Operational Art for the Replicator Initiative: Confessions of a Swarming Addict
My name is Ben Jensen and I am a shameless swarming enthusiast. I have been addicted to the idea for most of my professional life. Over the last 10 years I have done everything from model old tactics in new wargames and simulations based on horse archers and Mongolian steppe riders to replicating creative ways to bend cost curves and get adversaries to expend million-dollar missiles on thousand-dollar loitering munitions and decoys. This journey led me to work with the Defense Advanced Research Projects Agency on the mosaic warfare concept. Countless classes of marines have suffered as a result of my swarming addiction and desire to help them see war as a complex system made up of competing battle networks, human will, and operational art. And it now leaves me hopeful for the new Replicator initiative.
When I see calls for attritable all-domain autonomy, with thousands of systems ready in 18–24 months and the implied ability to regenerate salvos more rapidly than competitors can, I get really excited. At the same time, decades of studying military innovation leaves me conscious of the mix of bureaucratic challenges and antecedent conditions that need to be in place to change the American way of war.
Since all military innovation starts with ideas, it is important to begin a parallel dialogue to the Replicator initiative about operational art. This art is based on concepts for identifying and exploiting asymmetries and operational reach. These concepts will require creating an interoperable network with partners and allies and ensuring flexible, scalable payload options for the low-cost, attritable swarms. Most of all, operational art for swarming will require empowering military professionals at each echelon to experiment and develop supporting tactical concepts of employment and training iterations.
All war is asymmetric if done right. To be asymmetric means creating conditions where the adversary’s strengths become weaknesses, allowing friendly forces to adopt an economy of force posture — saving resources and energy to seize unpredictable windows of opportunity that arise when armies collide. In practical terms, most feints work along these lines. The enemy sees a small force, thinks it is larger, and takes the bait. The result: The enemy is out of position, off balance, and vulnerable to an attack in a different location.
The concept has a long tradition, and directly relates to seeing war as a complex system, a notion best captured in the U.S. Marine Corps’ rewriting of maneuver warfare in the 1990s. It finds affinities in enduring tactics like reconnaissance pull, spoiling attacks, turning movements, and engagement area development. In terms of principles of war, the logic is less about offense and objective and more about maneuver, economy of force, and surprise. The idea is that a Replicator swarm gains and exploits a position of advantage if it pulls the enemy off balance and retains the ability to adapt to changing circumstances.
Consider the current conflict in Ukraine. In a narrow sense, the Russian use of a mix of cruise missiles, electronic attack, and Shahed loitering munitions constitutes a swarming attack targeting critical infrastructure. The attacks only work when Russia can overwhelm air defenses.
Yet, this is a suboptimal form of swarming. First, the Russian military struggles with operational intelligence and fires integration, forcing it to target fixed structures (i.e., the energy grid or ports), so Ukraine can anticipate and adapt to these attacks. There is no surprise. Ukraine is not out of position tactically.
Second, swarms work best when they take advantage of tendency and potential in the system — that is, counterintuitively, when they are a second mover. The best attacks are those that first force the enemy to reveal the system and then retain enough slack to adjust to the revealed surfaces and gaps. Unlike traditional portraits of decisive points, this implies an injection of energy into the system to reveal it instead of attacking predetermined decisive points (i.e., principles of mass and objective). This is heresy to a profession that loves to tell itself stories about Schwerpunkte, decisive battles, and shock. It also runs counter to the entire concept of “fire effectively first” as a central idea of modern fleet engagements. The key pulse battles in a swarming attack are those that pull the enemy off balance and out of position through decoys, low-cost attritable munitions, and information effects. Thankfully, the Russian ability to take these actions is either contrary to their vision of joint fires or been denied by clever adversaries through a mix of deception and cyber operations.
The generation and manipulation of asymmetries is on display in the Black Sea. Ukraine uses a mix of special operations forces, decoys, intelligence operations, low-observable unmanned surface combatants, and missile strikes to hold the Russian navy’s Black Sea fleet at risk. There has been no decisive battle. Rather, there has been a series of pulse strikes combining multiple domains and attack trajectories that has forced Moscow to consider pulling its combatants away from Crimea. The small, many, and ambiguous forced the giant to pull back because — back to systems thinking — the whole is greater than the sum of its parts. When swarming generates and manipulates asymmetries it produces emergent effects and system change.
The center of gravity for swarming attacks like these is the ability to sense, make sense, and dynamically (re)direct salvos. Because war is a non-linear system it is prone to sudden and dramatic changes (i.e., regression to the tail). A military can only take advantage of these changes if it can out-cycle the adversary and adjust the swarm to the moment, similar to a flock of birds turning suddenly in unison.
What does this mean for Replicator? The Department of Defense’s communication infrastructure, dubbed Combined Joint All-Domain Command and Control (i.e., battle networks), has to work and form a machine-speed — but human judgment-informed — mesh network that allows swarms to adjust to emergent opportunities. If the department cannot connect its networks, link in partners, and run basic artificial intelligence and machine learning scripts to analyze patterns and synthesize large volumes of information, its Replicator swarms will just be overprized creeping barrages. If the targeting cycle cannot adjust as the system changes, it will not be able to support the operational concepts that the Department of Defense is now touting as critical for future combat operations.
Getting this interoperability right will require a new wave of experimentation along the lines of that already underway through the Global Information Dominance Exercises. It also means new approaches to training officers to work with basic data science and models as a complement to naturalistic decision-making. Absent a massive reform to professional military education, the service will be left with old men on horseback distrustful of technology and overconfident in their intuition (i.e., regression to bias). The military professional directing future swarms will need to balance different forms of reasoning and decision-making and not, like past generations, assume every battle is just a variant of the 2nd century B.C. Battle of Cannae.
No war is decided by a single battle. The shadow of the future hangs over every fire fight as commanders gauge how local events could produce larger risks and opportunities over space and time. Violence exists in the past, present, and future simultaneously. Soldiers remember training and past fights while taking immediate action drills in the present and trying to anticipate what happens next.
As a result, professionals think about operational reach and how to project and regenerate combat power. It is not just the current battle they focus on, but their ability to turn a series of tactical engagements into momentum in time and space. That requires a careful eye on logistics, readiness, morale, and how the adversary system is changing. Applied to swarming, it means the Replicator initiative needs the ability to regenerate swarms and project them to take advantage of changing circumstances. This logic reinforces striking second and “bait and ambush” approaches that stimulate the enemy system and take advantage of the new information. The swarm only works if it can exploit a gap and exploitation requires the ability to project effects over time and space.
The Ukrainian armed forces get high marks for creativity and generating combat power. They have unleashed a market revolution, bringing in the private sector into the mix far beyond any Department of Defense effort, through ideas like the Army of Drones Initiative. The Ukrainians have also demonstrated how to build a joint command and control system from the bottom up through initiatives like the Delta situational awareness system, and how to maintain a digital infrastructure for passing data that overlaps easy-to-use national systems like Diia. It is important to emphasize that the Ukrainian military have a more robust set of digital government services than the country that invented the internet. Ukraine has developed a way to generate low-cost combat power against all odds.
Ukraine is still in the process of experimenting with how to integrate these capabilities into ground maneuver beyond immediate tactical opportunities. They haven’t cracked the code on operational reach in terms of exploiting in depth through a swarming attack. Yes, the Ukrainian military has come up with ingenious ways to strike targets deep in Russia. Long-range kamikaze drones have political value and some operational benefit in terms of forcing Russia to keep air defense back, but they could be put to better use closer to the front to help create a breakthrough.
The same logic applies to the high volume of drones Ukraine uses for reconnaissance and striking targets of opportunity. The Ukrainian military is using their drones like skirmishers when they could launch swarm pulses coordinated with probing attacks along the front. If every small team of sappers and infantrymen trying to advance in Zaporizhzhia had 30 R18 octocopters (i.e., $3 million-worth, the same cost as one Bradley infantry fighting vehicle), the Ukrainian military could create more depth despite the obstacles, mines, and trenches. It is not hard to imagine entirely new forms of trench-clearing done by an autonomous horde of thousand-dollar DJI quadcopters using image recognition.
The challenge is that it is still costly to increase the maximum effective range of munitions. While the cost of computing power has declined, creating a revolution in precision, the cost of energy and projecting small munitions over distances hasn’t followed suit. To achieve operational reach in terms of distance, this implies a need to maneuver and establish intermediate staging areas for launching and recovering swarms into position and advances these positions as the front changes. The same effect can also be achieved by using motherships to get close, launch, and exfiltrate. These staging bases are a perfect concept to explore for expeditionary advanced bases manned by a mix of U.S. marines and special forces. Programs that convert current inventory aircraft and platforms into motherships are ideal. For example, efforts like Rapid Dragon illustrate how to use cargo aircraft to launch palletized effects ranging from cruise missiles to swarms of drones. Imagine a massive Liberty Lifter launching a mix of swarming unmanned surface vessels and aerial drones from over the horizon before flying back to pick up another pallet. There is also the ability to adapt the large U.S. drone inventory to launch air-launched effects — mixed swarm payloads combining lethal effects with electronic attack and intelligence. For example, the new Eaglet can launch from the MQ-1C and MQ-9 and be recovered for future missions. Operational art in the future will become the ability to combine these effects in space and time to generate operational reach.
What does this mean for Replicator? The United States should invest in the ability to generate operational reach with swarms in a cost-efficient way. If the Replicator initiative isn’t careful, industry will come back with exquisite drones the U.S. budget cannot afford. Each one will be tailored to the perfect mission rather than being capable of conducting multiple missions based on open software and swappable payloads (i.e., high explosive, electronic warfare, or surveillance and reconnaissance). Worst, they won’t be interoperable or capable of being fired from different air, sea, ground, and subsurface platforms. Open-architecture, flexible, cheap swarms are the name of the game.
Second, the Replicator initiative will need to start thinking about operational reach in terms of intermediate staging areas (i.e., bases of operations) and mothership constructs. These again should be flexible, take advantage of existing platforms and prototypes, and focus on cost. A sky filled with cheap delivery platforms is better than one exquisite aircraft when it comes to swarming.
Swarming is neither new nor novel, but it is cool — and could be leveraged to achieve key objectives in future wars. The replicator initiative has huge potential but only if communities like War on the Rocks and the silent military professionals tirelessly working across the services seize the moment to apply operational art and develop supporting concepts based on thinking about swarms as complex systems. Unlike some accounts, this is not a “high-risk, high-reward” gamble. It is a proven form of warfare with deep historical roots and an underlying economic logic. Outside of the bureaucratic challenges that accompany any effort to change the Department of Defense, adopting swarming to support warfighting and integrated deterrence requires a renaissance in the study of war and a commitment to experimentation at echelon.
As illustrated above, these supporting concepts and experiments need to focus on battle networks and operational reach. Generating and manipulating asymmetries — how a swarm produces cascading effects — requires the ability to pass data at machine speed to see opportunities in the making. This means the Department of Defense is going to have to get algorithmic warfare right and find ways of balancing human judgment with statistical pattern recognition. The efforts will also need to think about operational reach and how best to generate effects in time and space. That means keeping an eye on cost and thinking about how best to deploy swarms and regenerate pulse strikes as a new form of operational art. This reasoning will challenge deep ideas in the military profession. There has been a tendency to look down on attrition despite the fact that most wars find themselves in costly stalemates where cost curves and the ability to regenerate combat power are what produces operational advantage.
The Replicator initiative needs to live up to the promise. The United States can no longer afford to outspend its adversaries. The nation needs to find creative ways of linking systems together and generating swarms that bend the cost curve and take advantage of legacy equipment. These swarms should be interoperable with coalition partners, creating an economy of scale that the network of authoritarians challenging U.S. interests cannot match. America started both modern manufacturing and the information age. It is long past due for the United States to combine them to unleash swarms as a new form of operational art.
Benjamin Jensen, Ph.D., is a professor of strategic studies at the School of Advanced Warfighting in the Marine Corps University and a senior fellow for future war, gaming, and strategy at the Center for Strategic and International Studies where he runs the new Futures Lab. He is also an officer in the U.S. Army Reserve. The views expressed are those of a swarming addict and strictly his own and not indicative of any official government or industry position.