Human Machine Teaming FinalFormat
Human Machine Teaming FinalFormat
org 3
HUMAN-MACHINE TEAMING
FOR FUTURE GROUND FORCES
MICK RYAN
HUMAN-MACHINE TEAMING
FOR FUTURE GROUND
FORCES
MICK RYAN
2018
ABOUT THE CENTER FOR STRATEGIC AND
BUDGETARY ASSESSMENTS (CSBA)
The Center for Strategic and Budgetary Assessments is an independent, nonpartisan policy
research institute established to promote innovative thinking and debate about national security
strategy and investment options. CSBA’s analysis focuses on key questions related to existing and
emerging threats to U.S. national security, and its goal is to enable policymakers to make informed
decisions on matters of strategy, security policy, and resource allocation.
Mick Ryan is a Major General in the Australian Army. A graduate of Johns Hopkins University,
the USMC Staff College, and the USMC School of Advanced Warfare, he is a passionate advocate of
professional military education, global PME discourse, and lifelong learning. In January 2018 he
assumed command of the Australian Defense College in Canberra.
ACKNOWLEDGEMENTS
The author would like to thank Dr. Thomas G. Mahnken and the CSBA staff for their assistance in
publishing this report. In particular, thanks go to Rick Russo from CSBA for his significant
support, advice, and input—and sci-fi prowess. Special thanks go to Prof. Mike Evans, Prof. Mike
Horowitz, Dr. Deane-Peter Baker, Dr. Therese Keane, Lieutenant Colonel Tom McDermott, and
my wife Jocelyn for their comment and critique on various drafts of the report. Finally, the author
wishes to acknowledge the great mentorship, advice, and input into the report of Dr. Ross Babbage
and the Chief of the Australian Army, Lieutenant General Angus Campbell. The analysis and
findings presented here are solely the responsibility of the author.
CSBA receives funding from a broad and diverse group of contributors, including private
foundations, government agencies, and corporations. A complete list of these organizations can be
found on our website at www.csbaonline.org/about/contributors.
Cover Graphic:
U.S. Army Pacific Soldiers, 25th Infantry Division, move in formation while controlling unmanned
vehicles as part of the Pacific Manned-Unmanned Initiative at Marine Corps Training Area
Bellows on July 22, 2016. The photo is by Staff Sgt. Christopher Hubenthal.
©2018 Center for Strategic and Budgetary Assessments. All rights reserved.
Contents
THE NEAR FUTURE ........................................................................................... 1
INTRODUCTION ................................................................................................ 3
THE IMPERATIVE:
HUMAN-MACHINE TEAMING ON AND BEYOND THE BATTLEFIELD .............. 7
Strategic Drivers for Change in Ground Forces ............................................................................... 8
Implications for Future Ground Forces .......................................................................................... 11
Closing Thoughts............................................................................................................................ 15
CONCLUSION .................................................................................................. 51
ACRONYMS ..................................................................................................... 57
Figures
FIGURE 1: AUTONOMOUS TRUCK CONVOY 5
It wasn’t all good news. “Merlin thinks the ratio of friendly to enemy forces looks about right.
That’s the good news. But it looks like we’ll need to execute the urban breach and initial phases
earlier. Also, we’ll need to use more directed energy emitters and about twice the precision
guided munitions forward in the initial push.”
Bryant wasn’t looking forward to leading with the relatively delicate self-propelled directed
energy emitters. He was reassured though by the recommendation for more of the precision
guided munitions—they’d offset the risk of the temperamental directed energy emitters.
“Make sure it’s synchronized with the cyber, air swarm, and subterranean operations teams.
While you’re at it, grab the communications officer and find out where we are with the
cryptography update for all of our unmanned ground vehicles.”
Kinsley looked at his eye-activated Heads Up Display and visually toggled through the menus on
the orders board. It moved as fast as Kinsley could see and click on his hand set.
“Mmmm; Not done yet, sir. But looks like less than an hour.”
2 CSBA | HUMAN-MACHINE TEAMING FOR FUTURE GROUND FORCES
The security update for the Unmanned Ground Vehicle fleet was a critical part of the plan for
tomorrow morning’s operation to move into the fortified zone of the city. The fleet of unmanned
ground vehicles would provide a massed surveillance, reconnaissance, retransmission, and
engagement capability for their operations. Although generally reliable, they had been
susceptible to enemy interference in the past. Daily security updates had become a standard
procedure. Bryant handed Kinsley back the tablet.
“OK, confirm the amendments Merlin recommended and then publish the final order through to
the battle group. Virtual rehearsal starts in 45 minutes. Thanks, Steve.” Bryant twisted to exit the
rear of the vehicle. And then paused.
“Oh, and tell the team in the forward repair group that I want all of our exoskeleton battle suits
available this time. If I must, I’ll go to the boss for extra log-drones to fly them out here.”
Emerging from the armored command vehicle with low visibility camouflage, Bryant
approached the two kneeling figures positioned just outside the hatch.
“So, we go at zero three hundred. What do you have for me?” He listened as the Human
Intelligence and Social Media team leaders updated him on likely population and community
leader responses to his move into the city. He looked at their heat maps of population sentiment
and wondered how they’d look this time tomorrow. . .
www.csbaonline.org 3
Introduction
By the middle of the 21st century, ground forces will employ tens of thousands of robots, and the
decisions of human commanders will be shaped by artificial intelligence. Although the future is
impossible to predict, trends in technology and warfare make this a near certainty. Military
organizations must plan now for this new era of warfare. Governments must be prepared for the
political, strategic, and ethical dimensions of this shift in the character of war.
In his book On War, Carl von Clausewitz highlighted how failing to understand the character of
war leads to disaster. Discussing the Prussian defeat in 1806, he chastised Prussian generals for
using the old tactics of Frederick the Great against a Napoleonic army waging a new type of
warfare. 1 They had not appreciated the changes in how war was being fought or in the character of
war. The future development and deployment of human-machine teams and autonomous weapons
systems represents such a shift in the character of war.
In 2017, the United States Army published its strategy for the development and deployment of
robotic and autonomous systems. This strategy outlines activities for the U.S. Army that will
provide a wider range of robotic capabilities to secure U.S. national security objectives over the
next two decades. 2 However, the U.S. Army is not the only institution that can benefit from the
enhanced teaming of humans and machines in the future; the ground forces of U.S. allies and
partners can enhance their operational and institutional effectiveness through this approach as
well.
1 Carl von Clausewitz, On War, Michael Howard and Peter Paret, eds. and trans. (Princeton, NJ: Princeton University Press, 1984),
pp. 154–155.
2 U.S. Army, Robotic and Autonomous Systems Strategy (Fort Eustis, VA: U.S. Army Training and Doctrine Command [TRADOC],
March 2017), p. i.
4 CSBA | HUMAN-MACHINE TEAMING FOR FUTURE GROUND FORCES
This study examines the key drivers, opportunities, and challenges for ground forces in developing
future human-machine teams. It provides an intellectual foundation for the detailed analysis of the
personnel, equipment, training, education, doctrine, sustainment, and infrastructure required by a
future human-machine force.
Chapter one examines the rationale for ground forces investing in human-machine teaming. It
addresses the development of human-machine teaming for combat operations and its potential in
a wide range of ground force activities. This includes combat, preparing for combat,
modernization, doctrine, education, and training.
Chapter two analyses three areas that offer future ground forces a competitive advantage in war:
human-robot teaming, human-AI teaming, and human augmentation. 3 Each will pose challenges
to the training, resourcing, and culture of ground forces.
Chapter three reviews the key challenges in developing future human-machine teams. Strategic,
institutional, and tactical issues are considered in this section of the paper.
Chapter four proposes a strategy to improve future ground force performance through human-
machine integration. A key theme is that human-machine teams have applications beyond the
battlefield. These applications extend across the entire enterprise of raising, training, sustaining,
and employing ground forces. A strategy that includes all aspects of a ground force’s institutional
activities—from those on the battlefield to strategic decision-makers—ensures organizations can
develop a competitive advantage through human-machine integration.
3 Recent U.S. military efforts have focused on five lines of endeavor: autonomous deep learning systems; human-machine
collaboration; assisted human operations’ advanced human-machine combat teaming; and network-enabled cyber-hardened
autonomous weapons. See “Remarks by Defense Deputy Secretary Robert Work at the CNAS Inaugural National Security Forum,”
Center for a New American Strategy, December 14, 2015, available at https://www.cnas.org/publications/transcript/remarks-by-
defense-deputy-secretary-robert-work-at-the-cnas-inaugural-national-security-forum.
www.csbaonline.org 5
A convoy of driverless Army trucks makes its way through the Department of Energy's Savannah River Site in South Carolina, May 29, 2014. Photo Credit:
U.S. Army Tank Automotive Research, Development and Engineering Center.
6 CSBA | HUMAN-MACHINE TEAMING FOR FUTURE GROUND FORCES
www.csbaonline.org 7
Chapter 1
Western ground forces now similarly find themselves at a new crossroads. Large-scale operations
in Iraq and Afghanistan have mostly concluded and have been replaced by smaller “economy of
effort” operations to train indigenous forces. Concurrently, the world in which military
organizations have operated over the past two decades is rapidly changing.
Many Western governments and military organizations have attempted to identify prevailing
trends that will drive or influence strategy and national policy. Likewise, military organizations
around the world are studying the changing character of war to inform force structure and
procurement decisions. The uncertainty of the future security environment confounds precise
predictions of the future. Prudence demands that governments and military organizations outline
a range of probable futures based on prevailing trends to inform their planning.
8 CSBA | HUMAN-MACHINE TEAMING FOR FUTURE GROUND FORCES
Assessments of the future security environment from Canada, the United States, the United
Kingdom, Australia, and New Zealand 4 contain several common themes; significant changes in
demographics and urbanization, geopolitics, economics, the role of the state, the diffusion of
power, climate and resources, and emerging and disruptive technology are forecasted. These will
not only affect the policy and strategy of nations but will also drive changes in the character of war
and future ground force operations.
Geopolitics
With the military growth and modernization of revisionist powers such as China and Russia, the
character of military operations for ground forces will likely evolve. Both powers have invested
heavily in conventional military capabilities, including qualitative improvements in a range of
ground force organizations. For Russia, this investment was driven by the rapid military
modernization of its neighbors and the understanding that a strong nuclear force alone would not
secure Russia’s great power status. 5 For China, the ongoing investment in conventional forces is
part of the Chinese Communist Party’s stated aim of becoming a great power. 6
In this environment, preparing for large-scale ground operations is once again an imperative for
Western ground force design. This does not negate the need for ground forces to prepare for
smaller-scale contingencies. But even these smaller-scale engagements will be influenced by the
presence of weapon systems, sensors, and cyber capabilities developed for high intensity
4 Ministry of Defence (MOD) UK, Global Strategic Trends—Out to 2045 (Shrivenham, UK: Development, Concepts and Doctrine
Centre, June 30, 2014); MOD UK, Future Operating Environment 2035 (Shrivenham, UK: Development, Concepts and Doctrine
Centre [DCDC], December 14, 2015); David T. Miller, Defense 2045: Assessing the Future Security Environment and
Implications for Defense Policy Makers (Washington, DC: Center for Strategic and International Studies, November 2015);
Directorate of Future Land Warfare, Future Land Warfare Report (Canberra: Australian Army Headquarters, 2014); Army
General Staff, Future Land Operating Concept 2035: Integrated Land Missions (Wellington, NZ: Headquarters New Zealand
Defence Force, 2017); Vice Chief of Defence Force, Australia, Future Operating Environment 2035 (Canberra: Commonwealth of
Australia, 2016); and Canadian Department of National Defence, Designing Canada’s Army of Tomorrow: A Land Operations
2021 Publication (Kingston, Canada: Directorate of Land Concepts and Designs, 2011).
5 For an examination of Russia’s re-investment in conventional military capability, see Bettina Renz, “Why Russia is Reviving Its
Conventional Military Power,” Parameters 46, no. 2, Summer 2016.
6 The progress in developing sophisticated conventional forces, the modernization of equipment, and structural reform is charted in
a series of unclassified Pentagon annual reports to the U.S. Congress. The most recent report in 2017 describes a range of
initiatives to implement joint command and control systems, and a range of modernization activities. See Office of the Secretary of
Defense (OSD), Military and Security Developments Involving the People’s Republic of China 2017, Annual Report to Congress
(Washington, DC: DoD, May 15, 2017), p. 31, available at
https://www.defense.gov/Portals/1/Documents/pubs/2017_China_Military_Power_Report.PDF.
www.csbaonline.org 9
operations conducted by the large conventional forces of countries such as Russia, Iran, and
China–but could also be employed by their proxies.
This will affect both blue- and white-collar jobs. Manufacturing and high-speed financial trading
have already been affected by robotics and AI. As AI continues to improve, industries such as
transportation, healthcare, banking, and construction will see people displaced from their jobs,
although experts disagree about the pace at which this will occur. 8
This change in the global civilian labor market will eventually affect military personnel
management models. New technologies will permit the automation of many tasks currently
performed by soldiers. As automation and AI allow civilian business leaders to place humans in
different kinds of work, so too will military personnel planners be forced to think anew about the
recruiting and employment opportunities of a new global workforce approach. It is likely to drive
the creation of new military personnel models and in turn the designing of new ground force
structures. This, along with the disruptive technologies of robotics, AI, and human augmentation
could enable new operating concepts.
7 See James Manyika, “Technology, Jobs, and the Future of Work,” executive briefing, McKinsey Global Institute, May 2017,
available at https://www.mckinsey.com/global-themes/employment-and-growth/technology-jobs-and-the-future-of-work.
8 Aaron Smith and Janna Anderson, AI, Robotics, and the Future of Jobs (Washington, DC: Pew Research Center, August 6, 2014),
available at http://assets.pewresearch.org/wp-content/uploads/sites/14/2014/08/Future-of-AI-Robotics-and-Jobs.pdf.
9 For more on this topic, see Erik Brynjolfsson and Andrew McAfee, The Second Machine Age: Work, Progress, and Prosperity in a
Time of Brilliant Technologies (New York: W. W. Norton & Company, 2016); and Andrew McAfee and Erik Brynjolfsson,
Machine, Platform, Crowd: Harnessing Our Digital Future (New York: W. W. Norton & Company, 2017).
10 CSBA | HUMAN-MACHINE TEAMING FOR FUTURE GROUND FORCES
innovations for commercial advantage, military advances in robotics, AI, and augmentation will
largely be based on these developments in industry.
Gill Pratt, former DARPA Program Manager and CEO of the Toyota Research Institute, has argued
that technological and economic trends are converging to deliver a “Cambrian Explosion” of new
robotic capabilities. 10 This is an analogy to the history of life on Earth, specifically the period
roughly 500 million years ago in which the pace of evolutionary change, for both diversity and
complexity of life forms, increased significantly. Many of the foundational technologies for robots,
such as computing, data storage, and communications, have been progressing at exponential
growth rates.
Two more recent technologies, Cloud Robotics and Deep Learning, are likely to build upon these
earlier technologies in what Pratt has described as a “virtuous cycle of explosive growth.” 11 Cloud
Robotics permits each individual robot to learn from the experiences of all robots, in turn leading
to rapid growth of robot competence. Deep Learning algorithms allow robots to learn and
generalize their associations based on very large (and often cloud-based) training sets that often
include millions of examples. 12
Commerce currently leads the way for how robots and AI can be employed. For example, Amazon
currently employs approximately 80,000 robots in its logistic distribution centers, known at
Amazon as fulfillment centers. 13 Amazon also possesses its own robotics research and
development capacity, known as Amazon Robotics. 14 The mining industry applies autonomous
systems for many functions; Excavating and hauling vehicles have been equipped with vehicle
controllers, high-precision global positioning system (GPS) sensors, and obstacle detection. These
allow for safer operations through a complex load, haul, and dump cycle and enable integration
with other vehicles and people.
These are just some of the lessons from commerce that the military can and should learn for its
own employment of robotics and AI. 15 The applicability for these advancements in military
support and combat operations should be obvious. This is not to say that ground forces are readily
prepared to adapt these commercial models for military employment. Substantive adaptation is
10 Gill, A. Pratt, “Is a Cambrian Explosion Coming for Robotics?” Journal of Economic Perspectives 29, no. 3, Summer 2015,
available at https://pubs.aeaweb.org/doi/pdf/10.1257/jep.29.3.51.
11 Ibid.
12 Ibid.
13 Casey Coombs, “Amazon to Spend $200M for One of its Most Expensive Fulfilment Centers Ever (Video),” Puget Sound Business
Journal, June 9, 2017.
14 For more information on Amazon Robotics, visit https://www.amazonrobotics.com/#/.
15 Defense Science Board, Office of the Under Secretary of Defense (OUSD), Summer Study on Autonomy (Washington, DC: DoD,
June 2016), p. 8, available at https://www.hsdl.org/?view&did=794641.
www.csbaonline.org 11
required of the technologies and the military organizations employing them is required before the
full benefits of automation can be realized.
1. Teaming humans with robotic and AI capabilities can boost national military
power. The effects of robot/AI and military teaming could be seen as a force multiplier in
aspects of military capacity and strategic planning capability.
The number of combat-age males in a country is one of the elements that determine its
military potential. By using large numbers of robots, augmented humans, and AI, countries
with small, elderly, and declining populations might possess military mass beyond their
human population size. Although such a scenario is speculative, it is possible that a
technologically advanced country with a smaller population could build a significant
advantage using AI-based military systems and fielding large numbers of robotic warfighters.
This could also provide a deterrent effect for these nations’ national security strategies. 16
The use of AI by policymakers and military planners in decision support may also offer
advantages at the strategic level. AI can assess large amounts of data, challenge human bias,
and recognize patterns that humans may not comprehend. Although it will present ethical and
technical issues, the marriage of humans and AI in strategic decision-making will have wide
utility.
2. Teaming humans with robots and AI will improve individual and team
performance while reducing threats to humans. Augmenting human capabilities
potentially offers additional gains in performance and reductions in threats. The science and
technologies underwriting human enhancements are advancing quickly. 17 Unlike the
mechanical approach of robotics, augmentation seeks to create a super-soldier from a
biomedical direction, such as with drugs and bionics. As Lin and Abney have noted, “In
between robotics and biomedical research, we might arrive at the perfect future warfighter:
16 The issue of deterrence in a non-nuclear military organization is covered only briefly in the most recent military strategy for
Australia, described in its Defence White Paper. The only capability which is explicitly described as a deterrent are submarines.
See Department of Defence, Australia, Defence White Paper 2016 (Canberra: Commonwealth of Australia, 2016), especially p. 90.
17 Patrick Lin, Maxwell J. Mehlman, and Keith Abney, Enhanced Warfighters: Risk, Ethics, and Policy (San Luis Obispo, CA:
California Polytechnic State University, January 1, 2013), p. 86, available at http://ethics.calpoly.edu/greenwall_report.pdf.
12 CSBA | HUMAN-MACHINE TEAMING FOR FUTURE GROUND FORCES
one that is part machine and part human, striking a formidable balance between technology
and our frailties.” 18
3. Robots and AI can enable new operating concepts. The new and interdisciplinary
research areas of AI, complex adaptive systems, and swarm optimization indicates the
potential for self-organized robot swarms to be used in future conflict. 19 As conventional
enemy forces move to lower signature systems and operations, and as non-state actors
continue to hone non-linear and dispersed approaches, 20 the ability to cover more ground
becomes more challenging. 21 One potential solution for friendly forces, described by Robert
Scales in Future Warfare Anthology, is to saturate an operational area with small autonomous
systems that force an adversary to move, be detected, and be targeted by friendly forces. 22
From an institutional perspective, new operating concepts are vital in building a competitive
edge. This does not suggest that swarming become the only means of operating, but it does
offer additional options to military commanders within joint operations.
Even if friendly forces do not employ swarming, they will have to protect themselves against
hostile swarming approaches. Allied ground forces possess thousands of land vehicles and
helicopters. A high-quality quadcopter UAV currently costs roughly $1,000. In consequence,
for the cost of a single helicopter, a ground force might acquire tens of thousands of drones. In
the future, drones could be cheaper than some ballistic munitions today. How would a ground
force embarked on an amphibious task group respond to an attack from hundreds of aerial
18 Ibid., p. 87.
19 The opportunities and challenges of swarming robots is examined in detail in Andrew Ilachinski, AI, Robots, and Swarms: Issues,
Questions, and Recommended Studies (Arlington, VA: CNA, January 2017), pp. 105–131, available at
https://www.cna.org/CNA_files/PDF/DRM-2017-U-014796-Final.pdf; Sean J. A. Edwards, Swarming and the Future of
Warfare, Pardee RAND Graduate School dissertation (Santa Monica, CA: RAND corporation, 2005), available at
https://www.rand.org/content/dam/rand/pubs/rgs_dissertations/2005/RAND_RGSD189.pdf; and Jules Hurst, “Robotic
Swarms in Offensive Maneuver,” Joint Forces Quarterly 87, October 2017, pp. 105–111, available at
http://ndupress.ndu.edu/Portals/68/Documents/jfq/jfq-87/jfq-87_105-111_Hurst.pdf?ver=2017-09-28-093018-793.
20 The trend toward a more dispersed battlespace (physically but also including aspects such as cyberspace) has been identified in
multiple studies and publications over the past two decades. These include Trevor Nevitt Dupuy, The Evolution of Weapons and
Warfare, first printing (Indianapolis, IN: Bobbs-Merrill Co., 1981); Robert H. Scales Jr., Future Warfare Anthology (Carlisle, PA:
Strategic Studies Institute, U.S. Army War College, May 2001); Brian Nichiporuk and Carl H. Builder, Information Technologies
and the Future of Land Warfare (Santa Monica, CA: RAND Corporation Arroyo Center, 1995); Directorate of Army Research and
Analysis, Complex Warfighting (Canberra: Australian Army Headquarters, September 2009), p. 17; Andrew F. Krepinevich and
Eric Lindsey, The Road Ahead: Future Challenges and their Implications for Ground Vehicle Modernization (Washington, DC:
Center for Strategic and Budgetary Assessments, 2012), p. 18; and U.S. Army, The U.S. Army Operating Concept: Win in a
Complex World 2020–2040 (Fort Monroe, VA: TRADOC, October 31, 2014), pp. 10–11.
21 Edwards, Swarming and the Future of Warfare, pp. 7–12.
22 Scales, Future Warfare Anthology, p. 82.
www.csbaonline.org 13
explosive drones? Some of the major platforms and strategies upon which current military
forces rely might be rendered obsolete, or at least highly vulnerable. 23
5. Future adversaries will use these technologies. These technologies are highly attractive
to other national military organizations 26 and non-state actors. There is a proliferation of
unmanned capability around the world in national armed services with nearly 80 nations
either developing or deploying these systems. 27 The Russian Military Industrial Committee
has approved a plan that would have 30 percent of Russian combat power consist of entirely
remotely controlled and autonomous robotic platforms by 2030. 28 Other countries facing
demographic and security challenges are likely to set similar goals. And although the U. S.
Department of Defense has enacted restrictions on the use of autonomous and semi-
autonomous systems wielding lethal force, 29 hostile nations and non-state actors may not
exercise such self-restraint.
23 Greg Allen and Taniel Chan, Artificial Intelligence and National Security (Cambridge, MA: Belfer Center for Science and
International Affairs, July 2017), pp. 21–22.
24 Ronald C. Arkin, “The Case for Ethical Autonomy in Unmanned Systems,” Journal of Military Ethics 9, no. 4, 2010; and Ronald C.
Arkin, “Ethical Robots in Warfare,” IEEE Technology and Society Magazine 28, no. 1, Spring 2009.
25 These issues are examined in Amitai Etzioni and Oren Etzioni, “Pros and Cons of Autonomous Weapons Systems,” Military
Review, May-June 2017, available at http://www.armyupress.army.mil/Journals/Military-Review/English-Edition-
Archives/May-June-2017/Pros-and-Cons-of-Autonomous-Weapons-Systems/.
26 The use of unmanned systems by adversaries is an assumption in U.S. Army, Multi-Domain Battle: Evolution of Combined Arms
for the 21st Century 2025–2040, Draft Version 1.0 (U.S. Army, October 2017).
27 P. W. Singer, “The Global Swarm,” Foreign Policy, March 11, 2013, available at http://foreignpolicy.com/2013/03/11/the-global-
swarm/.
28 Samuel Bendett, “Get Ready, NATO: Russia’s New Killer Robots are Nearly Ready for War,” The Buzz, blog, The National Interest,
March 7, 2017.
29 DoD, “Autonomy in Weapon Systems,” Department of Defense Directive 3000.09, November 21, 2012.
14 CSBA | HUMAN-MACHINE TEAMING FOR FUTURE GROUND FORCES
In its Mosul operations of 2017, ISIS deployed a range of lethal unmanned ground and air
vehicles. 30 Russian forces have used unmanned ground and aerial vehicles in their operations
in Syria, 31 and they have used unmanned aerial vehicles extensively in the Ukraine. 32 No
existing ethical or legal framework, or international counter-proliferation framework, prevents
this.
In comparison to more expensive conventional capabilities, the low cost and accessibility of
robotics and AI make them highly attractive. As Brecher, Niemi, and Hill have recently
written, “Our adversaries care about the morals and ethics of lethal, autonomous systems only
insofar as those concerns give them a competitive advantage. If full autonomy gives them
supremacy on the battlefield, they will care little about what human rights lawyers think.” 33
6. Robots and AI can improve all institutional and support functions of ground
forces. The broad potential applicability of these systems means that ground forces should
adopt an enterprise approach to the employment of human-machine teams. Human-robot
teams can be used in training institutions, freeing up personnel to be re-deployed for other
operations. Advanced computing and analytical capacity may well be very useful in human-AI
strategic decision-making teams for capability development, resource allocation, and talent
management of personnel. Cognitive augmentation may be as useful for decision-makers at a
strategic headquarters as it is for soldiers deployed on operations.
30 The Battle of Mosul has been examined in multiple publications. Examples include “The Battle for Mosul,” Strategic Comments
23, no. 4, May 17, 2017; Gary Volesky and Roger Noble, “Theatre Land Operations: Relevant Observations and Lessons from the
Combined Joint Land Force Experience in Iraq,” Military Review, June 22, 2017; and Australian Army, “Immediate Lessons from
the Battle of Mosul,” Land Power Forum, June 25, 2007, available at https://www.army.gov.au/our-future/blog/land-
combat/immediate-lessons-from-the-battle-of-mosul.
31 Patrick Tucker, “Armed Ground Robots Could Join the Ukrainian Conflict Next Year,” Defense One, October 10, 2017, available at
http://www.defenseone.com/technology/2017/10/armed-ground-robots-could-make-their-combat-debut-ukrainian-conflict-
next-year/141677/?oref=d-river.
32 Patrick Tucker, “In Ukraine, Tomorrow’s Drone War is Alive Today,” Defense One, March 9, 2015, available at
http://www.defenseone.com/technology/2015/03/ukraine-tomorrows-drone-war-alive-
today/107085/?oref=search_drone%20war%20ukraine.
33 Joseph Brecher, Heath Niemi, and Andrew Hill, “My Droneski Just Ate Your Ethics,” War on the Rocks, August 10, 2016, available
at https://warontherocks.com/2016/08/my-droneski-just-ate-your-ethics/.
www.csbaonline.org 15
Closing Thoughts
As these technologies continue to develop, additional imperatives—beyond these six
propositions—are like to emerge. Civil and military organizations are still learning about the
potential of robotic systems and AI. Andrew Ilachinski has recently noted:
the military is on the cusp of a major technological revolution as it enters the Robotic Age, in which
warfare is conducted by unmanned and increasingly autonomous weapon systems, operating across
all domains (air, sea, undersea, land, space, and cyber), and across the full spectrum of military
operations. The question is not whether the future of warfare will be filled with autonomous, AI-
driven robots, but when and in what form. 34
The barriers for entry for this type of technology are lower than many other types of conventional
weapons and continue to fall. This is an advantage for conventional and non-state military
organizations, large and small. 35
Chapter 2
Future Human-Machine
Teams: Three Key Endeavors
Chapter I of this report explored a rationale for enterprise-wide adoption of human-machine
teams in ground forces. It is proposed that three human-machine endeavors will underpin the
development of future human-machine ground forces. Each is a distinct area of research,
development, and investment.
2. Teaming humans and AI. Most teams are likely to contain increasingly sophisticated types
of AI. The marriage of humans and AI for strategic and operational planning, as well as for the
analysis of future activities, are key applications. This requires analytical focus that is related
to, but distinct from, human-robot teaming.
3. Human augmentation. This mode of human-machine teaming is distinct from the other
two because humans and machines are combined as a single entity. It is focused on improving
36 Current examples of the types of robotic systems that are teamed with humans include explosive ordnance disposal robots, aerial
drones, and the seaborne Sea Knight surface patrol vessel. The Sea Knight is an unmanned sea patrol vessel developed by Israel.
37 Pierre Urlings, Teaming Human and Machine: A Conceptual Framework for Automation from and Aeronautical Perspective,
thesis (Adelaide: University of South Australia, December 2003), p. 84.
18 CSBA | HUMAN-MACHINE TEAMING FOR FUTURE GROUND FORCES
These three endeavors should be pursued concurrently. 38 The basic technology for each already
exists. Concurrent development would allow program interaction and would significantly improve
the deployability, lethality, and sustainability of future ground forces.
Human-Robot Teaming
In his 2009 book, Wired for War, Peter Singer examined the future of warfare with robots
becoming a pervasive element of military operations. This is not a new vision. Science fiction
writers have been writing about robots in warfare for nearly a century. Indeed, human automata
have been described and constructed for over 500 years. But it was only in 1921 that the Czech
writer Karel Čapek coined the term “Robot” in a play called Rossum’s Universal Robots. 39
Scientists, writers and industrialists have built and imagined robots for a wide range of functions.
From models for assembly line construction of motor vehicles to the more sophisticated models
employed on the International Space Station 40, early generation robots have assumed functions
that are either cheaper or safer than using humans.
A number of reports have described how robotic systems are likely to proliferate over the next 20
years. 41 As they become cheaper and easier to produce, these technologically advanced systems are
likely to be used widely, with developing states and non-state actors gaining increased access to
38 Recent U.S. military efforts have focused on five lines of endeavor: Autonomous deep learning systems; human-machine
collaboration; assisted human operations; advanced human-machine combat teaming; and network-enabled cyber-hardened
autonomous weapons. See remarks by Deputy Secretary of Defense Robert Work at the Center for New American Security Defense
Forum, December 14, 2015, quoted in George Galdorisi, “Designing Unmanned Systems for Military Use: Harnessing Artificial
Intelligence to Provide Augmented Intelligence,” Small Wars Journal, August 23, 2017 available at
http://smallwarsjournal.com/jrnl/art/designing-unmanned-systems-for-military-use-harnessing-artificial-intelligence-to-
provide-a.
39 Karel Čapek, R.U.R. (Rossumovi Univerzální Roboti) (written 1920), Paul Selver, trans., Rossum’s Universal Robots (London:
Oxford University Press, 1923).
40 Evan Ackerman, “Astrobee: NASA’s Newest Robot for the International Space Station,” IEEE Spectrum (online), February 11,
2017, available at http://spectrum.ieee.org/video/robotics/military-robots/astrobee-nasa-newest-robot-for-the-international-
space-station.
41 A selection of publications that have examined this trend include: MOD UK, Future Operating Environment 2035, p. 32; U.S.
Army, Robotic and Autonomous Systems Strategy; Australian Army, Directorate of Future Land Warfare, Future Land Warfare
Report (2014); and U.S. Army, The Operational Environment and the Changing Character of Warfare (Fort Monroe, VA:
TRADOC, 2017). See also The Atlantic Council’s Art of the Future project, available at http://artoffuturewarfare.org/; and the
World Economic Forum’s project on the relationship between the Fourth Industrial Revolution and International Security,
summarized in Anja Kaspersen, Espen Barth Eide, and Philip Shetler-Jones, “10 Trends for the Future of Warfare,” World
Economic Council, November 3, 2016, available at https://www.weforum.org/agenda/2016/11/the-4th-industrial-revolution-and-
international-security/.
www.csbaonline.org 19
them. As they become more capable, more reliable, and more trusted by humans, their
employment in all military services is likely to become widespread. 42
Robots in human-robot teams are likely to possess varying levels of autonomy. In their 2015 paper,
An Introduction to Autonomy in Weapon Systems, 43 Horowitz and Scharre offer useful definitions
of autonomous, semi-autonomous, and human-supervised autonomous weapon systems. As
shown below, these definitions could be applied to the various levels of autonomy that humans
might delegate to robots:
• A semi-autonomous robot is a one that incorporates autonomy into one or more functions
and, once activated, is intended to only undertake those tasks that a human has decided are to
be undertaken. 44
• An autonomous robot is one that, once activated, is intended to undertake tasks not
specifically tasked by humans. In the case of lethal robots, they might select and engage
targets where a human has not decided those specific targets are to be engaged.
The military has long held an interest in robotic capability. Remote controlled boats were used by
the Germans in the First World War. In the Second World War, the Germans employed
unmanned, Goliath tracked vehicle robots filled with explosives in France and Italy. More recently,
at the high point of the Iraq War in 2006, over 8,000 robots were in use by the U.S. military. 45
relationships. In logistics, robots will have utility in performing tasks that are dirty, dangerous, or
repetitive (such as vehicle maintenance and repair and basic movement tasks). 46
In addition to the better known unmanned aerial vehicle programs, there are many robotic ground
systems either in development or being deployed with military organizations. Since 2004, the
Defense Advanced Research Project Agency has held “DARPA challenge” competitions to
encourage the development and capabilities of unmanned ground vehicles. 47 Companies such as
Lockheed Martin have developed a range of autonomous vehicles for logistics at different levels.
This has included small squad-level unmanned ground vehicles to support larger autonomous
heavy trucks. 48
The use of human-robot teams during operations offers a solution to an enduring challenge for
ground forces—the building of mass. Potentially, each soldier might control a small fleet of ground
and air systems, providing an exponential increase in the capability of deployed forces. A 2016 U.S.
Department of Defense experiment saw over 100 micro drones released to form an autonomous
swarm. 49 If swarming concepts were embraced as one aspect of ground force operational design,
future land vehicle “carriers” might fill similar “mothership” roles.
A highly capable and sustainable land combat battlegroup in 2030 may consist of as few as 250–
300 human soldiers and several thousand robotic systems of various sizes and functions. By the
same token, many functions of artillery and combat engineer units, currently undertaken by
humans, might be better done by robots in human-robot teams. This has the potential to reduce
the size of these types of units by hundreds of combat arms personnel. This approach could free up
personnel for redeployment into areas where the art of war demands leadership and creativity—
enabling intelligence functions; training and education; planning; and, most importantly,
command and leadership.
For ground forces, human-robot teaming could provide the core for all future ground force design.
Select high-priority, focal areas of human-robot teaming might be chosen to provide benchmarks
46 This idea of employing robots for the “dirty, dangerous, and dull” tasks is examined in Max Boot, War Made New: Technology,
Warfare, and the Course of History: 1500 to Today (New York: Gotham Books, 2006), p. 442.
47 This effort has recently emphasized safer vehicles and more assured autonomy. “DARPA Assured Autonomy Seeks to Guarantee
Sargent of Learning-Enabled Autonomous Systems,” DARPA website, August 16, 2017, available at https://www.darpa.mil/news-
events/2017-08-16.
48 The Lockheed Martin Website has detailed information on their unmanned ground vehicle programs available at
http://www.lockheedmartin.com.au/us/products/amas1.html.
49 The Strategic Capabilities Office, partnering with Naval Air Systems Command, demonstrated the micro-drone swarms at China
Lake, California. The test, conducted in October 2016 and documented on Sunday’s CBS News program “60 Minutes,” comprised
103 Perdix drones launched from three F/A-18 Super Hornets. The micro-drones demonstrated advanced swarm behaviors such
as collective decision-making, adaptive formation flying, and self-healing. “Department of Defense Announces Successful Micro-
Drone Demonstration,” DoD media release, January 9, 2017, available at https://www.defense.gov/News/News-Releases/News-
Release-View/Article/1044811/department-of-defense-announces-successful-micro-drone-demonstration/.
www.csbaonline.org 21
for driving and measuring progress. Potential focal areas might include the ISR battle, breach and
break action, and close combat. These are activities that are demanding in the integration of
various ground and joint functions. Initial experiments in these three areas would focus on simple
human-robot integration activities. Subsequent experiments could expand in scope to include
human-AI integration and human augmentation. This is explored further in Chapter IV.
Human-AI Teaming
In 1899, diplomats from the world’s leading military powers convened in The Hague for a peace
conference. One of the outcomes of the conference was a five-year moratorium on all offensive
military uses of aircraft. Although the intention was to later make the ban permanent, it was
abandoned at the second Hague conference of 1907 once countries saw the irresistible potential of
aerial warfare. 50 Aerospace technology eventually became nearly synonymous with military power.
It is likely that the use of AI will chart a similar course. Businesses are choosing machine learning
because competitively they have no choice; so, too, will militaries and intelligence agencies feel
competitive pressure to expand the use of military AI applications. 51
Key to human-AI teaming is understanding the increasing pace and complexity of decision support
and decision-making. In future conflicts, military organizations will work across “multiple
domains with multiple partners, considering multiple dilemmas and options.” 52 Decision cycles
will ultimately become faster than the capacity of human cognition to process. 53
In this environment, military command and control will need AI that can process information and
recommend options faster (or of higher quality) for making decisions than can the enemy. There
are three primary areas in which AI might be applied in this context:
• Assisted intelligence, widely available today, improves what people and organizations are
already doing. A simple example is the GPS navigation programs in vehicles and aircraft that
offer directions to drivers and aircrew.
There is a wide debate among leading AI researchers about whether human-level AI is possible
and when it might appear. The nearest estimates are “in a few decades” with other experts
predicting “not this century” and even “not ever.” 55 This timeline for human-level AI prevents
against anything other than speculation when thinking about and planning for military capability
development. But scientists and experts have been wrong about the pace of scientific discovery in
the past. In 1933, Ernest Rutherford, one of the great nuclear scientists of his time, stated that
nuclear energy was “moonshine.” And in 1956, Astronomer Royal Richard Woolley described
discussions on space travel as “utter bilge.” 56 So, it is possible that human-level AI may appear
sooner than anticipated. But for the purposes of this study, it is given less focus than assisted,
augmented, and autonomous intelligence.
Partially autonomous and intelligent systems have been used by military organizations since the
Second World War. However, advances in machine learning and AI represent a turning point in
the use of automation in warfare. 57 This is a field where rapid advances will provide opportunities
for military organizations to re-think the conduct of planning, information gathering and analysis,
cyber security, logistics, and strategy development in war.
A key driver for the use of AI in warfare is the convergence of large numbers of advanced sensors,
extensive communication links, and an ever-increasing flow of information. As the quantity of
information continues to increase, the capacity of humans to deal with it is not increasing in a
commensurate manner. The slowest element in decision-making is becoming the human decision-
maker. In the competitive environment of war, the race truly does go to the swift. 58
54 “Workforce of the Future: The Competing Forces Shaping 2030,” Price Waterhouse Coopers slide presentation, 2017, available at
https://www.pwc.com/gx/en/services/people-organisation/workforce-of-the-future/workforce-of-the-future-the-competing-
forces-shaping-2030-pwc.pdf.
55 For discussion on the potential timelines for human level AI, see Max Tegmark, Life 3.0: Being Human in the Age of Artificial
Intelligence (New York: Alfred A. Knopf publishing, 2017), pp. 30–40.
56 Ibid., p. 40.
57 Allen and Chan, Artificial Intelligence and National Security, p. 5.
58 Richard Simpkin was one of the first authors to write about the implications of command and control in the 21st century. While he
wrote several books on various aspects of war, his 1985 book Race to Swift remains one of the best examinations of decision-
www.csbaonline.org 23
Developments in the civilian sector provide insights into potential military applications.
Microsoft’s Cortana, Google’s Now, Amazon’s shopping sites, Netflix’s streaming algorithms, and
Apple’s Siri all combine user input, access to large databases, and limited AI algorithms to provide
decision support for human users. These provide individually tailored options based an
individual’s previous decisions, as well as those of millions of other users. It represents a slow and
steady shift of authority from humans to algorithms. 59 Although there are pitfalls in this for the
unwary, it also offers the potential benefit to the military of more rapid decision-making in
complex circumstances.
Despite the speed and analytical capacity of computers and robots, creativity or imagination
remains a challenge for AI. 60 Advances in this area have been made, however; AI systems have
recently created songs and movie trailers. 61 This rapid advance in AI research and development
has the potential to result in new and disruptive innovations in the planning and conduct of
military operations. 62
The U.S. Department of Defense, in a study of the future operating environment, has recently
described artificial intelligence as “the most disruptive technology of our time. Big data techniques
interrogate massive databases to discover hidden patterns and correlations and are continually
leveraged for intelligence and security purposes by nation states and non-state entities alike.” 63
Efforts by institutions such as the United Nations and the Future of Life Institute 64 to impose a
moratorium on the use of AI in weapons are unlikely to succeed. As Ian Morris wrote in 2014,
making, speed, and 21st century warfare. Richard Simpkin, Race to the Swift: Thoughts on Twenty-first Century Warfare
(London: Brasseys, 1985).
59 In his book Homo Deus, Yuval Norah Harari conducts a detailed examination of the biological and technical aspects of these
trends. He notes that “the shifting of authority from humans to algorithms is happening all around us, not as a result of some
momentous governmental decision, but due to a flood of mundane personal choices.” Yuval Norah Harari, Homo Deus: A Brief
History of Tomorrow (London: Vintage Books, 2015), p. 402.
60 There are three types of creativity: novel combinations of familiar ideas, generation of new ideas by exploring structured
conceptual spaces, and generation of new and surprising ideas. Computer models do exist of creativity. Generally, those focused
on exploratory creativity have been most successful. Margaret A. Boden, “Creativity and Artificial Intelligence,” Artificial
Intelligence, no. 103, 1998.
61 “The Quest for AI Creativity,” IBM, available at https://www.ibm.com/watson/advantage-reports/future-of-artificial-
intelligence/ai-creativity.html#section2.
62 See George I. Seffers, “Commanding the Future Mission,” Signal, May 2016; James Canton, “Next-Gen Computers Will Soon
Transform Battlefield Intelligence,” National Defense Magazine, February 2017.
63 U.S. Army, The Operational Environment and the Changing Character of Warfare, pp. 8–9.
64 The most recent call for a ban on “killer robots” was issued by top robotics and AI companies through a letter posted on the
“Future of Life Institute” website on August 20, 2017. Ariel Conn, “Leaders of Top Robotics and AI Companies Call for Ban on
Killer Robots,” Future of Life Institute news, August 20, 2017, available at https://futureoflife.org/2017/08/20/leaders-top-
robotics-ai-companies-call-ban-killer-robots/.
24 CSBA | HUMAN-MACHINE TEAMING FOR FUTURE GROUND FORCES
“When robots with OODA loops 65 of nanoseconds start killing humans with OODA loops of
milliseconds, there will be no more debate.” 66 Weapons and other artifacts of war will incorporate
AI because military organizations will fear if they do not, their enemies will. 67
Modeling different equipment procurement options is another example of where human-AI teams
may prove more effective than human decision-making. Identifying leaner (and less targetable)
approaches to logistics are also possible. Other strategic functions, including talent management,
and personnel administration could be improved through the development of human-AI teams
that can sift data and undertake analysis with less heuristic bias.
By teaming human decision-makers with AI that can collate and present meaningful information,
military leaders may be able to establish decision superiority over an adversary—assuming that
adversary is not using similar systems. But even if there is a suitable marriage of humans and AI,
the speed at which AI is developing means that more functions will move beyond human
comprehension and may have to be delegated to autonomous systems out of necessity. 69
65 The OODA Loop—Observe, Orient, Decide, Act—was developed by U.S. military officer John Boyd based on his observations of
competitive behavior and time cycles. Grant T. Hammond, The Mind of War: John Boyd and American Security (Washington,
DC: Smithsonian Institution Press, 2001).
66 Ian Morris, War: What Is It Good For? (London: Profile Books, 2014), p. 374.
67 For example, one driver for the United States military will likely be the embrace of AI by its main strategic competitor, China. On
July 20, 2017, China released its Artificial Intelligence strategy called A Next Generation Artificial Intelligence Plan. The original
version can be found at http://www.gov.cn/zhengce/content/2017-07/20/content_5211996.htm. The translated version of the
strategy can be found at https://www.newamerica.org/documents/1959/translation-fulltext-8.1.17.pdf.
68 Deputy Secretary of Defense, “Establishment of an Algorithmic Warfare Cross-Functional Team (Project Maven),” memorandum,
April 26, 2017, available at
https://www.govexec.com/media/gbc/docs/pdfs_edit/establishment_of_the_awcft_project_maven.pdf.
69 Thomas K. Adams, “Future Warfare and the Decline of Human Decision-making,” Parameters, Winter 2001–2002. This rapid
progress in the development of AI has four main drivers: first, the exponential growth in computer performance; second, the
improved access to large datasets that can be used to train machine learning systems; third, continuing advances in machine
learning techniques; and, finally, rapidly increasing commercial investment in AI. Allen and Chan, Artificial Intelligence and
National Security, p. 7.
www.csbaonline.org 25
During operations, the provision of sophisticated AI decision support has potential to assist tired
planners in military headquarters. This might have applications at all levels of command—from a
battlegroup headquarters analyzing and wargaming various tactical options to a strategic-level
headquarters modelling optimal application of kinetic and non-kinetic activities and logistic
support. AI offers the potential to automate functions such as routine resupply, 70 network
management, and movement schedules. Importantly, AI embedded in autonomous and semi-
autonomous systems offers rapid responses to the actions of an adversary, should these systems be
trusted by their human operators.
Ultimately, the speed at which autonomous systems function may force humans to operate further
up the chain of command. As Adams notes, tactical warfare may become the business of machines
and not appropriate for people at all. 71 But if war is to remain a human activity, and we assume
that war’s nature is enduring, humans will still play a role in policy, strategy, and campaigning.
Human Augmentation
The augmentation of humans is likely to represent the ultimate expression of the human-machine
revolution. This is an extension of centuries of practice where people sought to become faster,
stronger, and smarter by using tools and machines. It is useful to look at contemporary human
augmentation efforts to project future capabilities in this area.
The U.S. military has been a major investor in this field, leading a variety of research projects that
seek to optimize human fighting capacities. DARPA’s Accelerated Learning program, for example,
seeks to apply the best practices of learning as demonstrated by neuroscience and statistical
modelling. 72 There are three approaches to human augmentation applicable to the military:
wearable computing augmentation, mechanical augmentation, and implantable augmentation.
70 Noting the likelihood of a more dispersed and lethal battlefield in the future, this approach is examined in Krepinevich and
Lindsey, The Road Ahead, pp. 50–51.
71 Adams, Future Warfare and the Decline of Human Decision-making, pp. 70–71.
72 “DARPA’s Accelerated Learning,” tDCSPlacements.com, available at http://tdcsplacements.com/placements/accelerated-
learning/.
73 Smita Jhajharia et al., “Wearable Computing and its Application,” International Journal of Computer Science and Information
Technologies 5, no. 4, 2014, p. 5700.
26 CSBA | HUMAN-MACHINE TEAMING FOR FUTURE GROUND FORCES
assist them to win a game of roulette in 1961. 74 Wearable computing now includes augmented and
virtual reality devices, as well as the wide range of watch-like devices including products from
Apple, Samsung, 75 and Fitbit. 76 The commercial market for smart wearable devices such as smart
watches, smart glasses, and wearable scanners is estimated to gross more than U.S. $60 billion by
2022. 77 The civil demand for such devices is driving rapid growth in research, development, and
innovation in this field. This will underpin development of wearables for military purposes.
Wearable computing in the military has been led by the efforts of the United States Army.
Launched in 1994, the Land Warrior program currently integrates wireless LAN, helmet-mounted
displays, communications, and a soldier control system. 78 Many ground forces now possess, or are
about to deploy, wearable computing capabilities. Functions normally deployed in wearable
computing systems include navigation, communications, tactical information, and reporting.
Future developments are likely to include the linking of weapon sensors and sights into wearable
computers as well as the incorporation of biometric functions monitoring. 79
Mechanical Augmentation
Humans have used simple mechanical augmentation such as artificial limbs for centuries.
However, this field has received revived attention over the past two decades with many amputees
returning from military operations in Iraq and Afghanistan. This has driven advances in the
development of more sophisticated artificial limbs. Concurrently, this has re-energized interest in
more complex mechanical augmentation with systems such as exoskeletons.
Several research organizations are developing exoskeletons to increase human strength and
endurance. Examples for military use include Lockheed Martin’s HULC, 80 Raytheon’s XOS, 81 and
74 Steve Mann, “Wearable Computing,” in Mads Soegaard and Rikke Friis Dam, eds., The Encyclopedia of Human-Computer
Interaction, 2nd edition (Aarhus, Denmark: The Interaction Design Foundation, 2013), available at https://www.interaction-
design.org/literature/book/the-encyclopedia-of-human-computer-interaction-2nd-ed/wearable-computing.
75 Samsung currently offers several wearable products; see http://www.samsung.com/au/wearables/gear-s2/features/.
76 Fitbit offers a large range of fitness-oriented wearable devices; see https://www.fitbit.com/au/home.
77 “ABI Research Forecasts Enterprise Wearables Will Top US$60 Billion in Revenue in 2022,” press release, ABI Research, August
10, 2017, available at https://www.abiresearch.com/press/abi-research-forecasts-enterprise-wearables-will/.
78 “Land Warrior Integrated Soldier System,” Army Technology.com, available at http://www.army-
technology.com/projects/land_warrior/.
79 Jhajharia, “Wearable Computing and its Application,” p. 5702.
80 Additional information on the HULC exoskeleton is available at
http://www.lockheedmartin.com.au/content/dam/lockheed/data/mfc/pc/hulc/mfc-hulc-pc-01.pdf.
81 Additional information about Raytheon’s XOS project is available at http://newatlas.com/raytheon-significantly-progresses-
exoskeleton-design/16479/.
www.csbaonline.org 27
the University of California Berkeley’s BLEEX. 82 These exoskeletons are likely to offer a range of
enhanced operational and training functions for ground forces.
The U.S. Army has funded multiple research programs in this field, 83 including USSOCOM’s
TALOS program. 84 Other nations have also developed combat suits that rely on exoskeleton
technology. The Russian military reportedly deployed its Ratnik (Warrior) suit in Crimea as early
as 2013. The 3rd generation suit, Ratnik 3, was announced by its designers on October 10, 2017,
and it contains a range of sub-systems including communications, life support, engagement aides,
and advanced optics. 85
Implantable Augmentation
Implantables—small devices that humans can have implanted in their bodies—are already
available for multiple functions. Cochlear implants have been available for over a decade to return
hearing to the profoundly deaf. 86 Cardiac pacemakers are another example of our longstanding
acceptance of implanting small machines into human bodies for medical reasons. And while
medical technology will continue to be an important driver in this field, convenience also provides
motivation.
In Australia this year, a Sydney man had the chip from an Opal travel card implanted to make
traveling on public transport more convenient. 87 A Swedish company called Epicenter offers
implants the size of a grain of rice to its workers that function as swipe cards to open doors,
operate printers, or buy drinks with a wave of the hand. 88 Over the next decade, developments in
implantables may include implantable smart phones, self-healing devices, smart pills for medical
diagnosis, and readable smart tattoos.
82 Additional information about the University of California Berkeley’s BLEEX project is available at
http://bleex.me.berkeley.edu/research/exoskeleton/bleex/.
83 One insightful review of the various U.S. Army programs is Annie Jacobsen, “Engineering Humans for War,” The Atlantic,
September 23, 2015, available at https://www.theatlantic.com/international/archive/2015/09/military-technology-pentagon-
robots/406786/.
84 “Current Iterations of TALOS,” information briefing, U.S. Special Operations Command, Tactical Assault Light Operator Suit
(TALOS) Program Office, May 28, 2014.
85 Tom O’Connor, “Russia Military’s Future War Gear Gets a New Nuclear-Resistant Upgrade,” Newsweek, October 10, 2017.
86 Additional information on Cochlear implants is available at http://www.cochlear.com/wps/wcm/connect/intl/home.
87 “Sydney Man has Opal Card Implanted into Hand to Make Catching Public Transport Easier,” ABC News, June 27, 2017, available
at http://www.abc.net.au/news/2017-06-27/sydney-bio-hacker-has-opal-travel-card-implanted-into-hand/8656174.
88 James Brooks, “Cyborgs at Work: Swedish Employees Getting Implanted with Microchips,” Associated Press, April 3, 2017,
available at https://www.apnews.com/4fdcd5970f4f4871961b69eeff5a6585.
28 CSBA | HUMAN-MACHINE TEAMING FOR FUTURE GROUND FORCES
Early research has found that brain waves can be interpreted using a machine for simple functions
such as thought-controlled movement. Applying this research offers the opportunity for more
advanced augmentation. One example of this is the U.S. Defense Advanced Research Project
Agency, which recently commenced examination of technologies that allow heads-up displays to be
projected via the human visual cortex. 89 There is a history of humans accepting implantable
technology to prolong life. It is not a big leap to accept that in future, if the technology is available,
military personnel might receive implanted augmentation that enhances their physical and
cognitive capabilities.
Efforts to this end are already underway. SpaceX and Tesla CEO Elon Musk is backing a brain-
computer interface venture called Neuralink. This work seeks to design and build devices that can
be implanted in the human brain with the intent of helping human beings merge with software
and keep pace with advancements in AI. These could also potentially improve memory or allow for
more direct interfacing with computing devices. 90
By 2030 ground forces are highly unlikely to have achieved Ray Kurzweil’s singularity, where AI
and machines can outperform the human mind. 91 Robots and AI will proliferate, but they are likely
to remain in the realm of “the science of war.” Computers are not at the point—yet—where they
might supplant what is currently understood as human creativity. 92 Therefore the “art of war” will
remain the domain of people, at least in foreseeable future. But it is probable that technology will
reach a point in the near future where AI, machines, and man-machine teaming will open a range
89 Kareem Ayoub and Kenneth Payne, “Strategy in the Age of Artificial Intelligence,” Journal of Strategic Studies 39, no. 5–6, 2016,
p. 814.
90 Rolfe Winkler, “Elon Musk Launches Neuralink to Connect Brains with Computers,” Wall Street Journal, March 27, 2017.
91 The technological singularity (also, simply, the singularity) is the hypothesis that the invention of artificial super intelligence will
trigger runaway technological growth, resulting in unfathomable changes to human civilization.
92 Margaret A. Boden, “Artificial Creativity,” MIT Technology Review, October 20, 2015, available at
https://www.technologyreview.com/s/542281/artificial-creativity/; and Will Knight, “Amazon Has Developed an AI Fashion
Designer,” MIT Technology Review, August 24, 2017, available at https://www.technologyreview.com/s/608668/amazon-has-
developed-an-ai-fashion-designer/.
www.csbaonline.org 29
of new and exciting possibilities. 93 These will enrich society but profoundly challenge how military
organizations think about the profession of arms.
A Soldier of the 25th Infantry Division remote controls a Kobra 710 during the Pacific Manned Unmanned-Initiative (PACMAN-I) at Marine Corps Training
Area Bellows, Hawaii, July 22, 2016. Manned-Unmanned Teaming is one of many new concepts that has been identified as part of the Army Warfighter
Assessment 2017 (AWA 17). AWA is the U.S. Army Training and Doctrine Command's premier event to evaluate concepts and capabilities that address the
Army's warfighting challenges and shape the future Army's force. Photo Credit: Kimberly Bratic, U.S. Army TARDEC.
93 U.S. Army, The Operational Environment and the Changing Character of Warfare, p. 9.
30 CSBA | HUMAN-MACHINE TEAMING FOR FUTURE GROUND FORCES
www.csbaonline.org 31
Chapter 3
This reaction to the development of a new means of war is illustrative of the ethical and legal
challenges that must be addressed in developing future human-machine teams. As a recent article
in the Harvard International Review notes,
Robotics and human enhancement also pose alarming prospects of ethical blowback. The
depersonalization of warfare lowers the stakes of declaring war in the first place. Furthermore,
biological and technological upgrades to the human body raise a host of concerns, such as risks to
health, the ability to reintegrate into civil society, and the use of enhancements outside of warfare. 95
94 Clay Blair, Silent Victory: The U.S. Submarine War Against Japan (Annapolis, MA: Naval Institute Press, 2001).
95 Veronica Ma, “The Ethics and Implications of Modern Warfare,” Harvard International Review, January 16, 2017, available at
http://hir.harvard.edu/article/?a=14494. See also Patrick Lin, George Bekey, and Keith Abney, Autonomous Military Robotics:
Risk, Ethics, and Design (San Luis Obispo, CA: California Polytechnic State University, December 20, 2008); Robert Sparrow,
32 CSBA | HUMAN-MACHINE TEAMING FOR FUTURE GROUND FORCES
These challenges must be considered by ground forces concurrently with the development of the
technical aspects of human-machine teaming. As a recent report by Lin and Bekey noted, “Given a
significant lag time between ethics and technology, it is imperative to start considering the issues
before novel technologies fully arrive on the scene and in the theatre of war.” 96 The multiple
challenges involved are the focus of this part of the study. These involve strategic, institutional and
tactical issues, each posing serious ethical dilemmas.
It seems likely that a future president would find it easier to deploy a heavily or completely robotic
unit and to keep it in the field for an extended time. This could help with deterrence and crisis
containment. But by making it easier to use force, a robot centric military could also tempt a future
president into conflicts and crises that the United States might otherwise avoid. . . . The Founding
Fathers intentionally made it difficult for the United States to use force. Robots, like airpower, will
erode this firebreak. 98
The counter argument to this is that the unease of political decision-makers in using robotic
systems may constrain their use. Further, recent initiatives in the United Nations to review the use
of lethal autonomous weapons may see the rise of new international conventions and norms that
govern their application in conflict. 99
“Robots and Respect: Assessing the Case Against Autonomous Weapon Systems,” Ethics and International Affairs 30, no. 1,
March 10, 2016; Will Knight, “Military Robots: Armed, but How Dangerous?” MIT Technology Review, August 3, 2015; and Jean-
Baptiste Jeangène Vilmer, “Terminator Ethics: Should We Ban Killer Robots?” Ethics and International Affairs, online exclusive,
March 23, 2015.
96 Lin, Mehlman, and Abney, Enhanced Warfighters, p. iii.
97 I am grateful to Dr. Deane-Peter Baker, from the Australian Centre for the Study of Armed Conflict and Society for his input and
comment on this issue.
98 Steven Metz, “Strategic Insights: The Land Power Robot Revolution is Coming,” Strategic Studies Institute, U.S. Army War
College Press, December 10, 2014, available at http://ssi.armywarcollege.edu/index.cfm/articles/Landpower-Robot-
Revolution/2014/12/10.
99 The most recent series of meetings was held in Geneva in April 2018. See “2018 Group of Governmental Experts on Lethal
Autonomous Weapons Systems (LAWS),” United Nations Office at Geneva, website, available at
https://www.unog.ch/__80256ee600585943.nsf/(httpPages)/7c335e71dfcb29d1c1258243003e8724?OpenDocument&ExpandSe
ction=7#_Section7.
www.csbaonline.org 33
Destabilizing Influences
Altman and Sauer have argued that the application of autonomous robots and AI by the military
will have a negative impact on strategic stability. Using Cold War history, they suggest that the
attractiveness of drone and AI technology to state and non-state actors will result in proliferation
of lethal autonomous weapon systems and a new arms race. Given the potential incentives for pre-
emptive operations using low-cost swarming drones, documented Cold War examples of
computerized early warning errors, and the possibility of stealthy drones holding nuclear delivery
systems at risk, they argue that the introduction of robotics and AI will result in a less stable
security environment. 100
Strategy Development
Future human-machine teaming may pose challenges to strategy development and
implementation. As Hew Strachan has noted,
It is patently absurd to deny that the impact of new technology can be strategically significant. . . .
The steamship, the manned aircraft or the rocket, have triumphed over geography, changing the
relationship between space and time, and thus have a geopolitical effect as well as a directly
operational one. 101
The introduction by the U.S. of the new technology of atomic weapons changed the strategic
calculus of major powers in the wake of the Second World War. As Colin Gray notes, “There is
general agreement among strategic theorists . . . that there was a phenomenon worth calling the
‘nuclear revolution.’ This revolution cast a shadow over all statecraft and strategy.” 102 Similarly,
robots, AI, and human augmentation applied in concert with new organizations and operational
concepts may have an impact on military strategy. 103
The application of AI as a strategic decision support tool may also address some of the human
fragility and bias inherent in strategy development and implementation. 104 AI is not subject to
physical issues such as fatigue and can be built to take account of other psychological dimensions
100 Jürgen Altmann and Frank Sauer, “Autonomous Weapon Systems and Strategic Stability,” Survival: Global Politics and Strategy
59, no. 5, September 18, 2017.
101 Hew Strachan, The Direction of War: Contemporary Strategy in Historical Perspective (Cambridge, UK: Cambridge University
Press, 2013), pp. 190–191.
102 Colin S. Gray, Modern Strategy (Oxford, UK: Oxford University Press, 1999), pp. 203–204.
103 Strachan, The Direction of War, pp. 190–192. In a recent article, Paul Scharre also proposes that autonomous robotics systems, or
drones, pose a challenge for foreign policy. Paul Scharre, “The Coming Drone Wars: A Headache in the Making for American
Foreign Policy,” The National Interest, July 25, 2017, available at http://nationalinterest.org/feature/the-coming-drone-wars-
headache-the-making-american-foreign-21662.
104 The issue of AI and strategy making is examined in detail in Ayoub and Payne, “Strategy in the Age of Artificial Intelligence.”
34 CSBA | HUMAN-MACHINE TEAMING FOR FUTURE GROUND FORCES
of strategy such as cognitive load, risk taking and aversion, and bias. It can assess large amounts of
data, challenge long-held human assumptions, and recognize patterns missed by humans.
However, AI is limited by programming and the data sets available to it. Further, AI may produce
strategies that breach the values, ethics, or strategic objectives of its human users. 105 Regardless, it
is apparent that AI will impact strategy and strategic decision-making. The degree to which it does
must be decided by policymakers and senior military leaders.
Recent incidents of AI misbehavior have demonstrated the need for such scrutiny. Instances of
algorithmic defamation have included Google search engine routines learning to make defamatory
or bigoted associations about groups of people. 106 A 2016 study of the use of algorithms to support
criminal sentencing in the United States found systemic racial bias in risk estimation. 107 A 2017
report from AI Now raised concerns with the use of invalidated or unreviewed algorithmic systems
in social systems such as justice, welfare, and other government services. 108
More infamous cases of algorithmic misbehavior are the 2010 “Flash Crash” 109 and the 2016
offensive public statements of Microsoft AI chatbot, Tay. 110 These indicate that we still have much
to learn about the shortcomings of AI and machine learning. One writer for The Verge online
technology magazine posed an important question in the wake of the Tay incident: “How are we
going to teach AI using public data without incorporating the worst traits of humanity?” 111
These misbehaving algorithms are the result of shortfalls in human specification and can lead to
incorrect, inequitable, or dangerous outcomes. 112 These incidents also underscore the need for a
greater understanding of AI’s shortfalls by decision-makers and more rigorous testing prior to
operational use. As Will Knight noted recently, “Knowing AI’s reasoning is also going to be crucial
if the technology is to become a common and useful part of our daily lives.” 113 Even if this
understanding is not resident in a majority of the military personnel, there will need to be a core of
personnel who possess sufficient knowledge and technical ability to audit the output of AI and the
various algorithms that will be used. 114
Responsibility
Some contend that human beings may not always be responsible for the behavior of machines
(technologies) because of artificial agents that have the capacity to learn as they operate. This is
described by Mathias as a responsibility gap, and he notes,
Presently there are machines in development or already in use which are able to decide on a course
of action and to act without human intervention. The rules by which they act are not set during the
production process, but can be changed during the operation of the machine, by the machine
itself. 115
However, as Johnson writes, “A responsibility gap will not arise merely from the technological
complexity of artificial agents. . . . Whether or not there will ever be a responsibility gap depends
on human choices not technological complexity.” 116 Horowitz and Schafer have also written,
“Weapons themselves do not comply with the laws of war. Weapons are used by people in ways
that comply with, or violate, the laws of war.” 117 Leaving humans in the loop and humans on the
loop will be critical in both the design and early deployment of lethal systems in human-machine
teaming.
It will be important for the issue of responsibility to decide who, or what, makes the decision for a
robot to kill. Some situations may develop so quickly and require such rapid information
processing that we would want to entrust our robots and systems to make critical decisions. If
112 Misbehaving algorithms is the subject of a recent report from RAND. Osoba and Welser, An Intelligence in Our Image.
113 Will Knight, “The Dark Secret at the Heart of AI,” MIT Technology Review, April 11, 2017, available at
https://www.technologyreview.com/s/604087/the-dark-secret-at-the-heart-of-ai/.
114 Osoba and Welser, An Intelligence in our Image, p. 25.
115 Andreas Matthias, “The Responsibility Gap: Ascribing Responsibility for the Actions of Learning Automata,” Ethics and
Information Technology 6, no.3, September 2004.
116 Deborah G. Johnson, “Technology with No Human Responsibility,” Journal of Business Ethics 127, no. 4, April 2015.
117 Michael C. Horowitz and Paul Scharre, “The Morality of Robot War,” The New York Times, May 26, 2015, available at
https://www.nytimes.com/2015/05/27/opinion/the-morality-of-robotic-war.html.
36 CSBA | HUMAN-MACHINE TEAMING FOR FUTURE GROUND FORCES
human soldiers monitor the actions of each robot as they occur, it may limit the effectiveness for
which the robot was designed in the first place. 118
It might be argued that this Rubicon was crossed when Predator drones were armed in the early
2000s and used widely to dispatch terrorists and insurgents with their missile armament.
Similarly, ISIS terrorists used ground and air unmanned vehicles to deliver lethal payloads in the
Battle for Mosul in 2017. But ultimately the decision to kill humans by these machines was made
by a human operator. At least in the case of the U.S. and UK, it was done in accordance with
endorsed rules of engagement. 119 But military organizations must look beyond this and examine
whether it is desirable to have robots able to kill humans based on automated processes and
without a human in the decision cycle.
Civil-Military Relations
The relationship between military organizations, and the societies they serve, may be stressed if
augmentation is applied to military personnel. It is likely that mechanical, and eventually
cognitive, augmentation will be expensive and beyond the means of most people. As Yuval Harari
has noted, this could see a differentiation in how society views augmented and non-augmented
people. Harari notes,
Splitting humankind into biological castes will destroy the foundations of liberal ideology.
Liberalism can coexist with socioeconomic gaps. Indeed, since it favours liberty over equality, it
takes such gaps for granted. However, liberalism still presupposes that all human beings have equal
value and authority. 120
In Western democracies, this poses profound questions about equality and the value of individuals
within society. Ground forces must examine this and make a compelling case to their governments
and wider societies if they are to implement human augmentation.
118 Lin, Bekey, and Abney, Autonomous Military Robotics, pp. 74–75.
119 For more on the debate on the use of armed unmanned aircraft and the laws of armed conflict, see Milena Sterio, “The United
States' Use of Drones in the War on Terror: The (Il)legality of Targeted Killings under International Law,” Case Western Reserve
Journal of International Law 45, no. 1, 2012; Kenneth Anderson, Daniel Reisner, and Matthew Waxman, “Adapting the Law of
Armed Conflict to Autonomous Weapon Systems,” International Law Studies 90, no. 1, 2014; and Lynn E. Davis, Michael
McNerney, and Michael D. Greenberg, Clarifying the Rules for Targeted Killing: An Analytical Framework for Policies Involving
Long-Range Armed Drones (Santa Monica, CA: RAND Corporation, 2016).
120 The liberal solution for social inequality is to ensure equal value is given to different human experiences instead of trying to create
the same experiences for everyone. However, will this still work if human augmentation creates real biological gaps? Harari, Homo
Deus, pp. 403–404.
www.csbaonline.org 37
This can be magnified by the bureaucratic inertia that is resident in every large organization.
Therefore, new incentive systems and career structures will be required to enable the institutional
changes that will be driven through human-machine teaming. Changing organizational culture will
need to be a focus for leaders at all levels in the development of more integrated human-machine
teams.
Combat Concepts
Employing human-machine teams will demand examination of how the ground force fights, both
in combined arms and as part of a joint force. This idea of how a future land force might fight is as
important, and potentially more so, than the new technology employed by its people. As Trevor
Dupuy wrote, “The importance of new or imaginative ideas in military affairs, as opposed to simply
new things, can best be gauged by the fact that new ideas have often permitted inferior military
forces to overcome forces that were larger and better equipped.” 122
Currently, all forms of unmanned systems are integrated in human-centric concepts of operations.
This limits the potential of human-machine teams by implicitly designing operations around the
limits in human performance. 123 Future warfighting concepts for a human-machine ground force
must move beyond this construct and be developed well in advance of major investment decisions.
121 The issue of resistance to change in military organization is examined in a range of books and other publications. Among the best
are: Stephen Peter Rosen, Winning the Next War: Innovation and the Modern Military (Ithaca, NY: Cornell University Press,
1991); James S. Corum, The Roots of Blitzkreig: Hans von Seeckt and German Military Reform (Lawrence, KS: University Press
of Kansas, 1992); Williamson Murray and Allan R. Millet, eds., Military Innovation in the Interwar Period, revised edition (New
York: Cambridge University Press, 1998); MacGregor Knox and Williamson Murray, eds., The Dynamics of Military Revolution
1300–2050 (New York: Cambridge University Press, 2001); David E. Johnson, Fast Tanks and Heavy Bombers: Innovation in
the U.S. Army 1917–1945 (Ithaca, NY: Cornell University Press, 2003); Michael E. O’Hanlon, Technological Change and the
Future of Warfare (Washington, DC: The Brookings Institution, 2000); Michael C. Horowitz, The Diffusion of Military Power:
Causes and Consequences for International Politics (Princeton, NJ: Princeton University Press, 2010); and Dima Adamsky, The
Culture of Military Innovation: The Impact of Cultural Factors on the Revolution in Military Affairs in Russia, the US, and
Israel (Stanford, CA: Stanford University Press, 2010).
122 Dupuy, The Evolution of Weapons and Warfare, p. 316.
123 Ilachinski, AI, Robots, and Swarms, p. xvi.
38 CSBA | HUMAN-MACHINE TEAMING FOR FUTURE GROUND FORCES
These issues, which incorporate the technical, the ethical, and the organizational, must be
addressed if effective human-machine teams are to be designed and employed during operations.
The persistent threat to friendly networks is well known and described in a range of official and
academic publications. 124 Any design for integrated ground force human-machine teams must be
founded on resilient, robust, and trusted networks that connect humans and machines within, and
between, teams. Defense research organizations such as the U.S. Army Research Laboratory, 125 the
United Kingdom’s Defense Science and Technology Laboratory, 126 and Australia’s Defense Science
and Technology Group 127 all have active programs seeking to provide more resilient and reliable
networks for future military operations.
This challenge is particularly acute when humans must trust their lives to decision support by
algorithms within networks. Their security must be assured. In hardening the networks that
enable human-machine teaming, security must not compromise the speed—or tempo—of
operations. To achieve secure networks, ground forces must ensure they don’t give away the
competitive advantage that is desired through the application of AI and robotics in networked
human-machine teams.
A new approach to education and training will be required in an integrated human-machine force.
Throughout history, military training has focused on teaching humans to achieve outcomes as
individuals and teams. In a human-machine ground force, this foundational approach to training
is challenged. Similarly, education for military leaders currently seeks to achieve their intellectual
124 Key national strategies that address this issue include the American The Department of Defense Cyber Strategy (April 2015); the
Australian Australia’s Cyber Security Strategy, 2016 (2016); and the UK’s National Cyber Security Strategy 2016-2021 (2016).
125 The U.S. Army is progressing a range of inter-connected networking security initiatives under their Information Sciences
Campaign Plan.
126 Among the UK programs is a Contested Electromagnetic Program; see more at https://www.gov.uk/guidance/contested-
electromagnetic-environment-programme.
127 The Australian research program in this area is conducted under the Defense Science and Technology Group’s Information and
Communications program; see https://www.dst.defence.gov.au/research-area/information-and-communications.
www.csbaonline.org 39
development in the art and science of war. If learning machines are added to this environment,
both institutional and individual professional military education must adapt.
Part of any new training and education approach will be developing the understanding of common
goals in human-machine teams. Humans and autonomous machines will need common goals and
share awareness if they are to work together effectively. Many commercial aircraft accidents
associated with automation occurred when the flight crew had one goal (for example, staying on
the glide slope during an approach) and the flight management computer had another (such as
executing a go-around). 128 Called “automation surprises,” this is where human operators have
failed to observe or intervene in machine behavior, resulting in undesirable outcomes. Deploying
future autonomous systems will demand good human training, as well as increasing each
machine’s awareness of what the human or humans are trying to achieve.
Career development and management of military personnel will also need to be adapted to the new
human-machine force. As robots replace humans in many “dirty, dull and dangerous” functions, it
is possible that many lower ranking soldiers may be displaced. This will necessitate a change to the
traditional career pyramids, where the mass of an Army is found in the lowest ranks.
Costs
The financial cost of robotic systems, AI, and augmenting humans will add to existing challenges
of institutional budgeting. The amount of investment likely to achieve a deeper level of human-
machine integration—in research, experimentation, and deploying systems—is likely to run into
tens of billions of dollars across Western ground forces. For example, the United States Defense
Advanced Research Projects Agency in Fiscal Year 2018 alone plans to spend over U.S. $735
million in two of its offices that undertake research into robotics and algorithms. 129 Ground forces
will need to decide on an appropriate level of investment in such an approach, as opposed to the
procurement of more traditional human-operated equipment over the next two decades.
Procurement Bureaucracy
Related to this is the challenge of reforming Defense procurement bureaucracies to support a shift
to a more integrated human-machine design. Procurement policy and processes in Western
nations are generally designed to minimize procurement risk for governments and government
departments and ensure that taxpayer resources are maximized with minimal waste. Such long
128 One useful study of multiple aircraft accidents that covers these issues with automation is Christine Negroni, The Crash
Detectives: Investigating the World’s Most Mysterious Air Disasters (New York: Penguin Books, 2016).
129 The DARPA Tactical Technology Office funds research into robotics—among other areas—and has an FY 2018 allocation of
US$343 million. The Information Innovation Office funds research into algorithms and cyber issues and has an FY2018 budget of
US$392 million. See the DARPA Budget at https://www.darpa.mil/about-us/budget.
40 CSBA | HUMAN-MACHINE TEAMING FOR FUTURE GROUND FORCES
procurement processes are unlikely to be suitable for an environment where AI developments and
their applications are moving rapidly.
There is evidence that the world is now in the midst of a 4th Industrial Revolution. 130 However, the
majority of Western military organizations are representative of 2nd and 3rd Industrial
Revolution 131 structures and processes. Unless the decision-making approaches of these military
organizations can be reformed and exploit the pace of technological change, they will face a
growing “modernization gap,” 132 as described recently by retired Major General Robert Scales. 133
Informed Consent
Consent by soldiers to the risks of working with robots and AI will be an issue that must be
addressed. In October 2007, a semi-autonomous cannon deployed by the South African army
malfunctioned, killing nine soldiers and wounding 14 others. 134 This is both a legal and an ethical
quandary; should soldiers be informed that unusual or new risks exist, such as working with
potentially defective robots? 135
130 A key proponent is Professor Klaus Schwab, but his themes have been applied more broadly to represent the changes in industry
and society being driven by a convergence of technological breakthroughs in areas such as nanotechnology, quantum theory,
biotechnology, autonomy, robotics, and artificial intelligence. For more see, Klaus Schwab, The Fourth Industrial Revolution
(New York: Crown Business, January 2017).
131 The 2nd Industrial Revolution saw the arrival of telephones, electric lights, and the internal combustion engine, and it ended in
1914. The 3rd Industrial Revolution heralded the digital age and saw the shift from analog to digital technologies, commencing in
the 1980s with new technologies such as personal computers and the wide availability of the internet.
132 Author’s term.
133 Scales wrote that “the art of predicting the course of war is made far more difficult by a quickening of the rate of change among
those variables most likely to influence conflict such as technology, domestic politics, and international events. While the pace of
influencing events is accelerating the capacity of militaries to build weapons and structures to accommodate change is slowing.
Thus soldiers today must cast farther and farther out to stay ahead.” Robert Scales, “Forecasting the Future of Warfare,” War on
the Rocks, April 9, 2018, available at https://warontherocks.com/2018/04/forecasting-the-future-of-warfare/.
134 Noah Shachtman, “Robot Cannon Kills 9, Wounds 14,” Wired Magazine, October 18, 2007, available at
https://www.wired.com/2007/10/robot-cannon-ki/.
135 Lin, Bekey, and Abney, Autonomous Military Robotics, p. 74.
www.csbaonline.org 41
Human Dignity
Philosophers such as Peter Asaro and Robert Sparrow, as well as non-government organizations
and the Vatican, have argued that delegating the decision to target and open fire to a machine
violates human dignity. At its core, this is a proposition that people have the “right not to be killed
by a machine.” 136 Technologists such as Stephen Hawking 137 and Elon Musk 138 have also spoken
out against AI. Others, accepting that autonomous systems and AI are likely to play a large role in
war and society in the future, have proposed concurrent efforts to develop technology and examine
the ethics of these systems. As Amir Husain notes, “If autonomous systems are to be a pillar of
future supremacy, then now is the right time to present a framework within which autonomy can
be enabled in an effective and technically viable, yet legal and moral, manner.” 139
Trust
Perhaps the most important challenge is establishing trust. The degree to which military leaders
should trust advanced analytics and artificial intelligence to make decisions about its people, and
potentially make decisions about saving and taking lives, must be established. The level of trust
people will place on objects implanted in their bodies to improve performance is also to be
established. Finally, for augmented humans, it is yet to be proven whether they will be trusted in
teams composed of augmented and non-augmented humans.
The current approach of many ground forces in employing mission command and decentralized
command approaches offers some insights into how these questions might be at least partially
answered. Mission command—the practice of assigning a subordinate commander a mission
without specifying how the mission is to be achieved 140—has established the foundational
relationships between commanders and subordinates that might be adapted for trusted
relationships with robots and AI. But it does not fully address the issues that must be examined
and resolved in building a hybrid human-machine ground force.
Mankind has wrestled with these types of challenges before. For centuries, the application of new
technologies in war has generated debate during, before, and after conflict. In 1139, Pope Innocent
II led the Second Lateran Council in banning the use of crossbows in war. At the time the crossbow
represented very advanced technology that required minimal training and little strength and
possessed unparalleled lethality. A hastily-trained peasant could use the weapon, challenging the
existing power structure in war. The Roman Catholic Church viewed the new weapon as a gross
transformation of the nature of war. 141 This reaction to a new means of war illustrates the type of
concerns that military organizations must address in adopting plans for human-machine
integration. Ground forces must possess realistic strategies that address these challenges if they are
to successfully exploit the future of human-machine teaming.
Chapter 4
Building a Competitive
Advantage: A Strategy for
Future Ground Forces
Carl Builder once wrote, “An Army requires a theory of an Army. There must exist something in
addition to its soldiers and tanks and guns—a concept, a strategy, a notion of who it is and what it
wants to be, and what it is about and what it wants to be about.” 142 In embracing a future vision of
human-machine teaming, ground forces’ future ideas of themselves will change. To fully exploit
the potential of human-machine teaming, ground forces will require strategies to describe desired
capability outcomes and priorities for investment. Concurrently, they must undertake doctrinal,
conceptual, and culture adaptation. To that end, this final part of the study offers a strategy for
ground forces to develop human-machine teams over the short and medium terms. 143
Any strategy that is founded solely on the technology of human-machine integration is likely to
fail. Multiple authors have examined innovation and sources of success over centuries of military
history. A key finding, by authors such as Williamson Murray, Trevor Dupuy, and Stephen Peter
Rosen, is that technology alone is unlikely to provide a sustained competitive advantage. But when
142 Carl Builder, The Masks of War: American Military Styles in Strategy and Analysis, a RAND Corporation Research Study
(Baltimore, MD: Johns Hopkins University Press, 1989).
143 While the Australian Army and the ADF lack such an approach, institutions such as the U.S. Department of Defense and the U.S.
Army have already released comprehensive strategies for their development and employment of autonomous systems and AI in
future human-machine teams. See U.S. Army, Robotic and Autonomous Systems Strategy.
44 CSBA | HUMAN-MACHINE TEAMING FOR FUTURE GROUND FORCES
new technology is combined with new operating concepts and new organizational models,
innovation is more likely to provide sustained advantage.
This was also a finding of the Pentagon’s Office of Net Assessment in its examination of the impact
of new technology in the late 1980s and early 1990s. Then Director of Net Assessment, Andrew W.
Marshall, wrote,
A large part of pre-eminence . . . will reside in superior ideas with respect to concepts of operation
and in organizational innovation. Indeed, being ahead in concepts of operation and in organizational
arrangements may be far more enduring than any advantages in technology or weapons systems
employing them. 144
The best way to begin a strategy is to describe the ends of that strategy. The U.S. Army has recently
described its ends in the employment of robotic and autonomous systems: “Army formations use
RAS to increase combat effectiveness and to maintain overmatch in combined arms operations
against capable enemies.” 145 This is a useful starting point, but a broader notion of systemic
employment of robotic systems, AI, and augmented humans is required. Therefore, for ground
forces, an appropriate statement of the ends might be:
Ground forces employ robotic systems, AI, and augmented humans in integrated human-
machine teams in preparing the land force at home and in combat to generate operational and
strategic leverage, balancing operational and enterprise effectiveness, affordability, and
institutional values.
Specific targets are also required. Initial targets would be the subject of experiments. Experiments
in human-robot teaming could provide an initial core around which all subsequent human-
machine teaming programs are built. Initial experiments in areas such as ISR battle, breach and
break action, and close combat would include modest human-robot integration activities. Lessons
from these experiments would provide the foundation for follow-on human-machine teaming
activities. Subsequent experiments could expand in scope and complexity for human-robot
teaming and then gradually include human-AI integration and augmented humans.
144 Andrew W. Marshall, “Some Thoughts on Military Revolutions—Second Version,” Memorandum for the Record, Office of the
Secretary of Defense, Office of Net Assessment, August 23, 1993, p. 3.
145 U.S. Army, Robotic and Autonomous Systems Strategy, pp. 11–12.
www.csbaonline.org 45
Monitor: Ground forces should possess programs that are focused on monitoring developments
in robotics, AI, big data analytics, and human augmentation. This will provide institutional
understanding of where best practice and the most effective approaches are being implemented. It
will keep ground force designers informed of potential capability acquisitions and it will encourage
an appreciation of what possible enemies might be doing in this field. This monitor function
should be a collaboration among Western ground forces, potentially with the U.S. Army as the lead
agency.
Design: Ground force modernization staff must also commence designing the next generation
human-machine teams. 146 This can be informed by teaming experiments, and collaboration with
academia, industry, and allies. It must also be informed by employing agreed-upon long-term
goals that stretch the imagination and push ground forces into new and innovative designs.
Ground force designers should not be afraid to produce and discard many different models in this
process. These designs must include the connective tissue of team—secure and assured networks.
This design phase must include a discourse on the various ethical issues involved. The military
must not be caught wrong-footed by some of the issues that are now facing technology companies
such as Facebook, Google, and Twitter. Established as small startup companies, they evolved
quickly with growing user bases. But, these companies are now facing questions from governments
in the United States and Europe concerning their influence on society. One co-founder of Twitter
has noted that “there wasn’t time to think through the repercussions of everything we did” in the
early days of the company. 147 Given the lethality of many of the systems involved, designers of
human-machine teaming for ground forces cannot afford to make similar mistakes.
Another element of this design effort must be a redesign of education and training models.
Training people to develop individual mastery and to be proficient within teams at different scales
is the current focus of the Army training model. This model must change when large numbers of
autonomous and semi-autonomous machines are introduced at multiple levels across the combat
146 One interesting recent design for a future land combat force—albeit at unit level—is contained in Jeff Becker, “How Lethal, Mobile,
Protected and Aware?” Small Wars Journal, July 24, 2017, available at http://smallwarsjournal.com/jrnl/art/how-lethal-mobile-
protected-and-aware.
147 David Streitfeld, “Changing the World, but Not Quite the Way They Had Imagined,” The New York Times, October 13, 2017, pp.
A1, A11.
46 CSBA | HUMAN-MACHINE TEAMING FOR FUTURE GROUND FORCES
force and wider enterprise. It will also drive a wider adoption of simulation to develop effective
human-machine teams.
At the individual and small unit levels, more integrated human-machine teaming will increase
cognitive loads on almost every soldier. For leaders, building best practices in employing AI will
result in changes to officer and junior leader development as well as institutional professional
military education programs. Where human augmentation is implemented, developing trust
between augmented and non-augmented team members will be an important outcome of training.
Overall, this will drive change in attracting and recruiting new personnel, educational programs to
enhance intellectual resilience, and the broader approach to education and training.
The design of this new integrated human-machine force will also drive changes in military
personnel management. Current military personnel models are constructed upon the all-volunteer
professionalized military model which emerged in the later stages of the Cold War. Automation
and AI is changing workforce patterns in civil industry. Military personnel experts will also have to
change their approach to personnel modelling and career development. Military personnel in this
new human-machine ground force will be involved in very different forms of work. Some old
specialties are likely to disappear, and new ones will be formed. Military personnel experts should
anticipate these changes and begin their strategic planning for this now.
Another aspect of design must be planning for organizational and cultural change that will
accompany the adoption of large numbers of robots, AI, and augmented humans. This should
blend new technology, new operating concepts, and innovative organizational design. There are
many guides to successful implementation of organizational change. For military innovation,
Murray and Millet, 148 Rosen, 149 and O’Hanlon 150 all offer lessons in adopting new ideas and
technologies. However, one of the most relevant guides to organizational change is a short 1983
article from Military Review entitled “To Change an Army.” In this article, retired General Don
Starry examined how the U.S. Army reformed itself after Vietnam. It offers several mechanisms
that would—in a military organization—provide the best chance for organizations to adapt
successfully. 151 Three are relevant in the design for organizational change:
148 Their three-part series, Military Effectiveness (1988), and the later Military Innovation in the Interwar Period are very fine
examinations of military innovation and organizational change. Allan R. Millett and Williamson Murray, eds., Military
Effectiveness, volume 1, The First World War (London: Unwin Hyman, 1988); Allan R. Millett and Williamson Murray, eds.,
Military Effectiveness, volume 2, The Interwar Period (London: Unwin Hyman, 1988); Allan R. Millett and Williamson Murray,
eds., Military Effectiveness, volume 3, The Second World War (London: Unwin Hyman, 1988); Murray and Millett, eds., Military
Innovation in the Interwar Period.
149 Rosen, Winning the Next War.
150 O’Hanlon, Technological Change and the Future of Warfare.
151 Donn A. Starry, “To Change an Army,” Military Review 63, no. 3, March 1983.
www.csbaonline.org 47
1. Institutional mechanisms must be established that are able to identify the need for change,
draw up parameters for change, and describe clearly what must be done where change is
required.
2. Someone at or near the top of the institution must be willing to hear out arguments for
change, agree to the requirement for change, embrace the new operational concepts, and
become a champion for that change.
3. There must senior advocacy for change. This advocacy must build a consensus that will give
the new ideas, and the need to adopt them, a wider audience of converts and believers. 152
These will be vital elements of the design for change—and for a ground forces’ overall design
process to build a more integrated human-machine land force.
Experiment: Finally, ground forces must experiment. There are sufficient robotic systems,
advanced analytic capabilities, early generation augmentable capability, and manned-unmanned
teaming concepts to commence experiments now. 153 These experiments should initially entail the
battlefield use of robots as a core function around which AI and augmentation could subsequently
be applied.
Where fully realized systems are not yet available, sophisticated simulation programs can be used
to underpin experimentation. Modeling will also be required on the fleet management of many
types of robots. As most robots will be employed in teams with other robots, the fleet management
of robots may take on many of the aspects of the extant personnel management processes,
combined with best practice logistic fleet management.
These experiments might be restricted to the ground force of a single nation, or they may be
conducted with allies or with industry and academia. For example, ground forces might partner
with indigenous universities and other research centers. Western ground forces might also
undertake collaborative experimentation with the U.S. Army as it implements its 2017 Robotics
and Autonomous Systems strategy to develop the capacities to reduce the number of humans in
harm’s way, increase the speed of decision-making in time-critical operations, and to perform
mission sets that are not possible for humans to perform. 154 Working with academia and with the
U.S. Army, Western ground forces could establish small teams to generate prototype robotic, AI,
and augmentation solutions. Regardless of the methods selected for experimentation, ground
forces are facing a profound change in how they design future structures. These experiments will
allow them to learn, to fail, and to inform successive ground force designs.
Decide: Early in the next decade, ground forces will need to make substantial decisions about
their future. These decisions will include the shape, size, and look of the land force. Other
decisions will include the balance of combat and non-combat work forces, and how training and
education systems must be adapted to this entirely new human-robot workforce construct. Ground
forces will also need to decide how comfortable they are with employing lethal autonomous
systems and AI systems making decisions about planning, intelligence, and logistics. These
decisions must be informed by discourse throughout military institutions as well as wider national
communities. Preferably, this discourse should be occurring now to allow time to work through the
ethical and other challenges involved in human-machine teaming.
Invest: Based on these decisions, ground forces will need to invest. Different nations will assess
their requirements for human-machine teaming differently. There is unlikely to be a single
approach to the level of investment applied by various Western ground forces. But, development of
human-machine teams may demand that ground forces also invest in the capability for developing
robots and artificial intelligence that is resident in their national universities, and larger corporate
research departments. A “sovereign capability” approach for each nation could provide a more
secure approach to developing key technologies—and a foundation for sharing different
approaches.
Investment will also be required in the education and training of people that will underpin this
more integrated human-machine teaming. The training of ground forces must be reoriented to
balance the development of people in individual mastery with building human-machine teams.
Similarly, investment in an improved professional military education will be required. This could
be focused on enhancing individual capacity to handle increased cognitive loads and building
highly effective processes for incorporating AI into planning and decision support at multiple
levels of command. Institutional professional military education programs will need to continue
emphasis on mastery of the intellectual aspects of the profession of arms. However, courses that
deliver learning outcomes in planning, tactics, and the execution of operations must incorporate
the employment of AI.
of the human nature of war complemented with a broad understanding of a wide variety of
technical elements of the profession.
There will be many who will be skeptical or even afraid of some of these human-machine systems.
The lethal, autonomous systems are likely to give pause to political and community leaders.
Ground forces must apply their strategic communications capability to inform these external
actors about its plans and force structure aspirations.
50 CSBA | HUMAN-MACHINE TEAMING FOR FUTURE GROUND FORCES
www.csbaonline.org 51
Conclusion
In the early twenty-first century, the train of progress is again pulling out of the station—and this
will probably be the last train ever to leave the station called Homo Sapiens. Those who miss this
train will never get a second chance. In order to get a seat on it you need to understand twenty-
first century technology, and in particular the powers of biotechnology and computer algorithms. .
. . Those left behind will face extinction.
Society has yet to reach the technological mastery of the robots described by Isaac Asimov and
Phillip K. Dick. 156 Nor has mankind achieved the levels of artificial intelligence shown in films such
as Stanley Kubrick’s masterpiece 2001 A Space Odyssey 157 or the cognitive augmentation
described by John Scalzi in Old Man’s War. 158 But a review of historical technology development
shows that exponential patterns of development extend beyond Moore’s law. Wireless capacity
doubles every nine months. Internet bandwidth backbone is doubling roughly every twelve
months. During the 20th century, the range and effectiveness of artillery increased by a factor of
twenty and antitank fire by a factor of sixty. 159 There is sufficient evidence to suggest robotics, AI,
and augmentation will chart a similar path in capability growth.
Human-machine teaming has the potential to make soldiers much more effective—on the
battlefield and while training at home. The capabilities of new robots and AI, and the potential of
augmentation, also offer a potential revolutionary shift in how ground forces plan, train, and fight.
It requires ground forces to be reoriented around the opportunities of the human-machine
revolution. But it is an area replete with complex issues. Technological aspects cannot—and must
not—be separated from ethical, moral, and legal issues.
At its core, human-machine teaming is not just about better technology for ground forces. It is
about how clever organizations might leverage advancing technology and combine it with new
ideas on warfighting and new institutional and warfighting structures. This has been the basis of
all successful historical step-changes in military capability. 160 It will demand a sustained
institutional and strategic focus on experimentation, research, and development. It will need
significant investment to build what may become a decisive military edge for allied ground forces
by the middle of the 21st century.
The broad potential applicability of these systems means that ground forces will need to adopt an
enterprise approach to the employment of human-machine teams. Human-robot teams can be
used in training institutions, potentially freeing up personnel to be re-deployed into other
functions. Advanced computing and analytical capacity may be very useful in human-AI teams for
strategic decision-making, capability development, resource allocation, and talent management.
Beyond investing in the force, Western armies may choose to support the development of
sovereign capabilities for developing robots and artificial intelligence. With the potential scale for
procurement, and inherent security risks of networked lethal robotic systems, allied nations
should aspire to sustain national robot, AI, and augmentation research, design, and manufacture
capabilities to support military and other applications.
This study offers a roadmap for more capable future ground forces. Ground forces should exploit
the utility of human-machine teams to generate advantage on the battlefield and improve the
effectiveness and efficiency of military activities at home. As one recent author noted, “Countries
who fail to adequately develop autonomous warfighting systems are more likely to be the victims
160 Andrew F. Krepinevich and Barry D. Watts, The Last Warrior: Andrew Marshall and the Shaping of Modern American Defense
Strategy (New York: Basic Books, 2015), p. 200.
www.csbaonline.org 53
of them. The autonomous arms race has begun. Liberal societies can either participate or risk the
loss of a military advantage that might one day endanger their sovereignty.” 161 This is a capability
that future ground forces must possess. They have the opportunity now, with prudent strategic
planning and investment, to build a winning capability advantage.
161 Jules Hurst, “Intervention and the Looming Choices of Autonomous Warfighting,” War on the Rocks, August 25, 2016, available at
https://warontherocks.com/2016/08/intervention-and-the-looming-choices-of-autonomous-warfighting/.
54 CSBA | HUMAN-MACHINE TEAMING FOR FUTURE GROUND FORCES
www.csbaonline.org 55
The Sergeant Major turned and spoke softly. “How’s the skipper doing?”
Kinsey pondered this for a minute. “The micro-drone that caught his scent nearly killed him. . . .
He still ended up with a decent head wound. The doctor bots are plugging him in. He should be
up and about in a few weeks.” The new cognitive implants installed by field hospital surgical
robots would replace most of the function lost by the shrapnel wounds to their commander’s
brain. With some intensive physical and cognitive therapy, he’d be walking again within a
month.
The two men both stared out across the now quietened city. The operation had gone generally
according to plan several days before. They had lost more unmanned ground vehicles than
Merlin had forecast, but they could be replaced relatively quickly from their field additive
fabrication facility. The human toll, while heartbreaking, had been much lighter than they had
anticipated.
They had employed the old Kilcullen ideas about the metabolism of the city and key node control.
Now their battlegroup and the hundreds of unmanned ground vehicles and aerial drones
accompanying them were, according to Merlin, deployed in an optimal operating disposition.
There were four more days before the relief-in-place with the follow-on forces would arrive.
“Well then sir, I need to do another circulation before hitting the sack. I will see you on the VR for
orders in the morning.”
The battlegroup operations officer watched the Sergeant Major move away, quiet as a wraith.
Major Kinsey turned and walked toward his low signature command vehicle. He had ten
56 CSBA | HUMAN-MACHINE TEAMING FOR FUTURE GROUND FORCES
minutes before the virtual planning group with the Brigade operations team for deploying the
new stealth unmanned ground vehicle motherships on their next mission in two weeks.
www.csbaonline.org 57
Acronyms
AI artificial intelligence
UK United Kingdom
VR virtual reality
JAPAN-FO FFI AL USE ONLY
OGIS
INTEROPERABILITY WITH THE USN PROJECT
FINAL REPORT
31 MARCH 2018
BRYAN CLARK, TIMOTHY WALTON, GILLIAN EVANS,
1667 K Street, NW, Suite 900
Washington, DC 20006
Tel. 202-331-7990 • Fax 202-331-8019
www.csbaonline.org
HARRISON SCHRAMM, AND RYAN BOONE
JAPAN-FOR OFFICIAL USE ONLY