Defeating the Swarm – Observations from Kambula.

[Dick Cheney.]

It is becoming increasing apparent that the most lethal battlefield tactic ever devised is the swarm. A staple of asymmetric warfare -and indeed hunter gatherer days- swarm tactics attempt to maximize target saturation by attacking in unison with as many units as possible, while also avoiding casualties through dispersion and sustained hit and runs [1]. The end result of a successful swarm is that all units are able to converge on a single target – who by definition of being assailed on all fronts (ideally) is also overwhelmed. A main drawback of using swarm tactics however is its prerequisite for multiple units. These units -while also acting independently- must then attempt to exploit its quantitative edge through synchronization and maneuver, which isn’t always possible due to terrain and anti-access technologies. As its numbers increase a swarm’s combat power also increases, but its coordination and mobility all go down. It also becomes laughably hard to field and support a large enough force that could realistically be called a swarm, especially within time and space. Regardless, if its prerequisites are met, the combat power of a swarm is indisputable. Targets enveloped by swarms become fixed, mobility is reduced, decision cycles are interrupted, concentration becomes impossible, outside communication is lost, and all points of contact, to include exposed units and weak points, are assailed and harassed - in chorus and simultaneously.

Swarm tactics and its associated formations are now poised to become the next revolution in military affairs, and perhaps the final evolution of battlefield tactics. A familiar tactic, which has historically allowed the tiniest of creatures to overwhelm the mightiest of foes, its awesome prerequisites that barred it from becoming a viable method for waging war are now being rectified through advanced weapons technology; to include computer networks, robotics, advanced sensors, automation, UAVs, and artificial intelligence.

How were swarms though historically defeated? And will there be any applicable lessons for future war planners before it is too late?

Below are some observations from the Battle of Kambula.





Brief Background

Horns of the Beast

Before diving into the Battle of Kambula however, a clear link between the Zulu Army and swarm tactics must be established.

Besides the fact that the Zulu fought essentially as elite skirmishers and could take advantage of fluid formations, the point that gains the most serious attention from swarm theorists was its regiment system. Known as the beast horns, a Zulu army was typically divided into four regiments, which formed the head and body of a buffalo. Two regiments would form the Horns, one regiment the Chest, and another the Loins. Together these regiments would maneuver into a mass swarm at the beginning of a battle. The objective of the Horns -who mainly consisted of young men- would be to encircle the enemy and prevent escape. Meanwhile the Chest, which usually consisted of the most experienced warriors, would make initial contact with enemy forces and push them into the Horns. The Loins, depending on the circumstance, would act as a reserve or could be placed at a strategic location behind the enemy and perform a sneak attack. In all, the point of the regiment system was to enable fast encirclement where the Zulu could converge from many directions, on the same target, at the same time.

Zulu Diagram:

Spoiler Alert, click show to read: 

The Battle of Kambula (29 March 1879)

Following the disaster at Islawanda, Lord Chelmsford orders his remaining columns in Zululand to dig in until a fresh invasion can be prepared. Fortunately, due to the troubles of assembling a large army, the Zulu are not able to make any more headways until two whole months have passed.

After a brief skirmish at Hlobane, the North Column led by Evelyn Wood prepares for the worst and takes up a defensive position around Kambula. Learning from Islawanda, the British defenses consist of a 360 degree perimeter along a hillside. Among the earthworks is a large laager formed with wagons and lined with trenches, in addition to a stone redoubt that covers the east entrance.

Armed with Martini Rifles and 7-lb artillery, 2,000 or so British soldiers await a massive assault of 22,000 Zulu warriors. Arriving first and to the north is the Right Horn of the Zulu formation, 6,000 men strong.

Wood, correctly predicting encircling attacks from the Left Horn and Chest, seizes the initiative and orders a detachment of his rifled cavalry to make a quick attack against the Right Horn, in hopes of provoking an early assault. The plan works, and the Right Horn advances on the north-side of the British positions – unsupported by both the Left Horn and Chest who are still too far out. Wood proceeds to defeat the Right Horn in detail with a combination of canister, shrapnel, and rifled musket fire. Those Zulu who do manage to make contact with the wooden palisades are met with bayonets.

Wanting to catch up with the Right Horn, the Left begins its assault on the south side of the British position. The south provides much more natural cover, and the Zulu are able to deploy what muskets they have on advantageous firing positions. Realizing he’s in trouble, Wood reroutes his artillery and advances a small infantry column to drive away the Zulu who are providing flanking fire. The British attack is successful, and together with the rerouted artillery, Wood creates a crossfire that devastates the Left Horn.

Arriving super late to the battle is the Chest, who are also mowed down with rifles and artillery. By the time the Right Horn is able to regroup and support the other Zulu attacks, it is too late. The Zulu army is decimated.

The battle closes with 2,000 to 3,000 Zulu dead to only 29 British.


Observation #1 – Retreat - The Option Not Taken

Not considering the strategic picture (which is not relevant to this discussion) the correct tactical move when facing a swarm is often to retreat. At Kambula the British were outnumbered 11:1 and were separated from the other British column by dozens of miles. Had they been facing a peer competitor at 11:1, retreat may have been the only option available. Retreat would have been the correct move at Roke’s Drift (had the 24^th foot had the means to do so) and would have saved the British Army at Islandwana (provided the army also had the time). Retreat also prevents an army from becoming encircled and coming under siege. Regardless, this point does not argue whether Wood could or should have retreated from Kambula (though I think he could of) only that retreat is a valid counter against a swarm. Reducing a terrible numerical disadvantage that Wood could expect -and the predicaments of becoming encircled- comes only with retreat (or alternatively asking for reinforcements).

Observation #2 – Perimeter Defense

Perimeter defenses showed their worth at Blood River and at Roke’s Drift, defeating a Zulu swarm twice. More important than thwarting piecemeal attacks, a perimeter defense allows defenders to defend in all directions – a must against a swarm. Beyond Kambula, the advantage of establishing a perimeter defense is noted with the British Square. In the center of a square are the commander and the army’s most valuable assets -such as supply boxes, ox carts, and ammunition. With their valuables secured, and flanks protected, defenders in a square are free to lay down suppressive fire from all sides. A tightly packed square also provides a solid front, which can keep unit cohesion together when facing attacks from all directions.

The last point that needs to be made, besides the protective cover that perimeter defenses had offered to the troops at Kambula (not to mention lines of fire and a smaller front), is the ability to funnel attacks.

At Kambula this was observed with crossfire from the stone redoubt, and again with bayonet parties who could mass together to fight off any breaches. Though perimeter defenses -such as star forts- are not able to prevent simultaneous attacks, they are able to clump attacking troops together, who in turn can be targeted with enfilading fire. Beyond this, prepared defense can also slow or deny different points of entry, the most obvious case study is with terrain, moats, and barbed wire. At Kambula, trenches and high ridge lines prevented the Zulu from getting too close.

Observation #3 – Intelligence

Following the fight at Hlobane, and defeat at Islandwansa, Wood knew what was coming. He also had advanced warning from scouting parties and could predict the Zulu’s method of attack. This of course allowed him to position his artillery, distribute ammunition, and prepare a 360 degree defense. Knowing he would be enveloped sooner or later, he also made the correct call to goad the Right Horn into a premature attack. Though the speed and mobility of a swarm is unparallel, Kambula demonstrates that its attack pattern is predictable.

Observation #4 – Counter-Swarm

Wood’s cavalry attack on the Right Horn of the Zulu Army deserves the most study, in that it mimicked a counter swarm. Though the decision to give chase and engage the British early was brought about by poor discipline, the fact that Wood’s cavalry was able to provoke a premature attack demonstrates a potential weakness in swarms – its reliance on synchronization between units for simultaneous attacks. By disrupting attack patterns, and providing another target for a portion of the swarm, Wood’s cavalry effectively disrupted the initial plan for all Zulu regiments to simultaneously attack British positions. The end result of course, was defeat in detail, which later enabled Wood to concentrate a majority of his forces to the south-side.

Observation #5 - Weapons Tech

Like so many battles in history, weapons tech ultimately deserves the most credit for victory at Kambula. Spears and shields are ultimately no match for Marti Rifles and 7 lb artillery shells. With the Marti-Henry, soldiers could fire 6-11 shots a minute, and could begin engagements at 1,000 yards. By the time hordes of Zulu reached within 100 yards of a British regiment, volley fire approached hit rates of 100% [2]. The very fact rifle and artillery fire can be sustained with some predictably, means this battle can be modeled. Swarm tactics -to include the greatest strategies in the world- mean nothing against overwhelming technology.

Observation #6 - Lack of Diverse Units

Swarms depend on sophisticated coordination between units. In case of the Zulu army, this was achieved through their regiment system and collective behavior. However, part of what made the beast horns possible -to include simultaneous attacks from multiple angles- was that all units were trained, equipped, and fought in the same nonlinear manner. While this ultimately helps with coordination -and indeed doctrine- it does not leave much place for combined arms. At Kambula, the Zulu had no counter for British rifles or their perimeter defenses, carrying no siege equipment and stand off weapons of their own. The small initiative they were able to gain with the Left Horn, was brought about by antique muskets. Though the advantage of everyone thinking the same and acting the same is unprecedented coordination and mobility, its cost is that one counter that works against one will work against all.

Conclusion

These are the observations I feel best prove why the British won at Kambula, despite being outnumbered 11:1. It doesn't mean, and I do not argue, that these would be applicable counters against swarms in the future. These observations are only meant to be seen as a starting point for discussion. In the future, we should expect swarms to be highly sophisticated, lethal, and high-tech. The double whammy of weapons tech mixed with thousands of autonomous vehicles and thousands of flying machines will be an unprecedented threat to future battlefield commanders. And yet despite the challenge, history is a good starting point.

Notes:

[1] “Swarming occurs when the scheme of maneuver is a convergent attack of several semi-autonomous (or autonomous) units on a target.”
For an excellent briefing on Swarms: https://slideplayer.com/slide/7933497/#

[2] Figures and Hit rates for Martini Rifles against Zulu Forces:
http://www.modsimworld.org/papers/20...Isandlwana.pdf