1.  Introduction

It is probably becoming clear to anyone who follows war lately, especially the Ukraine conflict that automatic weapons, especially drones are rapidly changing what weapons and tactics are effective.

The purpose of this article is to first consider how a fully autonomous force with current or near term tech would go against a state of the art conventional force, then to consider where the equilibrium would be on attack vs defense for autonomous forces using advanced but foreseeable tech.

1.1 Description of fully autonomous force using near term tech

The idea is to have a force with as few different types of units as possible controlled in a distributed, mesh network fashion. The goal is for them to take over unlimited land territory, even in the face of tactical/strategic nukes, and without air superiority against very fast or high flying aircraft. Units are assumed to be fully autonomous, and emphasis is on economics as well as capability. For example many units are so cheap that they cannot effectively be countered by conventional missiles.

1.2 Unit Overview

All units can fly, they are optimized to destroy all land based armor, slow flying aircraft, competing drones and humans. Combat units are supported by larger logistics units.

1.3 Communication and Defense

All units are expected to communicate with point to point links (e.g. laser) and are hardened to varying degrees against microwave attacks. Given they are autonomous this would make jamming very difficult and electronic warfare not very effective.

1.4 Unit details

1.4.1 Recon and targeting drone

The cheapest and smallest – it is a battery powered drone with video and targeting system. Recon units form a mesh network.

They coordinate with missiles to defeat countermeasure such as flares and chaff from slow moving aircraft (helicopters etc) by observing and transmitting the position of the target from somewhat further away and transmitting that info to their missile.

They spot opposing drones, troops, fortifications etc. They can spend a brief time in the air, before being recharged by a larger unit.

1.4.2 Suicide drone

This unit is optional – Sniper drones and Javelin type missiles may make it not needed. This is the drone we already see, however with AI, the same p2p comms system as above and with some protection against microwave and laser attack.

1.4.3 Sniper drone

Larger unit, still likely battery powered. Sniper gun is its weapon. This is intended to be more efficient than using suicide drones against soldiers. It is cheap enough to be uneconomic to target it with quality missiles. When in the air it normally flies in a jitter pattern at high speed and ~1,000 m high up to defeat automatic ground based targeting systems that fire unguided bullets. It uses guided bullets if they are technically possible. It can also be equipped with sensor blinding rounds to help defeats tanks or similar. These shells explode close to the target and release IR and flak to confuse Active Protections Systems, so that Javelin type missiles can get through.

1.4.4 Missile launching drone

This is heavier and launches a single missile against larger ground or air targets. That is the Javelin or MANPAD type. It spends most of the time hidden on the ground and only takes to the air when a target has been identified. It fires, then immediately hides again. These units will often coordinate for a swarm attack, i.e. 2-10 Javelin simultaneously against one tank. To take new territory, these units will first attack anti-drone systems, (armed with lasers or guns) in coordination with the sniper drones.

1.4.5 Logistics drone

The most expensive drone, capable of carrying ammo and fuel. Contains a fuel cell or similar so it can use chemical energy to power itself and charge up battery powered drones. It needs to be large for physical reasons related to carrying efficiency, however too large and it will be vulnerable to attack from missiles from fast aircraft (too fast or high for MANPAD to stop). Many aircraft missiles cost ~$1M, so if these units cost $100K that may be a good tradeoff.

Where possible it will move on land using wheels, however its flying ability would be essential. It has some defense against existing missiles by flying low to the ground and taking cover if an income missile is detected, perhaps accompanied with simple flares/chaff deployed the seconds around landing.

The challenge for this unit is supporting and resupplying a force across a large and somewhat contested territory, say 1000km. The mass required would be much less than a conventional army because there are no soldiers to support, drones won’t shoot unless they are very likely to hit, and there are no heavy armored units in this setup. Fuel can be further conserved by a system where only light recon drones go into the air in an initially contested area, with sniper and anti-armor only becoming airborne when targets are identified. These units can refuel each other and also operate as a mesh pattern. They may create static hidden caches to extend range.

2. Results when proposed system takes on a conventional force

2.1 Unarmored soldiers

It should be clear they are no match for sniper drones or suicide drones by now.

2.2 Ground Armor

The missile drones carry MANPAD or Javelin to destroy ground armor or low flying aircraft. Fast or high flying aircraft can be ignored. They cannot hit anything with bombs as all units are mobile, and missiles will be at least as expensive as any drone they can target. 

Lets consider attacking an armored anti-drone system (gun or laser) with active protection systems.

2.2.1 Attack details

1. Recon drones pinpoint the armor from a long distance, out of range of the system.

2. Sniper and missile drones get into position by staying very close to the ground and hopping from cover to cover while they are potentially in range of the system.

3. Sniper drones fly up from behind cover 500m-1K away and quickly shoot the system’s sensors, directly with bullets and indirectly with shells that explode with IR and chaff in a cloud surrounding it.

4. While the chaff cloud still surrounds the armor, missile launching drones coordinate an attack so all the Javelins arrive at the same instant from multiple directions when the target systems sensors are still degraded. 

This is expected to work against an anti-drone system with a gun, laser or combination.

2.3 Slow aircraft, helicopters

The above attack could also work, however aircraft are more maneuverable and can deploy flares, so the MANPADs could lose lock. Flares can be overcome by a mesh of recon drones somewhat close to the target that can give targeting information to the missile. A smarter missile with a reasonable flight time could also break off, circle round again when flares are detected and take another pass when they are gone.

2.4 Proposed system’s defense against other existing counter measures

2.4.1 Protection against jamming

1. Communication can be done with directional point to point laser. If the system is sufficiently directional, then a more powerful laser would not be able to jam it.

2. Jammers can be easily targeted. They have to emit a strong radio signal which can then be targeted.

3. AI autonomous systems would mean communications would be needed much less.

2.4.2 Microwave weapons

If radio/microwave comms are not needed then shielding can be cheap and effective. If radio comms is needed, then it can be active in safe areas – for example retractable antenna deployed when such weapons are not in range.

2.4.3 Guided Missiles

The idea is that almost all drones are cheaper than missiles, leading to the saying that if you have more drones than they have missiles, then you have air superiority! 

2.4.4 Unguided bullets

These cannot hit drones flying in a jitter or random pattern at medium range. Sniper drones should outrange a hand-held anti-drone system as a result. The sniper drone could be accurate from 1,000m away but out of range of the anti-drone system.

Rough calculations:

A drone can accelerate to 200Kph in 1 sec

A fast gun fires at 1,200 m/s. If the sniper drone was 1K away, moving at 200Kph, it can move approximately 50m in the time it takes for the bullet to reach it, so it can’t be effectively targeted.

A sniper drone can still accurately hit a stationary object from 1K away. It could hover for 500ms at a randomly chosen time, target and fire, then fly off again.

2.4.5 Laser

Drones can be hardened against a laser by reflective or ablative coatings and rolling when targeted so that the same spot is not hit. They could fire smoke flares or reflective chaff to help further. They can then target the laser like for armor – shells that explode near the laser to blind it, so anti-armor weapons can get through. Drone sniper bullets could hit the sensors or laser directly also.

3. Current situation

It seems to me that such a system could wreak massive havoc against a conventionally far superior force for a fraction of the cost. Russia has a massive incentive to develop such a system, as do Iran and others. Countries such as China could mass produce drones or components of such a system without getting directly involved and with somewhat plausible deniability. 

Ways to ensure international stability include preventing such weapons from being developed or quickly making sure the balance of power stays the same if everyone has them.

4. Drone army duel – how a fight between two similar systems would look

What would the balance of power look like if such systems where developed to their fullest? If two such systems where to fight each other, what would determine which side won.

4.1 Guided bullets

These would be a major advantage in a sniper drone dog-fight because both drones can change direction fast, having the bullet adjust and track the target drone would give a large increase in effective range. 

For two drones in the air, a small range advantage would enable one side to keep pushing the other back. Tech to target and guide the bullets would be the deciding factor.

4.2 Drones vs slow ground based armor revisited

If the armored vehicle is without support then it looks like the lighter drones can destroy it, even if the vehicle out-ranges it with faster better guided ammo. If somewhat out-ranged, the drones could stay on the ground for cover, and carry out the attacks above, they would sustain losses but it probably would be a cost effective attack. However if the armor has backup drones it could change things. For example a swarm of very small, cheap recon drones out of range of enemy sniper drones to detect javelin and missile launches and continuously track their position. Additionally rather than having the Hard Kill Active protection system attached to the vehicle, it could consist of larger drones flying nearby each capable of destroying a Javelin type anti-armor missile. 

What is the point of the armor in such a situation? It can be used to interrupt the logistics drones discussed above, and for area denial and protection. If the opposing side has local air superiority then the transport drone cannot get past – however flying high above the vehicle could still work.

It also could attack fortified positions or facilities, e.g airport hosting conventional fighters or hypersonic missiles.

4.3 Bombs are ineffective

That is anything that falls by gravity alone is too slow and predictable to reach the target. It would be intercepted by the anti-missile defense or simply a much cheaper suicide drone in mid-air.

4.4 Offense vs Defense - Strong side vs dug in weaker side?

It seems to me that fortifications, cover and dug in units would offer defense a large advantage. For example a sniper duel where one was on the ground and could hide behind cover would have a large advantage against one in the air. The outranged drone could still hold off the stronger one by only coming out from cover for a very short time to shoot, but still guide the bullet against the airborne one. Similarly for larger units, a unit that could shoot from within ground based tunnels would have a large advantage. The stronger side may be able to fly at high altitude over the weaker sides fortifications, however to be truly effective, they would need their logistics units to be able to do that un-opposed.

The defensive side tries to create a ring protecting their entire territory, especially from logistics units crossing. A laser within a protected tunnel system would also be more effective.

4.5 Railgun perfected

This would favor dug in defense – given the long range of a rail gun, any logistics attempting to overfly could likely be shot down by the guided slugs. Sniper drones would probably be too numerous and uneconomic for the railgun to take down most of them.

4.6 Water

Our current tech is very far from the limits allowed by moderate tech advances. Our drones don’t compare to the capabilities of sea creatures in speed, maneuverability, range. For example a water unit that could swim with the capabilities of a dolphin, recharge itself by coming to the surface and deploying floating solar panels, and carry explosive would be very disruptive. A swarm of them could attack a cargo ship, targeting the propellers or other weak points and disable a much larger more expensive vessel. Detecting such a swarm would be very difficult as they could spread out and just sit on or slightly below the surface until needed. They additionally could launch airborne units to attack the ship above water. In this situation, even a much weaker side could cause MAD destruction of seaborne cargo transport. Unlike in the red sea where choke points are most vulnerable, wide areas would be hard to defend. Even if the cargo ship had an escort say of 100 similar units, then 1,000 attacking units could swarm up to overcome them.

5. Conclusion

New weapons can change the balance of power – for nukes it became MAD that prevented one side launching an attack instead of an advantage to defense for the conventional warfare that existed before. For drones it looks like defense will be favored, however the two sides need to be somewhat matched for that to happen. No current military is close to such a capability so there is potentially a dangerous and unstable time as militaries quickly increase capabilities. War often breaks out when one or both sides do not know how they compare to the other, or have false beliefs.

6. Links Related to Current Progress

AI drones

AI drone for Ukraine

Sniper drones:

Sniper drone

Laser guided bullet

Anti-Armor

Tank vs drone

Overall picture

Military

LW 20 year lead