This characterization of the Ukrainian forces could describe many startups. Both are groups of individuals who commit to come together, and collaborate to solve difficult problems for a common cause.
In their war with Russia, Ukraine is bolstered by technologists that are motivated to contribute to the fight, and forces that are adapting quickly.
Whether it is outfitting UAS with autonomy capabilities or battlefield communications via Starlink, the Ukrainian army is deploying innovation in the war effort. The creative maneuvers that are necessary for a nation’s survival today stand to reshape defense tech going forward.
What follows is a personal account from Dan Madden, Principal for National Security at Squadra Ventures, where he leads the investment strategy for defense and dual-use startups. With a small group of technologists and venture capitalists, Dan traveled to Ukraine this summer to get an up-close look at the role innovation is playing in the conflict. Dan teaches at Columbia University, and previously served in the Marines, the US House of Representatives, the Office of the Secretary of Defense, NSIN, and at the RAND Corporation where he researched topics ranging from capability requirements to war plan reviews.
(Photo by Dan Madden)
Over the course of the next week, we spoke with government officials, startup founders and engineers, civilians and soldiers. This research was intended to inform Squadra’s national security investment thesis, develop deeper relationships with the Ukrainian tech ecosystem, and identify lessons learned for how democracies need to manage relations between their military and industrial base — particularly startups — in the face of an existential threat.
Conversations in the city streets and stories shared with us from the battlefield revealed that innovation during wartime is shaped by a nation of people who are steadfast in their will to fight and a collection of tactics and capabilities that are constantly advancing.
These discussions also clarified a hard but looming truth — that most U.S. and Western systems are having an important impact, but still massively underperforming expectations.
U.S. policy makers and DoD leaders must wrestle with this underperformance to understand how we got here, how the Ukrainians have adapted, and how defense and industrial policy must change going forward. Defense tech founders and investors should be — and are — paying close attention to identify the opportunities available for making a meaningful contribution to the security of the U.S. and our allies. It comes down to two things — the mobilization of talent and adaptation of technology.
On the train to Kyiv, a kindergarten teacher told me about her favorite T-shirt. It was black with a HIMARS rocket launcher printed on it. The HIMARS is an iconic system provided to Ukraine by the U.S. It was the same shirt that triggered a Russian woman to slap her while she was on a trip out of the country to visit her mother.
Kyiv couldn’t feel more different than Kabul. From a U.S. foreign policy perspective, some of the most important differences are actually the most subjective. Walking around Kyiv, I didn’t just feel safe, I felt a sense of unity and confidence. Walking around Kabul, I was worried about getting kidnapped, at best. That difference is symptomatic of two key factors in Ukrainian national will to fight — Ukrainian unity against Russian aggression, and alignment of U.S. interests with Ukraine’s in the current conflict.
That Ukrainian citizens fighting for their homeland would fight harder than Russian conscripts and recently released criminals is both inspiring and unsurprising — particularly in the context of Russian atrocities and President Zelenskyy’s wartime leadership.
What I found especially interesting was how this will to fight manifested itself throughout Ukrainian society. This was obvious in the sudden decline in spoken Russian in everyday life, sitting at 15% of the population down from 26% pre-war, but this fervor was especially striking in the tech sector.
This moment of extraordinary solidarity and social mobilization brought a tremendous amount of new talent to Ukraine’s defense, through both the military and industry. The evolving battlefield tactics and institutions are unifying the efforts of Ukrainian founders and technologists.
It has become a common trope to compare the current fighting in eastern Ukraine with WWI trench warfare. The deeply layered Russian defensive lines, extensive minefields, and reinforced trenches and command bunkers justify the analogy. Other analysts have suggested more optimistically that the fighting is closer to the WWII grinding attacks through the Normandy hedgerows, followed by more fluid and decisive allied operations. Explanations based on this analysis tend to focus on technology and terrain, and debate the dominance of offensive or defensive operations. All are plausible to me, but what I heard from acquaintances with recent combat experience pointed to a more human factor — training.
As a young Marine, I recall being told, “Fire without maneuver is wasted, maneuver without fire is fatal.” Both sides are doing an uneven job integrating fire and maneuver. Among Ukrainian forces, some felt that fires were being used as a substitute for maneuver. Drone and artillery attacks were conducted without supporting maneuver forces. Armored forces were operating without integrated infantry support.
During a war for national survival, citizen soldiers have compressed training timelines, even for those units earmarked for offensive operations. Reporting on Russian “human wave” assaults and skirmishes indicates their shortcomings, but also adaptation to variance in training levels.
At one level, it seems that both sides are relearning combined arms warfare in real time. But at another, the protracted character of the conflict has also resulted in both sides developing a deep understanding of one another’s tactics, driving effective leaders to creatively adapt, not simply employ, textbook tactics. There is a clear move-countermove competitive learning dynamic on the battlefield. Being predictable is a recipe for failure.
There has been ample reporting and analysis on how both sides are employing capabilities in Ukraine. From what we saw, capabilities involving technology and employment seem to be changing most rapidly. The scope of what I address here is constrained by classification and access – most of the conversations I had with operators were at a small unit level, and all of them were unclassified. Notably I don’t address cyber, remote sensing, space, mobility, air defense, or contested logistics capabilities here — nonetheless, these are clearly critical capability areas.
Many U.S. and Western military systems are massively underperforming expectations. Ukrainian forces have a revealed preference for simpler, cheaper systems that can be acquired and employed rapidly at scale. The more exquisite systems produced by the traditional U.S. defense primes, reflecting U.S. Department of Defense expectations about the future of conflict, have in most cases turned out to be irrelevant. The two most important dimensions driving this outcome have been the protracted, industrial scale attrition warfare, and the fiercely contested electromagnetic spectrum.
This capability performance deficit hasn’t been universal. I heard no complaints about dumb munitions or armor. The complaints only came for anything at the tactical level that required communications or precision location data — the foundations of the modern kill chain.
The contested electromagnetic spectrum shapes the employment of almost all other capabilities. Russians have deployed electronic warfare (EW) systems extensively, with counter-UAS EW systems reaching the platoon level, and signals intelligence (SIGINT) assets deployed across the front.
Communications are at constant risk of being jammed, decrypted, or used for targeting Ukrainian forces. It becomes impossible to communicate without a signal being detected. Employing legacy tactical radios with proprietary waveforms that claim to be “low probability of detection” is a great way to have your location prioritized by Russian fires. Even when available, these exquisite radios are often left behind on operations – or used as very expensive decoys in the face of overwhelming enemy signals intelligence and electronic warfare capabilities.
As a result, rather than seeking to avoid detection altogether, soldiers seek to avoid “being the most interesting thing on the battlefield.” Ukrainian forces are forced to use non-proprietary systems, typically commercial, to help them blend communications waveforms into the broader RF environment. This typically entails lower power, open frequencies with encryption, and waveforms that are ubiquitous. Motorola radios with a 256-bit encryption have become very common — on both sides. For similar reasons, Starlink has become the favored satellite communications solution, with more than 42,000 terminals in the country. The Russians themselves have shifted to exploiting Ukrainian civilian telecom networks for communications while on the defense.
Local digital coordination solutions like Diia, Kropyva, and Delta, similar to ATAK in the U.S., have enjoyed widespread adoption, enhancing coordination, and increasing the responsiveness of fires.
GPS signals are habitually jammed. This significantly reduces the efficacy of GPS dependent systems provided by the U.S. The effect is so acute that dumb artillery munitions, whose accuracy can be corrected using traditional means, such as bracketing the target, end up being preferred to precision strike systems in some cases.
The U.S. Air Force and Army are beginning to reconstitute their electronic warfare capabilities after decades of post-Cold War atrophy. The lack of sustained institutional focus may partially explain Western industrial underperformance. Given that China’s EW capabilities will be substantially more robust than Russia’s, this will have to be a major focus area for the DoD going forward.
Drones have become as iconic a system for the Russo-Ukrainian war as they once were for U.S. counterterrorism campaigns — only the use cases, variety, and employment have scaled massively. While here I focus on UAS sensing and strike missions, it’s worth noting that unmanned systems, to include ground and naval, are also being used for logistics, communications relays, mine detection, and remote weapon stations (RWS).
“Everyone knows where everyone is,” noted one operator. Drones, friendly and threat, are ubiquitous. UAS employ both optical and signals sensors. Drones might serve as a sensor to support artillery strikes, carry deployable munitions, or have munitions directly integrated into their design, such as “one way” attack drones. First-person view (FPVs) are not only used against infantry and trench lines, they’re increasingly being used effectively against armored fighting vehicles, with payload size and sophistication, like explosively formed penetrators, increasing with time.
FPV drones are prized at the front lines over longer-range one way attack drones. This might be an artifact of current offensive operations seeking to clear trench lines. It may also be that the earlier successful employment of Guided Multiple Launch Rocket Systems (GMLRS) in attriting Russian higher echelon command and control nodes within Ukraine, in combination with Russia’s transition to defensive operations, led Russia to adopt much stronger force protection measures including command bunkers and exploiting commercial internet for communications, that are less vulnerable to detection, indirect fires, and drone strikes. If Ukraine starts to prioritize deeper strategic targets, beyond current modest levels, long-range one way attack drones may become a higher priority as the conflict continues to evolve.
More broadly, FPVs are requested over more exquisite drones produced by traditional defense primes because of the scale that they can be produced and deployed at. Given the level of attrition that is occurring on the battlefield — 10,000 drones per month, half due to EW — Ukrainian forces have greater confidence in their ability to achieve their mission with a large volume of assets that may individually have a lower probability of effect than a given U.S. produced drone, but collectively have a higher probability. They’d rather have 1,000 $1,000 drones with reduced capabilities than have to depend on one exquisite $1,000,000 drone. Even the Iranian Shahed drones that the Russians have been using to strike at civilian and operational targets such as command posts and artillery sites are only $20,000 a platform, cheap relative to their $1M cruise missiles.
The Russians have honored the threat not only with their extensive deployment of EW systems, but also by prioritizing fires against Ukrainian drone operators. The Ukrainians have in turn adapted to Russian counter-drone operations with organizational, tactical, and technical innovations. Even before the 2022 invasion, nonprofits like Aerorozvidka were bringing talented technologists together to build and employ reconnaissance drones in support of Ukrainian forces. Now, Ukraine has established drone assault companies. One of the more interesting organizational innovations is Angry Birds, a hybrid operational and experimental unit that both employs UAS on the battlefield, collaborates with industry on emerging needs, and assesses new UAS systems for employment by Ukrainian forces. DevOps for drones.
Russian EW will typically disrupt drone operations by targeting communications links or GPS signals. Drone operators and builders sometimes respond by adopting different communications standards like control frequencies, but have also been exploring more AI-based solutions. Technologists are inserting computer vision and autonomy algorithms to solve for both navigation and terminal guidance challenges. These algorithms will exploit not only optical sensors and GPS when available, but also accelerometers and inertial navigation systems.
These autonomy solutions are important for enhancing the survivability of operators, as well. FPV controls can be disrupted not only by EW, but even by terrain features that UAS operators might use for cover or concealment — something as simple as a grove of trees. Operators regularly run into the problem where FPVs carrying lethal payloads lose control signal just as they’re entering terminal phase. They have a strong communications signal while high in the sky, but the signal becomes occluded as they drop towards a target. Standoff enhances protection.
Conceptually, the technical solutions that Ukrainian startups are using to add autonomy enhancements to drones aren’t extraordinarily novel. Engineers familiar with control theory and how U.S. systems solved for precision strike prior to the advent of GPS would recognize many of the solutions. What’s novel is the extraordinary compression of SWAP-C (Size, Weight and Power, Cost).
For a few hundred dollars, sophisticated drone operators can have a platform that can take out a modern tank.
For a few hundred dollars, a sophisticated engineer can augment a drone with sufficient autonomy to overcome the most effective EW counter UAS (CUAS) solutions available today.
These autonomy solutions haven’t yet been deployed at scale, but they’re being refined in the crucible of current operations in Ukraine. In conjunction with Moore’s Law, they will have an extraordinary impact on the character of war, and in the very near term.
Reporters have rightfully been impressed by the pace of military innovation in Ukraine. U.S. defense planners and policy makers have an opportunity to learn important lessons about military adaptation on the modern battlefield. But if the lessons are reduced to, “Russian centralization is bad, and Ukrainian decentralization is good,” mirroring older narratives about Soviet versus U.S. attributes, much will be lost. Ukraine has sought to strike a balance between centralized scaling and decentralized experimentation, but in practice the relationship between the two still seems to offer opportunities for improvement.
Ukraine’s General Staff provides the Ministry of Digital Transformation with guidance on current critical problems or capability needs. These aren’t typically detailed system requirements to generate a request for proposal. They tend to be high-level statements that give the Ministry of Digital Transformation the opportunity to explore a variety of technical solutions from industry. The Ministry of Digital Transformation will supplement the guidance they receive from the General Staff with interviews of servicemembers returning from the front, as well as more informal communications with soldiers forward deployed, sometimes through group chat apps.
The Ministry of Digital Transformation has several resources available to pursue these capability needs. The most clearly focused is Brave1, an accelerator for Ukrainian unmanned systems. Brave1 will assist Ukrainian startups and small businesses not only with grants, but guiding them through the authorization to field process. It should be noted that the authorization to field process can take several months, though progress is being made. Other resources include the Ukrainian Startup Fund, which focuses on commercial Ukrainian startups, United24, and coordination with the Ukraine Defense Fund.
In parallel to this top-down process, entrepreneurs, builders, and donors will directly engage with small unit leaders echelons below brigade to offer needed capabilities — unmanned systems, radios, and medical equipment were commonly mentioned. Entrepreneurs will sometimes maintain communications with forward deployed units that are using their systems to provide product support, but also to provide feedback on needed capability enhancements. Sometimes, the need is as complex as solving for navigation in a GPS-denied environment. Other times, it’s as simple as providing metal sheathing for power cords to protect them from mice, or improved carrying handles for a remote weapon station. Organizations like Angry Birds more formally connect and orchestrate collaboration between operators, technologists, and industry. Having flat communications between builders and operators greatly accelerates product development cycles, as well as enhances user adoption and capability effectiveness.
Builders also have easy access to experimentation sites for operating UAS in a realistic EW environment. One training site I visited on a typical Tuesday had several different UAS and aerostat companies flying simultaneously.
It remains unclear to me to what degree lessons observed from experimentation at the tactical level are communicated to higher echelons, how efficiently the Ukrainian General Staff is able to identify combat proven solutions for fielding at scale, or to what degree scaling is constrained by resource limitations, notably most UAS being employed are funded by donors.
The response of Ukrainian civil society to the Russian invasion and its will to fight has had an extraordinary expression in the influx of talent into Ukraine’s defense industrial base. We spoke with scores of founders and technologists that were building defense technology to protect Ukraine. Few had experience in defense. Prior to the invasion, most were working on B2B SaaS, adtech, agtech, or pursuing graduate degrees in electrical engineering. In some cases, teams formed in the early days of fighting around Kyiv, where they were living together and working around the clock to respond to warfighter and humanitarian needs. Counterintuitively, this crisis became a moment for rapid progress in public-private collaboration and digital citizenship.
This influx of talent has directly contributed to the rapid growth of Ukraine’s UAS industrial base. There are close to 200 Ukrainian UAS startups. We had the opportunity to speak with several of them, watch field experiments, and tour production facilities. The Ukrainian Ministry of Digital Transformation, Brave1, and the Eric Schmidt-backed team at D3, a Ukrainian drone accelerator, were all especially helpful.
Decentralization and entrepreneurship in defense has both positive and negative effects from a policy perspective. The upside has been greater resilience at both the platform and industrial level. The diversity of designs has made it less likely that the Russians can adopt a single, static counter UAS solution, driving continued investment, notably at some opportunity cost to offensive capabilities. The diversity of manufacturing locations reduces the probability of a catastrophic strike that cripples Ukraine’s ability to continue building UAS in support of current operations.
The downside is that each of these UAS companies are climbing the manufacturing learning curve from scratch, in undercapitalized facilities, with reduced ability to manage supply chains.
Ironically, supply chains are dominated by China. Ukraine is sensitive to the geopolitical risks of dependence on China, particularly as leaders criticize China for supplying Russia with commercial UAS for military applications. None of the alternative sources of supply have proved remotely competitive with Chinese component prices. Founders reported order of magnitude price differences for components like EO/IR cameras.
Though the decentralized character of the UAS industrial base offers challenges for scaling production, it may serve an important indirect role in public morale, providing citizens with a sense of agency and pride in supporting those going into harm’s way. It also strengthens the narrative of Ukrainian solidarity, innovation, and resilience. A 21st century victory garden for Ukrainian technologists.
The Ukrainians are fighting an extraordinary war to counter Russian aggression that has mobilized its entire society. The U.S. and its allies have important lessons to draw from the conflict, and Ukraine’s performance.
As we seek to continue advancing the critical edge for America through development and investment in new technologies, there are four major takeaways from my time in Ukraine that have significant implications for founders, investors and policy makers going forward.
If the velocity of adaptation on the battlefield and across the Ukrainian industrial base is the benchmark, significant cultural and structural hurdles remain to prepare the U.S. for its next major conflict.
Over the last several decades, the U.S. has progressively become more dependent on fewer, older, less ready, and more costly platforms and capabilities. This has made us an inherently more brittle force, and a force that our adversaries have ample time to prepare for. This has left us more likely to be surprised, than to surprise an adversary on the battlefield.
The U.S. and allies should be designing for adaptation. Startups like Hadrian and Prewitt Ridge are making the entire industrial base more agile and scalable with innovations in advanced manufacturing and digital engineering. Others, like Shift5, are providing key enablers to making these systems more resilient and adaptive by unlocking proprietary data and enabling modular open system architectures.
The electromagnetic spectrum that the modern kill chain depends on for sensing, communications, and precision strike has proven extraordinarily vulnerable. The ubiquity and disruptive effects of Russian EW and SIGINT systems on the battlefield is only a light foretaste of the kinds of capabilities the U.S. and its allies would encounter in a conflict with China. While there has been a great influx of talent trying to solve for autonomy in defense, there doesn’t seem to have been anything comparable on the EW or counter-EW side.
In part, this imbalance may be an artifact of investments in autonomy for commercial applications, and in part the opacity of EW use cases, much less data access. If DoD comes to the conclusion that this is indeed a critical area that requires innovation, much more material initiatives and investment will be required to cultivate a vibrant ecosystem of technologists and startups. Distributed Spectrum is a promising example of what’s possible. The communications tech ecosystem is more mature, with startups like goTenna and Fenix Group proving themselves in theater.
U.S. adoption of unmanned systems to support counterterrorism operations was a first move towards more affordable mass, but to be employable on the modern battlefield, these systems will need enhanced levels of autonomy. Important policy issues will continue to be wrestled with, but practical forward progress is needed.
Programs like the Air Force’s Collaborative Combat Aircraft are headed in the right direction. Deputy Secretary of Defense Kathleen Hicks’ Replicator Initiative is bold, and appropriately ambitious for the Department. Rep. Rob Wittman and Rep. Dutch Ruppersberger have been driving legislation to accelerate the adoption of autonomy capabilities across the Department.
Startups like Anduril, Shield AI, Primordial Labs, and Overwatch Imaging are poised to have a massive impact by providing the kind of autonomy solutions that platforms will need to be survivable and effective. Several startups are getting after the challenge of affordable mass, whether for UAS like Firestorm, USV like Saildrone, UUV like Vatn, or even hypersonics including Castelion, Specter Aerospace, and Hermeus.
The Ukrainians are learning and building under fire. The U.S. is not yet postured to do the same. If the U.S. found itself in a comparable fight demanding the real-time integration of novel capabilities, we’d discover that many of our current organizational missions and processes would fall significantly short of the mark. DIU is the most promising organizational innovation that the DoD has adopted, but rapid integration of emerging technologies can’t simply be outsourced to supporting agencies, it needs to be hardwired into the core processes and culture of the DoD.
Flat communications and rapid iteration between industry, senior government leaders, and forward deployed forces will be needed to achieve scalable wartime innovation. The trust required for collaboration can’t be quickly and effectively developed after the conflict has started. To counter the IED threat in Iraq, Joint IED Defeat Organization (JIEDDO) wasn’t established until 2006, and the MRAP wasn’t fielded at scale until 2007 — even then requiring the intervention of a Secretary of Defense.
Recent Combatant Command initiatives like CENTCOM’s Task Force 59 and INDOPACOM’s Joint Mission Accelerator Directorate are promising, particularly with Doug Beck committing DIU 3.0 to support this and other efforts, but will need to have more weight in driving Military Service priorities. The authorization of a hedge portfolio by Congress, an effort spearheaded by Rep. Ken Calvert, could be a critical enabler for a more adaptive force.
The character of the threat and technology environment has changed dramatically. The Ukrainian people have adapted heroically. The U.S. and its allies shouldn’t wait till the next war to learn from their example.
Ukraine did not choose this war, but it is being fought and mobilized by the decisions of individuals choosing sacrifices for the common cause.
Builders and operators. Soldiers and technologists. Leaders and citizens. They are all in a fight for survival, unified to do impossible things — and succeeding. A Startup Army.
—
Deep thanks to Porter Smith for organizing the visit, permitting me to learn from his network. I’d also like to thank the friends and colleagues that were kind enough to review earlier drafts of this paper, particularly Lieutenant General Michael Howard, Bob Gourley, and Jerry Lavely for their thoughtful feedback.
Disclaimer: Four of the companies mentioned above are portfolio companies of Squadra Ventures — Shift5, Primordial Labs, Overwatch Imaging, and Prewitt Ridge.