Moving Toward Defense as a Service

Defense Innovation Unit leads the Blue UAS Refresh at The Combat Center

History and pre-history are littered with the corpses of competitors who failed to adapt as quickly as their adversaries. From animals to nation-states, victory rarely depends on strength alone but comes down to survival and adaptation. This lesson is being driven home on the battlefields of Ukraine. The United States must take heed, or it risks being rendered irrelevant on the world stage. U.S.-provided Excalibur precision artillery shells were devastatingly effective when they first entered service in Ukraine, but within a couple of months, Russian adaptation had rendered those precision munitions dumb. In war, concepts of operations change hourly, software is revised daily, and hardware is iterated on weekly — this is ever truer as technology cycles accelerate.

The traditional defense acquisition model, often slow, inflexible, and designed to be largely top-down, will struggle to keep up with these shifts. What this means for the United States in practice is that in a major shooting war against a modern adversary, we may very quickly find that adversarial adaptation renders U.S. technology obsolete as soon as the dust settles on the opening weeks of a conflict. If the United States hopes to keep pace with the capabilities of its adversaries, particularly in its adversaries’ backyards, it should consider a trio of policies including moving from product-based contracts to outcome-oriented service contracts, embedding engineers at the frontline, and developing expeditionary manufacturing capacity.

As a venture investor operating in the national security space, I know that some may be tempted to disregard my suggestions as self-interested. However, as the suggestions that follow would all increase the operational complexity of building startups in defense, I am not asking for an easy button. For example, distributing manufacturing all over the globe is harder to finance and almost certainly less efficient than a giant centrally located factory, and forward-deploying engineers almost certainly increases employee overhead. The reason I make these suggestions isn’t to change the Department of Defense to suit the demands of industry — it’s about changing industry to suit the demands of modern war and offering the Department of Defense an opportunity to facilitate that much-needed transition. We’ve looked at thousands of companies over the last few years and no matter how cool their technology or great their teams, vanishingly few are capable today of adapting at the pace of the modern battlefield. We must change that.

 

 

The Changing Pace of Adaptation

Advances in commercially available technologies — such as drones, 3-D printing, and space imaging —enable adversaries to quickly adapt their strategies and capabilities using commercial off-the-shelf technologies. The air-to-air drone war in Ukraine is illustrative. Initially, Ukrainian forces used first-person-view drones to ram Russian drones. In response, Russian drone operators adjusted their tactics, forcing Ukrainians to equip drones with munitions and begin work on AI targeting systems. The Russians, in turn, experimented with dazzle paint schemes to evade visual or AI detection. This weekly cycle of innovation underscores the rapid pace of modern battlefield adaptation. There are similar cat and mouse games taking place across the entire spectrum of capabilities.

Consider this scenario playing out between the United States and one of its primary adversaries. They adopt a new technology and begin seeing major battlefield success. As American casualties mount, how long would it take the United States to adapt? An instructive historical example is the mine-resistant ambush-protected vehicle program, which was launched during the deadliest phases of the Iraq and Afghanistan conflict. The requirements process for this program took 20 months, at the conclusion of which the program needed to select winners, scale manufacturing, and deliver units. This program was driven by a life-and-death emergency and was a priority of the secretary of defense, yet the United States still took over three years to field a technology adapted to adversary action. While the mine-resistant ambush-protected vehicle program was largely considered a success, in a future conflict the United States will be in trouble if its adaptation cycle is measured in years but that of its adversary can be measured in weeks or months. Critics will argue that the United States does a good job of adapting in the field. I agree that U.S. forces are great at one-off field adaptation, but the challenge is that building a great custom solution in a barn is not something the United States can rely on in a conflict against a major adversary where adaptive solutions by the hundreds of thousands are required. As of October, Ukraine has the capacity to produce up to four million drones annually. These are being assembled by over 200 different companies and untold numbers of private citizens. This scale and diversity allow for incredible adaptive capacity. Traditional defense acquisition models, characterized by lengthy development cycles and bureaucratic processes, fail to address this reality and by the time evolved technology is deployed, it may be obsolete, resulting in missed opportunities to outmaneuver adversaries.

Acquiring Services over Products and Outcomes over Solutions

One of the challenges highlighted above is that the way the U.S. military acquires new solutions runs through a rigid requirements process. The process begins when someone recognizes that there is a problem in the world that needs to be solved either through a new operation or through the acquisition of a new widget. The Joint Capabilities Integration and Development System process, for example, takes, on average, 2 years. If it is determined that a new widget is required, a process is run to determine the exact specifications for the solution. The U.S. Army (in)famously spent over 9 years and wrote over 350 pages of specifications for a new sidearm rather than just purchasing a workable weapon off the shelf. While a rigorous requirements process is intended to ensure troops receive something they actually need, the rigidity of the resulting product (the solution specification) becomes problematic in a world where the solution specified is likely to be out of date before the ink has time to dry on the requirement.

The Joint Capabilities Integration and Development System isn’t the only way requirements can be generated. There are more agile processes such as working through Urgent Operational Needs, Section 804 Rapid Acquisition Authority, Middle Tier Acquisition Authorities, or the various Rapid Capabilities Offices, all of which can accelerate the fielding of new or adapted technology. However, many of these processes remain too slow for current technology cycles — Middle Tier Acquisition Authority, for example, takes two to five years to field a solution. That said, an Urgent Operational Need can lead to a solution in weeks to months, which begins to align with the modern battlefield reality. The problem is that keeping up with today’s adversaries will require almost exclusive usage of these exceptions. Once exceptions subsume the rule, we should ask ourselves if the rule still makes sense, or if we should reimagine it from the ground up.

The way forward must be to adapt the requirements process itself by changing the product it generates into something that is responsive to changing conditions. Rather than describing solutions to problems, the requirements process should describe desired end states. Using the air-to-air drone example above, rather than specifying an aerial drone type capable of targeting another aerial drone type over a specific distance carrying a specific payload, and utilizing specific sensors, the requirement should be for an outcome. In this case, the outcome could be the ability to clear friendly airspace of adversarial drones. These more general requirements can then be awarded to prime contractors who can both serve as direct performers or subcontract mixing and matching thousands of possible solutions. Essentially, this moves from a product-focused model toward a more service-based model, which allows for capabilities that can be adjusted and updated as needed — much like software updates — rather than fixed products that take years to develop. A service-based approach allows for the continuous improvement of capabilities, enabling the military to respond to changing conditions quickly.

Product capabilities delivered as a service may seem like a new idea to some, but the model is one that goes in and out of favor in government contracting. For example, during the Global War on Terror, Predator drones were operated by General Atomics contractors. More recently, the U.S. Coast Guard awarded Shield AI almost $200 million on an indefinite delivery, indefinite quantity contract to provide surveillance and reconnaissance as a service, for which the company will use its V-BAT platform. And Metrea, which describes itself as providing “effects as a service,” has “as a service” contract with U.S. Naval Air Systems Command that has it conducting aerial refueling operations for aircraft from multiple U.S. services as well as allies.

New processes like Replicator or programs like TITAN are great steps in the right direction. The genius of TITAN is in fielding hardware that is inherently software-defined, and therefore primed for rapid battlefield adaptation. To the extent feasible, more programs should look like TITAN. However, even there, the Department of Defense is still a little too solution-oriented for the future fight. The requirements and contract structure for TITAN should have laid the groundwork for replacing it with a battlefield-adapted system for a date, hopefully far in the future, when TITAN itself has outlived its usefulness. Only an outcomes orientation lays the foundation for infinite adaptation.

Engineers on the Battlefield: Real-Time Adaptation

While adaptive requirements and service-oriented contracts can be part of the solution, rapid adaptation also requires a real-time understanding for how an adversary has adapted their technology. The United States must embed engineers on the frontlines and not just combat engineers. This isn’t about bridges and minefields but is about placing the architects of adaptation in the thick of it so that they can help modify equipment in real time as battlefield conditions adapt. Contractors placing their engineers forward isn’t a completely novel concept. Palantir famously relies on forward-deployed engineers to help end users get the most out of their intelligence platform. Similarly, companies tasked with keeping American forward operating bases safe from adversarial unmanned systems should be on hand to witness how their systems perform, how end users interact with their systems, and how adversaries are adapting in real-time. As any product manager can tell you, end users — in this case, warfighters — rarely use a product in the way originally envisioned. The only way to really get this sort of feedback is for engineers to work plate carrier to plate carrier with the warfighter. Traditionally, engineers and developers work from secure facilities far removed from the frontlines, relying on feedback loops that can be slow and filtered through multiple layers of reporting. This model results in delays and a disconnect between those designing solutions and those using them in combat. As an example, while many of our founders are actively tracking developments out of Ukraine, when they return from trips to the frontline, they have a completely different understanding of the situation. Founders developing new battlefield technology view Ukraine as an irreplicable lab for learning.

Detractors of this idea may object to the risk this would impose on private engineers. While I don’t want to completely discount this risk, Palantir’s engineers have bravely shouldered it for decades, as have war correspondents. A future conflict with a peer adversary will require an all-of-nation effort, and that means the time when only 1 percent of Americans fought America’s battles must come to an end. This approach transforms engineers from distant, disconnected developers into integral parts of the tactical team, enabling a level of responsiveness that aligns with the unpredictable nature of modern warfare. It ensures that the technology evolves alongside the conditions of the battlefield, offering a significant strategic advantage.

Proximal Manufacturing: Closing the Gap Between Innovation and Implementation

To further enhance responsiveness, the manufacturing process must be moved closer to the frontlines. Traditional manufacturing, which often takes place far from combat zones, is slow and inflexible, lacking necessary feedback loops. Moving manufacturing capabilities closer to conflict areas allows for rapid prototyping, testing, and production of adapted technologies and shortens supply lines. This shift can dramatically reduce the time needed to get new solutions into the hands of the warfighter.

Forward-deployed manufacturing hubs can leverage advanced techniques like 3-D printing and modular production to manufacture customized solutions on demand. For example, if a new type of drone or sensor is needed to address a specific threat, a nearby manufacturing facility could assemble the necessary components and deliver them to the front within days instead of months. This proximity allows for quicker iterations, testing, and refinement, ensuring that adaptations can be implemented swiftly and effectively. The United States might consider outfitting something like decommissioned amphibious assault ships or similar vessels as dedicated expeditionary manufacturing facilities and logistics hubs, as the U.S. Navy is already testing 3-D manufacturing on ships. Similarly, the U.S. Air Force could test mini-manufacturing hubs on C5s. The specific platform could be flexible and will probably be theater-dependent, but the point is that moving these distributed hubs toward the point of greatest need could dramatically shrink the feedback loop from battlefield observation to adapted fielded equipment.

A related example is Firestorm (I am an investor in Firestorm and have a financial interest in the company’s success). Firestorm makes modular small unmanned aerial systems capable of being adapted on the fly like a Lego set for different mission parameters. What sets them apart, however, is that their systems are produced out of factories that are embedded in shipping containers, which can be prepositioned anywhere a need is anticipated. This allows for hardware and software modifications to be designed from anywhere in the world and pushed to an expeditionary manufacturing facility at the point of need. The future U.S. military must have similar adaptive capabilities across almost all platforms and mission sets because the force that wins is going to be the one that is most adaptable.

Conclusion

The rapid pace of change on the modern battlefield demands a shift in how military capabilities are developed, deployed, and adapted. Moving from a focus on acquiring static products to acquiring desired end states, embedding engineers on the battlefield for real-time feedback, and positioning manufacturing capabilities closer to the frontlines can all significantly enhance the military’s ability to respond to evolving threats. This approach allows for quicker adjustments, more relevant solutions, and tactical advantage over adversaries who are slower to adapt. The future force can’t be a static hammer we take out of the toolbox whenever we need to pound a nail — it must become a mutagenic blob, a Swiss army knife with adaptive tools that looks different each time an adversary sees it.

 

 

Jake Chapman has been in and around the venture industry for almost 20 years as a lawyer, three-time founder, and venture capitalist, including as managing director of the Army Venture Capital Corporation. Jake is an adjunct at the Rand Corporation, where he works on commercialization and national security. He is the managing director of Marque Ventures, a venture capital firm focused on national security technology. He writes all too frequently on Twitter as @vc and somewhat less frequently on LinkedIn.

Image: Lance Corporal Richard Perez Garcia via DVIDS.