&D [research and development] that's feeding into those programs and all the prototypes that are being built that don't have a direct alignment to a customer and aren't in line with … the leading edge, I think that's a problem.
NEWELL: I'll look at it from a warfighter's perspective. The way the world is today, it doesn't matter how good your kit is the day you show up for a war. What does matter is how quickly you change once the bullets start flying. People are finding ways to circumvent our best technology more rapidly than we can actually get it out on the battlefield. So offset in the future is really about speed, not about any one technology.
I've had a little bit of pushback from some of the bigger guys, and some others who said, "You know what? You're right. We have to figure out how to behave better in this arena, and we need to figure out how we're going to become part of that ecosystem." Today we work with some very large corporate clients who are hell-bent on figuring out how to do this. The us-versus-them thing between defense firms and startups is nice fodder for news articles, but it's not based on reality. Wasted energy, I think.
This article will be published in the January - March 2017 issue of Army AL&T Magazine.
U.S. Army Acquisition Support Center