08 February 2011

Drones, Robots and the Next War

Drones (i.e. unmanned vehicles that are remotely operated) and robots (i.e. computer controlled machines) are just on the verge of radically transforming the way that the United States conducts wars.

Unmanned Aerial Vehicles

MQ-1 Predator

The Pakistani side of the War on Terrorism is being carried out mostly by the CIA via armed drones similar to the MQ-1 Predator above which uses Hellfire missiles similar to those mounted on the U.S. Army's AH-64 Apache helicopter gunships (which are large enough to destroy a tank), and has purported led to the death of thousands of Taliban and terrorist group leaders and operatives, but also to significant numbers of deaths of innocent civilians, particularly family members of targets. A less powerful and lighter version of the guided missile, designed to allow more missiles to be fired with less collateral damage, called the Viper Strike, is also in use.

Drones in advanced stages of development promise to take on far more roles than the relatively primitive drones like the Predator (and several similar drones such as the MQ-5 Hunter) do today.


Some of the new drones, such as the flying wing reconnaisance drone called the RQ-170 shown above, utilize the same stealth technology that makes the planes invisible to radar that is found in the recently retired F-117 Stealth Fighter, the B-2 bomber, the F-22 fighter and the F-35 joint strike fighters which are currently being tested prior to entering military service.

Modifying the Air Force RQ-170 to drop "smart bombs," perhaps in places where there are anti-aircraft weapons that would make the mission unsafe for a manned fighter aircraft and radar systems that would make a non-stealth aircraft vulnerable, would be only a minor upgrade for which proof in principle already exists from the MQ-1 Predator. Indeed, in a simple mission to drop a bomb on a target, it wouldn't even have to be remotely controlled and could fly autonomously to its launch area, drop its bomb and return, without being vulnerable to having its telecommunications link to its human operator severed at a key point in the mission.


An experimental prototype of an unmanned stealth fighter plane intended for use on aircraft carriers and about the same size as existing carrier based F-18s, shown avove, has made its first test flights without a hitch.

All of these drones offer a variety of military advantage over manned warplanes. Weight doesn't have to be devoted to an ejection seat and life support systems for the pilot, and can often be smaller than a similar sized manned aircraft. A drone isn't limited by a pilot's need for food, water, bathroom breaks or the need to get up and stretch. The extra weight can translate into more payload or longer range. They can handle g forces in aerobatic manuevers that would cause a human pilot to pass out or at least be too distracted to be optimally alert. Drones can take a risk and get closer to a target to identify it better or increase its accuracy if needed. If it gets shot down, no American Airman dies, and operators as a result can accumulate their skills despite being sometimes shot down, a classic problem with fighter pilots who rarely have more than a handful of actual battles in their careers. Drone operators don't have to be located anywhere near harm's way, and can work in shifts (many current drone operators for drones in Pakistan do their work from Utah). It is easy to train operators with simulators without causing wear and tear to the real expensive thing. You can wrap a drone in plastic wrap in a warehouse until you need it without having to incur maintenance costs associated with training time until it is needed. The quality control difference between a 1% failure rate and a o.1% failure rate matters far less in a drone than in a manned aircraft and the last little bit of realiability is by far the most expensive part. Drones are cheaper than manned aircraft of comparable capabilities. There is every reason to believe that air refueling will be possible -- this function is already automated to some extent. Automated takeoff from a known airfield shouldn't be a problem either, although landings may require remote control by a drone operator.

In addition to the fact that the aircraft itself would be cheaper, it is much cheaper to train an drone (aka UAV) operator than it is to train a pilot of a manned fighter aircraft.

Today, a manned aircraft pilot completes undergraduate and graduate pilot training programs. The service spends more than $2.6 million to train a fighter pilot. Training for an airlift pilot, relatively speaking, is far less — about $600,000. . . . [The UAV training program starts with the manned pilot training program] “eliminating 20 unnecessary weeks of the current undergraduate pilot training program, deleting unnecessary graduate training on other aircraft,” and adding an eight-week UAV undergraduate course and 12 weeks of UAV graduate training. The cost . . . a little more than $135,000 per [UAV] pilot. . . .

Predator and Reaper flight hours have jumped in the past three years. Predator flights nearly doubled last year, according to the Air Force. UAV orbits over Iraq and Afghanistan — or around-the-clock combat air patrols — have gone from 11 in 2007 to 33 in 2009. Plans call for 50 CAPs by 2011.

Air Force leaders had to temporarily assign manned aircraft pilots to handle the higher number of flights. In 2008, 161 fighter, bomber, tanker and cargo pilots flew Predators and Reapers, the two UAV models. . . . “Developing a UAS-specific career path and eliminating unnecessary training will reduce costs by more than $1.5 billion with aviation fuel savings alone accounting for about $180 million."

There are missions for which remote control, rather than autopilot use of drones will be required (although even then, the planes can travel on autopilot to the immediate vincinity where they will carry out a combat mission). Pursuing moving targets on the ground, missions where it is necessary to look at the target before firing to confirm that it is the intended target, and air to air combat (i.e. dog fights), all involve this kind of situation.

While there are many possible missions of aircraft, reconnaissance and bomb dropping (sometimes in "strike" missions where there is not control of the airspace, and sometimes not) that drones are well suited to carry out have been the dominant missions for U.S. fighter and bomber aircraft. You can list every instance of air to air combat involve U.S. forces since Vietnam in a short blog post, and every instance of air to air combat in the world since then in a long one. With opponents who lack the ability to jam GPS signals and remote control telemetry, there is no reason that drones can't be effective in anti-tank and anti-vehicle warfare -- the missiles mounted on Predator drones were invented for that very purpose.

It may be necessary to have some manned fighter planes and bombers as a hedge against opponents who can jam GPS singles and remote control telemetry, or out of concerns that drone software could be hacked, or to deliver VIP passengers. But, drones seem to be well on track to make it unnecessary to have a manned successor to the F-22 or the F-35. Indeed, an operational version of the X-47 shown in this post may be possible within a few years of the time that the carrier based F-35C variant of the joint strike fighter enters naval service (and the F-35B variant has been put at the back of the line and may enter service around the same time as an operation version of the X-47). There is essentially nothing that a pilot does in a fighter aircraft in the kind of military environment that the United States military has ever encountered in the history of the fighter and bomber aircraft that couldn't be done by a remote operator of a military drone aircraft.

A notable corollary of this development is that it would allow for much lower impact foreign airbases. A drone base could have a skeleton crew to serve as security guards, maintain the fitness of the runway, the hangar/warehouse, the tools, the ammunition and supply stores, and a vacant dorm. The operators would at all times stay safe in the United States. The drones would sit silent in their hangar until actually needed for operations, avoiding the noise and activity that makes military air bases bad neighbors. In the event that tensions were rising in the region of the foreign base, and it looked like it might be necessary to deploy drones from a foreign airbase, drone maintenance technicians could be assembled at a U.S. base and flow in on a day or two's notice to maintain the drones while they were operating. Operates could train on simulators, and from time to time with actual drones, from the safety of U.S. bases.

If a few lukewarm U.S. allies were agreeable to having these kinds of lower impact drone bases on their territory, the distance of any plausible target in the world from the nearest U.S. base would be greatly reduced, making fast response times to threats possible without hypersonic aircraft speeds and the attendant sonic booms.

The drone concept isn't limited to replacing fighters and bombers.


Reconaissance missions are the main purposes of drones today, and allow U.S. forces to have a bird's eye view of the battlefield that is on the verge of becoming much more comprehensive (follow up link here). Current drones can look at one small spot at a time directly below it; camera systems in development that would be transported by existing drones could look at thirty points at a time in a five kilometer radius circle.

Already, the U.S. has a variety of means to put explosives of almost any desired capacity at any point in the world with extreme accuracy and drone aircraft will only improve this capability - call this omnipotence. The U.S. already basically has the capacity to see any particular point in the world that it wants and this capability is being greatly expanded with drones - call this omnipresence. The next step is to be able to process the incoming flood of data well enough to make sense of what is coming in so that what is seen everywhere can be dealt with as if the U.S. forces were actually present everywhere it can observe something - call this omniscience. But, since the data collection part of this process is something that already exists in the hear and now or is ready to enter operational service in the very near future, the real bottleneck at this point is the data analysis.

However, increased data analysis capacity, either through the brute force of assigning more people to watch the feeds coming in, or through computer image processing, or through a combination of the two, is essentially invisible in war zones. The data analysis can happen can be done by people comfortable offices in the United States and its autonomous image processing capacity can be enhanced with invisible to the user computer software and hardware in some anonymous military server farm.


A Berlin airlift type operation is just a bomb run with a different payload.

Cargo Transports

While military leaders might appropriate be concerned about using drones to transport people, at least apart from high pilot risk situations like medical evacuations in the middle of a fire fight, there is no reason why a UAV wouldn't be appropriate as a cargo plane. Indeed, it would be relatively easy, once the flight software is developed for a drone fighter aircraft, to retrofit existing cargo planes so that they could be operated remotely.

Staffing Impact

It takes about twenty active duty military personnel to keep an aircraft operating in the existing Air Force. And, most of those personnel, the maintenance crews, for example, would still be needed. But, the number of manned aircraft that need pilots could be reduced dramatically in a matter of two or three decades, and maintenace crews would be needed only where drones are being used for actual operations - not for the most part, for training duties.

Unmanned Ground Vehicles

The Black Knight (vehicle)

The drone trend isn't restricted to aircraft. For example, the U.S. Army is testing and evaluating the Black Knight, a prototype light tank that is remotely controlled, not so differently from the remote controlled vehicles you can buy for your kids at Target. It can go as fast as the Army's existing tanks over similar terrain, it has a 30mm cannon and a machine gun, and it can be carried in a C-130 cargo aircraft, unlike existing heavy military hardware like the M1A tank and the M2 Bradley Infantry Fighting Vehicle.

A smaller ground drone called the Gladiator, that has been in military service for three years, has a light machine gun and tracks and weighs three-quarters of a ton (think armed, remote controlled riding lawn mowner on steroids). Obviously, this isn't very useful for chatting up locals to determine who is siding with insurgents. But, it provides the kind of firepower that would otherwise require a soldier without the need to hide from income fire or be unduly concerned about IEDs.

Remote controlled ground drones are more prone to getting stuck than drone aircraft and have to contend with the messiness of surface terrain, but are near future possibilities with off the shelf technology for just about any military mission that doesn't involve interacting with human beings. More humble remotely operated vehicles could deliver supplies or have forklifts to move loads around warehouses.

Army efforts to develop unmanned artillery batteries, the Non-Line of Sight missile system component of the "Future Combat System" that has been cancelled after it failed to meet performance test, went over budget and was behind schedule are sidelined for now. But, since the problems were almost entirely a matter of electronic guidance systems and software, rather than the rest of the missile system, it is only a matter of time before the technology can catch up to the mission and be reintroduced into the Army's military arsenal.

The superfast reflexes of robots and computers have already been harnessed in a modified version of the Navy's "Phalanx Close In Weapons System" for Army use that uses bullets shoot incoming rockets and artillery shells identified by its sensors before an operator even knows what is happening. Systems that use lasers rather than bullets to shoot incoming ordiance have been developed and tested, but are so far inferior to the current system which shoots big bullets with little bullets.

An interesting modification of Phalanx that recently entered service with the US Army is the Counter-Rocket Artillery Mortar system (C-RAM), which is designed to combat the small mortar and rocket threat. This system grew out of a 2004 request from Army Chief of Staff General Peter Schoomaker, who wanted a better means of defending US troops based in Iraq.

The complete C-RAM system networks a ground-based version of Phalanx together with the Army's Lightweight Counter Mortar Radar (LCMR) and Q-36 Target Acquisition Radar (AN/TPQ-36 Firefinder Radar), which detects incoming rounds and determines their point of origin. When C-RAM detects an incoming round, it turns on a set of strobe lights to alert local personnel to take cover, authorizes the modified Phalanx to open fire with explosive bullets to destroy the projectile and dispatches a Hunter UAV equipped with Viper Strike laser-designated munitions to kill whoever fired it. Raytheon is currently (2007) offering a trailer-mounted version of Phalanx as part of C-RAM under the product name of Centurion™. Unlike the naval version, the C-RAM version of Phalanx does not fire solid tungsten penetrators. Instead, in order to reduce the risk of civilian and friendly casualties, this system uses self-destructing explosive bullets.

The first two C-RAM systems arrived in Iraq in May, 2005 and have been reported as having successfully knocked down incoming mortar rounds.

The Centurion unloaded from a cargo aircraft

As in the case of air drones, ground drones aren't very useful as a tool to dispatch people in a war zone. But, for combat and transport roles, particular in asymmetric conflicts (or even "near peer" conflicts) where electronic means of interfering with these military systems aren't available to opposing forces, they have the potentially to replace a large share of the systems current military systems.

The need to have an operator present with a military vehicle or weapon when it is being used in high threat environments could be greatly diminished in the time frame before my grandchildren are of military service age.


Suffice it to say that similar unmanned system concepts apply to surface and submarine forces.

A New Era In Conventional Warfare

The net effect of all of this is to make possible a world in which conventional warfare with large military systems of the kind that has dominated every full scale war from World War I to the early days of the Iraq War can be carried out on behalf of technologically advanced countries like the United States mostly by drones and robots that are extremely accurate and capable of managing in conditions unsuitable for soldiers with endurance and patience.

In conflicts with countries that don't have this kind of technology, or advanced electronics that can counter their control systems, heavy weapons systems from tanks to armored personnel carriers to military aircraft to artillery batteries to warships may become all but useless.

Tank squadrons and non-stealth manned fighter wings and surface navy fleets may become as obsolete in conflicts with technologically advanced militaries as the "light brigade" of horse riding cavalry was in the face of machine guns.

The main limitation on the use of military force will be not the capacity to deliver explosive force precisely when and where the military wants to do so, overcoming all static armor systems, but the ability to distinguish rapidly enough between friends, foes and neutrals.

While military planners have portrayed conflicts in Afghanistan and Iraq, where heavy conventional military opposition was utterly destroyed in days or weeks, followed by long mopping up counterinsurgency campaigns conducted mostly by infantry with low levels of heavier military equipment providing support, this may become the norm in future wars, at least those conducted against advanced military powers by nations that do not have comparable military technology.

Spy v. Spy: The Red Queen Effect

Of course, the other problem is that proliferation of this kind of technology is much harder to control than proliferation of ballistic missiles, nuclear weapons, or navy fleets, for example. The computing power and electronic components necessary to make servicable versions of these kinds of weapons systems are available on a commercial, off the shelf basis. This is the kind of stuff that a bright high school student or undergraduate engineering student can devise with a six figure budget. The copies may not be quite as capable as the original, and may come on line a few years later, but they aren't that far behind.

A terrorist cell may not be able to field a stealth bomber drone, but they may very well be able to put together an ugly hybrid of a radio controlled airplane, a cell phone with a videocamera, and a handgun that can be used as a remote controlled assassination weapon, and a full fledged terrorist organization might even be able to mass produce something like that in some lonely industrial park in Yemen or Afghanistan or Pakistan or Compton, a smuggle crates of them (perhaps with a legal radio controlled aircraft component and an illegal weaponizing upgrade kit distributed separately) all over the world. I can even imagine some U.S. court finding that such a device can be lawfully mass produced and distributed in the United States as a Second Amendment protected right, given that it is made up of components that are protected by the Second Amendment or are lawful, and given free speech rights to discuss technology, even military technology, that is not esoteric.

Similarly, a low budget "cruise missile" with a rocket propelled grenade's explosive as a warhead could transform a somewhat heavier than average small arm into a medium range guided artillery weapon.

A single dud Excaliber artillery round or smart bomb recovered from a battlefield, as the Chinese are believed to have recovered parts from an F-117 stealth fighter shot down in Balkans, could allow a decent engineer to reverse engineer most of the core concepts in the weapon's guidance system while attracting far less notice than an attempt to smuggle nuclear materials to a rogue nation might.

A country like North Korea which has managed with get noticed with very small numbers of very advanced military systems like nuclear bombs, medium range missiles and submarines probably doesn't have the industrial base to do more than the equivalent of throwing a one time punch that gives some other nation a bloody nose. Maybe it could destroy one or two medium sized cities in China or Japan or even the United States or Canada. But, it would then be hopelessly outmatched and obliterated. If its next leader is as crazy as its current one, it might even try to make that happen.

But, a somewhat less isolated and backward country with resticted access to the international arms markets, like Iran or [North] Sudan, could devote the resources necessary to develop a pretty credible military-industrial complex to develop drone and robot weapons systems. The examples of South Africa and Israel illustrates the fact that a nation doesn't have to be very large or have that huge of a GDP to have a quite credible defense contracting industry, making not quite state of the art weapons systems and military vehicles. Indeed, perhaps the best justification for allowing the arms sales of Russia and the United States to foreign nations of questionable character to continue is that this discourages those countries from developing the kind of industrial base necessary to make them domestically.

Exports from these unfriendly arms makers, in turn, could spread worldwide to all manner of regimes that might be quite interested in launching attacks on the United States or its interests or allies remotely. China, even if it backs down from hostile direct action against its neighbors, could prove to be more of an arms proliferation threat to the United State than it is a direct military threat.

Our progress in figuring out how to defend our nation from drone and robotic weapons systems is far less advanced than our capacity to make them. But, simply refraining from developing our own drones and robotic weapons systems won't be enough to prevent other nations from doing so with the technological foundations that are available to all.

Ultimately, the technologies inherent in drone and guided weapons may make the need to invade other nations that don't control access to that kind of military technology the norm, rather than exception that the Afghan war seemed to involve.


Dave Barnes said...

"The service spends more than $2.6 million to train a fighter pilot."

Has to wrong by at least an order of magnitude if you honest about the accounting.
$26M and up.

The unit cost of a F-22 is $150M. So, if only 1% of training aircraft are lost during training, then that adds about $1.5M to the average pilot training cost.

Wear and tear on training aircraft.
Support crew.

Andrew Oh-Willeke said...

FWIW, most training takes place with airplanes much cheaper than the F-22. Then again, at least one F-22 has crashed on what appears to have been a non-combat mission (given that it happened in Alaska) so far.