How to Lose the Robotics Revolution

July 29, 2014

For special access to experts and other members of the national security community, check out the new War on the Rocks membership.

The U.S. military is at the leading edge of the robotics revolution, with some of the most advanced systems on the globe like the autonomous X-47B carrier-based aircraft. But that lead is fragile and other nations are racing to catch up. Scholars warn of a “looming robotics gap,” driven in part by the explosion in widely accessible commercial robotics, some of which will have dual-use military purposes.

The biggest threat to the U.S. military’s lead in unmanned systems isn’t commercial sector innovation or even declining defense resources, however, but rather hidebound cultures and entrenched bureaucracies within the Department of Defense (DoD) itself. Unmanned systems have been embraced for niche missions like reconnaissance, bomb disposal, or cargo resupply, but resistance persists to their use for many military tasks and missions, even when use of force is controlled by a person “in the loop.”

While numerous DoD vision or roadmap documents spell out the advantages of unmanned systems, with few exceptions these visions are not funded. Only one out of every 20 research, development, and procurement dollars are spent on unmanned systems. As a result, the U.S. military risks falling behind in a critical emerging area.

Unmanned systems, like any new program in the military, must fight an uphill battle for funding against existing programs. Culture, however, also plays a major role in how unmanned systems are perceived by various communities within the military services. Technologies do not exist in a vacuum, and the relationship between an emergent technology and its military user is often far more important than the technology itself. Within the U.S. military, cultural views on which tasks are appropriate for unmanned and autonomous systems is leading to game-changing innovations being ignored or, in some cases, resisted.

Nowhere is the role of culture in shaping how unmanned systems are used more apparent than in the differences between how the Army and the Air Force use their unmanned aircraft. To a lay person, the Army MQ-1C Gray Eagle and the Air Force MQ-1 Predator are virtually indistinguishable. The underlying technology behind them is the same. But they are used by their respective services very differently.

Air Force Predators are flown by officers and Army Gray Eagles by enlisted personnel. Gray Eagle operators are not really “pilots.” The platform’s takeoff and landing are automated, and in the air it is controlled by a human operator who directs the aircraft where to go from a console. Air Force Predators, on the other hand, are flown by a pilot—in a flight suit, with a joystick, and sitting in a mock “cockpit” on the ground. These differences translate to extra costs for Air Force Predators—not insignificant in today’s budget environment—but other differences have very real operational costs.

Air Force Predators are flown remotely from the United States, with only a minimal crew forward deployed for launch and recovery. Army Gray Eagle operators, meanwhile, deploy forward to fly the aircraft from theater. As a result, the Army needs roughly two additional operators stateside for every one deployed so that it can rotate them through theater, a hefty personnel bill in today’s end-strength constrained Army. A 2011 internal Army study, directed by the Office of the Secretary of Defense, validated the operational efficiencies that are a function of remote operations. Additional surveillance coverage could be delivered at very low cost by adopting a hybrid approach, where the Army continues to deploy soldiers but also operates additional unmanned aircraft remotely with soldiers between deployments. But the Army isn’t doing it.

From the Army’s perspective, soldiers deploy to war. That’s what soldiers do. Pride, honor, and identity are on the line, and these are far more powerful forces than studies promising increased efficiencies. A hybrid approach wouldn’t break this paradigm and would allow Gray Eagle operators to continue deploying to war. It would also increase Gray Eagle coverage downrange at very low cost by using aircraft and personnel the Army already owns. But the Army isn’t doing that, perhaps seeing it as a slippery slope to full remote operations. Often unstated, but underlying the Army’s objections is a fear that if the Army were to emulate the Air Force’s operational practices, its unmanned aircraft would be transferred to the Air Force to manage. Given the Air Force’s recent cuts to its Predator and Reaper fleet—tools that have been tremendously valuable to ground forces—the Army’s fears that its unmanned aircraft wouldn’t be safe in the hands of the Air Force are not unwarranted. Culture can be a driver in shaping how technology is adopted, but so can bureaucratic turf wars.

Similarly, while the Air Force embraces remote operations, cultural hurdles hold back Air Force operations in other ways. Multi-aircraft control technology, which allows one person to control multiple aircraft at the same time, has the potential to yield significant cost savings and increase operational capacity delivered to the battlefield. Yet the Air Force is resistant to multi-aircraft control, with some seeing such aircraft as “out of control.” Simple multi-aircraft control concepts have been used in a limited fashion in real-world operations, but early experiments led to frustration with human-machine interfaces and task loading. Rather than improve these interfaces, however, the Air Force deemed developing better multi-aircraft control an “unfunded requirement.” In 2010, Secretary of Defense Bob Gates directed nearly $50 million to develop improved multi-aircraft control, but the Air Force has failed to comply with this guidance. The service’s recently released Remotely Piloted Aircraft (RPA) Flight Vector discusses multi-aircraft control, but the Air Force has no funded plan for developing the technology to achieve this vision. Informally, Air Force officials admit that multi-aircraft control is a low priority for investment—a “decade after next” investment.

While for Army officers deploying to war is central to their identity, for Air Force leaders flying planes is what defines a pilot. The most senior Air Force position has always been occupied by a pilot, and in fact only by a non-fighter or bomber pilot twice. The importance of this identity label explains why the Air Force has gone to such great lengths to attempt to rebrand unmanned aircraft as “remotely piloted,” and why automation that takes control out of the hands of pilots has been resisted.

For the Army, on the other hand, piloting is not central to their identity, which is why even though the Army resists remote operations, it has embraced multi-aircraft control. The Army’s new ground control station, scheduled to come online next year, will allow one operator the ability to control two aircraft at the same time, which is possible because of the high degree of automation in Army Gray Eagles.

Using officers instead of enlisted personnel and flying unmanned aircraft by stick and rudder instead of a keyboard and mouse can lead to waste and inefficiencies, but both the Air Force’s cultural resistance to multi-aircraft control and the Army’s to hybrid remote operations have real operational impact. By not using its non-deployed Gray Eagle operators and aircraft to fly additional missions remotely from the states, the Army is missing out on an opportunity to supplement forces in the field with additional support. Given the severe shortfalls in overhead surveillance, the fact that this resistance exists for purely bureaucratic and cultural reasons is troubling. Meanwhile, the Air Force’s resistance to multi-aircraft control means that it is not marching down the path towards true swarming in the future, a breakthrough capability where one person controls large numbers of unmanned aircraft.

Culture and bureaucracy can either be powerful drivers of innovation, or they can hinder it. The differing approaches of the Army and Air Force to unmanned aircraft show that the question of their adoption isn’t as simple as them being embraced or not within an organization. In both the Army and Air Force, unmanned aircraft are embraced for some missions and roles, but not for others.

Still, it is hard not to feel the shadow of Billy Mitchell looming in the background when one considers Army and Air Force resistance to some aspects of unmanned systems. Mitchell’s early and vociferous promotion of airpower in the 1920s challenged power structures within the Navy and Army (the Air Force did not exist at the time). The Army’s response—demotion and court-martial—shows that even institutions and people dedicated to the nation’s defense can nevertheless be blinded by cultural prejudice. No one is being court-martialed for advocating for unmanned systems today, but some of those blind spots still exist.

The Air Force’s recently released RPA Flight Vector carves out an ambitious vision for the future of unmanned aircraft, but the lack of funding to achieve that vision show where unmanned aircraft actually fall on the Air Force’s priorities. The mere fact that the Air Force insists on referring to unmanned aircraft as “remotely piloted” suggests that increased automation, multi-aircraft control, and swarming are still conceptually a bridge too far for many in the Air Force. It is telling that the author of the recent Air and Space Power Journal article advocating for development of a future unmanned air-to-air fighter, dubbed the FQ-X, is an MQ-9 Reaper instructor, not a fighter pilot.

For the Army, of course, unmanned aircraft are less threatening because the core of the Army’s identity doesn’t lie in aircraft but in ground fighting. But there are analogies in core Army missions. Army leaders have recently made strong statements in favor of robotic systems, but outside of small ground robots like the Packbot or larger tele-robots for bomb disposal, funding is lacking. Conceptually, Army personnel often talk about unmanned vehicles for cargo resupply or other support missions, but similar to the Air Force, rarely for core Army missions like ground combat and maneuver. Even more troubling are areas where unmanned systems are not held back merely by an implicit de-prioritization for funding, but by explicit policy prohibitions, as in the case of the Army’s prohibition against using unmanned vehicles for casualty evacuation.

Historical examples abound of resistance in some military circles to technologies that fundamentally changed warfare in a certain domain. Elements of the Navy initially resisted the transition from sail to steam-powered ships. Parts of the Army—the cavalry—resisted the transition from horses to tanks. And elements of the Army fought bitterly against the adoption of airpower, with the court martial of Billy Mitchell an extreme example.

Militaries are often caricatured as resistant to innovation as a whole, but such a view is not supported by a comprehensive historical review of military innovation, nor by the nascent experience of the U.S military with unmanned systems. The reality is that unmanned systems have been embraced wholesale for missions like reconnaissance or bomb disposal, but not for others like casualty evacuation or air-to-air combat. Even the common (and shortsighted) refrain that unmanned systems are useful for “dull, dirty, or dangerous” missions implies a useful, but limited role.

Why are unmanned systems embraced for some missions and not others? Or more generally, why are some innovations, like airpower, staunchly resisted while others are not?

Identity is a common theme. When the Navy transitioned from sail to steam-powered ships, what was at stake was not merely a new technology but in fact the very essence of what it meant to be a sailor—no longer climbing the mast and working the rigging, but working down in an engine room as an engineer. Similarly, for cavalrymen the horse was not merely a means of getting to and joining battle but a part of their identity, a legacy that lives on in the Stetsons and spurs of Army “cavalry” units today. And the advent of airpower raised fundamental challenges to the primacy of ground units in warfare, sparking a debate that continues to this day.

Recent psychological studies have demonstrated that otherwise brilliant people are capable of ignoring information that, if adopted and incorporated into their worldview, would challenge their core identity. This problem is not unique to militaries; it is a function of the human condition and arises when incorporating such information might put a person at odds with their core identity group. The researchers explain that so-called “identity-protective cognition” is a “psychic self-defense mechanism that steers individuals away from beliefs that could alienate them from others on whose support they depend in myriad domains of everyday life.” This human propensity to ignore information that is challenging to one’s identity group should be particularly troubling for militaries, where identity is closely aligned not only with a particular service (Army, Air Force, Navy, Marine Corps), but also the branch or specialty within that service (aviation, submarine, surface warfare, artillery, armor, infantry, etc.). Viewed in this light, it would be surprising to see officers embracing new technologies that undermined their occupational skill, even if they were more effective in combat.

And yet, the history of warfare is full of abrupt changes in warfare, and littered with the carcasses of militaries that failed to adapt and lost battles or even wars as a result. The DoD’s recent Quadrennial Defense Review trumpets innovation as a core theme, but many innovative approaches for adapting to future threats are not funded. DoD has chosen to prioritize modernization, but modernizing within existing paradigms for warfighting will not be sufficient to sustain the U.S. military’s advantage into the middle of the 21st Century. Disruptive technologies not only force new countermeasures, but fundamentally change the rules of the game. Already, states and non-state actors are using unmanned systems in novel ways that are different from how the U.S. military uses them.

In many ways, the U.S. military’s dominance in traditional warfighting is a liability in terms of adapting to disruptive change, since bureaucratic interests within the U.S. defense establishment are heavily invested in a particular way of fighting and may be resistant to alternative approaches. Adapting to these approaches will require experimentation, incremental technology development, and the willingness to take risks. This includes funding, but also support for culturally-challenging concepts like unmanned casualty evacuation, multi-aircraft control, or remote operations. When these concepts challenge core identities within a service, support from senior leaders in the Pentagon, the White House, and Congress will be required.

The Center for a New American Security’s recent report, “Robotics on the Battlefield – Part I: Range, Persistence and Daring,” outlines specific recommendations for action for the DoD to position itself to remain dominant in the face of disruptive technological change. The robotics revolution is coming and will enable new concepts and capabilities. This revolution will not be won simply by developing robotic technology first or even by developing the best technology, but by figuring out the best ways of using it.


Paul Scharre is a fellow and Director of the 20YY Warfare Initiative at the Center for a New American Security. From 2008-2013 he worked in the Office of the Secretary of Defense on policies for unmanned and autonomous systems. He is a former infantryman in the 75th Ranger Regiment and has served multiple tours in Iraq and Afghanistan.

We have retired our comments section, but if you want to talk to other members of the natsec community about War on the Rocks articles, the War Hall is the place for you. Check out our membership at!

10 thoughts on “How to Lose the Robotics Revolution

  1. Mr Scharre makes some excellent points with respect to the clash of cultural identity with emerging technology. While true for every society, Americans have demonstrated time and again a willingness to incorporate new technology, embrace it, and make it a part of their culture faster than most other nations. Yes, we still resist change, but I’ll bet we won’t be losing the Robotics Revolution against more-entrenched cultures (e.g., Chinese, Arab, Persian, & etc.), at least for the reasons stated here, within the foreseeable future.

  2. I am curious if the author would like to detail his reasons for proposing an unmanned ground MEV for the Army. I honestly can’t think of any other than it would reduce battlefield risks for medical peronnel, but it is an interesting proposition.

  3. These “scholars” who warn of a “robotics gap” – are they not tuned in to the historical irony of using the phrase “gap” with regard to weapons system? I refer to the autumn of 1960 when JFK made a lot of noise about a “missile gap.” Robert McNamara settles into office in January of ’61 and one of his early memos to the Oval Office mentions that there’s no such thing as a missile gap. Just another big lie in the interest of advancing an agenda.

  4. To know tactics, know technology – Fleet Tactics, CAPT W.P. Hughes

    “Viewed in this light, it would be surprising to see officers embracing new technologies that undermined their occupational skill, even if they were more effective in combat.” Were the author’s statement true, we would indeed still regularly be sailing ships or riding horses into battle. While some SF members are riding mules in Kunar, there are few sailing vessels in the naval register outside Charlestown. Rather, American naval history demonstrates that once innovative tactics or technology have demonstrated their military utility, they are often readily adopted, while eliminating the obsolete, albeit with a measured and studied approach.

    The author cites the transition from sail to steam as resistance to innovation. H.I. Chapelle chronicled in “History of the American Sailing Navy” that the commercial steamer was “becoming common in the [18]50’s and in this the Americans were intensely interested.” Further, the U.S. Navy’s transition to steam was fully completed within 15 years. How could this change be possible without the support of the Captains and Admirals of the naval bureaucracy?

    Vice Admiral A.W. Johnson, commander of the 1921 Naval Bombing Experiments (the event that Billy Mitchell used as a publicity stunt) recalled “It was Admiral Sir Percy Scott, the great gunnery expert of the Royal Navy who asserted that the new weapons [aircraft] made battleships obsolete. It was Admiral of the Fleet, Sir John Fisher, First Lord of the Admiralty, father of the Dreadnought type of Battleship, who said that the surface battleship would be superseded by a submersible battleship. It was Admiral Scheer, Commander of the German High Seas Fleet in the Battle of Jutland, who asserted that battleships were obsolete. Like General Mitchell, they were more of less right, but when did battleships really become obsolete?” Thusly, leading experts in their warfare area recognized the limitations of their own platforms and advocated for adopting new technology.

    There were no organic aviation admirals until the late 1930’s; The Bureau of Aeronautics was led by ADM Moffett, a Battleship captain, and later ADM King, a surface and submarine officer, and later CNO. How was naval aviation able to innovate so well during the interwar years? Why would the Navy Department fight so hard to resist GEN Mitchell’s single air service proposal if they feared or ridiculed such new technology?

    Today, the unmanned X-47B program is championed and managed by naval aviators. The current debate is about the nature of the first generation of technology and how to best integrate it into airwing and fleet operations, not that an unmanned aircraft shouldn’t exist.

    As counterintuitive as it may seem, successful innovation is enabled by the establishment, not the iconoclasts. We need to take the most promising of the new tactics and technology, integrate the best of today’s, and eliminate the obsolete.

  5. I’m terribly sorry, but I believe that this is quite frankly an incredible piece of over thinking, without understanding. Why do U.S. Air Force drone pilots remain in the U.S. while U.S. army drone pilots remain on the battlefield? It has nothing to do with culture, but mission, jurisdiction, and utility.

    U.S. Airforce drones are multimillion dollar pieces of weaponized equipment, requiring a greater responsibility on the user end, and are large enough, and posses the flight range of many modern piloted aircraft. Airforce drones are almost never “launched” from within the theater which they opporate, but from a nearby secured facility in another country. If the drone (i.e. aircraft) is based out of a non-combat location, it would be incredibly difficult to make the argument the pilot should be located in a combat environment to pilot it. Simplistic reasoningI understand. But let’s take a look at the primary U.S. Army drones, and their use.

    U.S. Army (and USMC) drones are launched by the end user, on the ground, in sheer numbers, the primary drones are no larger than an RC model plane. The “operator,” normally held as a secondary duty, on top of being an infantryman, etc. must launch the drone by arm strength alone, and can only fly it’s electric motor as far as the battery, and RC signal can reach. Therefore, it would be rediculous to even consider an alternative for the operator to be located somewhere else.

    U.S. Airforce Drones — primarily strategic level intelligence gathering, observing entire regions on a single flight.
    U.S. Army Drones — primarily tactical intelligence gathering, observing an open field, or the perimiter of a 100 man outpost, or surveiling a building occupied by enemy forces prior to a platoon of Soldiers entering.

    As an active duty infantry officer, my hand thrown drones proved invaluable in certain situations, just as, likewise, my RC electric motorized ground robots did, when the mission dictated their use.

    Take this article as you wish, but don’t over think it so much that you separate the concepts from practical reality. The entire military adopts and uses technology within an ordered structure. Many times, on the ground, I have up linked to a live U.S. airforce drone feed. We are not so compartmentalized by designators of Army or Airforce, that we do not share real time information. Nor are any higher brass concerned with their personal identity above accomplishing the mission.

    1. Please note the author’s entire premise relies on the indistinguishable difference between Army and Airforce drones, to the layman. Is it the perspective of the layman that should shape our military, and the use and functions of technology? No.

  6. Jack needs to do a little research. The US Army Grey Eagle is the same size as the USAF Grey Eagle. In fact, they are both made by General Atomics and use the same Ground Control Station (GCS) The biggest difference is the Army uses a lot of automation to launch and recover aircraft (They also suffer fewer L/R losses by the way).

    I believe the author is correct that the US military sees the use of RPVs as very limited and not a good career path. I do not believe a USAF RPV pilot will ever be promoted to Commanding Officer of an ACC. For the enlisted Sensor Pilot, it is another story.

  7. Excellent article! But I’m surprised the author, a former Ranger, didn’t highlight the obvious reason the Army deploys the enlisted “pilot” to a combat zone. Yes, they value deployment as a defining characteristic of what the Army is, but they also award a very visible status symbol for the combat deployment – a right shoulder unit patch, which tells the world you were there, and did that, and what team you are on. No deployment, no right shoulder patch, no status as a warrior.