Recent truly impressive improvements in the performance of commercial drones (a form of robot), task-specific artificial intelligence, and 3D printing require all national militaries take a hard look at how robotics are changing the modern battlefield. The growing use of sophisticated surveillance and strike remotely piloted aircraft represents only a tiny fraction of the impact these systems will have. Advanced robotics guided by task-specific AI and produced in massive numbers by advanced manufacturing—particularly 3D printing – will change how we fight in air, sea, and land domains as well as how we protect our civil societies.

In the short term, aerial drones present the greatest threat and opportunity for both state and non-state actors. Commercial and hobbyist demand has resulted in increasingly capable drones ranging from small hobbyist quad-rotors like DJI’s Phantom series to long-duration and long-range surveillance requirements for fishery and forest regulation like Aerovel’s Flexrotor (2000 miles range /40 hours endurance) and Defiant Lab’s DX-3 (900 miles/25 hours).[1]^,[2] Many of these commercial products boast GPS waypoint navigation and multi-spectral sensors. Soon they will have inertial and imagery navigation as well as effective target identification systems.

While current commercial drones carry small payloads that limitation can be overcome via three distinct approaches. The first and least technically challenging approach is to think in terms of “bringing the detonator.” The second, employing explosively formed penetrators (EFPs), requires a bit of technical expertise but most of it can be learned on line.^[3] The third is to use swarms of cheap drones and count on cumulative damage to accomplish the mission.

Drones’ Commercial Threats:

1-   

       “Bringing the detonator,” uses the drone to deliver a small initiating charge to the much larger supply of explosive material provided by the target. Even a few ounces of explosives delivered to the right point on the target can initiate a much larger blast. The 2013 explosion at a Texas fertilizer storage facility, which destroyed five hundred homes and killed fifteen people demonstrates the potential of this approach. However, the highest political and economic impact could be generated by attacking multiple passenger aircraft parked at terminal gates.

2-    

        EFP warheads, allows light drones to penetrate heavier targets such as a tanker truck or railroad tank car. With as little as 32 grams of explosives, a thumb-sized EFP can penetrate up to 1 ¼ centimeters of steel.[4] Increasing the size of the EFP to only a few pounds allows it to destroy even well-armored vehicles. In Iraq coalition forces found EFPs in a wide variety of sizes—some powerful enough to destroy an Abrams tank. Others were small enough to fit in the hand—or on a small drone.^[5]

3-     

Using swarms is enabled by task-specific artificial intelligence (AI) and 3D printing. Task specific AI can employ existing GPS systems to navigate to the designated target area and then achieve precision by using cell phone technology to identify and attack a specific target. As a result, there is no requirement for a pilot to guide any of the drones.

Aerial Drones

3D Printing or additive manufacturing provides the means to produce tens, and soon hundreds, of thousands of drones. Three years ago, a team at the University of Virginia printed a drone in a single day, then added a small electric motor, two batteries, and an Android phone for guidance to produce an $800 autonomous drone with a range of twenty kilometers.^[6] Today Carbon 3D sells a printer that can print 100 times faster than the University’s.^[7] A small factory with only a hundred such printers could make 10,000 drones a day. A 3D printing plant expanded to the 1,000 printers planned by UPS could print 100,000 drones a day.^[8] The limitation is no longer the printing but the assembly and shipment of products. Both processes can be automated with robots. In the near future, drones could be produced at a rate exceeding many types of ammunition – and often at less cost per round.

Today the U.S. Armed Forces are actively exploring the use of smart, coordinating swarms in the air and at sea.^{[9], [10]} While these programs are still developmental and thus use a limited number of drones, recent dramatic cost reductions in each of the needed technologies will increase the number by an order of magnitude. Once 3D Printing matures, states will be able to employ them in the thousands. The Chinese already have launchers mounted on medium trucks that contain eighteen Harpy long-range drones. The LOCUST program shows launchers can be configured into twenty-foot commercial cargo containers.^[11] Depending on the type of drone, a twenty-foot container could hold anywhere from twenty to several hundred. The use of these containers means every commercial truck that can carry a container becomes a potential weapons system. Every sea going vessel to include fishing boats becomes a potential weapons platform.

Swarms of attack drones guided by limited AI will provide relatively inexpensive but effective anti-access and area denial weapons. The Polish Army is already fielding 1000 dronese per year with a range of thirty kilometers armed with carry anti-tank, high explosive, high explosive anti-tank (HEAT) or thermobaric warheads.^[12]

Ground Drones

Both state and non-state actors have already built and are employing remotely operated land robots – and using them for basic patrols. In 2008, the U.S. Army purchased twenty four robot sentries but did not arm them.[13] In 2014, the Russian Strategic Missile Forces deployed mobile robots to guard five of its ballistic missile installations. The Russians armed them and gave them limited autonomy. Dmitry Andreyev stated, “These robots can detect and destroy targets, without human involvement.” The robot sentry has multiple sensors and a 12.7-millimetre heavy machine gun.^[14] In 2014, the South Koreans put autonomous robots on patrol along the DMZ – but they cannot fire without a human’s permission.[15]

As always, active warfare has stimulated creativity among the combatants. In Iraq, Shia militiamen have deployed remotely operated robots in the chaos of actual ground combat. Members of the Hashd al Shaabi (Popular Mobilization Units) in Iraq are using four homemade bots to fight ISIS. All four use cameras so the remote operator can see and aim. They are armed with a variety of weapons from rifles to medium machines guns to rocket propelled grenades.[16]

The Russians have gone a step further and plan to deploy their Uran-9 remotely operated mini-tank to Syria. “The Uran-9 is an unmanned armoured reconnaissance vehicle equipped with an automatic cannon, a machinegun, guided missiles and a wide array of surveillance sensors. The vehicle can automatically identify, detect and track enemy targets based on a pre-programmed path set by its operator.”[17]

Maritime

The maritime domain has not been neglected. Remote-controlled, inexpensive sea gliders have very long endurance (years), and require no crew or support ship (except while they’re being deployed or recovered).^[18] Michigan Tech has already produced a prototype of a $10,000 version.^[19] These systems clearly have the potential to be configured as self-deploying, relatively inexpensive, smart sea mines that could be launched hundreds of miles from the target area. Navies are also experimenting with vessels ranging from drones that can be launched from submarines to full size, long-endurance surface vessels.

Logistics

To support units in exposed locations, the U.S. Marine Corps used the K-MAX cargo-hauling heli-drone in Afghanistan.[20] Today, it is experimenting with expendable, wooden gliders that can be dropped from cargo aircraft or helicopters and glide 130 kilometers to precisely deliver a 450 kilogram payload.[21] European truck manufacturers have successfully tested truck “platoons” where a single driver operates the lead truck and up to eleven others driverless truck follow closely.[22] Several shipping companies are testing unmanned commercial ships ranging from simple ferries to large container ships.[23] In short, many logistics missions – particularly high risk ones – will be done by drones.

Strike

The Pentagon is seeking low-cost, high performance unmanned aircraft to accompany manned aircraft into combat. Kratos Defense and Security Systems Inc. has designed the XQ-222 and aims to sell it for about two million dollars a copy. The aircraft features high sub-sonic speed and a 2,500 kilometer combat radius (2 ½ times the F-35) with a five hundred pound payload. It can also be sent on a 5,000 kilometer one way mission. It has low-observable features as well as no requirement for an airfield to launch or recover. It takes off with a rocket assist from a stand and lands using a parachute.[24] But strike is not limited to million dollar systems. The U.S. Army and Marine Corps have been using the $15,000 Switchblade drone as an observation and strike platform in Afghanistan since 2011. ISIS has been using much cheaper models for several years. In early 2017, Iraqi forces reported ISIS’ first use of drones to deliver bombs against Iraqi forces.^[25] In one video, the drone is used to support a complex ambush. It distracted the local Iraqi forces while a suicide car bomber closed in and detonated his vehicle.^[26]

Looming Questions

The convergence of technologies mean the arrival of thousands of drones with limited autonomy on the battlefield is inevitable. That leads us to three key questions. The most basic is “How does a modern society protect itself?” States have invested billions in protecting their facilities against terrorist ground attacks but drones render most of these defenses obsolete. Defenders must not only protect fixed sites but also some types of mobile assets and must do so while not drastically interfering with the civilian communities. Even today governments are struggling to defeat drones. Since 2002, the U.S. Department of Defense has sponsored the annual Black Dart exercise to find ways to defeat drones. Anti-drone technology has taken a number of different paths:

--various kinetic kill technologies have been used to track, identify, and then destroy the target drone.             

--lasers have been used successfully to engage drones and, with continued investment, show great promise but are ineffective under some environmental                     conditions.  

--electromagnetic pulse has been used to burn out a drone’s electronics.

--software attacks have seized control of drone operating systems. 

--electronic jamming has blocked the command signal from the pilot to the drone.

--GPS jamming has misdirected autonomous drones.

In 2015 and 2016, the Russians used electronic jamming to defeat Ukrainian drones to include U.S. provided RQ-11B Raven drones. Ukrainians now assemble their own drones for only $20,000 to $25,000 apiece, often funded from private donations. Because they are newer, they are more jam-resistant than the more pricey Ravens.^[27] This is yet another case of civilian technology being ahead of corresponding military technology. The speed of development is being driven by commercial need.  Each of the underlying technologies appears to have passed the knee of the innovation curve and thus capabilities will expand exponentially. The autonomous drone competition is just beginning.

The presence of thousands of drones on the battlefield leads to two other immediate questions. The first question is “Does the tactical defense become dominant in the short term?”  Between 1863 (U.S. Civil War) and 1917 (World War I), the advent of rifled muskets, rapid fire artillery, and then machine guns allowed the defense to defeat most attacks through the sheer volume of accurate fire. Anything moving above ground with several kilometers of the enemy could quickly be killed. Masses of drones may create similar conditions where anything or person moving above ground within fifty kilometers of an enemy force will be seen and killed. Like the era of trench warfare, the defender will have the protection of a prepared position and, because he will not reveal himself until he fires, will generate less signature for targeting by enemy drones.

The rapidly expanding family of relatively inexpensive but long-range drones leads to the next question. Can swarms of cheap, long-range drones neutralize short-range fighter-bombers? Unrefueled, the F-35 has a combat radius of between 900 and 1100 kilometers depending on the version. A variety of vertical takeoff and landing autonomous drones already have ranges from 1500 to 5000 kilometers. While unrefueled drone ranges continue to increase rapidly, the F-35 and similar fighters would require major redesign to achieve similar ranges. Currently, they require large scale aerial tanker support to gain any additional range. Large numbers of relatively cheap drones will be able to reach the fixed fighter bases while fighter bombers will be hard pressed to reach any of the almost infinite number of launch sites VTOL drones will use. Long-range drones will not hunt advanced fighters in the air but on the ground at their operating bases.

Conclusion

This brief article barely touched the extensive range and incredible speed of advances in drone technology and employment. It is imperative nations explore how such systems may be employed and how they can be defeated. Directed energy – laser and electromagnetic pulse – weapons show great promise but have distinct limitations. Other creative approaches include the use of drones to intercept drones. It is likely to require a robust mix of tactics and technologies to neutralize drones. Militaries must conduct rigorous experiments and wargames to understand the impact the proliferation will have on the modern battlefield. This will be particularly challenging precisely because the evolving family of autonomous drones directly challenges the favored weapons systems of today – manned aircraft, large warships, and heavy armor formations. It will take intellectual honesty as well as significant civilian oversight to insure these new systems are fairly evaluated and optimally employed.

------------------

*Dr. T. X. Hammes is a Distinguished Research Fellow at the US National Defense University.  The views expressed are solely his own and do not reflect the views of the US government.