Combat forces’ movement has evolved throughout history—from melee to mass to the maneuver concepts that are enshrined in contemporary modern military doctrine. On the modern battlefield, however, the US military needs to prepare to fight without the assured communications it has relied on for the last eighty years. What comes next may be another fundamental change: swarming.
What is Swarming?
John Arquilla and David Ronfeldt first formalized swarm tactics in 2005, when they defined swarming as “engaging an adversary from all directions simultaneously either with fire or in force.” Sean Edwards built on their foundation, adding that swarms act in four stages: locate, converge, attack, and disperse. All three agree that swarming is an information-intensive form of maneuver.
A key advantage swarms bring to the battlefield is that they can self-organize without using electronic communications if they are enabled by stigmergy and rules-based decision making. Stigmergy—a mechanism used by ants and a range of other organisms—is indirect coordination through environmental cues, and rules-based decision making is the use of a predetermined set of responses for specific situations. Both concepts are borrowed from biology and swarm-based artificial intelligence and optimization programs.
Swarming’s Benefits: Rapid Decision Making
Swarming’s advocates usually focus on the advantages a swarm gains by attacking from multiple directions, which paralyzes an enemy’s ability to decide, respond, and concentrate efforts. This allows small units to disrupt and potentially destroy larger formations, a fact that should inform discussions about command structures and the best means of drawing out swarming’s potential advantages.
One of swarming’s less discussed—but similarly powerful—advantages is the reduction of the time required to make decisions and communicate. Every unit in a combat zone acts as an information processer, collecting information about its environment, processing it to better inform decisions to be made, and transmitting that information to subordinates and higher headquarters as necessary. Even with the most well-trained teams, each of those steps takes time. Units that can make effective decisions more quickly will gain an advantage over slower units. At the small-unit level—an infantry, platoon, for example—swarming enables this more rapid decision making.
The Army’s Infantry Platoon and Squad manual describes sixteen detailed steps that a traditionally structured platoon will take in order to execute a platoon assault. Working through those steps requires a minimum of seven and a maximum of nine decisions and entails leaders communicating with their subordinates a minimum of seven and a maximum of nine times. For the sake of this scenario, we can use a conservative assumption that each decision takes five seconds and each communication takes eight seconds. The time cost of a platoon assault’s decision making and communication is a minimum of ninety-one seconds and a maximum of 117 seconds prior to the squad assaulting the objective.
A swarming force is composed of agents whose size does not vary regardless of the size of the overall force. Each agent is given a set of rules that inform its use of stigmergy and rules-based decision making. For this example, assume agents are traditional rifle teams of one junior noncommissioned officer, a rifleman, an automatic rifleman, and a grenadier. Their rules are to move to the sound of gunfire if it is within three hundred meters, to attack when they are not decisively engaged, to retreat when decisively engaged, to maintain a distance of thirty meters from friendly forces, and to attack from a different direction than other friendly forces when possible.
Upon contact with adversary forces, teams would perform the following steps:
1. Determine if they are decisively engaged and the distance and direction of attack of adjacent friendly units.
2. Decide the assault route.
3. If not decisively engaged, attack.
4. If decisively engaged, retreat until no longer decisively engaged.
The swarming agents each have a minimum of four and a maximum of five decisions and will communicate within their teams a minimum of two and a maximum of three times. Using the same assumptions as above, the time cost of a swarming agent’s decision making and communication is a minimum of thirty-six seconds and a maximum of forty-nine seconds. In the abstract, swarming agents can cycle through all of the decision-making and communication cycles required for a platoon assault 2.5 times for every time a traditionally structured platoon performs the same action. In reality, this advantage is likely to be even greater: rules-based decision making lowers the cognitive burden of each decision, and reducing the number of times echelons need to communicate with each other dramatically reduces opportunities for friction.
A previous article included an example of two soldiers conducting a “floating box ambush”—using stigmergy and rules-based decision making against traditionally structured platoons. In training scenarios, the soldiers’ rapid decision making and initiative allowed them to quickly disrupt most platoons, and to even defeat some of them. The platoons they fought had far more firepower and forces to maneuver than the two soldiers. The traditionally structured units struggled because, by comparison, their decision making and communication took too much time. The traditionally structured platoons’ time cost and their confusion from being attacked simultaneously from multiple directions allowed the swarming agents to consistently maintain the initiative, preventing the traditionally structured platoons from bringing their considerable advantages to bear.
Swarming’s Benefits: Self-Organization
Rapid decision making on its own is not necessarily an advantage. Fast, independently made decisions can lead to disaggregated forces doing whatever they want, unable to link their efforts together to produce results. For swarming agents’ rapid decision making to have value, they also need to self-organize in a way that accomplishes their collective purpose. This allows swarms to make decisions quickly and function without electronic communications.
Self-organization has different definitions in several fields, but these definitions share a fundamental essence. In biological systems, self-organization is “a process in which pattern at the global level of a system emerges solely from numerous interactions among the lower-level components of a system. Moreover, the rules specifying interactions among the system’s components are executed using only local information, without reference to the global pattern.” Ilya Prigogine, one of the trailblazers in chaos theory, defines self-organization as a process by which order and organization spontaneously form from chaos in systems that are far from equilibrium. In his work on cellular automata, Stephen Wolfram describes self-organization as a system’s spontaneous transition from random conditions to conditions “that are not at all random.”
For self-organization to take place, systems need to have a large number of agents that interact with each other, control parameters that define the system’s patterns, the ability to shift energy, order parameters that that coordinate the system’s different components, and fluctuations. For swarming ground forces, rules about interacting with terrain, enemies, and friendly forces define system parameters and coordinate system components. The interactions between separate agents and between agents and enemies shift energy. Adversary actions and agents acting on their higher headquarters’ initial guidance create fluctuations.
Traditionally structured units, by contrast, use top-down guidance to organize. This works well under many circumstances, but has major disadvantages under others. As mentioned above, self-organizing systems can act more quickly than units that rely on top-down guidance. More importantly, traditionally structured units will struggle with changing events when they have lost the ability to communicate over the electromagnetic spectrum. Unless a force has reliably secure communications channels, self-organization’s inherent capacity to drive action without communication will offer significant advantages.
Swarming’s Benefits: Junior Leader Empowerment
The disaggregation of swarming units into team-sized agents taps into the immense potential for initiative found at the squad and team level. Unfortunately, in traditional command structures, much of that potential is lost during large operations. Often, the best way to slow down really great squad leaders is to put them in a company operation. Sometimes that is the result of leaders deciding not to implement a mission-command philosophy, but often it is the natural product of larger operations’ necessary constraints, like fire-control measures. Swarming capitalizes on the American military’s strong noncommissioned officers, and the advantages created by junior NCO education, empowerment, and tactical initiative.
Swarming also reduces the number of training tasks units need to accomplish. In theory, units need to certify each echelon before training the next higher echelon. A platoon that wants to certify its ability to conduct a platoon assault needs to conduct team and squad training exercises before platoon exercises. Units are sometimes challenged to perform team, squad, platoon, company, and battalion maneuver exercises before training at the brigade level during Combat Training Center rotations. Swarming units can focus on team-sized swarming agents and developing the junior NCOs that lead them.
Despite all of swarming’s advantages, traditional tactics have an advantage over swarming units in many environments. Traditionally structured units with fully functioning communications are able to quickly shift concepts of the operation and sometimes even missions. They can bring artillery and other potentially decisive firepower to bear quickly in a way that is not feasible for swarming, disaggregated units that are not actively communicating. Traditionally structured units can also observe, communicate, and exploit opportunities and commit reserves, allowing them to capitalize on local successes better than swarming units. On top of these, traditionally structured units implement fire-control measures that help reduce fratricide in a way that would be very difficult for swarms to implement.
All of these advantages, however, assume that traditionally structured units will maintain their ability to communicate. Without reliable communications, traditionally structured units will not be able to share new concepts of the operation or missions, coordinate fires during an engagement, share opportunities with higher headquarters, or actively update fire-control measures.
Swarming’s advantage lies in its ability, when rules and cues are properly selected, to allow agents to self-organize into more complex, collective actions without directly communicating. While this means accepting the loss of some of a traditionally structured unit’s advantages, those will be already lost if adversaries can disrupt American communications.
The Threat Environment
It is likely that some American units will operate in an environment with limited or nonexistent electromagnetic communications. The 2018 National Defense Strategy names China and Russia as two of the United States main threats. Both countries have developed or are developing the ability to threaten their adversaries’ use of the electromagnetic spectrum.
Russia’s military uses the electromagnetic spectrum to direct artillery fire and to disrupt its adversary’s ability to communicate. It uses electronic warfare assets, such as the RB-301B Borisoglebsk-2, to locate, intercept, and disrupt radio and cellular signals. Famously, Russian artillery fire destroyed two Ukrainian battalions at the Battle of Zelenopillya after locating their position via electromagnetic emissions. Unfortunately, the electronic-warfare technology the Russians used seems to be more effective than comparable American capabilities. The US State Department attempted to mitigate Russia’s electronic-warfare capabilities by providing the Ukrainian military Harris radios with mixed results.
China is developing a similar ability to counter American advantages with its Assassin’s Mace programs. The People’s Liberation Army and People’s Liberation Army Navy are investing in capabilities to disrupt American military strengths. Their programs include electronic-warfare attacks against communications systems, cyber attacks against digital systems, and smart weapons designed to destroy radio stations. Michael Pillsbury, author of The Hundred-Year Marathon, claims that war games indicate Assassin’s Mace technology will give China a distinct advantage if the United States does not prepare to counter it.
One of the reasons the Russian and Chinese governments created these capabilities is they believe they will allow them to prevent the United States from employing its conventional strengths. The Asymmetric Warfare Group notes:
The key cornerstone of U.S. and NATO methodology is maneuver warfare. Maneuver warfare depends on communication and synchronization of assets. The U.S. has communication infrastructure down to the four man Infantry Fire Team level and the ability to battle track those formations with almost real-time speed. When everything functions as designed, these smaller formations can achieve much larger effects than their Russian equivalent.
But when everything does not function as designed—when US forces lose the ability to communicate over the electromagnetic spectrum—the American military’s communication infrastructure will be severely degraded, and with it the ability to conduct maneuver warfare.
The United States needs to prepare to fight adversaries that will attempt to disrupt its communications along the electromagnetic spectrum. Unless it can rely on technological or behavioral solutions that will allow it to continue to consistently communicate, those preparations need to include new tactical concepts that do not rely on regular communications in some operations. Swarming may be that concept. Swarming, as shown above, has significant disadvantages, but it prepares American forces to fight without top-down direction. Just as importantly, it offers the ability for self-organizing groups of soldiers capable of rapid decision making to overwhelm their adversaries.
Justin Lynch is an Army infantryman and a member of the Defense Entrepreneurs Forum. The views expressed are those of the authors and do not reflect the official position of the United States Military Academy, Department of the Army, or Department of Defense, or any group with which the author is associated.
Image credit: Maj. Carson Petry, US Army
This article tells me swarming works great unless the radios and internet are jammed, which we sort of knew with or without the novelty of swarming.
Insect stigmergy is based on chemical cues that require an environmental trigger, such as the destruction of individual animals I mean, one way to tell you are in a mine field is when people start getting blown up, which doesn't require radio or internet to discern locally, but which does require a means for disseminating that data that overcomes any scale of a battle in excess of the 7 Years War, which is not presented here.
All things considered, "When everything functions as designed,…" remains the defining constraint. For example, it was mentioned in a different MWI article that America has lost certain operational arts, such as camouflage, a casualty of 73 years of air supremacy. The Soviet Red Army was the master of camouflage during The Great Patriotic War, but let the capacity atrophy in Afghanistan for the same reason. I have no direct knowledge of the state of play in these areas. but the alarm which these two articles reflect regarding Russian and Chinese capabilities suggests that asymetric warfare is already leveraging our material advantages against us and swarming is a desperate application of Sitting Bull's response to Custer when nothing functions as designed.
If you start over defining "stigmergy" in terms of horizons instead of boundaries, a practical organic figure might arise. In this regards, it might serve your purposes to review "Starship Troopers" regarding the problem of suppressing "stigmergic swarming" as an tactical phenomena. Heinlein drew upon Korean war after-action reports of the communist massed wave assault for his portrait of that sort of combat. As a Naval officer, he has no feel for terrain, but it's something to consider for your subject.
One would hope that Justin owns and has read ATTACKS (aka Infanteria Grieft An) published in 1937 by a rising young German officer named Rommel. It is also the tactics he used with tanks and that most heavy cavalry units could also execute until we disbanded them.
Alas and ALLONS
Swarming might work for platoons or even companies, but we will need to fight at echelons much higher than that. We need a solution that will work as well for Corps as for platoons.
I share the author's worries about our dependence on unhampered use of the EM spectrum (and I would add, the airspace). But it seems to me that for all the reasons he points out, rather than simply give up on the advantages of traditional C2 structures, we should simply plan to communicate in a contested or even denied EM environment. That might mean going back to the old ways like running line or physical dispatches, or it might mean harnessing technology by, for example, using highly directional burst transmissions. Maybe we can deliver orders by drone? I don't know the answer but I do know that if enough smart folks work the problem some answers will turn up.
This reminds me of MOUT and 1993 “Blackhawk Down” in Somalia. In 1993, the U.S. solider lacked combat experience so his performance may have been a factor compared to the GWOT. Fast-forward 26 years and is the concept of “Swarming” the same? The Somalis “Swarmed” GIs in 1993. The U.S. GI body armor, gear, and radios may be different, but essentially, the combat arms and US Army 2019 soldier remains basically the same since 1993. There IS NO “2020 U.S. Iron Man Suit.” Would “U.S. Army soldier swarming” work against a determined foe armed with AK-47s, RPDs, RPKs, or AK-74s?
The issue here is that miniature drones and UAVs have made the U.S. soldier more cautious. Chances are that GIs will deploy drones from behind cover instead of bum-rushing targets blind and attacking via “Swarm.” The “Run-and-gun” approach as evident in MOUT and “Blackhawk Down” proved that causalities occur with enough enemy fire being concentrated on the GIs. Body armor coverage protection remains about the same in 2019, 2004, and 1993.
That is why I believe the solution is armored cars, APCs, IFVs, and Secret Service tactics. You either fight from under truck, APC, or IFV armor, or you escape (via vehicles) like the Secret Service (car column). 1993 Somalia proved that unarmored HMMWVs had a hard time dodging and fleeing fire from RPGs and AKs. GIs got out and tried to attack but were pinned down.
“Swarming” by foot might work in certain situations, but 2020s will see more infantry riding (armored) wheels than on foot. The days of the U.S. Marine Corp mass-infantry tactics may be over against a determined armored foe due to the lack of organic heavy firepower carried by the boot Marines.
This article seriously oversells the concept of swarming, and particularly in a communications-poor environment. Rule-based action, and even (to a large extent) stigmergy are adequate for simple stimulus and response — think of a series of "if-then" statements. Complex "if-then" routines rapidly become either rigid or cumbersome, though, and even the simple-looking ruleset in the articles example breaks down pretty fast:
– Move to the sound of gunfire if it is within three hundred meters
– Attack when not decisively engaged
– Retreat when decisively engaged
– Maintain a distance of thirty meters from friendly forces
– Attack from a different direction than other friendly forces when possible
What happens if 1/3 of the enemy retreat under pressure, but the other 2/3 chose to engage friendly forces? Or the opposite?
What happens if two squads end up across from each other, with the enemy in the middle?
What defines the end of the action? Does a single surviving enemy element justify continuing the action with all forces?
Left as is, the action would devolve to chaos, heavy with fratricide, until one side was gone. That looks good in simulations, but it's rarely the desire result in real life.
Freedom of action aligned with the Commander's Intent (or Platoon leaders intent) has always been a hallmark of US small unit operations. The ability of our Junior leaders and soldiers to take the initiative and act to accomplish the mission isn't too new a concept. Good units build that culture. I'm not sure how different "swarming" is –when in contact, battle drills, training, dress rehearsals, mission, etc. frame the response to ensure unity of effort, purpose, and mission success. Technologically enabled weapons system utilization via swarming (drones) has potential in creating defensive challenges. That's a reality we'll have to somehow address. However, in a degraded communications environment, the impact on our operations is something that will have to be considered in training along with risk reduction measures (tank signal flags, whistles, smoke, star clusters, etc). Never underestimate the potential of small area effects. Decades ago, we used the Hand-Emplaced Expendable Jammer (HEXJAM) in shaping the local electromagnetic battlespace one time–it's effects (several of these use at just the right time and place) were too disruptive to use again during training. A lesson worth remembering when trying to coordinate via radio any maneuver, medevac, fires, etc. Brigade Combat Team Commander, Task Force Commander, Company Team Commander, Platoon Leader–What do you and your forces do when radios don't communicate? Have you practiced in that environment?