Swarming and Network Centric Warfare - Modern Snake Oil?

[TangoTwoBravo]

I've been chewing on this for a while. I see references made to "swarming" from time to time, and I have even been advised to try "swarming" with my troops. I also read references to network-centric warfare. I did some digging on swarming a while back to find whatever theoretical basis I could for the concept, since most of what I have met thus far has been hand-waving. I then came across a reading for a course that I am taking that suggested removing the term "command and control" and replacing it with Agility, Focus and Convergence. The reading mentioned how applicable it would be swarming. This got my interest so I went digging again. Perhaps I am being a reactionary military conservative, so I'm throwing this out there to see if I'm alone or perhaps missing something.

One paper on swarming that I've met a few times is Sean Edwards' Rand piece Swarming on the Battlefield published in 2000. It is easy to find on the web, and it is indeed a good read. He defines swarm tactics as "a scheme of manoeuvre that involves the convergent attack of five or more semi-autonomous our autonomous units on a targeted force in some particular place."  This sounds good, but the author does not allow a conventional attack with a frontal fixing and a flanking strike element to qualify as swarming. The key point seems to be the semi-autonomous or autonomous nature of the units involved. The author also dismisses "Blitzkrieg" tactics since the Germans massed their forces at key points.

The graphic example given in the article show a target unit wandering along as it is attacked by five other units from all sides. What I find interesting is that if I as a tank squadron commander array my tank troops in such as way to strike an advancing force from three sides I do not count as swarming. Leaving that aside, I propose a thought wargame of two equal forces. The first will employ traditional tactics and C2, while the other will "swarm" with a flat command and control structure. I'll give each one the same equipment and assume that the officers are skilled in their profession and that the troops are well trained and motivated. For laughs we'll pit a LAV battlegroup with three companies, a tank squadron, a recce squadron, an anti-armour platoon and a mortar platoon against a force with the same equipment but split into nine equal groups without any company commanders. The swarming team has Battle View version 1000 that allows them to look at their little screen and see everybody. The other team has maps, radios and grease pencils. They are advancing against each other to seize some key point (call it an airfield).

Team Conventional will most likely advance towards the objective with his recce squadron leading and covering his flanks. He'll have an advance guard of some kind (maybe a combat team) with the other companies trailing. Team Swarm will unleash his nine platoon groups towards the airfield.

At some point the lead recce troop from Team Conventional will meet one or more of the platoons from Team Swarm. The CO will make a plan of some sort, perhaps hitting one of the groups with his mortars and forming a screen with his TOW platoon while he moves around. The swarm platoons will now all see the conventional force through their Battle View, and will need to come up with a plan. Who is in charge? Do they all make their own plan or does their CO issue orders? If so, he needs to give orders to nine elements. This raises a span of control issue for swarms. The network-centric folks don't seem to like the word control, but there it is. If he doesn't give orders who decides on the lucky platoon or two that tries to fix the conventional enemy. Who re-allcoates resources such as tanks and AT that are going unused while other areas are overmatched? Who decides on who will get to occupy the best fire positions? Who decides on CSS priorities for all this?

Lets say that the other swarm platoons figure out a way to try and encirle the conventional force. As they move, however, the conventional force's recce elements on the flanks will detect this. The conventional CO can then choose to block with one company and the AT platoon while the rest of his force gangs up on the most isolated swarm platoon (attack one of the "wings"). The "pain train" of a tank heavy combat team then rolls up the swarm platoons one at a time. Two tanks, some LAVs and a TOW in each swarm platoon are no match for fourteen tanks on the warpath backed up by a LAV company. The bonus is that the other swarm platoon commanders get to watch their compatriots get destroyed in real time. The swarm CO might decide to try employ some leadership to stiffen his troop's resolve, but he can't be everywhere and he has eschewed Command and Control in his training for Agility, Focus and Convergence.

I still believe that mass/concentration of force will win. Dispersion can certainly work, but if the enemy has ten and you have one at a certain place then you are going to lose. My problem with swarming is that it pre-supposes a stupid enemy. It would be nice if the enemy would sit in a leaguer and allow himself to be picked apart by a swarm of our forces moving around him. I just don't think that he would oblige us. On the matter of network-centric warfare, even if they resolve the technical challenges this is a human endevour. People need motivating in order to kill and be killed. This comes from leadership, but command is a big part as well.

The authors of the paper I am reading for a course state that command and control is outdated, and that it should be replaced. They state that it is unsuited for coalition warfare and the realities of the information age. My first question is what was WW2 and WW1 if not coalition warfare? On the issue of the information age I suppose I am in the camp of Biddle. The Industrial Age indeed had a revolutionary effect on warfare. The so-called Information Age has given us GPS to know where we are and outstanding night-vision to fight after hours, but I don't see the same order of magnitude change that occured circa 1900-1918.

Should I just drink the kool-aid and get on with it?

 

[Old Sweat]

Iain

In my dottage I am quite befuddled by this approach. Developing sophistication in miitary thought involved incremenral increases in command and control. There may be a place for removing a level of command, but it surely is not at the company/squadron level. It seems to me that this concept reverts to the tribal charge, or the assault of the French chivalry at Crecy and Agincourt. Heck, I guess it worked at the Little Big Horn, but even there I am not all that sure that it would have worked against a more tactically cautious commander. I submit that if you had reduced the size of your conventional battle group, it still stood an excellent chance of prevailing against a roughly coordinated horde.

Is there more to the concept that I have missed? Surely no competent student of military affairs would propose this without offering an example of how it would be able to defeat the enemy by swamping his command and control. Maybe that is the ultimate aim, but it, at least to my point of view, underestimates the ability of a functioning command structure to process information and develop and implement a plan.

 

[TangoTwoBravo]

I suppose it could work against an enemy who allows himself to be surrounded and doesn't care about mutual support or security. It could work, I suppose, if your weapons could hit the enemy without any chance of the enemy hitting back (the Romans vs the Parthians at Carrae). Even then, we call the asiatic horse archers "swarms" but I bet that you could find a chain of command that could be traced back to someone who issued orders of some kind.

I'm trying to see how we swarm somebody who is in an defensive position with interlocking arcs of fire. Why would we anyway?

I find the academics view on command and control especially troubling. They seem to assume away the importance of somebody actually being in command and telling people what to do. I think that span of control is a real issue. The network-centric folks don't like the idea of control. They want everybody to see everything and figure it out for themselves. I believe in mission command. The C2 opponents offer that war is chaos and that it cannot be controlled. I offer that you try and muddle through the chaos as best you can, animated by the commander's plan and intent but taking such opportunities as you see fit.

 

[Infanteer]

Links to the Edwards RAND monograph and the Arqulia/Rondfeldt monograph, which are the two key documents in the "swarming" theory:

http://www.rand.org/pubs/monograph_reports/MR1100.html

http://www.rand.org/content/dam/rand/pubs/documented_briefings/2005/RAND_DB311.pdf

It's been sometime since I read these, but some problems become immediately apparent:

1.  Swarming is something that fell out of the "network centric" fad.  Unfortunately, like other fads such as Maneuver Warfare and Effects-Based Operations, these theories are "dorito chips"; attractive and tasty but empty of any useful nutritious value for the military professional.  Edwards states that something can only be a swarm if it involves an attack on 5 or more axes - what defines an axis?  If OC A Company sends 1 and 2 platoon to a line of departure, does each platoon constitute a separate axis?

2.  Unlike ants or bees, humans in warfare have relatively the same velocity on the battlefield; just look at the cover to the Arquila/Rondfeldt monograph and tell me how those infantry units are supposed to defeat each successive enemy element in detail from five different directions?  If one force is indeed faster than the other and uses to pick off the enemy, than this is an example of better mobility and effective concentration and not "swarming" (unless we just want to give it a fancy name to sound smart and cutting edge).

3.  The Network-centric fad seems to, as T2B pointed out, wish away some basics of military organization.  Terms like "flattening" and drawing parallels to Walmart are neat, but they completely ignore the fact that these systems fall apart under a battlefield dynamic.  Echelons of command and the chain of command have been around in some form for thousands of years due to the fact that they work.

As someone wiser than me once said, "swarming" is a matter of perception and not a doctrine.  Simply appreciating the dichotomy of concentration and dispersion and its application to the battlefield is more useful than chasing junk theory.

 

[a_majoor]

Since I am the guilty party here (and in the ADTB back in the day...) I will try to lay it out as I see it.

The example of the Little Big Horn is actually a good starting point. Modern reconstruction of the battle shows the Lakota, Northern Cheyenne and Arapaho warriors moved up using cover to envelope and overwhelm the skirmish lines of Custer's forces. The US Cavalry soldiers were also being suppressed by high volumes of fire from arrows and repeating rifles (while answering with single shot rifles). Individual native warriors firing and moving up under their own initiative, then moving in for the kill as US fire slackened would constitute a "swarming", and while there would be organized and semi organised warrior groups under various war leaders, there was not a coordinated attack the way we would understand it. Indeed, many of the warriors might have come seeking a fight after the initial attempt by Custer and Reno to cross the river and attack the village had been repulsed, and probably a large percentage would have been advancing to the sound of the guns without any definite orders.

So my take on a swarming attack is soldiers or sub sub units able to move towards the enemy and rapidly engage them in conjunction with nearby units without requiring detailed orders or coordination. The analogy of flocking or schooling reflects the fact that a flock of birds or school of fish can carry out seemingly coordinated actions without any central organizing function at all (although in nature, the prime purpose is to move the body of creatures in such a fashion as to confuse or outmanoeuvre predators).

The primary reason to suggest this might have advantages over hierarchical command and control relates to the local knowledge problem, since fleeting impressions and diffused and undifferentiated information available to the local commander can be quickly assimilated and exploited on the spot, but will take much more time to collate, transmit up the chain, be assimilated and then orders passed back down to be executed (this is also why the Wal Mart analogy makes sense on a larger scale, local knowledge can be exploited on the ground rather than waiting for orders and instructions from higher). Now this is something that  works on a tactical level to leverage speed of action before an enemy force can organize. I can imagine a company sized unit advancing to contact, and soldiers and sub sub units rapidly moving to converge on a contact to isolate and destroy it. Higher up the food chain, there still needs to be coordination of assets, effort and logistic support, but even there, we *might* see reorganization to allow or support swarming tactics. For example, imagine the battlegroup releases air assets to various "boxes" where they link directly to the shooters on the ground in the way SOF operators could directly access air assets against the Taliban in the 2002 invasion of Afghanistan.

In situations where speed cannot be leveraged, such as a dug in defense, swarming may not have any clear advantages (unless there are corresponding behaviours to flow around the obstruction and seek a way past while isolating the position).

If this seems slightly confusing, remember I am not a military theorist nor credentialed, so I am doing the big hand, small map version of swarming tactics as I see it.

 

[TangoTwoBravo]

Since I am the guilty party here (and in the ADTB back in the day...) I will try to lay it out as I see it.

The example of the Little Big Horn is actually a good starting point. Modern reconstruction of the battle shows the Lakota, Northern Cheyenne and Arapaho warriors moved up using cover to envelope and overwhelm the skirmish lines of Custer's forces. The US Cavalry soldiers were also being suppressed by high volumes of fire from arrows and repeating rifles (while answering with single shot rifles). Individual native warriors firing and moving up under their own initiative, then moving in for the kill as US fire slackened would constitute a "swarming", and while there would be organized and semi organised warrior groups under various war leaders, there was not a coordinated attack the way we would understand it. Indeed, many of the warriors might have come seeking a fight after the initial attempt by Custer and Reno to cross the river and attack the village had been repulsed, and probably a large percentage would have been advancing to the sound of the guns without any definite orders.

I'm not sure what you are guilty of!

Looking at Little Big Horn, I'm not an expert on that battle but there was something like 3,000 Indians against some 225 US Cavarly. Since they were armed pretty much the same (although the Indians actually had an advantage as you note), I'm not sure if we can draw too much from that battle beyond the danger of hubris. Of note, though, Custer himself split his command up in order to attack the Indian camp from multiple directions. As such his force was divided when they got hammered and were unable to consolidate (Benteen. Come On. Big Village.Be quick. Bring packs. Oh Crap).

I'm also not sure if the Indians were a swarm. There was a guy in charge, and I wouldn't be surprised if there was a command structure - just not one that we would instantly recognize. It would, though, probably be "flatter" than a traditional military command structure. Still, Crazy Horse had a plan and his troops executed it - part of his force fixed Custer while the other swept around the hill and enveloped the US Cavalry.

Heck, I'll take Little Big Horn as reinforcing traditional tactics and principles. The Indians had concentration of force while the US Cavalry did not. Still - not sure if it would have helped against those odds...

I was skeptical in 2000 reading about swarms and C2, since we are neither fish nor fowl. I'm all for initiative, but I think that the levels of C2 have a real purpose. In traditional tactics/C2 there is certainly scope for leaders on the ground to use their initiative in light of their understanding of the local conditions (mission command). I like the idea of marching to the sound of the guns, but at the same time some level of command and control linked to a simple coherent plan is needed.

 

[Infanteer]

Little Big Horn is an example of encirclement; nothing more, nothing less - why a theorist has to make more out of it is beyond me.  "Swarming" seems like a fancy word for concepts like "initiative" and "dispersion" - why make something up to replace these?  The concept of "spreading out" is neither new nor doctrine.  The lessons we need to take from Little Big Horn are not "networks" and "swarming tactics" but that of "arrogance" and "putting yourself in a bad position".

The problem with swarming is that it is like the Maneuver Warfare principle of "surfaces and "gaps"; it creates definitions which don't actually provide anything meaningful.  Why isn't an attack from four points considered a swarm?  Is there a certain differentiation of angle of attack that qualifies a swarm (ie: are attacks from 20, 23, 26, 29 and 32 hundred mils a swarm?)?  What is the amount of lead time swarm components are allowed to have before swarming before the attack is considered to be conventional, synchronized attack?  What level does swarming occur at?  If I attack with swarming platoons along a linear battalion frontage, am I really swarming?

All these questions point to the fact that someone really just cherry-picked some examples from history, threw some buzz-words on it and called it "cutting edge theory" proving "network-centric operations" are the reality of the future.  As a profession we are ignorant of many of the basic fundamentals of the profession but we eat up the dorito chips.

It's nice that birds move in flocks and Walmart can get maxipads on the shelf pretty quick, but those really don't have much to do with battle or the human dynamics that influence it. 

 

[a_majoor]

To my mind, the key difference between a swarming attack and a conventional encirclement lies in time and the amount of command and control involved.

On the Indian side, there were chiefs who were working out an overall strategy and assembling their people and supplies at the Little Big Horn. They also had patrols out and whatever passed for intelligence so they had some sort of idea where the American columns were and what the Americans were planning to do. This would be a relatively high level command function. At lower levels there would be war leaders who would organize their bands, provide overall leadership and inspiration as well as discipline. During the tactical battle, they had neither the time nor ability to pass orders or directly influence the battle beyond the "follow me!" approach to warriors in close proximity. Their followers were spreading out independently and carrying out fire and movement on their own to get to grips with the enemy. The better analogy than a swarm might be the launch of a volley of fire and forget missiles in the general direction of the enemy, knowing they will steer their own way to target without any further instruction from you. In one of my ADTB papers I argued that "[f]locking or swarming behaviour will manifest itself when certain conditions are met, as defined by a common rule set that all members of the team know and understand".

Now I have no deep attachment to ideas that you need to attack from "x" separate axis of advance to constitute a swarm, to me the prime issues are self organizing behaviours controlled by certain rule sets or conditions and the time it takes to initiate the action. The flock of birds swirling about to confuse a predator can be modeled in a computer simulation using very few rules (mostly to do with maintaining a spacing or zone of separation between members of the flock); I believe that something similar can be done with people. Flash mobs *might* be one of the human analogues; groups like the Black Bloc anarchists use this to confound the police and intelligence agencies of governments around the world.

 

[TangoTwoBravo]

A group of soldiers employing fire and movement is "swarming" then I fail to see how it is distinct from the modern-era tactics that we know and love. A combat team quick attach features a single commander giving radio orders. The team then launches. It might look like a self-directed swarm, but there is lots of C2 going on at low levels. The Troop Leaders and Platoon Commanders have important parts to play, and its not just keeping formation.

They key to me is that the combat team reacts to the orders of a single commander. That commander might have quickly consulted with a small group (the FOO and the supporting arms commanders), but the team moves to the will of one commander. They will use their initiative to meet the intent, and will very likely make improvements on the plan. The plan might even fall out of what the lower-level commanders started.

I fail to see how a set of rules that govern behaviour could improve on what we do now. We have formations and drills that speed things up and give us a start point for decision making.

The chain of command withs its fairly tight span of control (2 to 5 units for me) might seem slow and redundant, but it is also resilient. Those extra levels of command might seem inefficient, but they give the force layers of thought process that add value. I heard all about "its a section commander's war" or "its a platoon commander's war", but I truly saw the value of the company. That HQ brings certain talents to the table, and having the whole company present (not necessarily lined up together) at an action gives the force some capacity to absorb misfortune and keep on going.

 

[Old Sweat]

I am not sure if this is not getting away from the main thrust, but a major factor that has been ignored by the swarmistas is the increased probability of blue on blue engagements. These could be by direct fire, by indirect fire or by air or aviation strikes. I also would not want to organize and implement the combat team commander's fire plan if the troops and platoons were charging about in all directions and would be absolutely flummoxed trying to fire plan without any one maneuver commander being in charge.

 

[TangoTwoBravo]

Old Sweat,

Fratricide is one of the concerns that I have regarding swarming. The network-centric folks rely on near perfect situational awareness to prevent this happening. I'm not sure that they will be able to deliver this. Even if they do, you raise a point regarding fire support. Who will allocate those fires? While a Battle View screen shot might show where everybody is, if there is no plan how do we know were people will be in ten to fifteen minutes? How does the fire support guy anticipate requirements if there is no plan?

I studied a Korean War battle for a presentation to my Advanced Course with the US Army. A Chinese attack captured a piece of high ground. Two separate US counter-attacks formed, each one approaching from a different axis. This would have worked great in the age of spears, but in the age of firearms it wasn't so successful. The fire of one element effectively pinned down the other. You don't see this in simulation, but it certainly happens out in the rhubarb.

 

[Infanteer]

The fratricide comments by Old Sweat and T2B are on the mark - a flock of birds does not need to worry about running into a friendly beaten zone.  100% blue PA is at odds with the chaotic nature of combat - a deep flaw of "netcentric warfare" and thus of "swarming".

Self-organizing behaviors; what's that supposed to mean?  That I just do what I want and wander around looking for an objective.  Do swarming advocates sincerely suggest this?

I don't see how the idea of self-organizing behaviours presents anything superior to the synchronization provided by a 30 second frag order.  The natives at Little Big Horn may have fought as individuals, but they were guided onto the hill in an organized fashion, as Thucydides pointed out - 2000 braves did not just look and think "attack the cavalry!".  As well, Little Bighorn shouldn't be cherry-picked from the general experience of the Plains Wars; "self organizing actors" like warriors or mob participants are generally rolled up by organized forces.

As for time, that is a human factor that is grounded in experience and intuition.  Five swarming commanders are just as likely to be confused and have their decisions clouded by friction in the battlefield; in fact, I'd argue that having five independent decision makers instead of five synchronized by one is more prone to the effects of friction.  Anyone who's used a military radio understands this.

I'd be curious to see if the swarming advocates at RAND can show that the good parts of their theory aren't principles that already exist (and don't need renaming or retooling) and the other parts are actually feasible and not just doodles on a napkin.

 

[TangoTwoBravo]

To be fair to having a set of rules that guide behaviour in battle, I've always liked Nelson's "No Captain can do very wrong if he places his ship alongside that of the enemy." Its a different time, but Trafalgar has something to show here. Nelson sought a close range knife fight against a portion of the enemy fleet. One could try and bend that to swarming. He devised the plan, however, and his subordinates executed it. He gave them liberty to seek action on their own if things got confused, but he didn't just aim his fleet at the enemy and tell them to get stuck in.

 

[dapaterson]

He devised the plan, however, and his subordinates executed it. He gave them liberty to seek action on their own if things got confused, but he didn't just aim his fleet at the enemy and tell them to get stuck in.

Sounds like, oh, Mission Command.  Specify an objective and parameters, and let your subordinates achieve the mission (with co-ord from above).

 

[TangoTwoBravo]

Precisely. It can be hard to translate C2 styles from one element to another, but I think that it fits here. My point was that on the surface it could have been seen as a swarm, but in reality it was an aggressive plan executed boldly with lots of latitude. Thats how we should fight/C2.

 

[a_majoor]

My google fu is getting weak, but I recall reading about Israeli patrols having a difficult time in Lebanon due to "swarm" tactics used by the locals. Essentially, if anyone saw or even suspected an Israeli patrol was in the area, they would spontaneously go out to try and spot the patrol. More and more neighbours would join and start quartering the area, while calling for Hezbollah fighters and, if possible, directing them against the patrol. (This was prior to 2006)

This is a sort of COIN in reverse, where the insurgents have the advantage of the locals providing support against the outside force. The Israelis would not engage unarmed civilians, but were forced to abort the patrol. This seems closer to what I was thinking of; no centralized command and contol, people taking action based on a common rule set when certain conditions are met and everything happening on a timely basis.

This is another example from the 2006 conflict which can be used to illustrate the point (but also illustrates the importance of basic principles as well)

http://www.spiked-online.com/index.php/site/article/6321/

In October 2008, I visited him in Beirut and asked him what he had seen and heard. At the beginning of the 2006 war, Crooke told me, Hezbollah fighters were given general orders and were then broken down into tiny cells, each of which operated independently of any central command. They performed very well without employing high-tech communications or complicated military manoeuvres to help get around the battlefield. A specialist team was given high-tech listening devices and managed, according to Crooke, to intercept electronic communications flying to and fro between Israeli military personnel. The bulk of Hezbollah’s military units, however, were encouraged to avoid unnecessary electronic chatter; when unit commanders did need to pass on messages they relied on relatively primitive means such as motorcycle couriers.

Despite the fact that their information loop was not speeded up by new communications technology, Hezbollah’s fighters were much better equipped to swarm around their enemy with great agility. ‘Rather than have to react faster than the [Israeli] decision-cycle’, one early analysis of the war from Washington’s Center for Strategic and International Studies concluded, ‘they could largely ignore it, waiting out Israeli attacks, staying in positions, re-infiltrating or re-emerging from cover, and choosing the time to attack or ambush’.

Hezbollah’s commanders found that giving their fighters clear prior battle instructions was vastly more important than allowing them to liaise with each other electronically during the conflict. Faced with a technologically superior enemy, they seemed to understand, it was still possible to knock your enemy off-balance and confuse him. But this was only true if those under your command had a very clear idea of what was expected of them; only if they ignored the enemy’s electronic information loop; and only if they switched off their mobile phones and fell back on their own initiative.

 

[TangoTwoBravo]

An interesting vignette. Note that the statement that "this was only true if those under your command had a very clear idea of what was expected of them...only if they switched off their mobile phones and fell back on their initiative."

I think that Nelson would have approved, as would Stosstruppen commanders in 1918. Give clear direction to include intent, but make sure that your subordinates know that you expect them to use their initiative and not keep looking to you during the fight.

I find it interesting that that Swarm theorists do not include insurgents as being swarms, probably due to the lack of a network-centric piece.

I am all for using mobility and firepower to achieve concentrated attacks against isolated enemy elements. I want subordinates to leaning into the fight and not sitting back waiting for orders. I think that is important, however, that there is still a command and control piece that makes this happen.

 

[Loachman]

I am presuming that none of these thinkers have any real-world experience.

 

[TangoTwoBravo]

I came across an interesting point from the 2006 Lebanon fight regarding "swarming." An article by Dr Milan Vego "Systems vs Classical Approach to Warfare" mentions that an Israeli Division was told to execute "swarmed, multi-dimensional and simulataneous attacks" instead of actually assigning them a mission. Apparently it didn't work out too well.

The more I study the more I am convinced that not much has really changed since 1918.