NAVAL

POSTGRADUATE
SCHOOL
MONTEREY, CALIFORNIA

THESIS

AUTOMATION AND ARTIFICIAL INTELLIGENCE

FOR NAVAL ISR: U.S. NAVY VS. CHINA’S NAVY

by

Crystal S. Hong

June 2020

Thesis Advisor: John J. Arquilla

Co-Advisor: Douglas L. Van Bossuyt
Approved for public release. Distribution is unlimited.
THIS PAGE INTENTIONALLY LEFT BLANK
 Form Approved OMB
REPORT DOCUMENTATION PAGE No. 0704-0188
Public reporting burden for this collection of information is estimated to average 1 hour per response, including the time for reviewing
instruction, searching existing data sources, gathering and maintaining the data needed, and completing and reviewing the collection of
information. Send comments regarding this burden estimate or any other aspect of this collection of information, including suggestions
for reducing this burden, to Washington headquarters Services, Directorate for Information Operations and Reports, 1215 Jefferson
Davis Highway, Suite 1204, Arlington, VA 22202-4302, and to the Office of Management and Budget, Paperwork Reduction Project
(0704-0188) Washington, DC 20503.
1. AGENCY USE ONLY 2. REPORT DATE 3. REPORT TYPE AND DATES COVERED
(Leave blank) June 2020 Master’s thesis
4. TITLE AND SUBTITLE 5. FUNDING NUMBERS
AUTOMATION AND ARTIFICIAL INTELLIGENCE FOR NAVAL ISR: U.S.
NAVY VS. CHINA’S NAVY
6. AUTHOR(S) Crystal S. Hong

7. PERFORMING ORGANIZATION NAME(S) AND ADDRESS(ES) 8. PERFORMING

Naval Postgraduate School ORGANIZATION REPORT
Monterey, CA 93943-5000 NUMBER
9. SPONSORING / MONITORING AGENCY NAME(S) AND 10. SPONSORING /
ADDRESS(ES) MONITORING AGENCY
N/A REPORT NUMBER
11. SUPPLEMENTARY NOTES The views expressed in this thesis are those of the author and do not reflect the
official policy or position of the Department of Defense or the U.S. Government.
12a. DISTRIBUTION / AVAILABILITY STATEMENT 12b. DISTRIBUTION CODE
Approved for public release. Distribution is unlimited. A
13. ABSTRACT (maximum 200 words)
The U.S. Navy faces challenges as it moves toward automating the maritime battlespace and risks
falling behind its rising great power competitor, the People’s Liberation Army Navy (PLAN). How are the
U.S. Navy and the PLAN adopting automation to improve naval intelligence, surveillance, and
reconnaissance (ISR)? Results of this study indicate that the U.S. Navy is an innovator and early adopter,
while the PLAN has embraced automated systems and artificial intelligence (AI) as a late modernizer,
benefiting from knowledge of already relevant technologies. The U.S. Navy’s Aegis and Ship Self
Defense System and AI technologies enable maritime superiority; however, the PLAN is advancing in AI
technologies faster than the U.S. Navy. This thesis compares the two navies in their adoption of automation
and AI technologies for ISR. For purposes of this study, automation is defined as a process or specific,
task-oriented system that operates without immediate human control. AI goes deeper and includes advances
aimed at creating machines able to analyze, evaluate and optimize alternatives in pursuit of broader aims. I
employ Everett Rogers’ S-Curve model of the diffusion process as a framework for analyzing efforts
to increase efficiency of naval planners and decision makers as they ponder which automated and
AI technologies to adopt and how best to utilize them.

14. SUBJECT TERMS 15. NUMBER OF

automation, artificial intelligence, U.S. Navy, China, PLAN, People’s Liberation Army PAGES
Navy, Great Power Competition, AI tools, special operations, Liaoning, aircraft carrier, 157
Aegis, Dragon Eye, Type 346 16. PRICE CODE

17. SECURITY 18. SECURITY 19. SECURITY 20. LIMITATION OF

CLASSIFICATION OF CLASSIFICATION OF THIS CLASSIFICATION OF ABSTRACT
REPORT PAGE ABSTRACT
Unclassified Unclassified Unclassified UU

NSN 7540-01-280-5500 Standard Form 298 (Rev. 2-89)

Prescribed by ANSI Std. 239-18

i
THIS PAGE INTENTIONALLY LEFT BLANK

ii
 Approved for public release. Distribution is unlimited.

AUTOMATION AND ARTIFICIAL INTELLIGENCE FOR NAVAL ISR:

U.S. NAVY VS. CHINA’S NAVY

Crystal S. Hong
Lieutenant, United States Navy
BA, University of California - Los Angeles, 2015

Submitted in partial fulfillment of the

requirements for the degree of

MASTER OF SCIENCE IN DEFENSE ANALYSIS

(IRREGULAR WARFARE)

from the

NAVAL POSTGRADUATE SCHOOL

June 2020

Approved by: John J. Arquilla

Advisor

Douglas L. Van Bossuyt

Co-Advisor

Kalev I. Sepp
Chair, Department of Defense Analysis

iii
THIS PAGE INTENTIONALLY LEFT BLANK

iv
 ABSTRACT

The U.S. Navy faces challenges as it moves toward automating the maritime
battlespace and risks falling behind its rising great power competitor, the People’s
Liberation Army Navy (PLAN). How are the U.S. Navy and the PLAN adopting
automation to improve naval intelligence, surveillance, and reconnaissance (ISR)?
Results of this study indicate that the U.S. Navy is an innovator and early adopter, while
the PLAN has embraced automated systems and artificial intelligence (AI) as a late
modernizer, benefiting from knowledge of already relevant technologies. The U.S.
Navy’s Aegis and Ship Self Defense System and AI technologies enable maritime
superiority; however, the PLAN is advancing in AI technologies faster than the U.S.
Navy. This thesis compares the two navies in their adoption of automation and AI
technologies for ISR. For purposes of this study, automation is defined as a process or
specific, task-oriented system that operates without immediate human control. AI goes
deeper and includes advances aimed at creating machines able to analyze, evaluate and
optimize alternatives in pursuit of broader aims. I employ Everett Rogers’ S-Curve model
of the diffusion process as a framework for analyzing efforts to increase efficiency of
naval planners and decision makers as they ponder which automated and AI technologies
to adopt and how best to utilize them.

v
THIS PAGE INTENTIONALLY LEFT BLANK

vi
 TABLE OF CONTENTS

I. INTRODUCTION..................................................................................................1
A. RESEARCH QUESTION .........................................................................3
B. BACKGROUND ........................................................................................3
C. RESEARCH METHODOLOGY .............................................................5
D. THEORY OF TECHNOLOGICAL ADOPTION ..................................7

II. LITERATURE REVIEW ...................................................................................13

A. AI POLICY AND STRATEGY DOCUMENTS ...................................14
B. PROMINENT LEADERS OF AUTOMATION ...................................16
1. History of Automation from Prominent American
Leaders ..........................................................................................17
2. History of Automation by Chinese Prominent Leaders ...........20
C. ON DEFINING AUTOMATION AND AI ............................................22
1. Working Definitions of Automation and AI ..............................23
2. List of AI Definitions and Strategies ..........................................24
3. Conceptualizing Automation ......................................................29
4. Levels of Automation ...................................................................31
D. NAVAL ISR..............................................................................................33
1. U.S. Naval Leaders Pursue Automation and AI .......................34
2. PRC Naval Leaders Pursue Automation and AI ......................36
3. Technical Naval ISR ....................................................................38
4. U.S. Aegis ......................................................................................40
5. PLAN Type-348 Dragon Eye ......................................................42

III. METHODOLOGY ..............................................................................................45

A. S-CURVE METHODOLOGY................................................................45
B. 7 STEPS TO ADOPTING AUTOMATION AND AI FOR
NAVAL ISR..............................................................................................46
1. Start with a “Campfire Talk” .....................................................46
2. Categorize and Analyze the Competing Naval Leaders of
Other Great Powers .....................................................................46
3. Understand Each Intelligence Organization’s Technical
Capabilities ...................................................................................47
4. Select and Categorize Top-Performing AI through Social
Network Analysis (SNA)..............................................................47
5. Combine Relevant and Operator-Oriented and
Intelligence-Oriented Services ....................................................47
6. Review and Re-orient ..................................................................47
vii
 7. Publish and Make the Plan Accessible .......................................48
C. UNDERSTANDING THE S-CURVE THEORY OF
TECHNOLOGICAL ADOPTION .........................................................48
D. MILLENNIALS, GEN-Z, AND ZENIALS ...........................................50
E. ALTERNATIVE OUTLOOK TO ROGERS’ S-CURVE ....................51

IV. AI SUPERPOWER NAVIES ADOPTING AUTOMATION TODAY ..........53

A. OVERVIEW .............................................................................................53
B. DISCUSSION OF INTELLIGENCE AND TECHNICAL
CAPABILITIES .......................................................................................54
C. VISION AND STRATEGIC NARRATIVES: CATEGORIZE
AND ANALYZE COMPETING NAVAL LEADERS .........................57
1. U.S. Naval Leaders Pursue Automation and AI .........................57
2. PRC Naval Leaders Pursue Automation and AI ........................58
3. Chinese Naval Modernization: String of Pearls and the
New Maritime Silk Road ..............................................................58
D. PLAN ISR AND AI 2020: EXAMPLES OF ADOPTING
AUTOMATION AND AI ........................................................................60
1. PLAN’s AI Strategy and Development Plans..............................60
2. PLAN Uses Automation to Test Its First Laser System
Against U.S. P-8 Poseidon ...........................................................65
3. Modernizing PLAN Aircraft Carrier.........................................65
4. China’s Satellites and Technology Giants .................................66
5. Automation for Long-Range and Early Warning: Joint
Over-the-Horizon Radar .............................................................68
6. Future of Anti-Submarine and AI ..............................................70
7. China’s Observer Corps and AI Stalkers ..................................72
8. Chinese Maritime Militia (CMM) “Little Blue Defenders” .....73
9. Social Media and Open Source ...................................................73
10. PLAN Cyberspace Copy-Cats ....................................................74
E. U.S. NAVY ISR 2020: ADOPTING AUTOMATION AND AI ...........75
1. Ford-Class Aircraft Carrier: Aegis to SSDS Computers
for AI .............................................................................................76
2. Reconnaissance Satellites: ELINT, SIGINT, and Ocean
Surveillance ..................................................................................77
3. Long Range: Over-the-Horizon Backscatter Radar .................80
4. ISR Platforms and Laser Technology ........................................81
5. Cyberspace and Social Media .....................................................84
6. Big Data: Cloud Services and the Semiconductor
Industry.........................................................................................85
7. Naval Special Warfare Adopts AI for Naval ISR .....................87
viii
 8. Private, Public, and Commercial Way of Adopting
Automation and AI ......................................................................88
9. Neural Networks and Deep Learning: Heat Maps, Trees
and, GANs .....................................................................................90

V. SUMMARY AND RECOMMENDATIONS .....................................................93

A. INTRODUCTION....................................................................................93
B. SUMMARY OF FINDINGS ...................................................................94
1. AI Benefits an Autocratic Navy ..................................................95
2. Automation and AI: Changes in International Maritime
Governance ...................................................................................96
3. U.S. Navy’s First Mover Advantage: Semiconductors and
Software ........................................................................................98
4. An Automated Aircraft Carrier May Be Powerless in
Future of AI and Sea Power ........................................................99
C. THE AI GAP: MISSING AI PLATFORM TO CONNECT AI
ENGINEERS AND OPERATORS/USERS ..........................................99
D. RECOMMENDATIONS .......................................................................101
E. THE FUTURE OF SEA POWER AND FUTURE WORK:
HUMAN-MACHINE TEAMING ........................................................104
1. Team AI Health: COVID-19 .....................................................104
2. Quantum Computing .................................................................105
3. Ethics and AI ..............................................................................106
4. Human-Machine Teaming and Virtual Reality
(Augmented Reality) ..................................................................108

APPENDIX .....................................................................................................................111
A. USING AI TOOLS TO FIND AI TALENT ........................................111
B. RESULTS: ANALYSIS OF SPECIFIC NETWORKS ......................116

LIST OF REFERENCES ..............................................................................................119

INITIAL DISTRIBUTION LIST .................................................................................135

ix
THIS PAGE INTENTIONALLY LEFT BLANK

x
 LIST OF FIGURES

Figure 1. S-curve Model of the Diffusion Process ......................................................8

Figure 2. Human Decisions Supplemented by AI Along with AI Continuum .........30

Figure 3. Common Operational Picture for Situational Awareness for

Operators ....................................................................................................31

Figure 4. Automation and Task Entropy Continuum to Help Solve Problems

with AI .......................................................................................................33

Figure 5. Shandong (Type 001A) Aircraft Carrier ....................................................35

Figure 6. China’s Military Power - DIA’s 2019 PLAN Fleet Composition in

Support of Naval ISR .................................................................................37

Figure 7. Venn Diagram on ISR Visualization to Show Teamwork Between

Intelligence Collection (J2) and Operations (J3) .......................................39

Figure 8. S-curve Model of Diffusion Process ..........................................................49

Figure 9. Intelligence Collection Capabilities for ISR ................................................54

Figure 10. Manned and Unmanned Naval ISR platforms .............................................55

Figure 11. China’s New Silk Road Called “String of Pearls” to Build and
Expand Ports and Industrial Parks Across Southeast Asia, Including
Sri Lanka, Kenya and Greece ....................................................................63

Figure 12. China’s Anti-Access Area Denial Defensive Layers .................................64

Figure 13. China’s Long-Range Surveillance Capabilities Possibly Extend to

Guam. Quality or Discrimination of Specific Class, Types, or
Accurate Tracks is Undetermined ..............................................................69

Figure 14. LeoLabs, Open-Source Private Company Designed to Provide

Applications for Operating in Low Earth Orbit .........................................79

Figure 15. Kiwi SDR Automation Software Identifies Possible Chinese OTH-
Backscatter Radar Signal Pulses and Analyzes the Frequency .................81

Figure 16. Lockheed Martin’s Optical Dazzling Interdictor, Navy laser

Designed for Counter-ISR .........................................................................83

xi
Figure 17. Marine Traffic Tool Tracks Chinese Maritime Militia (Left),
“Liaoning” CV-16 Operating in North Sea Fleet (Right) ..........................84

Figure 18. Algorithms in Intersect Sentry Detects and Visualizes Routes of

Maritime Vessels .......................................................................................91

Figure 19. Heatmap and Trees Indicate Variables and Clustering of Colors ..............92

Figure 20. U.S. Navy Innovators and Adopters Compared to PLAN Later
Adopters Along the Logistic S-Curve........................................................95

Figure 21. Cornell University’s ArXiv Researchers: Computer Vision and

Pattern Recognition: Paper to Person 1,999 Nodes, Density,
Clustering Coefficient ..............................................................................112

Figure 22. Computer Vision and Pattern Recognition: Density, Clustering

Coefficient................................................................................................113

Figure 23. Girvan-Newman Clustering Group (Nodes with Highest Internal

Degree) .....................................................................................................113

Figure 24. K-Core Subgroup of CV and PR .............................................................115

Figure 25. K-Core Group Size and Large Group Statistics .......................................115

Figure 26. Top Connected AI and Machine Learning Researchers Derived

ArXiv and Visualized through ORA........................................................116

Figure 27. Social Network Analysis Shows Stephane Mallat as a Bridge

Between LeCun’s and Sapiro’s Networks ...............................................117

Figure 28. Low Density of CV and PR Researchers, But Well Connected

Internationally ..........................................................................................118

xii
 LIST OF TABLES

Table 1. Working Definitions for Automation and AI ..............................................23

Table 2. Definitions by Technical and Government/Public Officials ......................25

Table 3. U.S. and PRC AI Strategies .......................................................................27

Table 4. Levels of Automation ................................................................................32

Table 5. 13th Year Plan: PLA’s Defense, Dual-Use Plans, Science and
Technology Strategies ................................................................................61

Table 6. Capture the Flag: Hack a Newly Developed Cyber Defense for
Industrial Control Systems.........................................................................89

Table 7. Group 1: Top 10 AI Experts with an Internal Node Count of 158 ..........114

xiii
THIS PAGE INTENTIONALLY LEFT BLANK

xiv
 LIST OF ACRONYMS AND ABBREVIATIONS

A2/AD Anti-access Area Denial

ABI Activity-Based Intelligence
AGI Artificial General Intelligence
AI Artificial Intelligence
AIDP Artificial Intelligence Development Plan
AIS Automatic Identification System
ANI Artificial Narrow Intelligence
APT Advanced Persistent Threat
ASCM Anti-Ship Cruise Missile
AUV Aerial Unmanned Vehicle
BAT Baidu, Alibaba, Tencent
BRI Belt and Road Initiative
CASC China Aerospace Science and Technology Corporation
CCP Chinese Communist Party
CEC Circular Error Probable
CEC Cooperative Engagement Capability
CMC Central Military Commission
CMM Chinese Maritime Militia
CNO Chief of Naval Operations
COP Common Operational Picture
COSTIND Commission for Science, Technology and Industry for National
Defense
CPC Chinese Communist Party
CSL Common Source Library
DARPA Defense Advanced Projects Research Agency
DAS Defense Acquisition System
DCGS-N Distributed Common Ground Station-Navy
DIU Defense Innovation Unit
DOD Department of Defense
FONOPS Freedom of Navigation
xv
GANs Generative Adversarial Networks
GLONASS Global Navigation Satellite System
GPC Great Power Competition
GPS Global Positioning System
GPU Graphic Processing Unit
GWOT Global War on Terror
HF High Frequency
HTI Hardware Technology Insertion
I&W Indications and Warning
IoT Internet of Things
ISR Intelligence, Surveillance, and Reconnaissance
IT Information Technology
JAIC Joint Artificial Intelligence Center
JCIDS Joint Capabilities Integration and Development System
JEDI Joint Enterprise Defense Infrastructure
JORN Jindalee Operational Radar Network
LiDAR Light Detection and Ranging
MABI Maritime Activity-Based Intelligence
MARS Machine-Assisted Analysis Rapid-Repository System
MIDB Modernized Intelligence Database
MIIT Ministry of Industry and Information Technology
MMSI Maritime Mobile Service Identity
MOST Ministry of Science and Technology
MTA Middle-Tier Acquisition
NGA National Geospatial-intelligence Agency
NIST National Institute of Standards and Technology
NMCI Navy and Marine Corps Intranet
NOSTRADAMUS New Transhorizon Decametric System Applying Studio Methods
ODIN Optical Dazzling Interdictor, Navy
OTH-R/B Over-the-Horizon Radar/Backscatter
PEO Program Executive Office
PLA GAD People’s Liberation Army General Armaments Department,
xvi
PLAN People’s Liberation Army Navy
PLANAF People’s Liberation Army Air Force
PRC People’s Republic of China
QZSS Quasi-Zenith Satellite System
RADAR Radio Detection and Ranging
RIMPAC Rim of the Pacific
SASTIND State Administration for Science, Technology and Industry for
National Defense
SCS South China Sea
SLOC Sea Lines of Communication
SNA Social Network Analysis
SOF Special Operations Forces
SSDS Ship’s Self Defense System
UAV Unmanned Aerial Vehicle
UUV Unmanned Underwater Vehicle

xvii
THIS PAGE INTENTIONALLY LEFT BLANK

xviii
 ACKNOWLEDGMENTS

I sincerely thank my primary advisor, Dr. John Arquilla, for giving me the
opportunity to be his thesis student and for his grandiose guidance and mentorship as
both the former Defense Analysis Department chair and as a DA professor. I am grateful
for the opportunity both Dr. Arquilla and Vice Admiral and NPS President Ann Rondeau
gave me to be the moderator for Mr. Shantanu Narayen, the CEO of Adobe, for one of the
Secretary of the Navy Senior Guest Lecturer (SGL) series. I further thank Dr. Arquilla for
sharing magnificent opportunities such as Stanford University’s 2019 Human-Centered AI
Symposium and teleconference with the Atlantic Council’s Dr. David Bray to learn about
the future of AI by 2045. His patience, feedback, words of encouragement, such as
“Action this day…what have you done today to contribute to your thesis?” and teaching
me about military strategy and technology will be imprinted in my mind for the rest of
my life.
I thank my co-advisor, Dr. Douglas Van Bossuyt from the Systems Engineering
Department, for providing me superb structuring, course correction, and golden nuggets of
wisdom in this thesis process. His guidance, constant communication, networking
opportunities, and flexibility as a co-advisor from a different department encouraged me to
continue writing systematically. I thank Dr. Zachary Davis from Lawrence Livermore
National Laboratory and Dr. Dan Cunningham, Dr. Sean Everton, Brendan Knapp, and
Chris Callaghan from the DA Core Lab faculty for taking the time to teach me data
analytics and computer science programming. I thank my NPS Dudley Knox Library
Graduating Writing Center coach Cheryldee Huddleston for spending countless hours
helping me revise my thesis.

Most importantly, I thank my best friends and family who all were there for me
throughout both the thesis process and during challenging times of health issues that I
successfully overcame during my time in school. My thesis is the direct result of my loved
ones, brilliant professors, advisors, and researchers who have allowed my mind to
transcend borders. My thesis journey resonates with Carl Sagan’s belief and value “to
preserve and cherish the pale blue dot, the only home we’ve ever known.”
xix
THIS PAGE INTENTIONALLY LEFT BLANK

xx
 I. INTRODUCTION

Automation and the development of artificial intelligence (AI) will influence future
sea power. On July 8, 2017, President Xi Jinping and China’s State Council stated their
goal to build a domestic AI industry and to make the country an “innovation center for AI”
by 2030. 1 On February 11, 2019, President Donald Trump signed Executive Order 13859
to maintain leadership in AI and to reflect the United States values. 2 While these strategies
have been announced, the rate of adopting automation in naval affairs remains slow. In the
Chief of Naval Operations’ 2018 article, “A Design for Maritime Superiority,” the CNO’s
first priority is to design a future integrated naval force structure and to “Put All Hands On
Deck to make the USS Gerald R. Ford (CVN 78) ready as a warship as soon as practically
possible.” 3 Xi Jinping called upon the People’s Liberation Army (PLA) to evolve to future
“intelligentized warfare.” The AI superpower navies compete to adopt automation and AI
technologies for persistent intelligence, surveillance, and reconnaissance. This thesis
compares the technological adoption of two critical automation sensors—U.S. Aegis and
China’s Type-348 “Dragon Eye” and how they improve naval intelligence, surveillance,
and reconnaissance (ISR).

Since the Cold War, maritime superiority has come to rely on automated early
warning systems from interconnected ISR platforms. Automation for maritime Indications
and Warnings (I&W) was heavily embedded in conventional naval capabilities between
the United States and then-Union of Soviet Socialist Republics (USSR). After the fall of
the Soviet Union, China steadily rose and has been seeking the use of Artificial Intelligence
(AI) to augment its military’s cognitive capabilities at an ever-faster rate, especially with
respect to the future of sea power. The automated communications link, called Link-11 or

1 Flora Sapio, Weiming Chen, and Adrian Lo, trans., A New Generation of Artificial Intelligence
Development Plan, State Council Document [2017] No. 35 (Beijing: Chinese State Council, 2017),
https://flia.org/notice-state-council-issuing-new-generation-artificial-intelligence-development-plan/.
2 Donald J. Trump, “Maintaining American Leadership in Artificial Intelligence: Executive Order
13859,” Federal Register 84, no. 31 (February 14, 2019): 3967–72.
3 John Richardson, A Design for Maritime Superiority, Version 2.0 (Washington, DC: Department of
the Navy, 2018), https://www.navy. mil/navydata/people/cno/Richardson/Resource/Design_2.0.pdf.
1
Link-16, between aircraft carriers, naval aircraft, and submarines allows human operators
to send messages over internet computer systems instantaneously.

This study focuses on the adoption of automation-based platforms by the U.S. Navy
and People’s Liberation Army Navy (PLAN). Examples of naval ISR platforms include:
ship automation, manned and unmanned systems, aerial reconnaissance satellites,
computer systems that visualize the identification and location of naval platforms at sea.
This topic is important because of three key factors: policy on rapid AI technology research
and development; the return of Great Power Competition (GPC); and the need to mitigate
uncertainty and enhance judgment. However, prior geopolitical incidents show that
automated systems can be prone to erroneous, unstructured or unsupervised data due to
device misconfiguration and human error. Receiving the right data, at the right time, at the
right location is critical to understand the adoption of automation for future sea power.

The scope of this study regarding the key areas includes: strategy of automation;
aids to decision-making (speed); and critical intelligence systems in space that enable
automated tools to work. The overview of the impact in adopting automation is bounded
by national defense and policy strategy documents by the United States and the People’s
Republic of China (PRC). Strategic implications of adopting critical intelligence systems
will be discussed.

It is critical to adopt and harness today’s automated technologies. Technological

adoption of emerging technologies relies on frequent software updates and compatible
hardware to successfully operate; otherwise, the system will not operate as desired or
originally designed. Another way to conceptualize this process of adopting and integrating
a 2020 piece of technology is to think of integrating a modern Bluetooth speaker or radio
into a 1980s vehicle; possible, but the attempt to integrate incompatible software or
hardware will come at a cost and can likely break down and cause incompatibility is the
absence of redundant tests prior to fielding or employing an automation system or AI tool
on a naval ISR platform.

2
 RESEARCH QUESTION

How are the U.S. Navy and the People’s Liberation Army Navy (PLAN) adopting
automation to improve naval ISR with Aegis and Type 348 radar sensors for future sea
power?

BACKGROUND

Half a century ago, General William C. Westmoreland, the U.S. Army Chief of
Staff, had a vision of future warfare that rings true today. In his speech, given on October
14, 1969, to the Association of the United States Army, he stated,

On the battlefield of the future enemy forces will be located, tracked and
targeted almost instantaneously through the use of data-links, computer-
assisted intelligence evaluation and automated fire control…and with
surveillance devices that can continuously track the enemy, the need for
large forces to fix the opposition physically will be less important. 4

General Westmoreland encapsulated a vision of adopting constant surveillance,

tracking, and targeting with computers to instantaneously track the enemy. 5

The prominent Stanford University computer science professor, Andrew Ng, AI

developer and business executive believed just as electricity transformed almost everything
over a century ago, today it is hard to think of an industry that will not be touched by AI in
the coming years. 6 His vision seems to be coming true, and today’s software program that
provides maritime domain awareness is called the Common Operating Picture (COP). The
U.S. defense company Raytheon developed the COP to visually display naval air, land, and
sea platforms in near-real-time. Picture dots on a screen identifying and locating naval
platforms at sea based on their unique identifiers, or fingerprints, which is the manner in
which GPS displays on a map. This is a form of AI application in which automation
catalyzes the process and streamlines the data flow between aircraft, surface vessels,
submarines, ground stations, and satellites.

4 Frank Barnaby, The Automated Battlefield (New York: The Free Press, 1986), 1.
5 Barnaby,

6 Shana Lynch, “Andrew Ng: Why AI is the New Electricity,” Stanford News (14 March 2017),
https://news.stanford.edu/thedish/2017/03/14/andrew-ng-why-ai-is-the-new-electricity/.
3
 Since the 1950s, prominent leaders have researched and adopted pattern
recognition of maritime platforms and its associated radar signature or signals in naval
affairs. For naval ISR, a software program called Activity Based Intelligence (ABI) is “an
analysis methodology which rapidly integrates data from multiple intelligence sources and
sources around the interactions of people, events and activities, to discover relevant
patterns, determine and identify change, and characterize those patterns to drive collection
and create decision advantage.” 7

According to the Naval Postgraduate School (NPS) computer scientist Professor

Joshua Kroll, automation is an operation of a process according to a set of established rules,
which are referred to as a set of explicit and implicit specifications. These rules can be very
simple or highly complex and implemented by both humans and machines in a variety of
ways. These rules are implemented and coded in an “if, then” logic in a software program
and then can be coupled with a mechanical function and typically in tandem with humans. 8
Automation is beneficial because it automates a task, mechanically or visually, while
avoiding human weaknesses such as fatigue and inattention.

Automation and AI are imperfect and sometimes have specific errors. The systems
rules or the output from the rules can be incorrect or incomplete, it can lead to a catastrophic
event that would cause a stop or pause for research in automation. 9 Human error can have
a big part in system error as well.

Naval sensors that automate data flow at sea have propelled naval operators and
strategists to reshape the adoption of AI for the future maritime battlespace. The
Department of Defense’s (DOD) Artificial Intelligence strategy has accelerated “the
adoption of AI and the creation of a force fit for our time.” 10 Simply put, using legacy

7 Chandler Atwood, “Activity-Based Intelligence: Revolutionizing Military Intelligence Analysis,”

Joint Force Quarterly 77 (April 2015): 24–33.
8 Joshua Kroll, “Classifying Machines by Kinds of Learning: Automation,” slide 3, https://nps.edu/
web/ai-consortium/harnessing-ai-course.
9 Kroll, slide 21.
10 Office of Science and Technology Policy, Summary of the 2018 White House Summit on Artificial
Intelligence for American Industry (Washington, DC: White House, 2018), 4, https://www.whitehouse.gov/
wp-content/uploads/2018/05/Summary-Report-of-White-House-AI-Summit.pdf.
4
technologies that work is valuable, but understanding how the AI superpowers are adopting
game-changing technologies is critical to improving naval ISR. A fit way to adopt
automation revolves around knowledge, access, and time for relevant and accurate early
warning.

The second approach is defining and understanding the literature of automation and
AI, to include its perception within various sectors of society. There is not one accepted
and concrete definition of automation. My approach will be to outline the standard
definitions from government, military, public and private and compare their definitions to
technical experts like computer scientists. When one says or hears the phrase AI various
thoughts and images come to mind—images from lethal autonomous weapons destroying
the adversary or killer robots to less intimidating images. The perception of AI can range
from imagining sailors sitting in front of a computer screen on a ship to Apple’s Siri or
Amazon’s Alexa technologies. The perception of automation can be as simple as an
assembly line to streamlining signals automatically to control the flow of information and
incorporated on a digital map to show where ships, aircraft, and submarines are located.
The definition depends on whom you ask.

The third approach is analyzing automation and AI capabilities today concerning

naval ISR between the AI superpowers. China’s blue-water borders—also known as first,
second, and third island chains and the “string of pearls” in the Indian Ocean. The fourth
approach is conceptualizing and describing the design methodology for adopting
automation coupled with AI emerging technologies that can be recycled, re-used, and
turned around in less than six months to have the advantage of long-term strategy for
adopting automation for naval ISR. This irregular method will incorporate measures of
effectiveness and performance with a demanding timeline that encompasses aggressive,
timely, and ethical standards.

RESEARCH METHODOLOGY

My thesis methodology will be to develop an adaptive model for comparatively

analyzing automation and AI capabilities today between the United States Navy and
China’s Navy, to include their Navy’s Air Force component. The goal is to produce a

5
research methodology with precise and concise indicators and measurements for the
Chairman of the Joint Chiefs to observe and absorb heuristically, to act today, and re-orient
when AI software begins to outpace current technology. In Chapter II, a literature review
of automation and AI are defined and technical aspects of naval ISR are necessary to
discuss, conceptualize, and use a simple working definition throughout this thesis.

In addition to qualitative analysis, quantitative analysis will be conducted through

both network analysis of contributing industry companies (private and public) in pursuit of
developing an adoption framework for senior-level decision making and feedback systems.
The purpose of a mixed methodology is to determine practices of automation for
Intelligence, Surveillance, Reconnaissance. A thorough analysis of automation regarding
ISR will allow for a better understanding of current trends and constraints for ISR in the
maritime battlefield. It is important to note that the theory of adoption is usually based on
the organization or the user; thus, I will discuss both. I will use Everett Rogers’ S-curve
model of the diffusion process and examine the level of adopting automation for ISR within
the last 50 years. 11

This thesis will encompass three specific automation systems that can connect to
the two focus automation sensors and the cloud services both navies have adopted to store
ISR data. The systems include unmanned underwater vehicles (UUVs), unmanned aerial
vehicles (UAVs). In addition to the DOD enterprise-wide cloud solution, the United States
has integrated the Joint Enterprise Defense Infrastructure (JEDI) Cloud Program, which
can improve naval ISR support and stay abreast of China. 12 There is little known publicly
available information regarding the PLAN’s development plan for cloud services.
However, in 2017, China’s Ministry of Science and Technology, which collaborates with

11 Mary Cain and Robert Mittman, Diffusion of Innovation in Health Care (Oakland, CA: California
Health Care Foundation, 2002), http://kpworkforce.org/projects/include/DiffusionofInnovation.pdf.
12 Department of Defense, DOD Cloud Strategy (Washington, DC: Department of Defense, 2018),
https://media.defense.gov/2019/Feb/04/2002085866/-1/-1/1/DOD-CLOUD-STRATEGY.PDF.
6
the PLA, recruited the internet giants Baidu, Alibaba Group Holding, and Tencent
Holdings—also known as BAT—for cloud computing. 13

Everett Rogers’ 1962 S-curve model of diffusion process will help better
understand the current adoption methods. For example, the U.S. Defense Acquisition
System (DAS) and Joint Capabilities Integration and Development System (JCIDS), and
the 2020 Middle-Tier Acquisition (MTA) Pathway per DODI 5000.2 are U.S. acquisition
models for technological adoption. 14 I will also use quantitative measures to relate spatial
and temporal network analysis of relevant companies, private and non-profit, that can
positively and negatively contribute to the adoption of automation. The purpose is to
possibly reveal companies’ proprietary rights and the distance or degrees of separation
between them that may show weaknesses in the methodology and overall adoption process
for future naval ISR.

As a case study in Chapter III, I will provide an overview of Everett’s S-Curve of

technological adoption and a step-by-step process to help with the process of adopting
automation and AI for naval ISR. Other mentionable countries with AI strategies for
adopting automated systems or AI tools to briefly be discussed in this thesis will be the
United Kingdom, Russia, France, Iran, and South Korea.

THEORY OF TECHNOLOGICAL ADOPTION

The theory of adoption, within the context of technology, has been a focus of
researchers, commercial, and industry for decades. The modern term adoption is derived
from one of the Roman forms of adoption, adoptio. This late 15th century French and Latin
phrase adoptare comes from ad meaning ‘to’ and optare meaning ‘choose.’ 15 Simply put,

13 Meng Jing and Sarah Dai, “China Recruits Baidu, Alibaba and Tencent to AI ‘National Team,’”
South China Morning Post, November 21, 2017, https://www.scmp.com/tech/china-tech/article/2120913/
china-recruits-baidu-alibaba-and-tencent-ai-national-team.
14 J. Jerry LaCamera, Jr., “Rapid Acquisition - The Challenge to Accelerate” (NDIA 2019 Spring IPM
Division Meeting, Herndon, VA: National Defense Industrial Association, 2019), https://www.ndia.org/-
/media/sites/ndia/divisions/ipmd/2019-04-meeting/207-middle-tier-of-acquisition-lacamera-
190508215620.ashx?la=en.
15 Frederick Mish, “Adoption,” in Merriam-Webster’s Collegiate Dictionary (Merriam-Webster Inc.,
2004), 17.
7
one can choose an inanimate object, person, place, or thing or choose to change, or one
will not change.

The S-curve of technological adoption represents a theoretical framework to help

conceptualize how one or an organization can successfully or unsuccessfully adopt a
technology. According to American communication theorist and sociologist Everett
Rogers, the successful adoption of an innovative technology occurs in a distributed bell-
curve, S-shaped, which is derived from the diffusion curve with respect to time and
normality (Figure 1). The five categories of the adoption system are: (1) innovators, (2)
early adopters; (3) early majority, (4) late majority, and (5) laggards. 16 Understanding of
the stages within the S-shaped bell curve is important to the successful adoption of
emerging technologies.

Figure 1 depicts the rate of technology adoption in the form of a horizontal and
stretched “S” with adoption on the Y-axis and time on the X-axis. This means that when a
technology is initially adopted, it takes years for either the technology or market conditions
to adjust or develop to the point where it hits an inflection point to achieve rapid adoption.

Figure 1. S-curve Model of the Diffusion Process 17

When the market begins to become saturated, the curve flattens out and late adopters of
technology are too late to adopt emerging technology and integrate into a competitive

16 Joseph P. Schwieterman and Lauren A. Fischer, “The S-Curve of Technological Adoption: Mobile
Communication Devices on Commuter Trains in the Chicago Region, 2010–2015,” Journal of Public
Transportation 20, no. 2 (2017): 1–18, https://doi.org/10.5038/2375-0901.20.2.1.
17 Source: Schwieterman and Fischer, 3.

8
environment at the point of saturation, which is where the laggards reside. Naval leadership
will find this useful because it means it is best to be at the inflection point of the S-curve,
not at the beginning or the end of it.

As populations grow, automation and AI emerging technologies will grow. The

history of the S-curve idea began in the 19th century with the motivation of knowing more
about the growth of humans. Rogers’ S-curve theory of adoption is based on the logistic
function as a model of population growth that was first introduced by Belgian
mathematician Pierre-Francois Verhulst in 1839. Verhulst related this theory to population
growth and initially conceived his idea from the Englishman and political economist
Thomas Malthus’ “An Essay on the Principle of Population” in 1789. During this time,
people worried about population growth surpassing food supply and the future of feeding
the population. From 1950 to today, the population growth of the United States has grown
from 330 million U.S. citizens to 1.4 billion Chinese citizens; contributing to a current
population of 7.76 billion and counting. 18 The more the human population grows, the more
likely the thirst for knowledge not only humans but also automation and AI. People want
more automated systems and to adopt AI technologies because they are typically faster
than human beings, can consume more data, provide both intelligence and surveillance,
and sometimes certain technologies are simply fun to use.

An important corollary from the S-curve adoption theory is that the navies ought to
build companies on technologies that have grown and experienced failures to achieve
successful adoption. Some technologies include AlphaGo deep-learning system and IBM’s
DeepMind subsidiary of Google to play the strategy game of Go, or Weiqi. In 2016, the
computer system AlphaGo defeated South Korean Grandmaster Lee Sedol. 19 For both AI
superpower navies, this simulated game relates directly to how both the sensors and
computer systems between USN and PLAN learn about each other on the maritime

18 “Current Population Growth,” Worldometer, https://www.worldometers.info/world-population/.

19 David Silver et al., “A General Reinforcement Learning Algorithm That Masters Chess, Shogi, and
Go Through Self-Play,” Science 362, no. 6419 (December 2018): 140–44, https://doi.org/10.1126/
science.aar6404.
9
battlefield. The adoption and growth of technology translate well to the life cycle pattern
and measured as a cumulative growth over time.

The growth of automation technologies since the 1950s is not a surprise and is
imperfect. According to Modus the S-Curve, “is derived from the law that which states that
the rate of growth is proportional to both the amount of growth already accomplished and
the amount of growth remaining to be accomplished.” 20 The limitations are inherent to risk
and uncertainty over time. Like the life cycle—periods of birth, growth, puberty or
maturity, and death—the S-curve stages serve as the key mechanism to decelerate,
accelerate, shift gears, or stop completely and start fresh. Not every specific naval sensor
or computer system designed for ISR is going to achieve perfect knowledge of the maritime
battlefield.

This observation is recognized and utilized in the commercial sector. Former CEO
of Google and board member of the DOD Advisory board Eric Schmidt believes the
problems are straightforward. He believes computer vision is easily understood if we build
reliable systems in which we understand failure modes and error rates. 21 Receiving,
analyzing, and disseminating information and intelligence via Aegis with computer
systems connected to sensors with a time gap creates a recognized frustration among the
younger operators who envision computer systems and AI applications, like a high-
resolution video game, is a weakness in adopting automation.

Human beings use automation software applications to solve problems. It is

important to understand that automation is currently not capable of making intelligence
assessments. AI simulates cognitive functions of the human brain and computer vision
displaying data. In the United States Navy, the human operators behind the machine use

20 David Lindgren, “Global Remittances and Space-Based Cryptocurrencies: A Transformational

Opportunity for the Post-2030 Agenda,” in Post 2030-Agenda and the Role of Space, ed. Annette Froehlich
(Cham, Switzerland: Springer International Publishing, 2018), 79, https://www.springer.com/gp/book/
9783319789538.
21 Paul Scharre, Anthony Cho, and Eric Schmidt, “Eric Schmidt Keynote Address at the Center for a
New American Security Artificial Intelligence and Global Security Summit” (Washington, DC: Center for
a New American Security, November 17, 2017).
10
the rule-based automation system which displays radar signals. The data at rest—the data
in the computer not being used or manipulated—is limited to the amount of data input.

Change in adoption theory is inescapable. Similar to American historian Elting

Morison’s research and reflection process, much of the adoption of automation centers
around four distinct parts: the initial condition “at the point of origin of any mechanical
change; the character of the primary agents of change; the nature of those resistant to
change; and the means to facilitate accommodation to the changes introduced.” 22 Without
understanding and enduring these parts, one cannot rise above the threshold of existing
bodies of knowledge.

In 2020, we are in another time of tremendous technological development and

return of GPC by which maintaining future sea power depends on superiority in the fourth
dimension of cyberspace. Russia’s President Vladimir Putin once said, “Artificial
intelligence is the future, not only for Russia but for all humankind. It comes with colossal
opportunities, but also threats that are difficult to predict. Whoever becomes the leader in
this sphere will become the ruler of the world,” 23 Automating simple tasks and adopting
AI tools, even with legacy systems, demands a change in today’s GPC and national
security.

22 Elting Morrison, Men, Machines, and Modern Times (Cambridge, MA: MIT Press, 1966), 7.

23 “‘Whoever Leads in AI Will Rule the World’: Putin to Russian Children on Knowledge Day,” RT World
News, September 1, 2017, https://www.rt.com/news/401731-ai-rule-world-putin/..
11
THIS PAGE INTENTIONALLY LEFT BLANK

12
 II. LITERATURE REVIEW

The U.S. naval intelligence team on the carrier is searching for the last location of an
unknown, Chinese Kilo submarine. The radar returns visualized as automated dots appear on
shipborne computer screens which show two Chinese warships. There is some ambiguity in
their identification, current time, and location. The operator watch team on the aircraft carrier
using AI software sees automated dots in motion appear with unique identifiers of Chinese
fighter aircraft taking off from Hainan Island in the South China Sea (SCS). Chinese fighter
aircraft have a radar signature, such as identification, and then U.S. Navy ISR collection
platforms and sensors can collect the data to provide situational awareness. This
technological exchange of information is like how commercial aircraft can be identified or
identify other aircraft while flying. A naval intelligence team of six notices the automation
and do what is necessary—notifies the watch and immediately provide timely, relevant,
accurate indications and warning (I&W). The Carrier Strike Group (CSG) decision-makers
operating in the SCS are depending on an early warning. One of several computer screens
using automation tools, fueled by algorithms or software, shows pixilated dots on a computer
screen with a geographic map of SCS as the background. The tactical action officer, Surface
Warfare Officers, and the team of intelligence officers and specialists track dot(s) on
computer screens. The dot represents a Chinese destroyer, under the command of the
People’s Liberation Army Navy (PLAN), approaching USS Wayne E. Meyer (DDG-108)
conducting Freedom of Navigation Operations (FONOPS). Automation tools, powered by
AI algorithms, provide precise time, distance, course, and speed to allow situational
awareness and early warning to react.

The aforementioned vignette is representative of historical events in SCS that have

sent a clear strategic message of mutual power projection and presence between the United
States and China. Automation expedites maritime ISR and naval encounters at sea between
foreign navies for situational awareness for both adversarial and commercial platforms. In
2017, the U.S. Navy conducted at least six routine FONOPS, according to the Defense

13
Department. 24 Further, National Geospatial-Intelligence Agency (NGA) collects and uses
data for Maritime Activity-based Intelligence (MABI) software, which provides: geo
reference time and location similar to Google Maps; sequence and data neutrality for
unbiased data; and, integration before exploitation. 25 In September 2019, USS Wayne E.
Meyer—named after the father of the Aegis combat system—conducted FONOPS near the
Paracel Islands (located in the northern part of South China Sea). In response, Chinese
military vessels and aircraft attempted to deter the U.S. warship from sailing near the
Chinese-claimed islands. 26

The purpose of this chapter is to gain a better understanding of automation and AI.
This will be organized through five subsections: (1) South China Sea historical vignette; (2)
a literature review on prominent leaders from America and from China who are experts in
the field of AI and automation; (3) definitions of automation and AI; propose my working
definition of automation and AI-based on existing literature, publications, and doctrine
between technical, government, and public experts; (4) conceptualize automation and, (4) a
succinct technical overview understanding of naval ISR about the two sensors.

AI POLICY AND STRATEGY DOCUMENTS

For foundational purposes, I cite four national strategy and policy documents:
Department of Defense Directive (DODD) 27; Joint Publication 1-02 28 and 2–01 29; and, two

24 Caitlin Doornbos, “Freedom-of-Navigation Ops Will Not Dent Beijing’s South China Sea Claims,
Experts Say,” Stars and Stripes, April 4, 2019, https://www.stripes.com/news/pacific/freedom-of-
navigation-ops-will-not-dent-beijing-s-south-china-sea-claims-experts-say-1.575609.
25 Patrick Biltgen and Stephen Ryan, Activity-Based Intelligence: Principles and Applications
(Boston: Artech House, 2016), xx.
26 Jesse Johnson, “U.S. Warship Challenges Chinese Claims in Disputed South China Sea,” Japan
Times, September 14, 2019, https://www.japantimes.co.jp/news/2019/09/14/asia-pacific/u-s-warship-
challenges-chinese-claims-disputed-south-china-sea/#.Xhd7aMhKj_M.
27 Department of Defense, Autonomy in Weapon Systems, DOD Directive 3000.9 (Washington, DC:
Department of Defense, 2012), https://www.hsdl.org/?abstract&did=726163.
28 Joint Chiefs of Staff, Department of Defense Dictionary of Military and Associated Terms, Joint
Pub. 1–02 (Washington, DC: Joint Chiefs of Staff, 2016), https://fas.org/irp/doddir/dod/jp1_02.pdf.
29 Joint Chiefs of Staff, Joint Intelligence Support to Military Operations, Joint Pub 2–01
(Washington, DC: Joint Chiefs of Staff, 1996), http:// www.hsdl.org/?view&did=3737.
14
summaries of AI Strategy by the United States 30 and China. 31 The first is DODD 3000.9,
which “establishes DOD policy and assigns responsibilities for the development and use of
autonomous and semi-autonomous functions in ISR systems, including manned and
unmanned platforms.” 32 The second is JP 2-01 (Joint and National Intelligence Support to
Military Operations), specifically the ISR section on integrated operations and intelligence
activities and synchronization of OPINTEL and ISR automated visualization (of collected
data). 33 The third is the 2018 Department of Defense Summary of the 2018 Department Of
Defense Artificial Intelligence Strategy: Harnessing AI to Advance Our Security and
Prosperity, which discusses harnessing strategic and focused deliverable AI. 34 The fourth is
China’s 2017 State Council “New Generation Artificial Intelligence Development Plan”
(AIDP), which aptly focuses on science, technology, industry, market, and societal
perception. 35 These policy documents are the most important driving sources for AI strategy
and development.

Existing literature on automation for sea power suggests that time, demand, and
resources are critical factors in succeeding in adopting AI. According to the Fiscal Year 2020
DOD Defense Budget, $927 million is dedicated to expanding AI to expand military
advantage with the Joint Artificial Intelligence Center (JAIC) and Advanced Image
Recognition (Project Maven) and $3.7 billion for autonomous and unmanned vehicles. 36 The

30 Department of Defense, Summary of the 2018 Department of Defense Artificial Intelligence

Strategy: Harnessing AI to Advance Our Security and Prosperity (Washington, DC: Department of
Defense, 2018), https://media.defense.gov/2019/Feb/12/2002088963/-1/-1/1/SUMMARY-OF-DOD-AI-
STRATEGY.PDF.
31 Sapio, Chen, and Lo, A New Generation of Artificial Intelligence Development Plan.
32 Department of Defense, Autonomy in Weapon Systems.
33 Joint Chiefs of Staff, Joint Intelligence Support to Military Operations.
34 Department of Defense, Summary of the 2018 Department of Defense Artificial Intelligence
Strategy.
35 State Council, “New Generation Artificial Intelligence Development Plan [国务院关于印发新一代
人工智能发展规划的通知],” July 8, 2017, ttp://www.gov.cn/zhengce/content/2017- 07/20/
content_5211996.htm.
36 Office of Undersecretary of Defense (Comptroller)/Chief Financial Officer, Defense Budget
Overview: United States Department of Defense Fiscal Year 2020 Budget Request (Washington, DC:
Department of Defense, 2019), https://comptroller.defense.gov/Portals/45/Documents/defbudget/fy2020/
fy2020_Budget_Request_Overview_Book.pdf.
15
most recent 2020 White House budget document regarding the non-defense AI agencies
listed a $1,291,000,000 FY20 request compared to the FY19 estimate of $1,248,000,000.

However, both commercial and private industries are leading innovation in emerging
technologies. The Trump administration has identified non-defense AI spending as a
supplement to the president’s FY 2020 budget request. According to U.S. Chief Technology
Officer Michael Kratsios, the budget released in September 2019 shows $654.4 million
dedicated to the U.S. government’s non-defense spending by agencies related to AI. 37 This
supplemental document aligns with Executive Order 13589, signed by President Donald
Trump in February 2019, specifically to maintain leadership in AI in the federal research and
development sector.

The two institutions most relevant to AI for naval ISR spending are Defense
Advanced Research Projects Agency (DARPA) and Intelligence Advanced Research
Projects Activity (IARPA). According to the White House, IARPA oversees several AI
projects: developing AI models for cybersecurity, analyze autonomous systems policies,
improve the interface between humans and intelligent systems, and develop counter AI tools
(i.e., TrojAI). DARPA oversees next-generation AI and “explainable AI,” programs such as
developing machine learning techniques and enable human users to understand, trust and
manage AI partners. 38 However, DARPA budget figures for AI research and development
are not publicly available.

The strategies, policies, and budget for automation are the fundamental basis for
gaining a shared sense of understanding of automation and AI. Next, this literature review
will cover the history of thought leaders, a variety of definitions, and my working definition
to better grasp both terminologies and ultimately, improve the adoption process.

PROMINENT LEADERS OF AUTOMATION

In 1969, then-U.S. Army Chief of Staff General William C. Westmoreland stated,

“On the battlefield of the future enemy forces will be located, tracked and targeted almost

37 National Security & Technology Council, 11.

38 National Security & Technology Council, 15.

16
instantaneously using data-links, computer-assisted intelligence evaluation, and automated
fire control. With first-round kill probabilities approaching certainty, and with surveillance
devices that can continuously track the enemy, the need for large forces to fix the opposition
physically will be less important.” 39 His outspoken vision resonates today with 24-hour real-
and near-real-time ISR. 40 Because of General Westmoreland’s vision, automation has long
been a focus of the United States military to rapidly provide early I&W against adversaries
at sea.

1. History of Automation from Prominent American Leaders

Stanford computer science professor and recipient of the Association for Computing
Machinery Turing Award Edward Albert “Ed” Feigenbaum did much to pioneer AI. 41 In the
1980s, he defined AI research as the “part of Computer Science that investigates symbolic
reasoning processes and the representation of symbolic knowledge for use in inference.”42
Feigenbaum’s concept was an important contribution to the future success of AI.
Feigenbaum describes AI as a process to simply start and continuously research and develop
AI and then to think about how to best represent that knowledge. His definition describes the
best way to represent such knowledge is through a symbolic way such that the target
practitioner, or operator, can digest and use the technology.

Feigenbaum’s strategic design centers around a direct partnership between three

main entities for the efficient and effective flow of information. He emphasizes dialogue
between the expert and the knowledge engineer directly because this process marries expert
designers, knowledge engineers, and the user or operator. As a result, the early exchange of

39 Frank Barnaby, The Automated Battlefield (New York: The Free Press, 1986).
40 The term “real-time” is defined as current time and “near-real-time” are naval platforms that naval
operators manually add to an automation situational awareness tool; also known as Common Operational
Picture (COP).
41 Nils J. Nilsson, “Edward A (‘Ed’) Feigenbaum: United States - 1994,” Association for Computing
Machinery, A.M. Turing Award, 2019, https://amturing.acm.org/award_winners/
feigenbaum_4167235.cfm.
42Edward A. Feigenbaum, “Expert Systems in the 1980s” (Stanford, CA: Stanford University, 1980),
https://stacks.stanford.edu/file/druid:vf069sz9374/vf069sz9374.pdf.
17
information facilitates an environment to accomplish a specific task or goal at hand. 43
Sparking an early and continuous dialogue between the three main entities is crucial in
contributing to the success of AI.

From a technical perspective, the goal of the “expert system” is to design and write a
computer science program that will automate the problem faster. According to computer
scientist and founder of the International Journal on Artificial Intelligence Tools Nikolas
Bourbakis, AI helps achieve a high level of performance for difficult problems and difficult
for a human to solve on his or her own. 44 Therefore, expert systems produce an effective and
direct flow of information logically between the three main entities. The architecture of this
expert system, however, limits the relevant and accurate information. This is where the
knowledge engineer filters out irrelevant information and methods or mechanisms within the
field of AI.

Technological innovators believed after the industrial age and information age comes
adopting the automation age. In the year 2000, Admiral William A. Owens’ Lifting the Fog
of War, expressed that we live in the fourth “Age of Automation” 45 He believed the most
notorious example of new technology was Adolf Hitler’s 1940s blitzkrieg, or “lightning
war,” because its purpose was to bring about a swift victory and combine the use of radio
communication, air and land support, and tank division. The users in the “lightning war”
were operators with critical information that had to be shared with both the knowledge
engineer and experts to adopt the automated system on the land battlefield. The concept and
the testing of automated systems on land helped advance the adoption of automation.

British Army officer, military historian, and strategist J. F. C. Fuller writes about the
origins of flight, satellite connectivity, and wireless capability advancing the “Age of
Automation.” He first describes the first flight on December 17, 1903, at Kill Devil Hill,
Kitty Hawk, North Carolina, where two brothers, Orville and Wilbur Wright, flew a power-

43 Barnaby, The Automated Battlefield.

44Nikolas G. Bourbakis, ed., Artificial Intelligence and Automation (Hackensack, NJ: World Scientific
Publishing, 1998), 210, https://doi.org/10.1142/3079.
45 William A. Owens, Lifting the Fog of War (Baltimore: Johns Hopkins University Press, 2001), 80.

18
driven airplane for twelve seconds and the best flight was in 69 seconds. Fuller writes about
the second invention encompassing the space domain in 1887 when a German Physicist
Heinrich Rudolf Hertz was the first man to conclusively prove the existence of the
electromagnetic wave where an electrical spark propagates into space. He then describes the
third invention of wireless capability. Later in 1897, an Italian electrical engineer Guglielmo
Marconi invented a device to detect said electric spark or waves and wirelessly transmitted
a message over 3,000 miles in 1901. 46 These scientific inventions are the bedrocks that
revolutionized both military and naval affairs to adopt automated systems.

Prominent scientists and theorists inspired the U.S. military to take advantage of
flight, satellite connectivity, and wireless capability for future naval ISR capabilities. For
example, then-CNO Adm. George W. Anderson, Jr. oversaw the first U.S. nuclear-powered
aircraft carriers USS Enterprise (CVAN-65). Automation on nuclear-powered carriers today
could not evolve if it were not for the first aircraft carrier. In addition to advancements in
naval aircraft carriers, Research and Development Corporation (RAND) funded three Logic
Theorists—Allen Newell, Cliff Shaw, and Herbert Simon—to present a computer problem
solving-program. 47 In 1956, Stanford professor John McCarthy and Massachusetts Institute
of Technology (MIT) professor Marvin Minsky hosted the Dartmouth Summer Research
Project on Artificial Intelligence (DSRPAI). 48 The popular desire for AI in both the military
and academia realms increased the demand for adopting AI.

However, overhyped emerging technologies and huge expenses hindered the

advancements of automation. Computer scientists Peter Norvig and Stuart Russell wrote that
the AI headline news in the 1950s typically read “Electronic Super-Brains” and “Faster Than
Einstein,” but the hype in media was only part of the incubation for adopting automation.49

46 John Frederick Fuller, Armament and History: The Influence of Armament on History from the
Dawn of Classical Warfare to the End of the Second World War (Cambridge, MA: Da Capo Press, 1998).
47 Rockwell Anyoha, “The History of Artificial Intelligence,” SITN Blog (blog), August 28, 2017,
http://sitn.hms.harvard.edu/flash/2017/history-artificial-intelligence/.
48 John McCarthy et al., “A Proposal for the Dartmouth Summer Research Project on Artificial
Intelligence” (Hanover, NH: Dartmouth College, August 31, 1955), http://jmc.stanford.edu/articles/
dartmouth.html..
49 Stuart Russell and Peter Norvig, Artificial Intelligence: A Modern Approach (Englewood Cliffs, NJ:
Prentice Hall, 1995), 4.
19
Both Norvig and Russell understood that the demand was present, but two issues prevented
the adoption of automation or AI. In addition to a high price, they stated that computer
processors could not store or “remember” input data, and computers cost over $200,000 per
computer and simply too costly of an investment. 50

2. History of Automation by Chinese Prominent Leaders

The vision of automation and AI to modernize the Chinese Navy from 1958 to 1962
began with the Great Leap Forward, shifting its organizational structure. The leap
represented a catalyst because of the Sin-Soviet dispute; abandoned the previous balanced
development strategy; and, gave birth to a new social organization called the people’s
commune. 51 Under the Communist Party of China (CPC), then-Chairman Mao Tse-tung had
the vision to reconstruct the country from an agrarian economy to a communist society to
leap ahead of competing neighbors and adversaries. The Chinese sought to change its navies
like that of the Soviet Gorshkovian doctrinal changes which were then echoed by Admiral
Liu Huaqing. 52 Chinese naval thought leaders began to appreciate and take advantage of the
value of their navy for sea power and maritime domain awareness.

Two prominent PLAN admirals, Liu Huaqing and his successor Zhang Lianzhong,
contributed to the future of China’s naval force projection and rapid response capabilities,
according to James Lilley, former U.S. ambassador to China in 1989, and James Shambaugh,
George Washington University Professor on China policy. 53 As the Commander of PLAN
from 1982 to 1988 and then-Vice Chairman of the Military Commission, Admiral Huaqing
designed a strategy to promote the PRC’s 1982 naval maritime plan to move from coastal

50 Robert Garner, “Early Popular Computers, 1950 - 1970,” Engineering and Technology History
Wiki, January 9, 2015, https://ethw.org/Early_Popular_Computers,_1950_-_1970.
51 Roderick MacFarquhar, The Origins of the Cultural Revolution: Volume II, the Great Leap
Forward 1958–1960, 2nd ed. (New York: Columbia University Press, 1983).
52 Geoffrey Till, Seapower: A Guide for the Twenty-First Century, 2nd ed. (Abingdon, UK:
Routledge, 2009).
53 James Lilley and David L. Shambaugh, China’s Military Faces the Future (Abingdon, UK:
Routledge, 1999), 27.
20
defense toward a blue-water capability in incremental 10-year plans. 54 This vision led to
increasing demands for advanced technology in a three stage strategy for maritime defense.

In Artificial Intelligence and Automation by Nikolas Bourbakis, PRC Xi’an Jiaotong

University Professor Hongyi Wang discusses issues about AI related to organic systems to
knowledge development issues and unifying the gap between computers and human
brains. 55 This definition paves a pathway that starts knowing what the problem(s) are, drives
through development, requires structuring and/or visualization, and unification between a
human being and the computer. In a way, his point describes a marriage between two separate
entities designed to bring forth increased intelligence to help solve problems faster and
together.

PLAN seeks to end U.S. maritime superiority across both the Pacific and the Indian
Oceans by technologically advancing its fleet. According to the Office of Naval Intelligence
analysts, Adm. Huaqing, a member of the Chinese Academy of Science, assisted then-
President Deng Xiaoping and the Chinese Communist Party’s (CCP) Central committee’s
Politburo to help shape the future of PLA. 56 In 1988, Admiral Lianzhong designed “sea-
denial” out to the first chain of islands to about 150 miles from the mainland, otherwise
known as access and area denial (A2/AD). This defense strategy is known as “defense in
depth” and its purpose is to surveil and defense its sea borders with conventional and nuclear
submarines, long-range aircraft, and surface vessels. 57 China’s methodology for achieving
maritime superiority receives tremendous support from its thought leaders in academic
universities that brought forth its naval ISR vision to fruition.

In contrast to Feigenbaum’s “Expert Systems,” China’s approach heavily weighs the

initial development stage. Kai-fu Lee is very authoritative for this topic in technology

54 Andrew Erickson and Gabe Collins, “China’s Real Blue Navy,” The Diplomat, August 30, 2012,
https://thediplomat.com/2012/08/chinas-not-so-scary-navy/.
55 Bourbakis, Artificial Intelligence and Automation, 9.
56 Office of Naval Intelligence, China’s Navy 2007 (Washington, DC: Office of Naval Intelligence,
2007), 11–14, https://fas.org/irp/agency/oni/chinanavy2007.pdf.
57 Till, Seapower, 153.

21
between China and the United States. 58 He is the chairman and COE of Sinovation Ventures,
a Chinese-based technology-focused investment firm, and has held vital positions in Apple,
Silicon Graphics, Microsoft, and Google where he was president of Google China. In his
book “AI Superpowers,” he estimates automation will decimate up to 40–50 percent of jobs
worldwide. 59 This is the trade-off in adopting more automated systems and AI technologies
that decision-makers can consider in the adoption process.

In July 2017, the State Council of China released the “New Generation Artificial
Intelligence Development Plan,” which is a strategy to build a domestic AI industry worth
nearly U.S. $150 billion and to become the leading AI power by 2030. 60 AI Industry
Development Alliance, a co-sponsored alliance between more than 200 enterprises and
agencies nationwide to develop China’s AI industry. The 2016–2018 Chinese Three-Year
Guidance for Internet Plus Artificial Intelligence Plan was written for socioeconomic
development. Three-Year Action Plan for Promoting Development of a New Generation
Artificial Intelligence Industry (2018–2020) reinforces the AI development plans previously
mentioned.

In the adoption process, understanding and discussing the vision set by naval
leadership can help decision-makers who are in the process of adopting automation and AI
in the coming decades. The U.S. Navy leverages major university researchers and computer
scientists to adopt automation and AI. The PLAN’s Three- or Five-Year development plans
and AI strategy lay out the adoption process and executing each step in a rapid pace. The
next section will cover the various definitions of automation and AI.

ON DEFINING AUTOMATION AND AI

This section will explore the evolution of automation and AI over the last 70 years.
The purpose of discussing the definitions of automation is to shed light on the contrasting
definitions, derived from both the U.S. and China’s AI strategy documents to gain a better

58 Kai-fu Lee, AI Superpowers: China, Silicon Valley, and the New World Order (Boston: Houghton
Mifflin Harcourt, 2018).
59 Lee.
60 Sapio, Chen, and Lo, A New Generation of Artificial Intelligence Development Plan.

22
understanding of the history of adopting automation of naval ISR and the current adoption
progress.

Humankind derives from homo sapiens—wise man—because our cognitive

capabilities are important in our everyday lives when making decisions. Human beings in
the private, public, and commercial centers have constantly developed new superior tools
and technologies to solve a problem faster; they create it and then the Navy acquires and
applies the tools.

There is no worldwide accepted definition of AI, according to prominent AI

researchers Kirsch, Allen, Hearst and Kirsh, Brachman, Nilsson, Bhatnagar, and Monett, and
Lewis. 61 Depending on whom one asks from different communities, categorizing the terms
“AI” and automation and learning the levels of complexity will help design the fittest
definition for the current strategies between the two superpowers.

1. Working Definitions of Automation and AI

The following table depicts my working definition of automation versus AI drawn

from the elements of automation and AI experts since 1950. Based on the evidence from a
diverse group of researchers and fields of study, automation and AI can be simple, complex,
or extraordinary, ranging from prespecified to unpredictable and manual to automatic.
Simply put, both the tools and techniques solve problems and attain goals. My working
definitions are based on lists of definitions written by government personnel, academic
researchers, and computer scientists (see Table 1).

Table 1. Working Definitions for Automation and AI

Automation Artificial Intelligence

Automation is a technique of making a computer, a AI is an evolutionary and combined field of study to
process, or a system operate automatically by a create intelligent machines through a set of
mechanical or electronic device that takes place of established rules in the form of algorithms to achieve
the human labor to accomplish a task or goal. a specific goal or set of goals.

61 Pei Wang, “On Defining Artificial Intelligence,” Journal of Artificial General Intelligence 10, no. 2
(2019): 1–37, https://doi.org/10.2478/jagi-2019-0002.
23
 AI has become an increasingly popular topic, but the term AI is something of a
misnomer, according to NPS Professor John Arquilla. 62 Computer software and algorithms
are only productive and made “intelligent” when created, typed, and used by human beings.
By software automation, we “rely on computer systems as much as possible in software
development, in other words, to generate programs from information requirements
automatically.” 63 To simply put, without its creator, the computer itself or the data at rest
cannot produce intelligence on its own in a way that human beings can, nor can it empathize
or have a beating heart like a human being. Although a well-defined and acceptable definition
differs and goals to achieving computer systems with these human-like characteristics and
capabilities, automation is still not intelligent independent of the human being. However, to
remain competitive and gain or maintain knowledge on the maritime battlefield, naval ISR
is worth the effort to give this topic the time and space.

2. List of AI Definitions and Strategies

Table 2 shows a diverse list of definitions by technical experts compared to

government and public experts on automation and the following table on AI. The
International Society of Automation (ISA), a non-profit professional association founded in
1945 defines automation as “the creation and application of technology to monitor and
control the production and delivery of products and services.” 64 In “Artificial Intelligence:
A Modern Approach,” UC Berkeley professors Stuart Russell and Peter Norvig define AI as
“the designing and building of intelligent agents that receive percepts from the environment
and take actions that affect that environment.” 65 This definition by its nature builds upon an
idea similar to the human mind at first glance; however, unites the subfields of the following
computer science skills: machine learning, pattern recognition, computer vision, speech
recognition, etc.

62 John Arquilla, “The New Face of Battle” (lecture, Naval Postgraduate School, Monterey, CA,
December 4, 2019).
63 Bourbakis, Artificial Intelligence and Automation, 289.
64 “What Is Automation?,” International Society of Automation, accessed January 28, 2020,
https://www.isa.org/about-isa/what-is-automation/.
65 Russell and Norvig, Artificial Intelligence:

24
 Tables 2 shows a list of definitions of automation and AI between technical experts
and military and government professionals. Table 3 displays the United States AI strategy
and PRC’s development plan.

Table 2. Definitions by Technical and Government/Public Officials

Technical Experts:
Computer Scientists, Data Scientists
“The study is to proceed on the basis of the conjecture that every aspect of learning or any other feature of
intelligence can in principle be so precisely described that a machine can be made to simulate it.” 66
“By ‘general intelligent action’ we wish to indicate the same scope of intelligence as we see in human action:
that in any real situation behavior appropriate to the ends of the system and adaptive to the demands of the
environment can occur, within some limits of speed and complexity.” 67
“AI is concerned with methods of achieving goals in situations in which the information available has a certain
complex character. The methods that have to be used are related to the problem presented by the situation and
are similar whether the problem solver is human, a Martian, or a computer program.” 68
“A useful definition of intelligence… should include both biological and machine embodiments, and these should
span an intellectual range from that of an insect to that of an Einstein, from that of a thermostat to that of the most
sophisticated computer system that could ever be built.” 69
“AI becomes a science (1987—present)” (Russell and Norvig, 2002), which was later changed to “AI adopts the
scientific method (1987—present)” (Russell and Norvig, 2010), because “It is now more common to build on
existing theories than to propose brand-new ones, to base claims on rigorous theorems or hard experimental
evidence rather than on intuition, and to show relevance to real-world applications rather than toy examples.” 70
“I suggest we replace the Turing test by something I will call the ‘employment test.’ To pass the employment
test, AI programs must be able to perform the jobs ordinarily performed by humans. Progress toward human-
level AI could then be measured by the fraction of these jobs that can be acceptably performed by machines.” 71
“The creation and application of technology to monitor and control the production and delivery of products and
services.” 72

66 McCarthy et al., “A Proposal for the Dartmouth Summer Research Project on Artificial
Intelligence.”
67 Allen Newell and Herb Simon, “Computer Science as Empirical Inquiry: Symbols and Search,”
Communications of the ACM 19, no. 3 (March 1976): 116, https://doi.org/10.1145/360018.360022.
68 John McCarthy, Formalizing Common Sense: Papers, ed. Vladimir Lifschitz (New York: Ablex
Publishing Corporation, 1990), 246.
69 James S. Albus, “Outline for a Theory of Intelligence,” IEEE Transactions on Systems, Man and
Cybernetics 21, no. 3 (June 1991): 474, https://doi.org/10.1109/21.97471.
70 Wang, “On Defining Artificial Intelligence,” 14.
71 Nils J. Nilsson, “Edward A (‘Ed’) Feigenbaum: United States - 1994,” Association for Computing
Machinery, A.M. Turing Award, 2019, https://amturing.acm.org/award_winners/
feigenbaum_4167235.cfm.
72 Peter Mendel et al., “Interventions in Organizational and Community Context: A Framework for
Building Evidence on Dissemination and Implementation in Health Services Research,” Administration
and Policy in Mental Health 35, no. 1–2 (March 2008): 21–37, https://doi.org/10.1007/s10488-007-0144-9.
25
 Government/Public: Military, Private, Public, Commercial
“The theory and development of computer systems able to perform tasks normally requiring human
intelligence, such as visual perception, speech recognition, decision-making, and translation between
languages.” 73
“A branch of computer science dealing with the simulation of intelligent behavior in computers. The capability
of a machine to imitate intelligent human behavior.” 74
“Artificial intelligence (AI), the ability of a digital computer or computer-controlled robot to perform tasks
commonly associated with intelligent beings.” 75
“The field of computer science dedicated to solving cognitive problems commonly associated with human
intelligence, such as learning, problem solving, and pattern recognition.” 76
“‘Create smarter, more useful technology and help as many people as possible’ ‘from translations to healthcare
to making our smartphones even smarter.’” 77
“The ability of a machine communicating using natural language over a teletype to fool a person into believing
it was a human. ‘AGI’ or ‘artificial general intelligence’ extends this idea to require machines to do everything
that humans can do, such as understand images, navigate a robot, recognize and respond appropriately to facial
expressions, distinguish music genres, and so on.” 78
“It is the science and engineering of making intelligent machines, especially intelligent computer programs. It is
related to the similar task of using computers to understand human intelligence, but AI does not have to confine
itself to methods that are biologically observable.” 79
“Advancing the file of machine intelligence and are creating new technologies to give people better ways to
communicate.” 80
“Fundamentally exists to substitute work activities undertaken by human labor with work done by machines,
with the aim of increasing quality and quantity of output at a reduced unit cost.” 81

73 “Artificial Intelligence,” in Oxford Reference, accessed June 8, 2020,

https://www.oxfordreference.com/view/10.1093/oi/authority.20110803095426960.
74 Merriam-Webster Inc., “Artificial Intelligence,” accessed February 12, 2020, https://www.merriam-
webster.com/dictionary/artificial%20intelligence.
75 B.J. Copeland, “Artificial Intelligence,” in Encyclopedia Britannica, March 24, 2020,
https://www.britannica.com/technology/artificial-intelligence.
76 “What Is Artificial Intelligent?,” Amazon Web Services, accessed June 5, 2020,
https://aws.amazon.com/machine-learning/what-is-ai/.
77 Bernard Marr, “The Key Definitions of Artificial Intelligence (AI) That Explain Its Importance,”
Forbes, February 14, 2018, https://www.forbes.com/sites/bernardmarr/2018/02/14/the-key-definitions-of-
artificial-intelligence-ai-that-explain-its-importance/#532664514f5d.
78 This quote is attributed to Dr. Matt Mahoney, a data compression expert; Daniel Faggella, “What Is
Artificial Intelligence? An Informed Definition,” Emerj, December 21, 2018, https://emerj.com/ai-glossary-
terms/what-is-artificial-intelligence-an-informed-definition/.
79 Maki K. Habib, Revolutionizing Education in the Age of AI and Machine Learning (Hershey, PA:
IGI Global, 2019), 6.
80 Marr, “The Key Definitions of Artificial Intelligence (AI) That Explain Its Importance.”
81 Mark Muro, Robert Maxim, and Jacob Whiton, Automation and Artificial Intelligence: How
Machines Are Changing People and Places? (Washington, DC: Brookings Institute: Metropolitan Policy
Program, 2019), https://www.brookings.edu/wp-content/uploads/2019/01/
2019.01_BrookingsMetro_Automation-AI_Report_Muro-Maxim-Whiton-FINAL-version.pdf.
26
 Government/Public: Military, Private, Public, Commercial
“Artificial intelligence (AI) refers to the ability of a computer or a computer-enabled robotic system to process
information and produce outcomes in a manner similar to the thought process of humans in learning, decision
making and solving problems.” 82
(1) “Any artificial system that performs tasks under varying and unpredictable circumstances without
significant human oversight, or that can learn from experience and improve performance when exposed to
data.
(2) An artificial system developed in computer software, physical hardware, or other context that involves that
solves tasks requiring human-like-perception, cognition, planning, learning, communication, or physical
action
(3) An artificial system designed to think or act like a human, including cognitive architectures and neural
networks.
(4) A Set of techniques, including machine learning, that is designed to approximate a cognitive task
(5) An artificial system designed to act rationally, including an intelligent software agent or embodied robot
that achieves goals using perception, planning, reasoning, learning, communicating, decision-making, and
acting.” 83

Table 3. U.S. and PRC AI Strategies

United States People’s Republic of China

“Artificial intelligence (AI) is one such “Artificial intelligence has become the new focus of
technological advance. AI refers to the ability of international competition. Artificial intelligence is
machines to perform tasks that normally require thought to be the strategic technology leading the future,
human intelligence—for example, recognizing the world’s major developed countries regard the
patterns, learning from experience, drawing development of artificial intelligence as the major
conclusions, making predictions, or taking action— strategy to increase national competitiveness and enhance
whether digitally or as the smart software behind national security, therefore they intensify the introduction
autonomous physical systems.” 84 of plans.” 85
“AI is already a popular concept, but there is not yet a
universally accepted definition for it. The traditional
approach to AI development is to study how human
intelligence occurs and create machines that imitate
human thinking and behavior.” 86

82 Mike Quindazzi, “Artificial Intelligence and Robotics 2017: Leveraging Artificial Intelligence and
Robotics for Sustainable Growth,” ASSOCHAM India, April 4, 2017, 7, https://www.slideshare.net/
MikeQuihumndazzi/artificial-intelligence-and-robotics-in-2017.
83 “National Security Commission on Artificial Intelligence Act of 2018,” § S. 2806 (2018), 2,
https://www.congress.gov/115/bills/s2806/BILLS-115s2806is.pdf.
84 Department of Defense, Summary of the 2018 Department of Defense Artificial Intelligence
Strategy.
85 State Council, “New Generation Artificial Intelligence Development Plan [国务院关于印发新一代
人工智能发展规划的通知],” July 8, 2017, ttp://www.gov.cn/zhengce/content/2017- 07/20/
content_5211996.htm.
86 China Institute for Science and Technology Policy, China AI Development Report 2018 (Beijing:
China Institute for Science and Technology Policy at Tsinghua University, 2018), 9,
http://www.sppm.tsinghua.edu.cn/eWebEditor/UploadFile/China_AI_development_report_2018.pdf.
27
 The technical experts divide the definitions between theory and practice which have
evolved since the 1950s. Computer scientists McCarthy, Newell, Simon, and Minsky
believe in computers achieving goals to solve a specific problem that a machine can
simulate in each environment. 87 The government, private, public, and commercial trends
toward productivity. This study shows that both technical experts and government
researchers are more concerned about automation and AI performing a task to solve a
problem expeditiously. AI experts and organizations such as Matt Mahoney, Facebook,
Congress specifically point out the desire for AI to imitate human intelligence for better
communication.

In recent years, a revitalization of AI has erupted, partly due to the hope of

successful new techniques such as machine learning and deep learning. Various categories
of AI include Artificial Narrow Intelligence (ANI), 88 Artificial General Intelligence
(AGI),, and Artificial Superintelligence (ASI). 89 Technical experts develop ANI, or
“weak” intelligence, to perform a singular task such as human beings playing chess online,
marketing trends, autonomous cars, and speech and pattern recognition. These new labels
show that the meaning and definition of “AI” has changed or evolved beyond mainstream
AI and is being re-branded, though the ideas of AI are not new.

Besides the attempt to define AI based on its original meaning, though under new
names now, the speculative nature of terms such as “superintelligence” by Kurzweil or
“singularity” by Bostrom aim to achieve higher goals beyond human intelligence in
machines., 90 Technical experts are attempting to develop AGI, or “strong” intelligence, to

87 Bourbakis, Artificial Intelligence and Automation.

88 “What Is Narrow AI?,” Narrow AI, accessed April 27, 2020, https://deepai.org/machine-learning-
glossary-and-terms/narrow-ai.
89 Ben Goertzel and Cassio Pennachin, Artificial General Intelligence (New York: Springer, 2007);
Pei Wang and Ben Goertzel, “Introduction: Aspects of Artificial General Intelligence,” in Proceedings of
the 2007 Conference on Advances in Artificial General Intelligence: Concepts, Architectures and
Algorithms, ed. Ben Goertzel and Pei Wang (Amsterdam: IOS Press, 2007), 1–16.
90 Ray Kurzweil, The Singularity Is Near: When Humans Transcend Biology (New York: Penguin
Books, 2006); Nick Bostrom, Superintelligence: Paths, Dangers, Strategies (Oxford, UK: Oxford
University Press, 2014).
28
a level of imitating the human brain. ASI is the goal of AI in which a robot demonstrates
all human-like cognitive and physical being, like the human-like robot named Sonny in the
2004 film “I, Robot” starring Will Smith.91 This study does not consider these concepts as
realistic and will not analyze them in this thesis. Authors such as Wang, Liu, and Doughtery
have argued that even though computer programs or algorithms can replicate human tasks,
AGI is not the same as human flesh or human experiences and will not lead to singularity. 92

The 2018 DOD AI Strategy explicitly states that AI one technological advancement
that refers to machines performing a task that a human being would perform. 93 China’s
New Generation AI Development Plan—along with Made in China 2025 released in May
2015—does not have a clear definition but, rather, focuses on the development stage within
the science, technology, research, and development stage with the core AI policy on
intellectual property and intellectual rights. 94 Both nations with different values will likely
have different definitions and perceptions of automation.

3. Conceptualizing Automation

The first step is to categorize automation and then separate the definitions into three
levels of complex applications: simple, complex, or very complex, according to the
consulting firm PricewaterhouseCoopers (PwC). 95 Simple automation can be as simple as
commercial use to manage Navy resources or personnel. Figure 2 conceptualizes how to
adopt automation and AI because of the varying levels of complexity ranging from
augmentation, assisted, and autonomous intelligence.

91 Rosario Girasa, Artificial Intelligence as a Disruptive Technology: Economic Transformation and

Government Regulation (New York: Springer International Publishing, 2020), 10–11.
92 Pei Wang, Kai Liu, and Quinn Dougherty, “Conceptions of Artificial Intelligence and Singularity,”
Information 9, no. 4 (2018): 1–2, https://doi.org/10.3390/info9040079.
93 The United States government has a clear definition in its AI strategy whereas China uses a
development plan.
94 China Institute for Science and Technology Policy, China AI Development Report 2018, 76.
95 Scholoer Consulting Group, “What Is Artificial Intelligence?,” Understanding Artificial
Intelligence, accessed January 27, 2020, http://www.schloerconsulting.com/understanding-artificial-
intelligence.
29
 Figure 2. Human Decisions Supplemented by AI Along with AI
Continuum 96

This graph was developed by PwC as an analytical tool to display how the AI
continuum is different from each category of automation. The three main differences are
assisted intelligence, augmented intelligence, and autonomous intelligence. In naval
affairs, the U.S. maritime battlefield decision aid called the COP (see Figure 3), is an
assisted intelligence tool that automates tracking of naval platforms radiating their unique
identifiers at sea or on land. 97

96 Source: Mike Quindazzi (@MikeQuindazzi), “Human decisions supplemented by

#ArtificialIntelligence along the #AI continuum; 1-#augmented, 2-#assisted, 3-#autonomous. #pwc,”
Twitter, March 22, 2017, 9:15pm., https://twitter.com/mikequindazzi/status/844764424917286912
97 Gary Koch, “NORAD-USNORTHCOM C2 Systems Overview for AFCEA JC2C: Evolving the
GCCS FoS” (AFCEA Course, Fairfax, VA, March 25, 2010), slide 22, https://www.slideshare.net/
gdkochjr/afcea-jc2c-gccs-presentation.
30
 Figure 3. Common Operational Picture
for Situational Awareness for Operators 98

4. Levels of Automation

The human operators behind automated computers and offer machines can better
adopt automation through understanding the various levels of automation. The Sheridan
and Verplank’s Scale of Human-Machine Interaction on automation, shown in Table 4,
describes levels 1 through 10 on systematic work that has been done to evaluate the key
levels of automation. The highly automated systems such as the U.S. Aegis combat system
and Dragon Eye improve automation for decision making and allows the human operator
to make decisions within each level and make the final decision.

Table 4 conceptualizes the degree of automation one requests from entirely manual
to entirely automatic (along the x-axis) and, the level of task entropy from prespecified task
to unpredictable future requirements along (along the y-axis).

98 Source: Koch, “NORAD-USNORTHCOM C2 Systems.”.

31
 Table 4. Levels of Automation 99

Automation Automation Description

Levels
1 The computer offers no assistance: human must take all decisions and
actions
2 The computer offers a complete set of decisions/actions alternatives, or
3 Narrows the selection down to a few, or
4 Suggests one alternative, and
5 Executes that suggestion if the human approves, or
6 Allows the human a restricted time to veto before automation execution, or
7 Executes automatically, then necessarily informs humans, and
8 Informs the human only if asked, or
9 Informs the human only if it, the computer, decides to.
10 The computer decides everything and acts autonomously, ignoring the
human.
The concept of automation levels ranges from manual to full autonomy. 100

Based on research by Sheridan and Verplank, automation can override human operators to
execute a task and may not be the fittest in a changing environment. 101 While automation
levels are designed for consistent quality control and performance checks, the levels of
automation also assist with re-evaluation and adjustments for improvements in a system or
management systems that can range from manual to fully automated. Thus, if he or she
understands the various levels of automation, then the human operator can adapt to
accomplish the mission (depending on the level of failure) when failure within automation
may occur.

Sheridan and Verplank’s entropy scale from Figure 4 can helps decision-makers
gain awareness and recognize the intricate level between entropy and automation and
applied appropriate can better assist commanders in the business of naval ISR.

99 Thomas Sheridan and William Verplank, Human and Computer Control of Undersea Teleoperators
(Cambridge, MA: Massachusetts Institute of Technology, 1978).
100 Heather M. Roff and David Danks, “‘Trust but Verify’: The Difficulty of Trusting Autonomous
Weapons Systems,” Journal of Military Ethics 17, no. 1 (2018): 20, https://doi.org/10.1080/
15027570.2018.1481907.
101 Sheridan and Verplank, Human and Computer Control of Undersea Teleoperators.

32
 Figure 4. Automation and Task Entropy Continuum
to Help Solve Problems with AI 102

This study shows that there is a variety of definitions and interpretations of automation and
AI across numerous studies. Most scholars, researchers, and commanders can readily agree
that techniques can be automated, through a set of rules, for faster results by computers;
normally a human being would take longer to accomplish several analytical or
mathematical problems using a computer.

NAVAL ISR

Part one of this research methodology will be a qualitative research design in five
parts. First, I will discuss how naval leadership is adopting aircraft carriers, ships, and
aircraft and to compare Western versus Eastern naval leadership. Second, I will breakdown
the main intelligence domains within which automation and AI are used: IMINT, SIGINT,
OSINT, MASINT, HUMINT. Third, I will cover the current development of hardware and
software on U.S. Aegis, Ship’s Self Defense (SSDS) and Type-346 sensors, and relevant
radars used for ISR.

102 Source: Sheridan and Verplank, 4–7.

33
 1. U.S. Naval Leaders Pursue Automation and AI

In the naval business of ISR, Western leadership has always had a keen eye and
thirst to know and understand the adversary. Abraham Lincoln once said, “I do not like
that man. I must get to know him better. I do not like that person, I must get to know them
better.” 103 Naval ISR is a profound art of war in the pursuit to know and understand the
adversary and noticing a pattern of life at sea and abnormal or amalgam within a naval
pattern of life; therefore, it contributes to understanding the Navy that which is better than
one’s own Navy.

Policy plays a key role in improving naval ISR. In Carl von Clausewitz’s On War,
he states that “war is nothing but the continuation of policy with other means.” 104 United
States Naval policy on ISR is promulgated in the 2017 Joint Publication 2-01. 105 Naval
ISR is an instrument to be used in times of war and peacetime where the sea is a highway,
as Alfred Thayer Mahan put, “over which men may pass in all directions…familiar and
unfamiliar dangers of the sea, both travel and traffic by water” has been easier and
cheaper. 106 Knowing the adversary or competitor requires the collection of intelligence
and knowledge of what lies on the highways or sea lanes. The policy set by acting CNO
Mike Gilday and his predecessor Adm. John Richardson is to maintain maritime
superiority; the mission is to protect America from attack, preserve strategic
influence…and to modernize the U.S. Navy. 107

Over the next decade, both the Ford-class nuclear-powered aircraft carrier and the
Chinese Type 001 “Liaoning,” Type 002 “Shandong,” as shown in Figure 5, and future
aircraft carriers will remain symbols of maritime superiority and project power, but may

103 Daniel Coenn, Abraham Lincoln: His Words, ed. Maureen Harrison and Steve Gilbert (New York:
Barnes & Noble, 1994).
104 Carl von Clausewitz, On War, ed. Michael Howard and Peter Paret, Reprint (Princeton, NJ:
Princeton University Press, 1989), 69.
105 See Joint Chiefs of Staff, Joint Intelligence Support to Military Operations, Appendix B-1.
106 Alfred Thayer Mahan, The Influence of Sea Power Upon History, 1660–1783 (New York: Dover
Publications, 1987), 25.
107 Richardson, A Design for Maritime Superiority.

34
not be the main line of effort at sea. 108 The aircraft carrier represents “National Command
Authority and warfighting Commanders-in-Chief with a flexible force to respond to a wide
variety of international challenges…four and one-half acres of sovereign-and mobile-
American territory that can project U.S. power whatever it might be required.” 109
However, part of the ISR situational awareness calculus will be greatly improved with both
unmanned aerial vehicles (UAV), unmanned undersea vehicles (UUV), and manned naval
platforms. Unmanned and manned ISR platforms in addition to the Navy’s nuclear aircraft
carrier and amphibious aircraft carriers—used to deploy troops on the ground—represent
maritime superiority.

Figure 5. Shandong (Type 001A) Aircraft Carrier 110

Where there is the naval theorist Mahan in the West, there is Sun Tzu in the east.
Unlike the Western concept of not politicizing the military, the East has the opposite
concept, which is to politicize the military because the party controls the military. 111 The

108 “The Aircraft Carrier Plan: A Dual-Track Strategy,” Aircraft Carriers, accessed February 21, 2020,
https://fas.org/man/dod-101/navy/docs/vision/carriers.htm.
109 Federation of American Scientists.
110 Source: Ronald O’Rourke, China Naval Modernization: Implications for U.S. Navy Capabilities—
Background and Issues for Congress, CRS Report No. RL33153 (Washington, DC: Congressional
Research Service, 2020), 10, https://crsreports.congress.gov/product/pdf/RL/RL33153/233.
111 Richard McGregor, “5 Myths about the Chinese Communist Party,” Foreign Policy, January 3,
2011, https://foreignpolicy.com/2011/01/03/5-myths-about-the-chinese-communist-party/.
35
PLAN is the party’s naval force and continues to build aircraft carriers. PLAN platforms
today consist of 86 Missile Patrol craft, 28 Corvettes, 26 Medium Landing Ships, 28
destroyers, and 41 frigates. 112 The PLAN have more small missile boats, torpedo boats,
and long-range weapons than aircraft carriers, which shows that China does not need
aircraft carriers to operate effectively.

2. PRC Naval Leaders Pursue Automation and AI

Naval ISR can be viewed as an element of spy-craft on par with the livelihood of
water and in execution, achieve great results. In Sun Tzu’s book “Art of War,” his chapter
on the “Use of Spies” states, “Hence it is only the enlightened ruler and the wise general
who will use the highest intelligence of the army for purposes of spying…Spies are a most
important element in war, because on them depends an army’s ability to move.” 113 Chinese
naval leadership likely uses ISR as spies in support of moving the army when necessary.

The PLAN is currently modernizing its naval battle group structure. In 2012 at the
18th Party Congress, then-President Hu Jintao called for China to become a “maritime
power.” In April 2018, President Xi-Jinping echoed this vision and stated that “the task of
building a powerful navy has never been as urgent as it is today.” 114 China’s 2019 Defense
White Paper states:

The PLA Navy (PLAN) has extended training to the far seas and deployed
the aircraft carrier task group for its first far seas combat exercise in the
West Pacific. It has organized naval parades in the South China Sea and the
waters and airspace near Qingdao, and conducted a series of live force-on-

112 Defense Intelligence Agency, China’s Military Power (Washington, DC: Defense Intelligence
Agency, 2019), 68, https://www.dia.mil/Portals/27/Documents/News/Military%20Power%20Publications/
China_Military_Power_FINAL_5MB_20190103.pdf.
113 Sun Tzu, Sun Tzu on the Art of War: XIII. The Use of Spies, trans. Lionel Giles, 1994,
https://fas.org/man/artofwar.htm#13..
114 China Power Team, “How Is China Modernizing Its Navy?,” China Power, December 17, 2018,
https://chinapower.csis.org/china-naval-modernization/.
36
 force exercises codenamed Mobility and systematic all-elements
exercises. 115

The PLAN’s modernization and extended sea training, live-fire exercises, and red
versus blue exercises with its aircraft carrier task group indicate great improvement of its
automation capabilities in the littorals (see Figure 6). This trend, set by President Xi-
Jinping, indicates the PLAN will likely develop and test automation and AI analytical tools
to increase productivity in its training environment, in exercises, and operations to the far
seas. Figure 6 depicts an approximation of PLAN’s order of battle in each theater of
operations that can support naval ISR.

Figure 6. China’s Military Power - DIA’s 2019 PLAN Fleet Composition in

Support of Naval ISR 116

The major plans and timelines between the two navies illuminate their
organizational differences. With China, its desires for rapid modernization is clear but
focused more on developing automation. In Mao Tse-tung ‘s 1936 “Problems of Strategy

115 Anthony H. Cordesman, China’s New 2019 Defense White Paper: An Open Strategic Challenge to
the United States, But One Which Does Not Have to Lead to Conflict (Washington, DC: Center for
Strategic & International Studies, 2019), https://csis-prod.s3.amazonaws.com/s3fs-public/publication/
190724_China_2019_Defense.pdf.
116 Source: Defense Intelligence Agency, China’s Military Power, 68.

37
in China’s Revolutionary War,” the first chapter revolves around the laws of war being
developmental at all levels in conjunction with China’s Five-Year Plan, which
demonstrates a similar strategy of development for AI today. 117 This plan specifies a major
scientific and technological infrastructure construction plan through 2020 and “Defense
and Dual-Use Plans and Strategies.” 118 Today, both U.S. CNO and PLAN Commanders
strongly emphasize automating naval ISR for situational awareness.

3. Technical Naval ISR

Intelligence, surveillance, and reconnaissance are three separate but vital functions
that represent the defense trinity. The U.S. Navy-Marine team and the People’s Liberation
Army Navy team orchestrate ISR for situational awareness in which automation plays a
vital role in national security and national defense. On January 14, 1991, the Joint
Surveillance and Target Attack Radar System, a new system called JSTARS, was the first
operational employment and deployment during Operation Desert Shield (ODS) in the
Persian Gulf. 119 There was a strong belief and justification for ISR capabilities because
locating and destroying Scud missiles in Iraq was a vexing problem not solved in the ODS.
ISR via JSTARs was meant to fill this gap that could provide wide-area, long-range
surveillance through a moving target indicator (MTI) and a battle management
technological aid.

JP 1-02, Department of Defense (DOD) Dictionary of Military and Associated

Terms defines intelligence, surveillance, and reconnaissance (ISR) as: “An activity that
synchronizes and integrates the planning and operations of sensors, assets, processing,
exploitation, and dissemination systems in direct support of current and future

117 Mao Tse-tung, Problems of Strategy in China’s Revolutionary War, Selected Works of Mao Tse-
Tung (Peking: Foreign Language Press, 1936), https://www.marxists.org/reference/archive/mao/selected-
works/volume-1/mswv1_12.htm.
118 Katherine Koleski, The 13th Five-Year Plan (Washington, DC: U.S.-China Economic and Security
Review Commission, 2017), 43, https://www.uscc.gov/sites/default/files/Research/The%2013th%20Five-
Year%20Plan_Final_2.14.17_Updated%20%28002%29.pdf.
119 Nordin Yusof, Space Warfare: High-Tech War of the Future Generation (Skudai, Malaysia:
Penerbit Universiti Teknologi Malaysia, 1999), 222.
38
operations.” 120 This joint intelligence operation consists of global naval ISR platforms
that, when tasked, starts the first stage of a battle to locate and identify adversary forces or
a specific target. As shown in Figure 7, the ISR process synchronizes with both intelligence
collection management and current operations.

Figure 7. Venn Diagram on ISR Visualization to Show Teamwork Between

Intelligence Collection (J2) and Operations (J3) 121

Given the mission and task, the first stage of a battle is to locate, identify, and
collect data from remote sensors such as Aegis automation system, visualize the data via
computer software, such as GCCS; and, deliver relevant and accurate exploited and
analyzed intelligence to the customer. ISR is quite the iterative process, slow or fast.

120 Joint Chiefs of Staff, Department of Defense Dictionary of Military and Associated Terms, 116.
121 Source: Joint Chiefs of Staff, Joint Intelligence Support to Military Operations, III–33.

39
 4. U.S. Aegis

There are two significant advanced automated weapons systems in the U.S. Navy
and the PLAN, the west and the east, which accelerate and bring automation to life—Aegis
and the Dragon Eye combat systems. First, Aegis has played a pivotal role over the last
sixty years on adaptive ship’s self-against aircraft, anti-ship missiles, surface threats, and
subsurface threats— “an integrated collection of sensors, computers, software, displays,
weapon launchers, and weapons named for the mythological shield that defended
Zeus.” 122 Aegis was initially developed in November 1963, which was called the
Advanced Surface Missile System (ASM) Project. Later in 1969, ASM was renamed to
Aegis with an awarded contract to Radio Corporation of America (RCA), which is the
legacy industrial corporation of Lockheed Martin. The only streamlined consistency was
the comfortable jumper, bell-bottom style uniforms—called coveralls—that the Sailors
wore while using this combat system.

Historically, it has taken decades to field and successfully use automation for ship
self-defense at sea. About 10 years after initial development, the Army-Navy SPY-1 radar
associated with Aegis successfully tracked its first target at the land-based test site and then
operated at sea onboard USS NORTON SOUND (AVM-1). It took nearly 20 to 30 years
of research, development, and testing to successfully use this automation tool. 123 For
example, in 1982 the U.S. Navy conducted two successful intercepts with SM-1 missiles
via TICONDEROGA-class cruisers (also onboard Arleigh Burke-class destroyers) and
then in 1991, USS SAN JACINTO fired its first Tomahawk land-attack missile (TLAM)
to strike Baghdad in Operation Desert Storm in the Red Sea. 124 The COP, fueled by Aegis
and SPY-1 radar, is significantly important for indications and warning for ship’s self-

122 Ronald O’Rourke, Navy Aegis Ballistic Missile Defense (BMD) Program: Background and Issues
for Congress, CRS Report No. RL33745 (Washington, DC: Congressional Research Service, 2018),
https://crsreports.congress.gov/product/pdf/RL/RL33745/189.
123 James C. Bussert and Bruce A. Elleman, People’s Liberation Army Navy: Combat System
Technology, 1949–2010 (Annapolis, MD: Naval Institute Press, 2011).
124 Joseph T. Threston, “The AEGIS Weapon System,” Naval Engineers Journal 121, no. 3 (October
2009): 85–108, https://doi.org/10.1111/j.1559-3584.2009.00205.x.
40
defense and executing a presidential order to launch and re-launch missiles on a prioritized
target.

Launching a strike on Baghdad set a precedent for launching a missile from the sea
and human operators watching this launch on a screen similarly displayed as a video game
console. The easiest way to track an aircraft or ship is by the navigational radar, which falls
Naval law, according to the United Nations Convention on the Law of the Sea and Naval
Warfare Planning 1–14M (Law of Naval Operations). 125 This brought forth meaning and
necessity for the U.S. DOD and the U.S. Navy, to have a justification to adopt naval ISR.

The U.S. Navy is adopting automation technologies for battlespace awareness.

“The battlefield technologies of electronic warfare, intelligence, and battlefield
automation, as well as precision-guided munitions, thus generated responses in each of the
services: Aegis for the Navy, AWACS for the Air Force, and CEWI for the Army.” 126 For
the navy, automating high energy laser weapons onboard Aegis Baseline 10, is being
developed and testing onboard the future USS Jack H. Lucas (DDG-125) with SPY-6 radar
by 2023. 127 Lockheed Martin received a USD 150 million contract to integrate High
Energy Laser Weapon Systems development, manufacture, and delivery for ISR. 128 Ship
automation combined with target acquisition benefits naval ISR by automating tasks for
situational awareness and self-defense.

Today’s Aegis program management is faster than the previous procurement and
fielding timelines. For example, the Office of the Chief of Naval Operations N9 department
(Warfare Systems) implemented a $294 million in contracts for the “2-4-6” program for

125 Department of the Navy, The Commander’s Handbook on the Law of Naval Operations
(Washington, DC: Office of the Chief of Naval Operations, 2017), https://www.hsdl.org/?view&did=
806860; United Nations Convention on the Law of the Sea, opened for signature December 10, 1982, 1833
U.N.T.S. 3, 397; 21 I.L.M. 1261 (1982) (entered into force Nov. 16, 1995) [hereinafter UNCLOS].
126 Kenneth Allard, Command, and Control, and the Common Defense (Fort McNair, Washington,
DC: National Defense University, Institute for National Strategic Studies, 1996), 148, 163.
127 Ben Werner, “Aegis Combat System Baseline 10 Set to IOC in 2023,” USNI News, January 15,
2019, https://news.usni.org/2019/01/15/40397.
128 “Lockheed Martin Receives $150 Million Contract to Deliver Integrated High Energy Laser
Weapon Systems to U.S. Navy,” Lockheed Martin Newsroom, March 1, 2018,
https://news.lockheedmartin.com/2018-03-01-Lockheed-Martin-Receives-150-Million-Contract-to-
Deliver-Integrated-High-Energy-Laser-Weapon-Systems-to-U-S-Navy.
41
modernizing existing Aegis ships, specifically cruisers that primarily serve as the air
defense command. 129 “No more than two of the cruisers are to enter the modernization
program each year, none of the cruisers is to remain in reduced status for modernization
for more than four years, and no more than six of the cruisers are to be in the program at
any given time.” 130

5. PLAN Type-348 Dragon Eye

Since the 1950s, the CCP inherited a large number of American and Japanese navy
ships, and rest were provided by the Soviet Union after its collapse in 1991. 131 This is the
reason why PLAN Liaoning aircraft carriers look quite like the Russian Kuznetsov-class
aircraft carrier. In 1998, the Chinese Research Institute in Nanjing developed a prototype
Type 346 and upgrades to the Dragon Eye phased array radar utilizes the S-band to further
improve surveillance radar for satellite communications, air traffic control, weather radar,
and surface ship radar.

Over the past twenty years, the PLAN has developed plans to make progress by
combining foreign and domestic military equipment. According to China’s 13th PLA “Five
Year Plan,” its Navy is moving toward an “Intelligentized Age.” 132 To reach this level of
intelligence and apply it to the Navy, the PLAN built two domestically designed naval
destroyers (DDGs) with Aegis-like phased-array panels. The Mineral Me “Band Stand”
one of PLAN’s data track and weapons control system with a passive range of 242 nautical
miles (nm) and active range of 97 nm (some say up to 134 nm). 133 As for PLAN’s data
link system, the Chinese integrated datalink system is called the Joint Service Integrated
Datalink System (JSIDLS), which plays the key, if not the most important role, in

129 David B. Larter, “With $294 Million in Contracts, the U.S. Navy Keeps Its Promise to Upgrade
Cruisers,” Defense News, August 24, 2018.
130 O’Rourke, Navy Aegis Ballistic Missile Defense (BMD) Program, 1.
131 Bussert and Elleman, People’s Liberation Army Navy: Combat System Technology, 1949–2010.
132 Graham Winstanley, ed., Artificial Intelligence in Engineering (Tiptree, Essex, UK: Courier
International Ltd, 1991).
133 Sarah Kirchberger, Assessing China’s Naval Power: Technological Innovation, Economic
Constraints, and Strategic Implications (Berlin: Springer-Verlag, 2015), 192.
42
connecting networks and communication for PLAN naval ISR platforms. 134 Mineral Me
and JSIDLS are very comparable to Aegis and Link-16. AI venture capitalists Kai-fu Lee
would reinforce this behavior as “China copycats.” 135 The challenge for the PLAN is not
their automation and Dragon Eye sensor capabilities but, rather, their ability to adapt and
operationalize their ISR equipment and platforms in a joint environment.

The Chief of Naval Operation’s (CNO) Maritime Design 2.0 aims squarely at
competing with Near Peer competitors such as China and Russia and the U.S. Navy must
be agile to keep pace with technology. Automation is important because it helps us learn
more about ourselves and how the U.S. Navy operates compared to the PLA Navy. First,
it provides situational awareness for trend analysis or early I&W. Second, automation is
more than your average fascination because it can automatically perform a task faster than
a human being can. Third, algorithms written by humans help fill the human gap of
understanding and recognize trends more quickly than a human can. While predicting the
future in detail is near impossible, automation can assist human beings in recognizing
patterns of behaviors or trends to increase the chances of forecasting what the opponent’s
next move might be. Thus, continuously making the effort to continuously learn about
automation combined with AI software can significantly contribute to the adoption process.

134 Kirchberger, 201.

135 Lee, AI Superpowers.

43
THIS PAGE INTENTIONALLY LEFT BLANK

44
 III. METHODOLOGY

Many theoretical and systems engineering design methods exist to illustrate a

complex adaptive process of implementing automation and AI technologies. Less is
known, however, about the dynamic process of adoption, including related factors about
decisions as to how to increase or decrease adoption of automation and AI pertaining to
naval ISR. According to researchers from George Washington University and New York
University, the need for adoption becomes apparent because of a knowledge gap, a search
for solutions, an initial decision to a specific technological solution, and an implementation
of the solution. 136 The overall goal is to identify emerging technologies with Artificial
Intelligence across the S-Curve adoption framework that are possibly modifiable, and thus,
demonstrate ways that improve naval ISR for sea power.

S-CURVE METHODOLOGY

To explore the adoption of automation in naval ISR between the two AI superpower
navies, this thesis will use Everett Rogers’ 1962 S-curve model of the diffusion process.
This proposed step-by-step process would encourage the Department of the Navy to
analyze both automated systems and AI for naval ISR every quarter within a fiscal year.
Quarterly reports can help improve the adoption process with updated or upgraded tactics,
techniques, and strategies, to harness new AI technologies. This “how to” format will be
illustrated through a systematic and flexible design model that can be applied and re-
applied quarterly because of the emerging technologies that outpace current naval
programs. Quarterly reports are critical feedback and auditing tool necessary to conduct
performance measures and stay competitive.

136 See various adoption models for different complex adoption problems by the following sources:
Fariborz Damanpour and Marguerite Schneider, “Phases of the Adoption of Innovation in Organizations:
Effects of Environment, Organization and Top Managers,” British Journal of Management 17, no. 3
(September 2006): 215–36, https://doi.org/10.1111/j.1467-8551.2006.00498.x; Michael J. Gallivan,
“Organizational Adoption and Assimilation of Complex Technological Innovations: Development and
Application of a New Framework,” The DATABASE for Advances in Information Systems 32, no. 3
(Summer 2001): 51–85, https://doi.org/10.1145/506724.506729; and Peter Mendel et al., “Interventions in
Organizational and Community Context: A Framework for Building Evidence on Dissemination and
Implementation in Health Services Research,” Administration and Policy in Mental Health 35, no. 1–2
(March 2008): 21–37, https://doi.org/10.1007/s10488-007-0144-9.
45
 7 STEPS TO ADOPTING AUTOMATION AND AI FOR NAVAL ISR

There are three major categories when applying the S-curve of technological
adoption: (1) Construct: an idea or theory containing subjective and conceptual elements,
(2) Pre-adoption phase: awareness of innovation (early innovators and early adopters), and
(3) Adoption phase: adopter’s commitment to the decision. 137

1. Start with a “Campfire Talk”

Beginning with a “campfire talk” can build a holistic way of understanding a

competitor’s vision and where successful emerging technologies are applied along the S-
Curve. American strategist and the University of Virginia Professor Jeanne Liedtka
encourages her readers to have an open mind of what the world might be, specifically a
“‘purposeful space’—virtual rather than physical—in which particular activities,
capabilities, and relationships are encouraged” to produce particular behaviors. This step
is essential in building a process of design. 138 The most important aspect relevant to initial
campfire discussions can fundamentally be getting to know each other, and definition and
perception of automation and AI priorities, and write the tasks that require solutions down.

2. Categorize and Analyze the Competing Naval Leaders of

Other Great Powers

In this step, begin with the vision and strategic narratives for great sea power, and
dive into the organizational and cultural factors. The rationale for choosing this method of
analysis is that those in power make or create change in a complex and adaptive system.
Thus, understanding the leaders will likely contribute to building awareness. By building
awareness, the cognitive mind can recognize the subtle similarities and differences between
competing naval powers.

137 Jennifer P. Wisdom et al., “Innovation Adoption: A Review of Theories and Constructs,”
Administration and Policy in Mental Health 41, no. 4 (2014): 480–502, https://doi.org/10.1007/s10488-
013-0486-4.
138 Jeanne Liedtka, “In Defense of Strategy as Design,” California Management Review 42, no. 3
(Spring 2000): 9, https://doi.org/10.2307/41166040.
46
 3. Understand Each Intelligence Organization’s Technical
Capabilities

An easy approach is to use the SWOT technique, which is understanding Strengths,

Weaknesses, Opportunities, Threats (SWOT). Specifically, this requires selecting a current
situational awareness computer system, its associated radar capabilities, such as Aegis and
Type-346, and the physical naval platform. If there are more weaknesses and threats, then
the program or plan can be re-organized or canceled in order to build strength and create
more opportunities to compete in leading in AI for future sea power.

4. Select and Categorize Top-Performing AI through Social

Network Analysis (SNA)

This step gears toward identifying and leveraging key talent quickly. Key talent
includes available or already-built software applications, hardware, and relevant
researchers and developers on the market. A place to start is with open-source databases
with subject matter experts, intelligence professionals, and operator professionals who
publish his or her work online. For SNA, programming platforms such as RStudio, ORA,
or Gephi can assist human operators in identifying well-networked and hidden talents
worldwide (see the appendix).

5. Combine Relevant and Operator-Oriented and Intelligence-

Oriented Services

This step helps generate or catalyze the flow of adoption from pre-adoption among
the early innovators to adoption phase by promoting awareness to achieve automation and
AI goals within each intelligence category.

6. Review and Re-orient

Review and reorient for feedback, with oneself or with a diverse team, and return
to the specific step to be addressed. Receiving senior level feedback within 10 weeks prior
to quarterly reports is critical to competitive because the awareness and knowledge of
specific automation or AI tool that shows a lack of progress, productivity, or compatibility
degrades the design process in long-term strategy if the problem or set of problems are not
identified throughout the process.
47
 7. Publish and Make the Plan Accessible

Lastly, finalize the plan and make it accessible upon publishing the quarterly
measures of performance and effort from the commands and encourage flexible and
changeable plans.

These are the elements necessary for identifying and detecting where people or
emerging technologies may or may not fall within the five categories of Rogers’ S-Curve
model. Similar to American historian of technology Elting Morison’s research and
reflection process, much of the adoption of automation centers change or resistance to
change, and understanding why changes occur. 139 With the steps in mind, I will use
Morison’s research and reflection process to analyze the combined information and
intelligence.

UNDERSTANDING THE S-CURVE THEORY OF TECHNOLOGICAL

ADOPTION

The history of the S-curve idea began in the 19th century with the motivation of
knowing more about the growth of humans as a logistic function. In 1839, a Belgian
mathematician Pierre-Francois Verhulst, who related this theory to population growth,
introduced the logistic curve, or “S-curve,” theory. Verhulst initially conceived his idea
from the Englishman and political economist Thomas Malthus’ “An Essay on the Principle
of Population” in 1789. In this time, food supply was declining and therefore, feeding the
greater population became difficult.

However, since the 1950s, the growth of automation technologies has not been
surprise. According to Modus the S-Curve, “the rate of growth is proportional to both the
amount of growth already accomplished and the amount of growth remaining to be
accomplished.” 140 Understanding categories of people and their key or hidden talents can
be applied to the S-curve. According to American communication theorist and sociologist
Everett Rogers developed the diffusion of innovation theory in 1962, there are five

139 Morrison, Men, Machines, and Modern Times, 7.

140 Lindgren, “Global Remittances and Space-Based Cryptocurrencies,” 79.

48
categories of the adoption (see Figure 8). 141 The successful adoption of an innovative
technology occurs in a distributed bell-shaped curve, S-shaped. The S-Curve is derived
from the diffusion curve with respect to time and normality. The point of inflection or
saturation to avoid is where the two lines intersect between early majority and late majority.

Figure 8 depicts the rate of technology adoption in the form of a horizontal and
stretched “S” with adoption on the Y-axis and time on the X-axis.

Figure 8. S-curve Model of Diffusion Process 142

The five groups can be shown through the S-curve to understand who, at what
sequence, and at what speed adopts emerging technologies. The first group of innovators
are the first people to adopt an emerging technology without consumers or users testing it
and account for an average of 2.5 percent of innovators. The second cohort is made up of
early adopters and accounts for 13.5 percent. The early adopters are the second wave of
purchasers. The third group is the early majority and account for 34 percent when the
product has become mainstream products by this time. The fourth group are the late

141 Schwieterman and Fischer, “The S-Curve of Technological Adoption.”

142 Source: Schwieterman and Fischer, 3.

49
majority and represent 34 percent and cautious adopters. The fifth group are the laggards
and consist of 16 percent; similar to people who start to use the internet or smart phones
today. These percentages can form a bell curve however, the sum of waves these groups
adopting an emerging technology over time becomes in the shape of an S-curve in terms
of time.

The first group is the incubation period in which a technology has potential for
growth or improvement but has significant problems to overcome before becoming
mainstream. The next stage is where the technology has shown rapid improvement and
cost efficiency. After this stage, the technology matures and shows limitations. Examples
in modern society include Airbnb or Uber services and associated user-friendly AI
applications. Typically, this growth can lead to the stereotypical name of disruptive
technologies.

Rogers’ theory can serve as a pivotal methodology of adoption. It opens a way to

begin designing and strategizing how to adopt innovative technologies. For example, to
stimulate strategy and game theory, there are strategy games such as AlphaGo deep-
learning system and IBM’s DeepMind subsidiary of Google to play the strategy game of
Go or Weiqi (pronounced “way-chee”). In 2016, the computer system AlphaGo defeated
South Korean Grandmaster Lee Sedol, which demonstrated an innovative technology that
can beat a human being at a strategy game. 143 This directly relates to a computer system
that beat a human being with the strategic advantage of AI having both different strategic
options and speed.

MILLENNIALS, GEN-Z, AND ZENIALS

In the context of future sea power, the naval leaders in the next 10 to 20 years will
be millennials and generation Z, also called Gen-Z. These are the groups of people who
grew up with automation and AI technologies. Michael Dimock, a Pew Research Center
researcher on U.S. politics, policy, demographics, and social trends, defines millennials as
“anyone born between 1981 and 1996 (ages 23 to 38 in 2019)” and those born after 1997

143 Silver et al., “A General Reinforcement Learning Algorithm.”

50
are Gen-Z. For example, the U.S. intelligence specialists operating and maintaining the
automation tools, ages 18 to 30, grew up with technologies; the tactical operators are also
culturally known as millennials or Gen-Z. Both generations can be the best advocates to
discuss awareness, recognize, and apply technological AI tools for ISR. It is important to
note that early innovators, who can be in any age group, are typically found more frequently
in the younger population because they used emerging technologies as a child and can
identify and reveal inoperable or non-compatible issues quickly.

Commercial and private sectors recognize automation and AI as engineering

problems. Former CEO of Google and board member of the DOD Advisory board Erich
Schmidt believe the problems are straightforward engineering problems that need to be
“done right.” He believes the biggest issue remains with multiple systems not being
compatible with one another. One of the solutions he proposes is to build reliable systems
in which we understand failure modes and error rates. 144 For example in the Navy,
receiving, analyzing, and disseminating data via Aegis and multiple other computer
systems inevitably creates a recognized frustration among the younger operators. Given
the Navy’s antiquated system, this reveals a weakness in the adopting automation because
envision computer systems and AI applications, like a high-resolution video game or Apple
product; the adoption process here is a little behind “the curve.”

ALTERNATIVE OUTLOOK TO ROGERS’ S-CURVE

Lessons learned from the past can be applied to the S-Curve model to show how
and how not to adopt automation for naval ISR. As an alternative method, I will also use
quantitative measures to show relational, spatial, and temporal network analysis of relevant
AI researchers who are researching AI. Data will be from open-source and public databases
such as Stanford University’s arXiv database and Clarivate Analytics’ Web of Science
database. 145 This will contribute to purely the awareness automation and AI research today
to stay ahead of the curve of adoption.

144 Scharre, Cho, and Schmidt, “Eric Schmidt Keynote Address.”

145 “Web of Science,” Web of Science Group, 2020, https://clarivate.com/webofsciencegroup/
solutions/web-of-science/.
51
THIS PAGE INTENTIONALLY LEFT BLANK

52
 IV. AI SUPERPOWER NAVIES ADOPTING AUTOMATION
TODAY

The ultimate impact will depend not just on how the principles actually
shape DOD investments in AI, but in how the broader AI community
perceives DOD actions. 146

—Professor Michael C. Horowitz (University of Pennsylvania)

Great powers with a grand narrative about global affairs often have a gravitational
pull. For example, after World War II, the U.S. government put forth all efforts to contain the
Soviet Union. In the aftermath of 9/11, the Global War on Terror (GWOT) became a grand
narrative. More recently, the U.S. and PRC have become driven to some extent by efforts to
outpace or leapfrog over the other in the AI realm. We find ourselves—as shown from both
governments’ national strategy documents, media coverage, think tanks, universities, public
and private companies—in an era of “Great Power Competition” and the narrative today
between the great powers of the United States and China increasingly relates to the arms race
associated with the military application of Artificial Intelligence.

OVERVIEW

This chapter will survey how the United States and China, as two AI superpower
navies, are adopting automation and AI tools and where their naval ISR falls along the Rogers’
S-Curve of technological adoption. Specifically, it attempts to answer the following two main
questions: How are the two AI superpower navies adopting automation for naval ISR in the
domains of space, cyberspace, air, sea, and undersea? The purpose of this chapter is to develop
a strategy and design method to adopt AI for naval ISR to stay on a healthy and competitive
path for innovation.

This chapter will begin by discussing AI and automation today for both USN and
PLAN through the step-by-step process developed from chapter 3. In this qualitative research
methodology, there will be three main parts: First, I will discuss the various intelligence

146 Patrick Tucker, “Pentagon to Adopt Detailed Principles for Using AI,” Defense One, February 18,
2020, https://www.defenseone.com/technology/2020/02/pentagon-adopt-detailed-principles-using-ai/
163185/.
53
domains within which automation and AI are used for ISR: IMINT, SIGINT, OSINT,
MASINT, HUMINT. Second, I will discuss the narrative and vision between the two naval
leaderships. Third, I will select examples of automation and AI capabilities from each ISR
domain: cyber, air, land, surface, and undersea.

DISCUSSION OF INTELLIGENCE AND TECHNICAL CAPABILITIES

Automation represents an important emerging military technology and network

infrastructure across the main disciplines of intelligence. There are five necessary ways of
collecting intelligence that fuel automation in naval ISR, called “intelligence collection
disciplines” or the “INTS” (see Figure 9).

Figure 9. Intelligence Collection Capabilities for ISR 147

Open Source information is derived from publicly available information. Human Intelligence
is the collection of information from human sources; Measures and Signals Intelligence is the
collection discipline concerning weapons, overhead and airborne imagery, telemetry, and
electronic intelligence. Signals Intelligence is the collection and interception of electronic
transmission from communications, electronic, and foreign instruments that can be collected
from ISR platforms. Lastly, Imagery Intelligence refers to the collection and analysis of

147 Source: Naval War College Library.

54
imagery, or photos, and geospatial information.148 Each intelligence agency has authorization
and responsibility for each of the intelligence discipline and are key organizations to
understand and consult with to improve collection, analysis and dissemination processes for
ISR activities. Figure 10 depicts examples of manned and unmanned systems necessary for
ISR collection from intelligence community satellites to underwater arrays.

Figure 10. Manned and Unmanned Naval ISR platforms149

148 Naval War College Library, “Intelligence Studies: Types of Intelligence Collection,” Intelligence
Studies Research Guide, March 5, 2020, https://usnwc.libguides.com/c.php?g=494120&p=3381426.
149 U.S. Navy, “US Navy Maritime ISR Roadmap: Industry Day” (2010 Navy Information
Dominance Industry Day, Heritage Conference Center, Chantilly, VA, June 22, 2010),
https://www.afcea.org/mission/intel/documents/IndustryDay-Kraft.pdf.
55
 Active and passive sensors are two types of remote sensing. Active sensing means
that the “sensor emits and subsequently receives energy,” to scan, detect, and then measure
the radiation that is reflected or backscattered from the object.150 Active sensing is analogous
to a cellphone that when turned on and sending text messages reveals the location of the
cellphone. Satellites represent the critical part of the sensing process. When naval ISR
platforms are operating at sea, satellites are the sole means to connect. Two disadvantages of
active sensing, however, include giving away the radar’s location and the possibility of
experiencing radio frequency or light interference.

While active sensing carries these vulnerabilities, passive sensing merely receives,
detects, and tracks information or objects by processing reflections from objects of
illumination in the maritime and space environment, such as communications signals or
broadcasts. Passive remote sensing utilizes third-party transmitters and the time of the signal
sent and its signal reflected off an object occurs in a triangular fashion to locate a target.
Passive systems use an automatic trigonometric function of a triangle that measures the time
difference of arrival or time distance of arrival (TDOA) from the emitted signal and the
receiving signal from the reflection of the object through radar or light.

Sensor systems are: Radar (Radio Detection and Ranging) and LiDAR (Light
Detection and Ranging). Both types can determine the range to a target where the time delay
between emitting a radio wave (the signal) and return (receipt) is measured to establish
kinematics (i.e., time, location, speed, altitude, and direction) of an ISR platform. For
example, an aircraft sends out pulses and then reflects off the earth’s surface, which provides
a position like GPS. LiDAR uses light (visible or infrared), both of which can provide
elevation, range, and bathymetric surfaces.151 LiDAR advanced technology allows for
signals to penetrate water and produce terrain mapping. While the plethora of data collected
for terrain mapping exceeds the cognitive load of a human being, these automation systems
can fuse radar and LiDAR information within a computer to provide the best battlespace
picture at a faster rate for decision-makers.

150 Mark M. Lowenthal and Robert M. Clark, eds., The Five Disciplines of Intelligence Collection
(Thousand Oaks, CA: CQ Press, 2015), 130.
151 Lowenthal and Clark, 130.

56
 VISION AND STRATEGIC NARRATIVES: CATEGORIZE AND ANALYZE
COMPETING NAVAL LEADERS

Western leadership has always had a keen eye and thirst to know and understand the
adversary. Abraham Lincoln once said, “I do not like that man. I must get to know him better.
I do not like that person, I must get to know them better.”152 While philosophers of military
affairs like Alfred Thayer Mahan and Carl von Clausewitz were imbued with Western
political thought, which drives the U.S. Naval leadership today, Sun Tzu’s idea imbues with
the ethos to translate Mao Tse-tung and the People’s war content to the sea. This is important
to understand in order to avoid “mirror imaging” sidestepping stereotypical assumptions about
Chinese naval leadership and how they would conduct operations compared with the
American way of naval warfare. This is a profound aspect of the art of war in the pursuit of
knowledge: to know and understand the adversary; and, noticing “pattern of life” at sea and
any divergences from the normal naval pattern of life.

1. U.S. Naval Leaders Pursue Automation and AI

To pursue the adoption of AI for naval ISR, policy, and doctrine are the first steps in
implementing changes. Policy plays a key role in improving naval ISR. Prussian General and
military theorist Carl von Clausewitz, in On War, stated that “war is nothing but the
continuation of policy with other means.”153 United States Naval policy on ISR is
promulgated in the 2017 Joint Publication 2-01.154 Knowing the adversary or competitor
requires the collection of information by ISR platforms and knowledge of what lies on the
highways, or sea lanes. Naval ISR is an instrument to be used in times of war and peacetime
where the sea is a highway, as Alfred Thayer Mahan put is, “over which men may pass in all
directions…familiar and unfamiliar dangers of the sea, both travel and traffic by water” is
easier and cheaper.155

152 Coenn, Abraham Lincoln.

153 Clausewitz, On War, 69.
154 See Joint Chiefs of Staff, Joint Intelligence Support to Military Operations, Appendix B-1.
155 Till, Seapower, 1.

57
 CNO Mike Gilday and his predecessor Adm. John Richardson prioritized designs for
maintaining maritime superiority and strategic direction toward GPC with “cyber sentries.”
Then-Acting Secretary of the Navy Mr. Thomas Modly stated in his memorandum SECNAV
Vector 11, “Take seriously your own role as a guardian of the digital information…Everyone
in the DON enterprise must become a Cyber Sentry.” 156 Mr. Modly believed the Navy
personnel should guard their digital information and to do so as a sentry on guard, or as naval
personal call it “on watch,” but through computers. He sent a strong message that contributed
to the notion that the future of naval affairs will be increasingly automated and to prepare in
the meantime to become a guardian of cyber as if it were a physical post.

2. PRC Naval Leaders Pursue Automation and AI

The PLAN represents the party’s naval force while being useful as an element of spy-
craft. In fact, the PLAN politicizes the military because the party controls the military. 157 It
is highly politicized because the party strictly controls the military. In the chapter,
“Employment of Secret Agents,” in Sun Tzu’s book, Art of War, he states, “Hence it is only
the enlightened ruler and the wise general who will use the highest intelligence of the army
for purposes of spying and thereby they achieve great results. “Spies are a most important
element in water because on them depends an army’s ability to move.”158 When necessary,
Chinese naval leadership can use ISR as spies in support of maneuvering naval forces while
building and modernizing its forces.

3. Chinese Naval Modernization: String of Pearls and the New

Maritime Silk Road

While the United States has continually modernized its navy over the last 200 years,
China is only now beginning seriously to modernize its naval power. In 2004, defense
contractor Booz Allen Hamilton (BAH) coined the term “String of Pearls” in a report to the

156 Thomas B. Modley, “SECNAV Vector 11,” official memorandum (Washington, DC: Secretary of
the Navy, February 14, 2020).
157 McGregor, “5 Myths about the Chinese Communist Party.”
158 Sun Tzu, Art of War, trans. Samuel B. Griffith (Oxford, UK: Clarendon Press, 1964), 149,
https://suntzusaid.com/book/3/18.
58
Office of Net Assessment for the DOD to describe China’s long-term strategy to build
partnerships from the Middle East to the South China Sea. China’s new maritime silk road
plan increases the PLAN’s capabilities to improve naval ISR through a “String-of-Pearls”
strategy. 159 Since 2008 PLAN has developed its antipiracy operations and exercises in the
Gulf of Aden—its first naval operation beyond China’s periphery to expand its interests in the
far seas.160 Despite China’s lack of historical naval development, they are rapidly catching
up to the U.S. Navy maritime superiority on the global stage.

PLAN leaders were once open to a division of labor across the Pacific to avoid future
miscalculations or incidents at sea. Since 2006, then-Commander of the PLAN Vadm. Wu
Shengli called for safeguarding and securing China’s maritime rights, interests—near and
abroad—and international sea lines of communications (SLOCs). Consequently, Admiral Wu
led a major shift in naval operations to extend the PLAN navy beyond the littorals in the
Pacific and the Indian Ocean. In May 2007, he proposed to then-Commander of PACOM
Adm Timothy Keating, a division of labor where PLAN would defend the waters west of
Hawaii and U.S. Pacific Fleet would protect the Pacific east of Hawaii.161 Chinese naval
leaders desired to divide the Pacific Ocean and defending it between PLAN and USN naval
assets. The PLAN’s leadership aspires to achieve great results through a layered defense using
AI and ISR platforms to increase awareness of the maritime battlespace.

Although “dividing and conquering” seemed fit for defending international waters
together, this was not the priority of the U.S. CNO. On July 9, 2019, then-CNO Richardson
and current PLAN Commander Vadm. Shen Jinlong agreed on “reducing strategic,
operational and tactical risk between the two navies to minimize the possibility of any

159 Jessica Drun, “China’s Maritime Ambitions: A Sinister String of Pearls or a Benevolent Silk Road
(or Both)?,” Center for Advanced China Research Blog (blog), December 5, 2017,
https://www.ccpwatch.org/single-post/2017/12/05/China%E2%80%99s-Maritime-Ambitions-a-Sinister-
String-of-Pearls-or-a-Benevolent-Silk-Road-or-Both.
160 Drun.
161 S. Mahmud Ali, China’s Belt and Road Vision: Geoeconomics and Geopolitics (Cham,
Switzerland: Springer International Publishing, 2020), 275.
59
misunderstanding or miscalculation” over video teleconference.162 This dialogue indicates
that the U.S. priority is safety and risk reduction and while dividing the labor of naval defense
across the Pacific might be a mechanism to keep safety, it was not implemented by PLAN or
the USN. Automating the amount of information flow and adopting AI analytical tools for
relevant and accurate results can help shape and attain the goals that both navies are competing
for.

PLAN ISR AND AI 2020: EXAMPLES OF ADOPTING AUTOMATION

AND AI

This section explores a collection of automation and AI application examples related

to naval ISR used by the PLAN. This exploration and analysis will encompass Over-the-
Horizon (OTH) radar, Global Positioning System (GPS) and satellites used for receiving and
analyzing radar signals, the Dongdiao-class spy ship, autonomous undersea vessels (AUVs)
for anti-submarine warfare, the advantageous use of social media and hacking, and undersea
sonar “listening” devices. Part of the analysis will describe where the PLAN situates along
the S-curve of technological adoption.

1. PLAN’s AI Strategy and Development Plans

The PRC and PLAN leadership are aggressively AI. On October 23, 2018, at the
Beijing Xiangshan Forum, senior executive of NORINCO—China Ordnance Industries
Group Corporation, China’s third-largest defense company—Zeng Yi said in a speech that
there will be no people fighting in the future battlegrounds and by 2025, lethal autonomous
weapons would be a commonplace because of the increasing use of AI. 163 In 2016, Google’s
deep-learning system AlphaGo defeated the South Korean weiqi world champion Lee Sedol.
After its victory, the program was re-named AlphaZero, which was a “break-through”
technology and surpassed the inflection point of the S-curve of technological adoption

162 Chief of Naval Operations Public Affairs, “CNO Holds Video Teleconference with Chinese
Counterpart Vice Adm. Shen Jinlong,” Navy News Service, July 9, 2019, https://www.navy.mil/submit/
display.asp?story_id=110159.
163 Gregory C. Allen, Understanding China’s AI Strategy: Clues to Chinese Strategic Thinking on
Artificial Intelligence and National Security (Washington, DC: Center for New American Security, 2019),
https://www.cnas.org/publications/reports/understanding-chinas-ai-strategy.
60
 because it incorporates analysis of past recorded games and data from the computer playing
itself.164 This is an exemplary technology that became even more attractive to the military to
employ both deep learning and neural networks as automation tools for naval ISR.

The major plans and timelines illuminate Chinese organizational structure and with
Three- or Five-Year Plans. In Mao Tse-tung ‘s “Problems of Strategy in China’s
Revolutionary War,” the first chapter revolves around the laws of war being developmental
at all levels in conjunction with China’s Five-Year Plan.165 The latest version of this plan
specifies a major scientific and technological infrastructure construction program through
2020 (see Table 5). Research for defense and strategic emerging technologies will likely
continue to increase precision and range for Dragon Eye (Type-346) phased-array radar and
its associated 3D air search radar “Top Plate” and therefore, expand the capabilities of the air
and surface range for identification on future “mobile airbase.”166

Table 5. 13th Year Plan: PLA’s Defense, Dual-Use Plans, Science and
Technology Strategies 167

Name (English) Name (Chinese) Issue Date Sector/Area Time Responsible

Span Agencies
836 High-Technology 国家高技术研究发 March 1986 Strategic 1986‒ MOST,
Research and Development technologies present COSTIND/
Plan (863 Plan) Civilian 展计 划(863 计划) SASTIND, PLA
Component GAD
New High-Technology 新型高科技武器计 1999 Defense 1999- Central Military
Weapons Plan (995 Plan) technology present Commission
划 (995 计划) (CMC),
PLA GAD
2006-2020 Medium- and 国防科技工业中长 May 2006 Defense 2006- COSTIND/
Long-Term Defense 期科学和技术发展 Research & 2020 SASTIND
Science and Technology 规划纲要(2006‒ Development
Development Plan (MLDP) 2020)
Civilian Explosives Industry 民用爆炸物品行业 October Explosives 2016- MIIT
Development Plan (2016– 发展规划（2016－ 2016 2020
2020) 2020年）

164 John Arquilla and Peter Denning, “Automation Will Change Sea Power,” Naval Institute
Proceedings 145, no. 6 (June 2019): 34.
165 Tse-tung, Problems of Strategy in China’s Revolutionary War.
166 Kirchberger, Assessing China’s Naval Power, 211.
167 Koleski, The 13th Five-Year Plan, 41–43.

61
 With support from President Xi Jinping, the Belt and Road Initiative (BRI) leverages
global support and thus increases the need for ISR at sea (Figure 11). In 2013, BRI was
designed for China to partner with Central Asia, the Middle East, Africa, and Europe to build
and rent out ports by building infrastructure in over 30 countries (i.e., rail lines, roads, bridges,
oil and gas pipelines, and port facilities). 168 Both the government and private investors, such
as Silk Road Fund and the Asia Infrastructure Investment Bank (AIIB), fund the Chinese
government and its BRI at an estimated $200 billion, which is projected to reach $1.2 to $1.3
trillion by 2027, according to Morgan Stanley. 169 The goal is to gain worldwide public
support and acceptance of China’s maritime plans in exchange for building infrastructure in
over 30 countries and use automated systems to communicate.

168 The Belt and Road Initiative is the latest term in a series of terms to describe China’s global
infrastructure building and trade initiative. Other terms used are One Belt, One Road (OBOR), the
Maritime Silk Road, the Economic Belt, the Silk Road Economic Belt, and the 21st Century Maritime Silk
Road. While this list is not all-encompassing, these terms are often used interchangeably and are used to
describe the same phenomena. The term Belt and Road Initiative is the most recent term used by China and
will be used throughout this thesis
169 Andrew Chatzky and James McBride, “China’s Massive Belt and Road Initiative,” Council on
Foreign Relations, January 28, 2020, https://www.cfr.org/backgrounder/chinas-massive-belt-and-road-
initiative.
62
 Figure 11. China’s New Silk Road Called “String of Pearls” to Build and
Expand Ports and Industrial Parks Across Southeast Asia,
Including Sri Lanka, Kenya and Greece 170

China’s Anti-Access/Area-Denial (A2/AD) doctrine relies upon strategic

surveillance provided through a three-layer defense strategy designed to control maritime
approaches (Figure 12).

170 Source: Jeremy Page, “China Sees Itself at Center of New Asian Order: Beijing Builds Roads,
Pipelines, Railways and Ports to Bind Itself to Region,” Wall Street Journal, November 9, 2014,
https://www.wsj.com/articles/chinas-new-trade-routes-center-it-on-geopolitical-map-1415559290.
63
 Figure 12. China’s Anti-Access Area Denial Defensive Layers 171

The first, outer-most, layer includes anti-ship ballistic missiles, diesel, or air
independent propulsion (AIP) submarines, and Over-the-Horizon Radars (OTHR). 172 The
second layer consists of long-and mid-range aircraft (with or without weapons) and
submarines for strategic messaging, which can operate beyond China’s periphery. The
third, inner-most layer, is closer to the littoral region and combines all ISR navy, air, land,
and sea platforms. The geographical constraint from neighboring countries represents the
main disadvantage for China, specifically, Taiwan, the Ryukyu (Japanese) archipelago,
northern Philippines, and Borneo. 173

171 “What Is Anti-Access/Area Denial (A2/AD)?,” Missile Defense Advocacy Alliance, accessed
February 27, 2020, https://missiledefenseadvocacy.org/missile-threat-and-proliferation/todays-missile-
threat/china-anti-access-area-denial-coming-soon/.
172 Anti-Access/area denial (A2/AD) strategic concepts were first used by the U.S. DOD 2001
Quadrennial Defense Review (QDR), and in 2018 was replaced with the National Defense Strategy (NDS).
See Office of the Secretary of Defense, Quadrennial Defense Review, (Washington, DC: U.S. Department
of Defense, 2010), 31–33.
173 China’s three-layer defense strategy caused U.S. strategy and policy to develop the 2010
Quadrennial Defense Review report as a national defense response.
64
 2. PLAN Uses Automation to Test Its First Laser System Against
U.S. P-8 Poseidon

PLAN strategies above exemplify its mission and capability to adopt automation
sensors for situational awareness. On February 17, 2020, PLAN Luyang III-class destroyer
(Type 052D) reportedly lased a U.S. Navy P-8 Poseidon maritime patrol ISR aircraft
assigned to the VP-45 squadron about 329 nautical miles west of Guam in the Philippine
Sea. This was the first reported instance of a Dragon-Eye radar tested against its main
competitor, the U.S. Navy. This radar technology aligns with its China’s 995 Plan and
potential future execution of laser and explosive technologies, which naval ISR automation
can identify, detect, and counter-lase faster than a human operator.

3. Modernizing PLAN Aircraft Carrier

The modernization of naval China’s forces, including aircraft carriers, will continue
automating naval warfare. In 2012 at the 18th Party Congress, then-President Hu Jintao
called for China to become a “maritime power.” 174 In April 2018, President Xi-Jinping
echoed this vision and stated that “the task of building a powerful navy has never been as
urgent as it is today.” 175 China’s 2019 Defense White Paper states:

The PLA Navy (PLAN) has extended training to the far seas and deployed
the aircraft carrier task group for its first far seas combat exercise in the
West Pacific. It has organized naval parades in the South China Sea and the
waters and airspace near Qingdao, and conducted a series of live force-on-
force exercises codenamed Mobility and systematic all-elements
exercises. 176

PLAN’s modernization and extended sea training, live-fire exercises, and red-
versus-blue exercises with its aircraft carrier task group indicate great improvements in its
automation capabilities. In unit level execution, the Luyang II–class (Type 052C) destroyer
Haikou engaged a floating target with surface guns by using both radar and optical

174 China Power Team, “How Is China Modernizing Its Navy?”

175 China Power Team.
176 Cordesman, China’s New 2019 Defense White Paper, 3.

65
guidance. 177 According to Zhang Junshe, a senior researcher at the PLA Naval Military
Studies Research Institute, Liaoning carrier operations “fired more than 10 air-to-air, anti-
ship and air defense missiles” and used air defense radar to hit targets at sea with both
Chinese destroyers, frigates, one replenishment ship, and one attack submarine in the Bohai
Gulf, which is west of the Korean peninsula. 178 This trend indicates that the PLAN will
likely continue to mimic the U.S. Navy’s actions in the far seas. This form of mimicking,
or “mirror imaging,” is a type of cognitive trap the PLAN could be exemplifying. Mirror
imaging is when human beings consciously or unconsciously assume the competitor thinks
or acts like us because of how difficult it is to imagine someone else’s perception being
different from our own perception. 179

4. China’s Satellites and Technology Giants

Since the 1996 Taiwan Strait Crisis, when Chinese missile tests, intended to
intimidate, were suspected of failing in flight, much has changed. 180 For example, Beidou,
a navigational satellite system, is a centrally controlled satellite constellation that provides
all-weather, accurate positioning, navigation and timing information, and leveraged by
PLAN forces. The Beidou satellite constellation—manufactured by China Academy of
Space Technology (CAST) and operated (China National Space Administration) CNSA—
serves as an alternative to U.S. GPS satellites. It is interoperable with existing global
navigation satellite systems (GNSS) and increases accuracy to precisely determine
position, and includes 35 satellites as of 2020. 181 However, the China Academy of Space
Technology CAST and operated by CNSA—which works across AI, surveillance,
telecommunications, satellites, and the internet—discloses little open-source information

177 Dale C. Rielage, “Chinese Navy Trains and Takes Risks,” Naval Institute Proceedings 142, no. 5
(May 2016), https://www.usni.org/magazines/proceedings/2016/may/chinese-navy-trains-and-takes-risks.
178 Zhao Lei, “CNS Liaoning Leads Live-Fire Drill,” China Daily, December 16, 2016,
https://www.chinadaily.com.cn/china/2016-12/16/content_27685144.htm.
179 Zachary Shore, Blunder: Why Smart People Make Bad Decisions (New York: Bloomsbury USA,
2015), 162.
180 U.S. GPS interference and the development of Global Positioning System (GPS) has proliferated,
causing possible situations for signal interference to occur at sea or land.
181 “BeiDou-3 Navigation Satellite System,” Aerospace Technology, accessed March 7, 2020,
https://www.aerospace-technology.com/projects/beidou-3-navigation-satellite-system/.
66
regarding policy, data, security, privacy, and censorship. By having radar stations,
receivers, and increasing posts overseas, including in Djibouti and Australia, Beidou
precisely collects and receives ISR data, and thereby provide precise positional data to
PLAN operators at sea.

In an echo of Rogers’ S-Curve, China was a late adopter of the global positioning
system (GPS). Innovative nations who began developing GPS since its birth in 1973
include: U.S. GPS, Japan’s Quasi-Zenith Satellite System (QZSS), the European Union’s
Galileo GNSS, Russia’s Global Navigation Satellite System (GLONASS), and India’s The
Indian Regional Navigation Satellite System (IRNSS), (operationally called NavIC or
Navigation with Indian Constellation). 182 The advantage of later adoption or being a
laggard is observing, collecting, and analyzing the past lessons of other global powers
building their own GPS system and then applying only the most advanced, successful
satellite technical and business practices while avoiding past issues and mistakes other
nations may have endured. 183

Since 2006, Yaogan satellites launched by China have improved naval ISR and
reinforced China’s pursuit of great power status. According to the China Aerospace
Science and Technology Corporation (CASC) and media outlets, on 29 July 2019, at 0347
UTC, China launched three Yaogan-30 (Weixing-30) remote-sensing, possible SIGINT
satellites into Chuangxin-5 (CX-5) constellation using a Long March-2C carrier rocket.
Since 2006, this Earth-observing satellite constellation in Lower Earth Orbit (LEO) can
potentially detect ships via radio emissions and related technical tests. LEO satellites can
triangulate multiple sources to enhance the procedures of automation faster in a near-
continuous manner for naval ISR.

At the operational level, satellite technology boosts naval ISR capabilities to

provide near-continuous ISR collection, exploitation, and analysis of signals that support
the defense-in-depth strategy. At the tactical level of the Chinese satellite systems, the

182 Mahashreveta Choudhary, “What Are Various GNSS Systems?,” Geospatial World (blog),
November 20, 2019, https://www.geospatialworld.net/blogs/what-are-the-various-gnss-systems/.
183 Alexander Gerschenkron, Economic Backwardness in Historical Perspective (Cambridge, MA:
Harvard University Press, 1962).
67
“architecture of 12 satellite constellation suggests that the purpose is to achieve a near-
continuous ELINT surveillance of the regions between 35 degrees North and 35 degrees
South Latitude” to detect and locate incoming adversaries from its littoral waters through
various access routes from the Indian to Pacific Oceans, according to the International
Strategic and Security Studies Programme. 184

5. Automation for Long-Range and Early Warning: Joint Over-

the-Horizon Radar

PLAN integrates long-range surveillance with its over-the-horizon backscatter

(OTH-B) radar to provide long-range, 24/7 persistent coverage (see Figure 13). According
to Janes, an intelligence and consultant company for defense and national security sectors,
this conventional radar coverage extends up to 1,350 to 1,889 nautical miles (2,500 to
3,500km), but is limited by the horizon, or curvature of the earth, and atmospheric
conditions. 185 OTH-B allows both the PLAN and the PLA to collect quality data from
China while also leveraging the data for early indicators of where the U.S. Navy could
potentially maneuver next (i.e., South China Sea, East China Sea, North Sea, or head
straight to the Persian Gulf). OTH-B is an automated tool that represents a strategic
advantage over the U.S. Navy and assisted by reconnaissance satellites.

184 S. Chandrashekar and N. Ramani, China’s Space Power & Military Strategy – the Role of the
Yaogan Satellites, ISSSP Report No. 02–2018 (Bangalore, India: International Strategic and Security
Studies Programme, National Institute of Advanced Studies, 2018), 2, http://isssp.in/wp-content/uploads/
2018/07/Chinas-Space-Policy_July2018.pdf.
185 Andrew Tate, “China Integrates Long-Range Surveillance Capabilities,” Jane’s Intelligence
Review, November 1, 2017, Janes.
68
Figure 13. China’s Long-Range Surveillance Capabilities Possibly Extend to
Guam. Quality or Discrimination of Specific Class, Types, or
Accurate Tracks is Undetermined 186

In favorable atmospheric and other weather conditions, this High Frequency (HF)
signal can possibly detect incoming naval platform as far from the Chinese coast as Guam
and triangulate a possible position using China-owned Beidou satellites to locate, identify
and track maritime or naval platforms with radar technology, as depicted by the large blue
swath (see Figure 13). This automation system, though not intelligent, provides a highly
competitive edge and strategic defense against incoming ballistic missiles.

Employing the S-curve, the PLA and PLAN OTH capabilities falls within the early
adopters group. In 1945, the Harvard economic historian Professor Alexander
Gerschenkron expressed the notion of “economic backwardness.” 187 His notion
exemplifies “the tendency on the part of backward countries to concentrate at a relatively
early point of their industrialization on the promotion of those branches of industrial

186 Source: Tate.

187 Gerschenkron, Economic Backwardness in Historical Perspective.

69
activities” with rapid technical productivity. 188 China’s “backwards strategy” to adopt
specific ISR technologies releases tension from pre-industrialization conditions and waits
for industrialization to overcome existing obstacles or technological challenges within an
already progressing and industrial environment. Other nations that have adopted and
exercise OTH capabilities: U.S. Skywave Radar, Australia’s Jindalee Operational Radar
Network (JORN), France’s New Transhorizon Decametric System Applying Studio
Methods (NOSTRADAMUS), Russia’s Container, and Iran’s Sepehr. 189 Few details are
known of these systems; however, the transmission of these radars can cause radio
frequency interference.

6. Future of Anti-Submarine and AI

To date, the PLAN has deployed and experimented with autonomous surface
vessels and unmanned underwater vehicles. Notably, the Sea Wing, or Haiyi, glider,
developed by the Chinese Academy of Sciences (CAS) Shenyang Institute for Automation,
has high potential to be used as a stealthy underwater spy drone to scour the seas within
the First Island Chain at a depth of 3.2 nautical miles (6 km). The Sea Wing is a buoyancy-
driven autonomous underwater vehicle (AUV). 190 Sea Wing is advantageous because of
its low acoustic signature and its ability to traverse large deep blue water with minimal
energy to depths of 11,034 meters along the Marian Trench. 191 Sea Wing has been known
to operate in the South China Sea and the Indian Ocean to support special operations by
detecting submarines and gathering both intelligence and bathymetric data for future

188 Maurice Dobb, “Economic Backwardness in Historical Perspective: A Book of Essays. by A.

Gerschenkron,” Economic Journal 74, no. 296 (December 1964): 1001–1003, https://doi.org/10.2307/
2228863.
189 Engineers Australia, Jindalee: Over-the-Horizon Radar (Barton ACT: Engineers Australia
Engineering Heritage Victoria, 2016), 64–70, https://portal.engineersaustralia.org.au/system/files/
engineering-heritage-australia/nomination-title/Jindalee.Nomination.V14.June%202016.pdf.
190 Jamie Seidel, “Beijing Releases 12 ‘Haiya’ Sea-Wing Surveillance Drones into the Disputed South
China Sea,” News Corp Australia Network, July 28, 2017, https://www.news.com.au/world/beijing-
releases-12-haiya-seawing-surveillance-drones-into-the-disputed-south-china-sea/news-story/
16bd0faf5195f31fdc7c3afeb5b826c2.
191 Stephen Chen, “PLA Navy Eyes China’s Deep-Sea Underwater Glider After Successful Test
Shows It Rivals U.S. Vessel,” South China Morning Post, September 1, 2016, https://www.scmp.com/
news/china/article/2012137/pla-navy-eyes-chinas-deep-sea-underwater-glider-after-successful-test.
70
operations. 192 The risk of deploying AUVs occurs if a competitor seizes the drone in
territorial or perhaps international waters, as in mid-December 2016 when China seized an
America underwater drone. 193 PLAN has the advantage of collaborating with Chinese
universities to develop and invest in automation for undersea warfare.

In the polar regions, the PLAN has taken advantage of AI as a more viable option
for operating in harsh and sub-freezing conditions. In November 2017, PLAN deployed
the Snow Dragon in company with the unmanned M80B seabed boat in an Antarctic
expedition and through Ocean Alpha Co. PLAN’s adoption of AI and advanced algorithms
in Sea Wing and Snow Dragon provides the strategic advantage in two distinct and opposite
environments for remote sensing. AI ensures accuracy in the undersea vessels sailing route
with GPS mapping to avoid obstacles. In freezing environments, M80B does not freeze
until it is at -35℃ and can withstand this harsh condition.

PLAN’s growing fleet with unmanned and autonomous vessels and swarm
technologies represents current attempts to augment maritime swarm tactics both undersea
and in the polar regions. For instance, PLAN attempts to aid in marine swarm tactics
through a project for naval ISR called the 912 Project. In 2018, 56 miniature unmanned
boats released a video demonstrating coordinated boat swarm tactics off the coast of the
Wanshan Archipelago in the South China Sea. 194 These special operations swarm tactics
and undersea ISR platforms are in line with the 912 Project, a classified project to develop
military underwater robots developed by Shenyang Institute of Automation of the Chinese
Academy of Sciences, which coincides with China’s hundred-year anniversary of the
Chinese Communist Party in 2021. From an “undersea great wall” in the polar regions to
the Mariana-Trench, the 912 Project improves maritime domain awareness in an innovative
way for swarming technologies.

192 Elsa B. Kania, “The AI Titans’ Security Dilemmas,” Governance in an Emerging New World, no.
218 (October 29, 2018), https://www.hoover.org/research/ai-titans.
193 Lyle J. Goldstein, “Robot War in the South China Sea?,” National Interest, October 11, 2018,
https://nationalinterest.org/feature/robot-war-south-china-sea-33146.
194 Phillip Smith, “Oceanalpha Debuts Maritime Drone Swarm,” Drone Below, June 6, 2018,
https://dronebelow.com/2018/06/06/oceanalphas-debuts-maritime-drone-swarm/.
71
 7. China’s Observer Corps and AI Stalkers

PLAN utilizes its special surveillance vessels like Dongdiao and Dadie Auxiliary
General Intelligence (AGI) ships and Chinese Maritime Militia (CMM) as automated
observers for early warning. The PLAN intelligence ship has the electronic capability to
use AI to identify and locate maritime vessels or aircraft. The data-absorbing machine,
PLAN’s Dongdiao-class AGI (Type 815G) spy surface ship, is purpose-built to intercept
electronic and communications signals. 195 In May 2018, the Chinese navy, and specifically
intelligence gathering ships like Dongdiao, were disinvited from participating in Rim of
the Pacific (RIMPAC) exercises, in response to China weaponizing the SCS. 196 Joint naval
exercises represent the optimal and complex tactical intelligence event for the PLAN to
observe, gather data electronically, and then use AI to analyze other navies in action.

As later adopters, PLAN has taken advantage of fusing the strategic design of an
observer corps with automation and AI at sea with automation systems onboard ISR
platforms to give some degree of advanced warning of adversary approach or attack.
During WWII, the Royal Observer Corps, traditionally the eyes and ears of the RAF—a
powerful partner to the Chain Home radar installation along the British coast—were
designed to listen to the sound of incoming German aircraft using parabolic concrete sound
mirrors and make visual sightings, both of which were then transmitted from Observer
Posts to Group and Sector controls. 197 PLAN has transcended this idea of an augmented
observer complex system for early warning by utilizing CMM as possible observers, South
China Sea islands as observer posts, and people using social media to exploit the possible
location of naval vessels at sea.

195 In the July 2019, the Chinese surveillance ship reportedly monitored the international and joint
Talisman Saber naval war games, which involved American, Australian, and Japanese naval forces.
196 Ryan Pickrell, “A Chinese Warship Is Believed to Have Been Sent to Spy on War Games between
the U.S., Australia, and Japan,” Business Insider, July 7, 2019, https://www.businessinsider.com/australia-
tracks-chinese-spy-ship-watching-talisman-saber-war-games-2019-7.
197 Nick McCamley, Cold War Secret Nuclear Bunkers: The Passive Defence of the Western World
During the Cold War (Barnsley, South Yorkshire: Pen and Sword Military, 2007), 123.
72
 8. Chinese Maritime Militia (CMM) “Little Blue Defenders”

China uses CMM vessels to develop its great maritime power. As an external force
and security force, CMM can be leveraged as a political tool to enforce disputed maritime
claims of its islands, rapid responders toward contingencies at sea, and normalizing
administrative control of the seas along the nine-dash line. 198 CMM reconnaissance
detachments can serve as early warning networks with 32 “mother ships” acting as third-
party nodes of surveillance in both low-intensity and high-intensity conflicts. The “Little
Blue Defenders” help China dominate in AI with its overt ability to collect data and
information around disputed islands (i.e., Spratly Islands, Scarborough Reef, Paracel
Islands, Senkaku Islands) and provide the data to the PLAN.

Due to the multi-role and nature of its special mission, the use of CMM falls within
the category of later adopters. CMM serves as a strategic asset in peacetime and to preserve
maritime rights and protect local fishermen. In a high-intensity conflict, CMM can support
larger PLAN forces with mine laying, fuel and food replenishment, deceptive
transportation of troops and ammunition, concealment, sabotage, etc. “Little Blue
Defenders” strongly improve ISR coverage, potential gaps, and can serve as a targeting
asset to report signals or electronic early warning of incoming competitors or contacts of
interest at sea. The concept of using non-naval warships as early warning assets was also
employed by Russian fishing trawlers during the Cold War. CMM is not innovative, but
the PLAN has a unique advantage to easily operate in the littorals, along one coastline.

9. Social Media and Open Source

In social media and unclassified sources, PLAN servicemember using social media
platforms is now subject to AI and automated bots monitoring their text messaging, social
media posts, or geo-tracking their military movements. According to a consultant at
Intelligent Biology and scholar at Georgetown University Medical Center Nicholas Write,
“in order to prevent the system from making negative predictions, many people will begin
to mimic the behaviors of a ‘responsible member of society’ to improve social control and

198 Dennis Blasko, The Chinese Army Today, 2nd ed. (Thousand Oaks, CA: Routledge, 2012), 23.

73
change the way they think.” This massive surveillance AI tool can be used to exploit the
location or the deployment cycle of naval operations when PLAN servicemembers are
posting or geotagging on Facebook, Instagram, or Twitter. 199

China can track U.S. naval forces through open-source sites that collect AIS data
or filtering through Facebook and Instagram posts of sailors preparing to deploy. For
example, in Marine Traffic, a public website, one can type and search “US Gov” or “US
GOV VLS.” China can leverage AI tools like Marine Traffic to locate U.S. naval surface
vessel its Automatic Identification System (AIS) locators are turned on. 200 China can
leverage this data by finding the unique maritime mobile service identity (MMSI) and
Google search MMSI 303891000.

10. PLAN Cyberspace Copy-Cats

China’s role in cyberspace espionage occurs through state-sponsored Advanced

Persistent Threat (APT) groups. 201 APT groups focus on and are designed to steal data,
disrupt operations, destroy infrastructure while collecting and prioritizing relevant
information. APT1 cyber activities stem from the 2nd Bureau of the PLA General Staff
Department (GSD) 3rd Department, also known by its Military Unit Cover Designator as
Unit 61398. APT40, code name Periscope, targets countries of great strategic importance
to China, like the United States. China dedicates about $14.9 billion (RMB 100 billion) to
the Internet Investment Fund for Chinese Internet of Things and high-speed fiber optic
cables, according to China’s 13th Five-Year plan in the USCC annual report. 202 The
evidence for IP theft and economic investment suggests that China is aggressive in its use

199 Nicholas Wright, “How Artificial Intelligence Will Reshape the Global Orde,” Foreign Affairs,
July 10, 2018, https://www.foreignaffairs.com/articles/world/2018-07-10/how-artificial-intelligence-will-
reshape-global-order.
200 “Marine Traffic Live Map,” Marine Traffic, accessed March 5, 2020,
https://www.marinetraffic.com.
201 FireEye Mandiant, M-Trends 2019 (Milpitas, CA: FireEye Inc., 2019), https://content.fireeye.com/
m-trends/rpt-m-trends-2019.
202 Katherine Koleski, The 13th Five-Year Plan (Washington, DC: U.S.-China Economic and Security
Review Commission, 2017), 13, https://www.uscc.gov/sites/default/files/Research/The%2013th%20Five-
Year%20Plan_Final_2.14.17_Updated%20%28002%29.pdf.
74
of cyber espionage. 203 On the other hand, the U.S. 2018 Cyber Security Strategy
emphasizes that China is “persistently exfiltrating sensitive information from U.S. public
and private sector institutions” and has responded with adopting automation and AI for
naval ISR to outpace China and PLAN’s AI strategy. 204

U.S. NAVY ISR 2020: ADOPTING AUTOMATION AND AI

Despite the PLAN’s efforts to modernize rapidly, the United States continues to be
the world’s foremost naval force. American aircraft carriers represent “national command
authority and warfighting Commanders-in-Chief with a flexible force to respond to a wide
variety of international challenges…four and one-half acres of sovereign-and mobile-
American territory that can project U.S. power whatever it might be required.” 205 The
aircraft carrier constitutes a small and mobile American airport to sail and defend the seven
seas and when deployed, the U.S. carrier strike groups can maintain its networks and
rapidly respond to a crisis or in defense of international shipping with its naval network
infrastructure. All of which makes an aircraft carrier an appealing target.

The U.S. Navy has continued to advance its network infrastructure within carrier
strike groups and in a joint environment. In 2016, then-CNO Adm. Richardson proposed
the Joint Concept for Access and Maneuver in the Global Commons (JAM-GC) to
strategize and design countermeasures against PLA’s defense-in-depth threat. 206 The
concept of network-centric warfare links today’s network systems in local and wide-area
networks. Networks that link SSDS as a whole have three main components: Cooperative
Engagement Capability (CEC), which is a virtual “spiderweb” of various networks to
connect and share data; Navy and Marine Corps Intranet (NMCI), which is a localized
ship-centric, encrypted network; and Link-16, which connects all of the systems.

203 FireEye Mandiant, M-Trends 2019, 3–20.

204 Donald J. Trump, National Cyber Strategy of the United States (Washington, DC: White House,
2018), 2, https://www.whitehouse.gov/wp-content/uploads/2018/09/National-Cyber-Strategy.pdf.
205 Federation of American Scientists, “The Aircraft Carrier Plan.”
206 Stephen Daggett, Quadrennial Defense Review 2010: Overview and Implications for National
Security Planning, CRS Report No. R41250 (Washington, DC: Congressional Research Service, 2010),
https://fas.org/sgp/crs/natsec/R41250.pdf.
75
 While CEC is not specifically an ISR program, this is the vein of the naval anatomy
and network-centric body that allows many Navy sensors—from space to undersea
systems—to create an automated picture of the battlespace. The Navy’s CEC has been
successful in interconnecting multiple systems; from USS John F. Kennedy carrier battle
group to today’s Nimitz-class and the upcoming Ford-class carrier strike groups. For
automation and U.S. naval ISR data, the foundational system to pay attention to is the
Distributed Common Ground Station Navy (DCGS-N). DCGS-N combined with CEC,
intranet, and Link-16 forms networks within networks that allow identification,
classification, and targeting.

1. Ford-Class Aircraft Carrier: Aegis to SSDS Computers for AI

In today’s Ford-class aircraft carriers, the U.S. Navy will keep the 50-year legacy
Aegis computer system and advance its automation capabilities with a computer system
called Ship’s Self Defense (SSDS) automation. The U.S. Navy has selected Lockheed
Martin’s Ship’s Self Defense System (SSDS) Mk 2 5 Systems Engineering Agent (CSEA)
today for next-generation Ford-class aircraft carriers. While Lockheed Martin made an
initial bid in August 2017, it did not issue a press release until two years later, in August
2019, which stated, “SSDS is a Combat Direction System that provides capability to defend
against Anti-Ship Missile (ASCM) attacks…developed the Common Source Library
(CSL) which enables efficient deployment of common software solutions across the
Surface Navy, with variation techniques to customize for particular configurations.”207
However, despite this delay from initial bid to press release and launch, both the company
and the DOD needed time and space to develop the product and increase the
interoperability of a diverse list of AI software and applications into one system.

In adopting automation and AI across the Navy, pre-testing, and testing phases are
essential phases to test compatibility between multiple networks and future software
upgrades. For instance, the SSDS computer makes it easier for compatibility. This “plug

207 “Future Warfighting Capabilities Secure with Lockheed Martin Engineering Experience,”
Lockheed Martin Newsroom, August 20, 2019, https://news.lockheedmartin.com/future-warfighting-
capabilities-secure-with-lockheed-martin-engineering-experience.
76
and play” unified software automates within one program rather than multiple systems,
which is like the concept of Microsoft or Apple products being compatible with in-house
products. This is a key and critical computer program to consider adding to the pre-testing
and testing phases when adopting emerging technologies across both naval ISR systems
and branch ISR platforms.

The improved SSDS and HTI systems contribute to the Internet of Things (IoT)
and suggest a continuous increase in automation in naval systems. The Hardware
Technology Insertion (HTI) 16 infrastructure is a new software update to assist in
battlespace awareness. 208 For air contact, the Enterprise Air Surveillance Radars
automates tracking for human operators, according to Jim Sheridan, Vice President of
Naval Combat and Missile Defense Systems at Lockheed Martin. 209 HTI updates for the
Ford-class carrier will also be adopted on several other commissioned ships. SSDS
Advanced Capability Build 20 (ACB 20) will be delivered to aircraft carriers and
amphibious ships: USS George Washington (CVN 73), the amphibious assault ship USS
Boxer (LHD 4) and the amphibious platform dock ships USS San Antonio (LPD 17) and
USS Fort Lauderdale (LPD 28), according to Seapower magazine. 210 Of note, the only
U.S. naval platforms not to receive this update are the Zumwalt-class guided-missile
destroyers. Building and implementing a network infrastructure to encompass both SSDS
and HTI signifies a “leap over” any possible chasm or “fall-off” point from early majority
adopters and getting closer to the inflection point along the S-curve of technological
adoption.

208 Richard R. Burgess, “Four Navy Ships Set for Delivery of Newest SSDS Configuration,”
Seapower, January 14, 2020, https://seapowermagazine.org/four-navy-ships-set-for-delivery-of-newest-
ssds-configuration/.
209 Marty Kauchak, “Naval Channel SNA 2020: ‘A Whole New Heat and Light,’” Mönch Publishing
Group, January 15, 2020, https://monch.com/mpg/news/naval-channel/6606-sna-2020-lockheed-ssds-acb-
20.html.
210 Burgess, “Four Navy Ships Set for Delivery of Newest SSDS Configuration.”

77
 2. Reconnaissance Satellites: ELINT, SIGINT, and Ocean
Surveillance

Reconnaissance satellites, combined with AI that automates detection and analysis

of electronic signals, survey the oceans to identify and locate naval forces at sea at specified
times. The time depends on when the reconnaissance satellites are overhead an area at a
particular time. Satellites are categorized into four main orbits at specific altitudes: Low
Earth Orbit (LEO) (372–621 miles), Medium Earth Orbit (MEO) (6,213—12,427 mi),
High Earth Orbit (HEO) (310–29,825 mi), Geostationary (GEO) (22,236 mi), and
Geosynchronous (GEOS) (22,236 mi). Each constellation has a capability and purpose that
requires knowledge of three key capabilities: bandwidth, throughput capacity, and
overhead or passing over a certain area and how frequently the satellite can pass over a
specific area. Today, online companies develop AI to ingest and analyze data from radars
and reconnaissance satellites to provide answers to these top three categories of required
knowledge. In turn, open-source AI tools online provide a strategic advantage over China.

Human operators can leverage open-source commercial AI websites to possibly

determine the Chinese satellite trajectory and when it could be above a target area. For
example, the Chinese satellite constellation called Yaogan can be used online through the
commercial company LeoLabs. LeoLabs has developed an algorithm for overhead sensing
times and locations using open-source data and three privately owned land radars (Poker
Flat Incoherent Scatter Radar (Alaska, USA), Midland Space Radar (Texas, USA), and
Kiwi Space Radar (New Zealand). 211 Figure 14 shows the ability for anyone to view the
website and view the orbital mechanics, available data, previous passes, next planned
passes, recent state vectors, and other forms of metadata.

211 “Three Ways to Use LeoLabs,” LeoLabs Platform for Operators and Developers, accessed March
8, 2020, https://platform.leolabs.space/.
78
 Figure 14. LeoLabs, Open-Source Private Company Designed to Provide
Applications for Operating in Low Earth Orbit 212

As placed on the S-curve, U.S. private companies fall within the categories of
innovators and early adopters. This AI tool has specific algorithms that provide visual
analytics to show the human operator location and tracking data from open-source
information on Chinese satellites, such as the Yaogan constellation in LEO orbit (Figure
14). The available data includes metadata, particularly a satellite’s name, catalog number,
country, and observed passes. Notably, this is a sophisticated AI tool may be leveraged to
proactively avoid open-ocean surveillance. Part of the adoption S-curve that is critical to
achieving success is investing in a capability that will provide a return on investment; and
thus, Department of the Navy, however, will have to purchase the data before using the
software to use this AI tool.

Among these orbital reconnaissance satellites, to use AI tools, electronic signals

can tell the identity of a platform based on its radar signature. ELINT is a means in which
a machine and another machine sends signals to each other, and electronic signals are a
type of SIGINT. The signal is the propagation of electromagnetic waves at specific
frequencies. Each pulse has a signature or set of characteristics like a fingerprint which
includes pulse width, pulse repetition frequency or pulse repetition interval, scan time, and

212 Source: “LeoLabs Platform for Operators and Developers.”

79
scan type; this automation software tool can potentially identify and determine the
characteristics of the signals. AI can automatically analyze low to high-frequency signals
at a faster rate than human beings. This is important for quick analysis of signals collected
by naval ISR platforms for decision-makers.

Electronic signals from active military radars illustrate how naval forces are
operating. According to the Navy’s Credentialing Opportunities Online (COOL) program,
naval cryptologists analyze radar signals from worldwide technical ELINT data to support
naval ISR and national intelligence priorities. 213 An automation ELINT tool called GALE-
LITE can provide rapid retrieval of data, track animation, histogram, scattergram, and other
analysis tools; “this information is used to design weapons to penetrate the enemy’s
defenses…electronic countermeasures to frustrate the enemy’s weapons” in a vicious
cycle. 214 The ability to gather intelligence via reconnaissance satellites can help better
understand the other navy from short to long-range frequencies and ultimately, for early
warning. Improvement to the adoption of an AI tool such as GALE can be upgrading or
modifying notifications to make it user-friendly.

3. Long Range: Over-the-Horizon Backscatter Radar

Long-range early warning from OTH radar supports naval ISR in a global and
multi-system strategic environment. The first long-range and passive operating system is
the Army-Navy/FPS-118 Over-the-Horizon Backscatter Radar System (OTH-B) from the
1970s. OTH-B is used “for a frequency modulation/continuous wave (FM/CW) radar
capable of detecting and tracking objects at over-the-horizon range.” 215 The purpose is to
provide long-range surveillance of aerial approaches to the United States. Specific
automation and AI software include Kiwi SDR, which can detect OTH radars and then
automatically analyzes the active or passive signals as far as over the Pacific Ocean. Figure

213 Naval Reserve Intelligence Program, “Module 8—Intelligence Automated Data Processing (ADP)
Systems: L. Gale Lite,” in Ready-for-Sea: Modular Course & Handbook (Naval Air Station North Island,
CA: Naval Reserve Intelligence Program, 1999), https://fas.org/irp/doddir/navy/rfs/part08.htm.
214 Barnaby, The Automated Battlefield, 30–31.
215 “AN/FPS-118 Over-The-Horizon-Backscatter (OTH-B) Radar,” FAS Strategic Air Defense
Systems, June 29, 1999, https://fas.org/nuke/guide/usa/airdef/an-fps-118.htm.
80
15 depicts how AI has been adopted for breaking apart each bit of a Chinese OTH signal:
possible frequency, type (naval surface platforms) and duration of the signal.

Figure 15. Kiwi SDR Automation Software Identifies Possible Chinese OTH-
Backscatter Radar Signal Pulses and Analyzes the Frequency 216

Kiwi SDR falls within the category of early innovation of a highly advanced nature.
Building the codes to implement this software coupled with human operators having to
learn how to use and understand this program signifies innovation and early adoption.
Although the entire process of Aegis or SSDS collecting and analyzing data for possible
target acquisition and be complete in a few thousandths of a second and no single human
being, the analysis of long-range detection radars can benefit from AI tools such as the one
used by the company Kiwi SDR. 217 Overall, the software helps human teams better
determine a pattern or possibly detecting an anomaly. The technicalities or reviews on user-
friendliness is unknown about the Kiwi SDR.

4. ISR Platforms and Laser Technology

High-altitude aircraft to aircraft and automated systems organic to the carrier

provide strategic theater awareness off-board the ship (inorganic ISR asset). Inorganic ISR
platforms onboard the aircraft carrier include the following: P-3 Orion and P-8 Poseidon
provides high-altitude submarine-hunting capabilities with onboard detection and ability

216 Source: Tate, “China Integrates Long-Range Surveillance Capabilities.”

217 Barnaby, The Automated Battlefield, 32–33.

81
to drop sonobuoy listening devices (active and passive radars); EP-3 can collect signals
emanating from other maritime platforms, Global Hawk are all inorganic assets and MH-
60R for hunting submarines, E-2 Hawkeye provides air early warning with a massive disc-
shaped radar to share data with other aircraft beyond the carrier; F/A-18 Super Hornet
provides visual identification with sensor-to-shooter and targeting capabilities, and the E/
A-18G Growler is the specialized version of the Super Hornet that provides tactical and
electronic jamming. This complex mobile airport system that uses organic and inorganic
automation allows the U.S. Navy to extend its reach to classify, identify, engage, and
respond to crises or threats beyond a mere 200 nautical mile distance.

Lasers provide an organic naval capability designed to counter ISR against threats
or naval platforms being lased but may cost more than it is worth. For counter-ISR, the
U.S. Navy has developed and installed the first Optical Dazzling Interdictor, Navy (ODIN),
laser weapon system to blind or disrupt unmanned aerial systems threatening U.S. forces
(Figure 16). 218 This technology requires a high level of directed energy and power—
sometimes enough to power a small city—with a low circular error probable (CEP) if
automated calculation to send high directed energy is accurate. The probability of hitting a
target manually can be extremely difficult and the system requires extensive locational data
and in real-time from the ship’s sensors. The major threat and necessity for AI within a
computer system is to decrease the CEP while being able to react against other incoming
threats. Other incoming threats can include future swarms of AUVs or countering other
technologies that may lase ISR cameras or optic sensors.

218 “Navy Leverages Workforce; Delivers C-ISR Capability Rapidly to Surface Fleet,” Naval Sea
Systems Command, February 20, 2020, https://www.navsea.navy.mil/Media/News/SavedNewsModule/
Article/2089079/navy-leverages-workforce-delivers-c-isr-capability-rapidly-to-surface-fleet/.
82
 Figure 16. Lockheed Martin’s Optical Dazzling Interdictor, Navy laser
Designed for Counter-ISR 219

When refueling aircraft, AI software within semi-autonomous aircraft reduces

human labor and provides surveillance capabilities on the maritime battlefield. Northrop
Grumman’s X-47B Unmanned Carrier Air System (UCAS) conducted its first flight in
February 2011 and from 2012 through 2015 conducted touch and goes from USS Harry S.
Truman and USS George H.W. Bush. Because of the iterative software tests, the first-ever
airborne refueling between a Boeing 707 and UCAS occurred in April 2015 and the AI
software inside this testbed became the new AI configuration for the Navy’s MQ-25A
Stingray. In addition to surveillance, because there are only nine air wings for 10 aircraft
carriers, this brilliant automation example can offload human labor and extends the range
of an F/A-18 Super Hornet beyond its 500 nautical mile range. 220 This AI-infused
unmanned aircraft can conduct aerial refueling and surveillance, which will further
automate naval warfare while extending the distance of naval aircraft range.

At sea, the fully autonomous ship “Sea Hunter” provides antisubmarine warfare
capabilities with AI applications and is cheaper than the P-8 Poseidon. Sea Hunter, which
was developed by DARPA and U.S. defense companies Leidos and Sonalysts, can patrol
the maritime domain for up to 70 days and range of possibly 10,000 nautical miles, and

219 Source: Harry Lye, “US Navy Deploys ODIN Counter-UAS Laser,” Naval Technology, February
20, 2020, https://www.naval-technology.com/news/us-navy-deploys-odin-counter-uas-laser/.
220 David Axe, “Stealth Refuel Drone: Meet the Navy’s New MQ-25 Stingray,” National Interest,
March 12, 2020, https://nationalinterest.org/blog/buzz/stealth-refuel-drone-meet-navys-new-mq-25-
stingray-132302.
83
only at $20 million per unit, according to The National Interest. 221 With one exception,
this aircraft provides ISR to monitor Chinese submarine activity with less human labor and
less in price. 222

5. Cyberspace and Social Media

Social media improve naval ISR missions in terms of identifying, tracking, and
locating naval platforms worldwide. For example, MarineTraffic and FleetMon, which
automate ship tracking and maritime intelligence, are websites generated by AI programs
to openly track radars actively emitting at sea (see Figure 17). The data from open-source
companies are gathered from a network of coastal AIS-receiving stations, satellite
receivers, and algorithms to identify, integrate, and track shipping, trade and logistics
worldwide. 223 The algorithm within this analytical tool can show past tracks and
background information. Purchasing the data in bulk can help fill the gap of unknowns and
improve the quality of AI tools or resources.

Figure 17. Marine Traffic Tool Tracks Chinese Maritime Militia (Left),
“Liaoning” CV-16 Operating in North Sea Fleet (Right) 224

221 Lyle J. Goldstein, “How China Sees the U.S. Navy’s Sea Hunter Drone,” National Interest,
January 31, 2017, https://nationalinterest.org/feature/how-china-sees-the-us-navys-sea-hunter-drone-19264.
222 One exception in which the vessel may not be able to withstand rough seas autonomously because
it may damage the computer system and robot that would render it inoperable
223 “Marine Traffic Live Map.”
224 Source: “Marine Traffic Live Map.”

84
 However, three major trade-offs are associated with this AI tool. First, maritime
tracking algorithms cannot provide continuous ship tracking if AIS is not transmitting data
or if the human operator does not have the exact name to search for a vessel. Second,
increasing automation and associated radar systems, AIS can be active and useful for
military tracking, particularly if one knows or can find the key name or MMSI is known or
can be found. Third, data must be available, accessible, bought, or easily transferable
through user-friendly databases or cloud-based services. Purchasing bulk data, integrating
a mix of open-source AI software tools can benefit ISR tracking compared to using several
separate resources.

The Navy, Special Forces and DOD writ large can utilize AI software, such as
Primer.ai, to quickly filter through online documents regarding the current geopolitical
situation and the competitors’ intent, strengths, and weaknesses. The Silicon Valley-based
AI software called Primer.ai reads and writes thousands of open-source documents with
machine learning assistance. CEO and founder of Primer Sean Gourley, Vice President of
Solutions Architect Ben Van Roo, and Director, National Security Group Brian Raymond
have developed an AI-assisted organization, summarization, and report updates on global
targets. 225 Key events, information, or tone are extracted among relevant documents. The
overall strategy and purpose of this software can provide quick indications and warnings
of the competitors next move based on thousands of open-source news, reports, and
documents. The overload of human cognitive labor and time in finding, researching, and
analyzing data is minimized.

6. Big Data: Cloud Services and the Semiconductor Industry

Maintaining a competitive edge with AI technology increasingly depends on

computer chips and semiconductor resources. The buzz terms “Big Data” and “the cloud”
are phrases that encompass both the memory capacity to store terabytes, and massive
amounts, of data in one network infrastructure (not multiple storage networks). The keys
to high-performance AI computing are semiconductors, made of silicon and germanium,

225 “Primer.Ai,” accessed March 14, 2020, https://primer.ai/.

85
and Graphical Processing Units (GPU), for data storage and to not overload a system to the
point of disruption or freezing. Although rarely found in nature, silicon can be found in
silicate or oxide minerals like quartz (looks like a crystal) and is the second most-abundant
element on Earth following oxygen. Germanium is a lustrous, gray-white, brittle metalloid.
According to Statistia, since 2019, the top three countries that produce silicon in million
metric ton are China with 4.5 million metric tons, Russia with 600 million metric tons,
Norway with 370 million metric tons, and the United States in fourth with 320 million
metric tons. 226 These are the two critical elements necessary to maintain a competitive
advantage. AI companies have been able to build websites or platforms online and be
competitive based on data quality, data quantity, and accessibility, whether the data is real
or synthetic. Russia, a great power competitor, is in the running with silicon production,
but this discussion goes beyond the scope of this thesis.

Synthetic data can be used for simulations. naval ISR aircraft all collecting data at
sea does nothing if one’s goal is to conduct FONOPS in the vicinity of the disputed islands
of the South China Sea. Some AI applications use synthetic data, which are real data, or
based on real historical data, and synthetic, or fake, data used for simulations. 227 A person
who trains and writes machine learning algorithms is computationally intensive, but high-
performance computing is necessary to maintain a competitive edge over the adversary.

Today’s intelligence databases or cloud services storing Big Data are outdated, but
the new Machine-Assisted Analysis Rapid-Repository System (MARS) may be a good
initial adoption of cloud services. The current 20-year old intelligence database called
Modernized Intelligence Database (MIDB) stores data globally for the defense intelligence
organization, but with the proliferation of data and sensors, today makes MIDB insufficient
and unreliable. MARS was developed for storage, cloud computing, and machine learning
AI applications that will allow human operators to use a system capable of ingesting and

226 M. Garside, “Major Countries in Silicon Production from 2014 to 2019 (in 1,000 Metric Tons)*,”
Statista, February 12, 2020, https://www.statista.com/statistics/268108/world-silicon-production-by-
country/.
227 AlphaGo in 2015 used historical data from human and human go matches; AlphaGo Zero in 2017
data was entirely trained with synthetic data from matches AI played against itself. Synthetic data can be
viable for wargame or peace game scenarios, but is not viable for all AI applications.
86
managing massive amounts of data and possibly simulate courses of action quickly for
operators. 228

7. Naval Special Warfare Adopts AI for Naval ISR

NSW and SOCOM combined continue to adopt AI in innovative ways and are true
early adopters. For example, on February 20, 2020, NSW contracted the company Aery
Aviation, LLC (“Aery”) to provide air-to-ground ISR services in Full Motion Video
(FMV), specialized sensors, and communications equipment with a C-208 reconnaissance
aircraft. 229 This new contract and capability encompass training and research and
development exercise. The previous DOD and Google AI called Project Maven, formerly
known as Algorithmic Warfare Cross-Functional Team, helps identify missiles in images
or crowds through a process of “trial and error by the computer as it ingests and processes
huge amounts of data.” 230 While this project may have received ethical or privacy rights
isuses, unmanned systems remain and innovative methodology to discriminate between the
population and objectives between civilians and the military, which supports a doctrine set
by the Geneva Convention (AP1). 231 AI programmed into automation systems makes it
easier to conceive of patterns of hostile or combatant behavior is occurring and thus,
provide both early warning and review battlespace decisions.

Special Operations Forces (SOF) are attractive units for rapid fielding and
prototyping of automation and AI and re-structuring doctrine. According to American
political scientist and counselor to the United States Department of State (2007-2009) Eliot
Cohen, elite units can serve as laboratories or incubators for the broader force while

228 DIA Public Affairs, “DIA’s Vision of MARS: Decision Advantage for the 21st Century,” DIA
News, May 23, 2019, https://www.dia.mil/News/Articles/Article-View/Article/1855910/dias-vision-of-
mars-decision-advantage-for-the-21st-century/.
229 Heather McAfee, “Aery Aviation, LLC Wins Naval Special Warfare Contract,” PR Newswire,
February 20, 2020, https://www.prnewswire.com/news-releases/aery-aviation-llc-wins-naval-special-
warfare-contract-301008746.html.
230 Lori Robinson, “The Evolution of Artificial Intelligence and Future of National Security,”
National Interest, March 13, 2020, https://nationalinterest.org/feature/evolution-artificial-intelligence-and-
future-national-security-133032.
231 Protocol Additional to the Geneva Convention of August 12, 1949, and relating to the Protection
of Victims of International Armed Conflicts (Protocol 1), June 8, 1977.
87
providing “fresh thinking into the mainstream of military thought” 232 and to shape a more
efficient AI doctrine. In the book Special Ops by former Navy SEAL and former head of
USSOCOM Admiral William McRaven, SOF defies conventional wisdom and via
constant repitition in training and rehearsals, is the key link in achieving a simple mission
in planning phase with the elements of surprise and speed in the execution phase. 233 Thus,
special operations and elite units identify, utilize, and report deficiencies among each group
of innovators along the S-Curve when tasked with specific problems and specific
objectives.

8. Private, Public, and Commercial Way of Adopting Automation

and AI

While China may produce the world’s largest amount of silicon in the world,
American private and commercial custom-designed chips and GPU innovations derive
from American companies in further advancing memory capacity on a computer. For
example, Google’s AI chip is called Tensor Processing Unit (TPU), which can offer
superior performance in AI computing over GPUs for AI software. While a company called
NVIDIA designs most GPUs in the United States and manufactured by TSMC in Taiwan,
acquiring semiconductor design in conjunction with custom-design chips may be the next
move in Big Data and cloud computing to stay ahead of the AI game. 234

Industry exemplifies staying ahead of the AI game in a tiered system to fit the needs
of the people in teams and the development cycle before deployment. For example, in June
2019, Booz Allen Hamilton, an American management and information technology
consulting firm, hosted a HACKtheMACHINE digital experience event to help the Navy
solve a critical future issue, which is seeking vulnerabilities and the ability to hack a ship’s
system. 235 Some examples of cyber challenges and opportunities given to academic or

232 Eliot A. Cohen, Commandos and Politicians: Elite Military Units in Modern Democracies
(Cambridge, MA: Center for International Affairs, Harvard University, 1978), 31–32.
233 William H. McRaven, Spec Ops: Case Studies in Special Operations Warfare: Theory & Practice
(Novato, CA: Presidio Press, 1996), 14.
234 Allen, Understanding China’s AI Strategy, 17–18.
235 Fathom5 and Booz Allen Hamilton, “HACKtheMACHINE 2019,” HACKtheMACHINE, June
2019, https://www.hackthemachine.ai/.
88
industry teams are listed in Table 6 who were able to achieve these AI challenges with AI
applications such as SharkWire; hacking only within the cyber bubble of this private event.

Table 6. Capture the Flag: Hack a Newly Developed Cyber Defense for
Industrial Control Systems

Low Threat Medium Threat High Threat

Take Advantage of the ship’s Spoof a waypoint that causes the Simultaneously move the rudder,
position auto-pilot system to change spoof the heading, and change
course GPS coordinates
Crack the ship’s WEP and WPA2 Remotely access websites for Create a fake catostrophic
Wifi Passwords Human-Machine-Interface weather event over the wire to
(HMI) virtually or send fake cause the ship to change course
distress signals over VHF Radio

If an individual can conduct a cyber attack, these opportunities and challenges pose
serious threats to naval ISR systems and platforms demand red-teaming by any team
worldwide to test naval systems, if given the opportunity or challenge in a series of events.
According to Dr. John Arquilla,

Unless there is a willingness to try innovative recruitment methods for

seeking out those with the necessary talents…One creative way to proceed
with recruiting would be to convince skilled IT industry techs to join up and
click for their country. This need not be a typical recruitment requiring
several years of active duty. Instead, the focus could be on bringing talented
men and women into Reserve and Guard formations, perhaps even forming
up new, purpose-built cyber units. These could be sited strategically, near
IT hubs. 236

Setting up quarterly AI challenges to find the fittest AI talent and recruit “cyber
sentries” from U.S. Reserve units to protect naval networks and systems can help make
great strides toward adopting automation and AI and avoid the saturation point. 237

The decentralized top DOD organizations that adopt AI may not be as connected,
posing gaps and weaknesses. Top organizations include the Defense Innovation Unit
(DIU), DARPA, SOFWERX, JAIC, NavalX. DIU is a DOD organization that contracts

236 John Arquilla, “Uncle Spam Wants You!,” Foreign Policy, February 5, 2013,
https://foreignpolicy.com/2013/02/05/uncle-spam-wants-you/.
237 Arquilla.

89
with commercial companies to solve national security problems and located in Silicon
Valley. 238 DARPA invests over $2 billion in new and existing AI programs in their “AI
Next” campaign with more than five decades of research and development in automation
in which systems can acquire new knowledge through algorithms. 239 SOFWERX is a
public-private military entity located in Tampa, Florida that combines DEFENSEWERX
and the United States Special Operations Command (USSOCOM) to fuse academia,
civilian companies, and other non-traditional DOD partners. 240

Specifically, the process of measuring the adoption of AI for naval ISR or for naval
automation writ large remains unclear. In February 2019, another organization called
NavalX, was solely designed for the Department of the Navy as a “super connector.”
NavalX creates social Tech Bridges between start-ups, academia, commercial, public and
private companies. 241 Another major AI center called JAIC, which is designed to harness
AI, likely faces significant challenging shortfalls as an organization trying to coordinate all
AI efforts. The plethora of options to adopt AI without a central strategy and design can
produce gaps and missed opportunities for adopting algorithms and connecting the right
people to the right organization. NavalX may consider partnering with Match.com to
quickly connect human operator and systems needs in a user-friendly and rapid way to
become aware of specific companies like a baseball card or swiping mechanism.

9. Neural Networks and Deep Learning: Heat Maps, Trees and,

GANs

Coherent change detection algorithms for terrain mapping are “hot commodities”
that use machine learning and neural network techniques to determine a pattern or

238 “Accelerating Commercial Technology for National Security,” Defense Innovation Unit, accessed
March 14, 2020, https://www.diu.mil/.
239 “AI Next Campaign,” DARPA - Defense Advanced Research Projects Agency, accessed March
14, 2020, https://www.darpa.mil/work-with-us/ai-next-campaign.
240 Michael Bottoms, “SOFWERX: A Smart Factory of Innovation Helping the Warfighter,”
USSOCOM - United States Special Operations Command, February 2, 2018, https://www.socom.mil/
Pages/SOFWERX--A-smart-factory-of-innovation-helping-the-warfighter.aspx.
241 Assistant Secretary of the Navy for Research, Development, and Acquisition (ASN RDA),
“Welcome to NavalX,” NavalX, accessed March 14, 2020, https://www.secnav.navy.mil/agility/Pages/
default.aspx.
90
anomalies. Raytheon has developed Intersect Sentry to detect change and generates heat
maps to determine change detection, tracking, and averting adversarial activity (Figure 18).

Figure 18. Algorithms in Intersect Sentry Detects and Visualizes Routes of

Maritime Vessels

The significance of a heat map is that the algorithm can calculate the probability
distribution to determine course, speed, and distance to determine patterns or anomalies.
Heat maps help with detecting multi-dimensional anomalies and situational awareness.
Some useful open-source websites with this capability include: Global Data on Events,
Location and Tone (GDELT), satellite imagery, AIS data, and more. 242 To fuse all the
data, the tree diagram is one useful AI technique to categorize and visualize maps.

Tree diagrams form alternative ways to determine the top two teams, instead, the
algorithm creates multiple teams and categorizes them per color along the light spectrum.
To create the heat map, a computer program first divides the area or target of interest into
smaller parts, then the frequency of unique vessels in each part is color-coded, and lastly,

242 GDELT is “supported by Google Jigsaw, the GDELT Project monitors the world’s broadcast,
print, and web news from nearly every corner of every country in over 100 languages and identifies the
people, locations, organizations, themes, sources, emotions, counts, quotes, images and events…”
https://www.gdeltproject.org/.
“ADS-B improves safety and efficiency in the air and on runways, reduces costs, and lessens harmful
effects on the environment,” https://www.faa.gov/nextgen/programs/adsb/; “OpenStreetMap is a map of the
world, created by people like you and free to use under an open license,”
https://www.openstreetmap.org/#map=4/38.01/-95.84; “The National Oceanic and Atmospheric
Administration is an American scientific agency within the United States Department of Commerce that
focuses on the conditions of the oceans, major waterways, and the atmosphere,” https://www.noaa.gov/.
91
the density in traffic of vessels is calculated based on the number of vessels per unit area
(Figure 19). 243

Figure 19. Heatmap and Trees Indicate Variables and Clustering of Colors 244

A strategy for rapidly training and prototyping machine learning is to integrate

Generative Adversarial Networks (GANs) into AI platforms for naval ISR. According to
Ian Goodfellow, Yoshua Bengio, a researcher in machine learning and deep learning,
proposed GANs, which is a new framework for estimating generative models through an
adversarial process. To illustrate possible anomalies, this algorithmic technique helps
detect, classify, and discriminate the input attributes. Examples of anomalies include false
positive or false negative objects or vessels produced by sensors onboard ISR platforms.

China’s fleet is growing and modernizing rapidly to the U.S. Navy with its strategy
of late adoption and late modernization, but the U.S. Navy still deploys a larger, heavily
armed, and greater military force. According to the Office of Naval Intelligence, in the last
two decades, PLAN has been on track to increase its naval capability by 2020 and AI is
helping its fleet reach its goals to surpass both the U.S. Navy. 245

243 Christine Nezda, “Machine Learning for Patterns of Life,” Technology Today, no. 1 (2018): 26–
29.
244 Source: Nezda, “Machine Learning for Patterns of Life.”
245 Office of Naval Intelligence, The PLA Navy: New Capabilities and Missions for the 21st Century
(Washington, DC: Office of Naval Intelligence, 2015), 13, https://www.oni.navy.mil/Portals/12/
Intel%20agencies/China_Media/2015_PLA_NAVY_PUB_Print.pdf?ver=2015-12-02-081247-687.
92
 V. SUMMARY AND RECOMMENDATIONS

INTRODUCTION

In this thesis, I have attempted to compare the different ways in which the American
and Chinese navies have adopted automation and AI for naval ISR. I conceptualized how
specific automation tools—Aegis, SSDS, and Type-346—improve the modern maritime
battlespace. Chapter II conceptualizes the term automation—a technique, process, or
system by which a machine achieves a specific goal— and AI, a more ambitious concept
that posits analytic and assessment capabilities. Chapter III describes Rogers’ S-Curve of
technological adoption and suggests a total of five categories of adopters to standardize the
usage of adopter categories. This methodology is important because, without it, matching
ideas discussing and categorizing appropriate technologies to its appropriate timeline is
critical in the adoption process. Chapter IV displays and analyzes ways in which both
navies have adopted automation and AI naval ISR platforms and software. Although the
U.S. Navy has maintained dominance in maritime superiority and air superiority, the PLAN
is fast-approaching in dominating AI for future sea power.

This thesis contains a seven-step strategy and design process applicable to simple,
complex, and extraordinarily complex automation or AI applications for technological
adoption. This theoretical process can help identify primary demands and material
necessary for successful naval ISR practices in the maritime battlespace. Human beings,
the material, and physical space represent key elements. The theoretical framework in this
study identified five key technological adoption capabilities: early and quarterly “campfire
talks”; visionary and thought leaders; a grand strategic narrative; a design an AI
development plan; lead in computer technology material resources; and, build an
instantaneous and user-friendly platform for AI engineering experts and naval knowledge
experts.

This chapter covers the summary of findings, the AI gap, ethics and AI,
recommendations for the U.S. Navy, and future work for automation and AI tools that
involve human-machine teaming for naval ISR.

93
 SUMMARY OF FINDINGS

Frank Barnaby reminds us that “military technology is automating warfare.” 246

Current trends show undoubtedly automated processing, fusion, and product delivery for
command and control the maritime battlespace. 247 I began this investigation into
automation and AI as applied to naval ISR to better understand the adoption mechanisms
and categories used by both navies that not only make each navy a great competitor but
also what characteristics and technologies make their forces successful naval AI powers.

The findings from the S-Curve of technological adoption for naval ISR could be
important if a commander becomes inundated with information or intelligence and can,
instead, stimulate a new culture to adopt a hybrid way of adopting automation and AI that
folds in early adopters with later adopters. According to former CEO of Google Eric
Schmidt, “The DOD has an innovation adoption problem” and should adopt a DevOps, or
Development Ops, culture for software systems and focused on “customer adoption.”248
Cultural development centers on the user or tactical operators and AI engineers that serve
in both navies. For example, the PLAN has a development culture of adopting technologies
later. Later adopters or laggards can benefit from early innovators’ successes and failures
because this category of adopters focuses on adopting and “copy-catting” successful
technologies and has observed what technological failures not to adopt.

Figure 20 demonstrates that both navies as near-peer competitors and are successful
at adopting automation; however, they are on opposite ends of the curve. This means that

246 Barnaby, The Automated Battlefield, 21.

247 See U.S. Navy Information Dominance Corps, U.S. Navy Information Dominance Roadmap
2013–2028 (Washington, DC: Department of the Navy, 2013), 33,
https://defenseinnovationmarketplace.dtic.mil/wp-content/uploads/2018/02/
Information_Dominance_Roadmap_March_2013.pdf “Assured C2 is ultimately about ensuring a
commander’s ability to command assigned forces to achieve the tactical, operational or strategic objectives
established by the chain of command. Navy’s future information infrastructure must be able to maintain
essential network and data link services across secured segments of the EM spectrum in high-threat
scenarios to transport, share, store, protect and disseminate critical data and combat information required by
forward deployed units and on-scene commanders.”
248 Eric Schmidt, “Promoting DOD’s Culture of Innovation: Testimony before the Committee on
Armed Services” (2018), https://armedservices.house.gov/2018/4/promoting-dod-s-culture-of-innovation.
94
the U.S. Navy represents innovation and early adoption, and PLAN represents later
majority and laggards of adopting technologies for naval ISR missions.

Figure 20. U.S. Navy Innovators and Adopters Compared to PLAN Later
Adopters Along the Logistic S-Curve

Both great power competitors appear to co-exist and complete the wavelength within a
forward or backward thinking way of adopting technologies, but the political system can
impact the rate of adoption whether it is by an innovator or later adopter.

1. AI Benefits an Autocratic Navy

The rate at which automation and AI technologies are adopted tends to benefit an
autocratic society. According to CCP’s AI strategy, China’s plan is to become the global
innovation center by rapidly developing AI, reinforcing China’s lead in adopting
automation for naval ISR. 249 China’s population is over 1.4 billion citizens, with
approximately 290,000 PLAN personnel, which roughly equates to 12.6 percent of PLA’s
2.3 million personnel. Since 2018, China has led the world in publishing approximately
42.64 percent of top AI papers, approximately 52 percent of AI global patents, and places
second in AI companies with over 1,011 companies compared to the 2,028 of U.S. AI

249 Sapio, Chen, and Lo, A New Generation of Artificial Intelligence Development Plan.

95
companies. 250 The authoritarian system creates an environment that values security over
privacy for China’s massive population because the government and AI companies are
ultimately controlled by the Chinese Communist Party. In the private sector, China
leverages Baidu, Alibaba, and Tencent (collectively referred to as BAT) for fueling the
growth of AI with fewer privacy restrictions compared to the West. China has the
advantage of a higher population in an open society to assist with cleaning data, to develop
or copy algorithms, and to collect and surveil naval platforms at sea with open-source data
and social media. Perhaps, the powerful feedback loop between corporations developing
AI and the CCP is a marriage made in heaven for PLAN; alternatively, this is a match made
in hell for the U.S. Navy. 251

2. Automation and AI: Changes in International Maritime

Governance

PLAN has since been able to take advantage of and lay claim to a new model of
maritime governance and is rapidly approaching its dominance in AI. China has already
begun to overtake the U.S. Navy in adopting automation. In China’s Tsinghua University’s
recent work on AI development, stage five (2017 to present) of China’s national AI policy
evolution shows the characteristics of next-generation AI among the military, industry, and
the civilian sector. 252 Since 2013, PLAN and China built and militarized its self-made
islands (i.e. Fiery Cross, Subi Reef, Mischief Reef) in SCS with cruise missiles and air

250 China Institute for Science and Technology Policy, China AI Development Report 2018, 3–4.
251 The ease of crosspollination between the People’s Liberation Army (PLA) and its private, public,
and commercial center can help build compatible hardware and software faster. Other cost-effective,
structure IT architecture (i.e. Wide Area Network connections with Cloud Services), social media, and
more.
252 Since the time of release of key AI policy documents, China’s AI policies divided into five stages:
“Stage 1 (before 2013), of potential development, where few policy documents were released and AI was
not specified as a national priority; Stage 2 (2013-2015), of preliminary development, where the
importance of AI began gaining recognition across all circles of society; Stage 3 (2015-2016), of rapid
development, where a lot of policies documents were released and AI was elevated as a national strategy;
Stage 4 (2016-2017), of stable development, where understanding of AI R&D and industry development
was increasingly mature and policy documents came out stably; and Stage 5 (2017 to the present), of steady
iteration, where all sectors have a more pragmatic understanding of AI and related policies are more
specifically targeted.”
96
defense systems. 253 Within the first island chain, the PLAN has the “home-court”
advantage to surveil and maintain its 12-nautical-mile zone around the disputed islands in
the SCS. PLAN continues to use Dongdiao-class AGI spy ships to learn about U.S. naval
affairs. As later adopters, the PLAN will likely be able to “copy-cat” successful naval ISR
tactics and operations demonstrated by the U.S. Navy.

The PLAN is taking the lead in AI over the U.S. Navy with its spy ships,
reconnaissance satellites, and OTH radars by collecting massive amounts of data. Moore’s
Law, where the number of transistors per silicon chip—computing processing—power
doubles every two years, leads to rapid growth and productivity. 254 The PLAN is taking
advantage of collecting massive data from sea to space assets to produce high quality data.
If computer processing doubles every two years, it will be less challenging to filter through
with high-quality data collected over time and easier to locate in a central data storage (i.e.,
the cloud).

As the PLAN continues to build more network infrastructure locally and globally,
it shows its willingness to take short-term risks for long-term advantages. Another long-
accepted factor, Metcalfe’s Law, asserts that the value growth of the fully connected
network is proportional to the square of the number of compatibly communicating
devices. 255 For example, if a network has five nodes fully interconnected with each other,
the inherent value of each being x, its power is 100x (due to its ten distinct links) versus
the 5x power of a single-linked network. While the value of the networks increases for
compatible communication, it also increases cybersecurity issues if a competitor disrupts
or hacks one of those nodes. The PLAN’s is willing to take the risk of computer network
attacks to dominate in AI for long-term strategy.

253 David Axe, “How China Is Militarizing the South China Sea with a Ton of Missiles,” National
Interest, March 23, 2020, https://nationalinterest.org/blog/buzz/how-china-militarizing-south-china-sea-
ton-missiles-136297.
254 “Moore’s Law,” in Encyclopaedia Britannica, July 21, 2011, https://www.britannica.com/
technology/Moores-law.
255 “Metcalfe’s Law,” in Cambridge Dictionary, accessed April 2, 2020,
https://dictionary.cambridge.org/us/dictionary/english/metcalfe-s-law.
97
 3. U.S. Navy’s First Mover Advantage: Semiconductors and
Software

However, the U.S. leadership in AI already extends beyond the U.S. littorals, has
more naval personnel than the PLAN, and leads material resources necessary to build
network infrastructures. While the U.S. population is 329.4 million, 256 lower than China’s
overall population, the U.S. Navy has 339,448 personnel, which is more than the number
of personnel in the PLAN. 257 The U.S. Navy has the first-mover advantage to adopt AI
technologies not simply because it has more people than the PLAN, but because the United
States leads in top AI companies. Specifically, U.S. AI companies focus on enterprise
software, semiconductors, and quantum computing.

However, the marriage between the U.S. Navy and industry or private companies
can present significant issues and friction with AI engineers who do not or do not
consistently support the military with AI. Great American talent derives from Silicon
Valley, the Joint AI Center, DOD’s programs such as DARPA, DIU, and NavalX.
However, the top AI companies such as Google, Facebook, and Amazon have received
criticism for privacy rights issues and public opinion of the weaponization of AI creates
issues of seamless collaboration. Even though the U.S. Navy and the DOD appear to lead
in some aspects of AI when collaborating with Silicon Valley to prototype and field
commercial AI tools within 60 to 90 days, 258 this does not apply to all naval ISR platforms.
USSOCOM may be able to take advantage of commercial AI tools with Silicon Valley-
based AI engineers, but conventional U.S. Navy may likely years or decades to fully
automate and adopt AI tools with legacy, conventional platforms.

256 “U.S. and World Population Clock,” United States Census Bureau, April 3, 2020,
https://www.census.gov/popclock/.
257 “Status of the Navy,” U.S. Navy, April 3, 2020, https://www.navy.mil/navydata/
nav_legacy.asp?id=146.
258 “Defense Innovation Unit: Who We Are/Our Mission,” Defense Innovation Unit, accessed April
22, 2020, https://www.diu.mil/about.
98
 4. An Automated Aircraft Carrier May Be Powerless in Future of
AI and Sea Power

ISR technologies make it nearly impossible to hide naval platforms, and massive
amounts of data can deceive the competitor. In his article, “Artificial Intelligence on the
Battlefield,” Dr. Zachary Davis, a Senior Fellow at Lawrence Livermore National
Laboratory and Professor at the Naval Postgraduate School, believes, “AI could erode
stability by increasing the perceived risk of a surprise attack,” and AI-supported ISR
platforms on one side can confuse other exquisite ISR. 259 With many naval ISR operators
working long hours and unable to fully vet the relevancy and accuracy of locating mobile
targets, and also strike with speed and precision, Davis claims ISR operators today are
easily victimized by the massive amounts of data and the over-collection of data with which
the competitor is all too eager to manipulate or deceive the other competitor.

THE AI GAP: MISSING AI PLATFORM TO CONNECT AI ENGINEERS

AND OPERATORS/USERS

This study reveals that there is no apparent central automation or AI tool with
instantaneous notification techniques connecting the AI engineer to the knowledge expert
or naval user to fix issues faster. 260 Bridging the gap automatically can significantly help
gain customer and competitor insights. According to a RAND study on assessing data
analytics, a particular challenge is associated with the collection and use of unstructured
data, and not in fixed locations such as a relational database. 261 Automating an AI platform
for AI engineers and naval users directly and within one central database is vital for the
U.S. Navy to resolve and master. Adopting a new platform to connect AI builders and naval
users will help surpass the inflection point of technological adoption.

259 Zachary S. Davis, Artificial Intelligence on the Battlefield: An Initial Survey of Potential
Implications for Deterrence Stability and Strategic Surprise (Livermore, CA: Center for Global Security
Research, Lawrence Livermore National Laboratory, 2019), 14–16, https://cgsr.llnl.gov/content/assets/
docs/CGSR-AI_BattlefieldWEB.pdf.
260 Refer to Chapter 1, pages 6–7 regarding Stanford University Professor Feigenbaum’s “Expert
System.”
261 Philip S. Anton et al., Assessing the Use of Data Analytics in Department of Defense Acquisition
(Santa Monica, CA: RAND Corporation, 2019), https://www.rand.org/pubs/research_briefs/RB10085.html.
99
 Despite the increased collection of information, this study reveals that descriptive
and diagnostic styles, specifically visual analytics and trend analysis, are utilized. AI
provides information about what has happened in naval ISR (i.e., platforms and sensors)
and diagnostic AI pinpoints the exact problem or issues quickly. As mentioned in Chapter
IV, the U.S. Navy adopted an OTH visual analytical tool for early warning of incoming
naval platforms. Descriptive Both descriptive and diagnostic styles of AI are fundamental
processes to master prior to adopting more complex predictive and prescriptive AI, typified
by neural networks, pattern recognition, machine learning, and deep learning.

Predictive AI is lacking in the U.S. Navy’s arsenal of data analytics. Predictive AI

provides data as to what will likely happen—where a naval platform may go next or
possibly why it is going to specific locations next. Examples of predictive AI techniques
are machine learning, pattern recognition, and statistical modeling; this approach to AI
programming uses historic and statistical data to conduct trend analysis of naval activities.
Predictive and prescriptive AI are advanced techniques necessary to maintain a competitive
posture toward China.

Prescriptive AI is also lacking. Prescriptive AI provides data on what will happen

and what could happen better if naval ISR platforms conduct x, y, or z maneuvers at sea,
based on historic data, and provide recommendations. Examples of prescriptive AI include
supervised learning algorithms such as “random forest,” which creates and merges decision
trees into one “forest” based on previously collected data or models. This can help support
future heat mapping and Generative Adversarial Networks (GANs) techniques for war-
gaming or “red-teaming” against competitors like the PLAN. Simply put, whether the
navies adopt early or after all other navies have adopted specific automation or AI
technologies, perfecting the basics and developing a culture of innovation and AI
operations at each level of command remains critical to the success of leading in AI.

100
 Another AI platform the U.S. Navy lacks is one that distinguishes possible
abnormalities or anomalies from “normal” patterns of life at sea. 262 To develop and adopt
technologies to distinguish abnormalities or anomalies at sea, the currently available
material, such as semiconductors and software talent from Silicon Valley, should be
accessed and tested consistently and frequently. Issues of software or hardware deficiency
can lead to latency in updating, which may ultimately pose issues in identifying and
locating PLAN forces operating at sea.

The commercialization of AI and military technologies can be the next game-

changing threat. 263 Drones sold and used by the commercial industry can lead to the
competitor using swarm tactics against critical sensors such as Aegis, SSDS, or Dragon
Eye radars. Each level of automation and AI adds a layer of complexity to diagnose,
predict, and prescribe the next move by the PLAN. Moreover, if PLAN develops AI with
the commercial sector, it creates less control by the U.S. Navy over the specific software
or hardware updates. Less control over specific automation or AI tools can lead to other
competitors or adversaries buying, selling, and adopting the tools and using them directly
against the U.S. Navy.

RECOMMENDATIONS

This study provides four major recommendations. First, the U.S. Navy should
develop a popular platform for direct feedback and change the culture to streamline the
connection between AI engineers and tactical operators, like Amazon’s instant feedback
loop. Another exemplary community that is already adopting emerging technologies
quickly is the Special Operations Force (SOF), specifically Special Operations Command

262 Pattern of life software exists to analyze large amounts of data from many different sensors
represented on a Common Operational Picture or large graphs to find distinct behavior patterns that may
indicate a deviation from normal modes of operation, malintent, or anomalous patterns. Any deviation from
expected pattern sends immediate “red flags” that can tremendously help understand the competitors
normal or abnormal behavior, help prevent collisions, and proactively prepare appropriate and professional
responses
263 Zachary Davis, Michael Nacht, and Ronald Lehman, eds., Strategic Latency and World Power:
How Technology Is Changing Our Concepts of Security (Livermore, CA: Center for Global Security
Research, Lawrence Livermore National Laboratory, 2014), 11, https://cgsr.llnl.gov/content/assets/docs/
Strategic_Latency.pdf.
101
(USSOCOM). DOD organizations such as DIU, NavalX, and USSOCOM can give,
according to NPS Professor Leo Blanken, “a simple and cost-effective way to improve
existing innovation efforts in the field: aligning military graduate researchers with
deployed special operations units to rapidly prototype. concepts and technologies.” 264 The
SOF community provides an attractive testbed for rapid prototyping and instantly connect
tactical users to AI engineers.

Second, the U.S. Navy should invest in and adopt a “DevOps” culture to maintain
a competitive advantage over the PLAN. In a hierarchical organization, cultivating and
practicing DevOps will inherently start with an AI school or time for education in the pre-
and post-deployment cycle. This can only be indoctrinated and inspired by senior officers
in charge. Arguably, the U.S. continues to lead in innovation and software technology and
talent but can further improve by co-locating U.S. naval commands or headquarters nearby
commercial and private AI companies to cultivate and adopt the next AI breakthrough.

Third, while implanting more automation and AI into the aircraft carrier, the
corresponding doctrine should consider swarm tactics and cyberspace. Just like hives of
bees can swarm a bear and overtake him, the autonomous drones can easily swarm a carrier
strike group, particularly Aegis or SSDS and its associated antennas. 265 The consequence
of not incorporating swarm technologies with operational and tactical ISR missions can
lead to the competitor destroying Aegis or Type-346 with very little cost and tremendous
damage to naval ISR.

Fourth, both navies share cyberspace. JP 3-12 defines information dominance as

“the degree of dominance in cyberspace by one force that permits the secure, reliable
conduct of operations by that force, and its related land, air, maritime, and space forces at
a given time and place without prohibitive interference by an adversary.” 266 Cyberspace

264 Leo Blanken, Philip Swintek, and Justin Davis, “Special Operations as an Innovation Laboratory,”
War on the Rocks, February 25, 2020, https://warontherocks.com/2020/02/special-operations-as-an-
innovation-laboratory/.
265 John Arquilla, In Athena’s Camp: Preparing for Conflict in the Information Age (Santa Monica,
CA: RAND Corporation, 1997), 465.
266 Joint Chiefs of Staff, Cyberspace Operations, Joint Pub 3–12 (Washington, DC: Joint Chiefs of
Staff, 2018), GL-4, https://www.jcs.mil/Portals/36/Documents/Doctrine/pubs/jp3_12.pdf.
102
is the Achilles heel of automated naval platforms because automation and communication
between naval ISR platforms rely on sharing the information space and deconflicting
frequencies to communicate.

In this thesis, I discovered five points that should be considered when adopting
automation and AI for naval ISR:

1. Develop a strategic narrative from day one and discuss the AI strategy at
each level of command. A strategic narrative is a special kind of story that
an organization can clearly understand and engage with the story. This is
important because it tells other maritime competitors who you are, where
you have been, where you are going, and where you are. Each command
should want to inspire its military service members, excite allies and
partners, and attract users such as the AI engineer(s) and human operators.

2. The country and navy that has control of the technological materials
necessary to build automation and AI, such as semiconductors, silicon,
and germanium, will dominate in AI in future sea power. Is the market
ready to support the military’s solutions in naval ISR?

3. Success or failure of adoption of automation relies significantly, if not

entirely, on the control of the electromagnetic spectrum and ability to
deconflict frequencies.

4. There should be an automated AI feedback platform so that military users

and AI engineers can text and speak directly. A winning and better end-to-
end product requires great user experience for efficient workflow and
instantaneous reporting.

5. The significant insight between early adopters and laggards shows that the
U.S. Navy and PLAN are on opposite ends of the wavelength of
technological adoption and competition. In Rogers’ work on the S-Curve
of technological adoption in Chapter III, both American and Chinese naval
leaders falling on opposite ends of the S-Curve suits each of their
organizational and doctrinal structures.
103
 THE FUTURE OF SEA POWER AND FUTURE WORK: HUMAN-
MACHINE TEAMING

Humans and machines teaming together is the key to the adoption of automation
and AI in naval ISR affairs. This section covers future work and long-term considerations
in the adoption process of automation and AI. These factors are: automate health indicators
for the operator; take advantage of research in quantum computing; and, the increase of
various reality technologies (virtual, augmented, and mixed).

1. Team AI Health: COVID-19

The center of gravity for Naval ISR and adopting automation or emerging
technologies depends on healthy human operators on naval platforms to operate the
systems. If an epidemic disease spreads throughout one aircraft carrier such as USS
Theodore Roosevelt (CVN 71), then all ISR in support of carrier operations halts and the
U.S. Navy fails to adopt AI for surveillance—with consequences that could lead to life or
death situations. 267 AI can provide significant human-assisted tools that can automate clear
indicators.

Commanders will find it useful to strategize various designs and design a strategy
to integrate health practices and AI for naval operators. An automated stop-light chart and
daily tracker with notification symbols that blink, or flash will help indicate a diagnosed
problem or trigger for commanders and decision-makers. A simple model to automate
indicators related to biological event-related social disruption is the Wilson-Collman Scale
of four stages of the increased likelihood of a biological event: favorable conditions,
unifocal or multifocal biological events, severe infrastructure constraint, and depletion of
local response capacity, and then social collapse. 268 A second complex model is Johns

267 Lucy Craymer, “Virus Grounds a U.S. Aircraft Carrier as Crew Quarantined in Guam,” Wall
Street Journal, April 1, 2020, https://www.wsj.com/articles/virus-grounds-a-u-s-aircraft-carrier-as-crew-
quarantined-in-guam-11585736476.
268 See James M. Wilson et al., “A Heuristic Indication and Warning Staging Model for Detection and
Assessment of Biological Events,” Journal of the American Medical Informatics Association 15, no. 2
(April 2008): 158–71, https://doi.org/10.1197/jamia.M2558 can be used to “determine the level of concern
warranted, such as whether the pathogen in question is responding to established public health disease
control measures, including the use of antimicrobials or vaccines; whether the public health and medical
infrastructure of the country involved is adequate to mount the necessary response.” .
104
Hopkins Coronavirus interactive map. 269 Another database to extract or utilize models
from is Cornell University’s arXiv database for top AI researchers in the world
submitting to Stanford University’s prominent repository (see the appendix).270
Automating lists of SWOT in an interactive way can benefit naval ISR. This is important
because having a clear defined list of tasks and problems helps diagnose problems as
early as possible and finding the right AI researchers quickly.

2. Quantum Computing

Further research into quantum computing will contribute to the U.S. Navy’s
successful adoption of AI. Quantum computing goes beyond binary digits (bit value of
either 0 or 1), which is the smallest unit of data in a computer, and uses what researcher
Thomas Campbell calls “qubits, in which an individual bit can be in one of three states: on,
off, and, uniquely, both on and off simultaneously.” 271 In the U.S. National Quantum
Initiative of 2018, Congress identified the National Institute of Standards and Technology
(NIST) and the Department of Energy’s Quantum Information Science Research Centers
as top leading departments for quantum computing. Thus, the U.S. Navy should
continuously collaborate with NIST and DOE closely. 272 In this case, Congress is one of
the top keys in financing and putting forth bills to initiate the adoption of emerging
technologies within the Navy. While in its infant stages, quantum computing can increase
AI speed, particularly in machine learning, in order to spot patterns rapidly and filter
through massive datasets; it will cast the long shadows of future sea power.

269 See Johns Hopkins University, “COVID-19 Dashboard by the Center for Systems Science and
Engineering (CSSE) at Johns Hopkins University (JHU),” Johns Hopkins Coronavirus Resource Center,
accessed May 26, 2020, https://coronavirus.jhu.edu/map.html.
270 arXiv, “Artificial Intelligence,” accessed September 11, 2019, https://arxiv.org/list/cs.AI/recent.
271 Thomas A. Campbell, Artificial Intelligence: An Overview of State Initiatives (Evergreen, CO:
FutureGrasp, LLC, 2019), 11, http://www.unicri.it/in_focus/files/Report_AI-
An_Overview_of_State_Initiatives_FutureGrasp_7-23-19.pdf.
272 National Quantum Initiative Act, Pub. L. No. 115–368, 132 Stat. 5092 (2018),
https://www.congress.gov/bill/115th-congress/house-bill/6227.
105
 3. Ethics and AI

The United States and China have both released “principles” rather than focusing
on the controversial issue of whether automation and AI are ethical, morally permissible,
or not morally permissible. In June 2019, China’s Ministry of Science and Technology
published the “Governance Principles for a New Generation of Artificial Intelligence:
Develop Responsible Artificial Intelligence,” which lists eight principles of AI
governance: (1) harmony and friendliness; (2) fairness and justice; (3) inclusiveness and
sharing; (4) respect for privacy; (5) security and controllability; (6) shared responsibility;
(7) open cooperation; (8) agile governance. In February 2020, the DOD’s Defense
Innovation Unit spent 15 months prior to announcing its adoption of five principles of AI
ethics: They should be Responsible, Reliable, Equitable, Governable, Traceable. 273
Releasing principles like the ethical AI guidance suggests that both nations desire
flexibility, but also responsibility in the development of AI. With or without principles,
major controversies over the use of AI quickly become relevant to issues on AI
misidentification and the possibility of weaponizing automation and AI tools.

First, ethical issues exist related to AI and the military that have been portrayed
both in the movies, such as Terminator or I, Robot, which began as thoughtful short stories
by Isaac Asimov, 274 and demonstrated in real-life situations. Some argue that military
technology or the use of AI in the military should not continue, due to the possibility that
computer systems could act on their own accord or make their own lethal decision based
on popular movies and social media. If AI tells a human operator that an aircraft is hostile,
only seconds exist for the Commander to decide to respond or not respond based on the
information available. However, since 1988, the adoption of automation and AI engineered
in ISR platforms has increased transparency and avoided incidents like USS Vincennes,

273 C. Todd Lopez, “DOD Adopts 5 Principles of Artificial Intelligence Ethics,” DOD News,
February 25, 2020, https://www.defense.gov/Explore/News/Article/Article/2094085/dod-adopts-5-
principles-of-artificial-intelligence-ethics/.
274 Isaac Asimov, I, Robot (New York: Bantam Books, 1950).

106
which was a human error and not a machine error. 275 Arguably, automation helps for
defense against friendly or enemy fires and for national security.

Second, in human-machine teaming, humans are moral agents and machines are
not, but can be built to provide ethical choices. In Aristotle’s Nicomachean Ethics, a
virtuous person does not hold an openly virtuous attitude but, rather, is one who acts in
certain ways in specific situations, with a range of reasons. Humans can consciously make
separate choices; whereas machines today simply assist human beings in making decisions.
In the U.S. Navy, ethical matters likely depend on a traditionalist or legalist perspective, or
possibly a revisionist view of an idea with a more inquisitive or possibly utilitarian
mindset. 276

In the PLAN, future work on ethics and the adoption of automation can be studied
through the ethical decisions made by the PRC and CCP. In China, ethics relate closely to
virtue ethics, concerning how one ought to live, and consequentialism, concerning the
benefit of all involving material goods. 277

While AI machines today are autonomous to a certain extent, AI machines could

eventually present consequences based on specified moral ethical codes quicker than a
human can think about them. Two serious ethical issues are: can autonomous systems be,
or will they be categorized as moral agents in the future? What happens when there is a
lack of accountability regarding amoral agents—autonomous systems—that violate human
laws of war in life or death situations? 278 These ethical questions posed by NPS Professor
Bradley Strawser are worth considering, but go beyond the scope of this thesis.

275 Dirk Jan Barreveld, ed., Air Crash Investigations: Killing 290 Civilians - The Downing of Iran Air
Flight 655 By the USS Vincennes Kindle Edition (Morrisville, NC: Lulu Enterprises, Inc., 2016).
276 Seth Lazar, “War,” in Stanford Encyclopedia of Philosophy, ed. Edward N. Zalta (Stanford, CA:
Metaphysics Research Lab, Stanford University, 2020), https://plato.stanford.edu/entries/war/.
277 David Wong, “Chinese Ethics,” in Stanford Encyclopedia of Philosophy, ed. Edward N. Zalta
(Stanford, CA: Metaphysics Research Lab, Stanford University, 2018), https://plato.stanford.edu/entries/
ethics-chinese/.
278 Bradley J. Strawser and David Whetham, Killing by Remote Control: The Ethics of an Unmanned
Military (Oxford, UK: Oxford University Press, 2013), 16–17.
107
 4. Human-Machine Teaming and Virtual Reality (Augmented
Reality)

While I share great enthusiasm for automation and AI’s potential to greatly improve
human wellbeing, the development of machines with intelligence superior to humans could
cause cognitive dissonance within the mind of a human operator and possibly be used as a
deterrent mechanism. 279 Albert Einstein warned the power of the atom could change our
modes of thinking. The atomic bomb led to the nuclear catastrophes of Hiroshima and
Nagasaki. Then nuclear power turned into the idea of a “nuclear Armageddon,” which is a
theoretical scenario involving the use of nuclear weapons to cause widespread destruction
and eventually, the collapse of civilization. Today, former Secretary of State Henry
Kissinger, among many others, believes AI threatens our consciousness and our way of
understanding truth and reality. 280 The notion that AI can be used as a deterrence tool is
an important question that can be further studied in future research.

Based on this study, the more automated the naval ISR platforms and the increase
of AI tools, the more virtual naval affairs and sea power will become. Recent reports like
the U.S. “Extended Reality Applications in U.S. Defense Training, 2020” shed light on the
daily impact of digital transformation on human beings and has extended to Virtual Reality
(VR), Augmented Reality (AR), and Mixed Reality (MR) technologies that could be the
next disruptive technologies. The commercialization of various forms of VR may become
a threat against naval forces. 281 This study has not addressed reality technologies, VR, AR,
or MR can be extremely useful to keep human beings safe from harm or as a deterrence
tool.

With Einstein’s and Kissinger’s warnings on the catastrophe of a nuclear war and
AI threatening our human consciousness, respectively, AI could possibly trend toward a

279 For instances, person’s reality may come into question when operating at sea if the automation
system tells the human operator a ship or submarine is operating in one location, when in fact it could be an
autonomous beacon replicating the same characteristics as a naval platform.
280 Graham Allison, “Is China Beating America to AI Supremacy?,” National Interest, December 22,
2019, https://nationalinterest.org/feature/china-beating-america-ai-supremacy-106861.
281 Frost & Sullivan, Extended Reality Applications in U.S. Defense Training, 2020 (San Antonio,
TX: Frost & Sullivan, 2020), https://www.researchandmarkets.com/reports/5007728/extended-reality-
applications-in-us-defense.
108
“cognition war.” A cognition war would be the use of automation and AI on the maritime
battlefield, virtually controlled by human beings not physically at sea. As a deterrence tool,
AI can be used to threaten the competitor with the possible existence of naval threats at sea
with “fake injected data” when in reality, no machine is present or has been present.

Humanity is at the edge of information dominance in a revolution driven by

automation and AI. From ancient times, fleets at sea have sought to know what lay beyond
the horizon, where the enemy might lurk. A century ago, the rise of radio and radar
mounted on piloted aircraft gave navies the ability to extend surveillance far beyond
horizons. Ironically, the ultimate effect of AI solving problems and managing information
may be the transformation of human reasoning, intelligence, and decision making. Today,
automation is both extending its reach globally and interpreting the flood of information to
solve problems and create full transparency more than ever before. Truly, automated ISR
is changing human knowledge, perception, reality, and, ultimately, the face of 21st Century
sea power.

109
THIS PAGE INTENTIONALLY LEFT BLANK

110
 APPENDIX

USING AI TOOLS TO FIND AI TALENT

The acceleration for research and development of Artificial Intelligence (AI) has
globalized social networks worldwide. AI social networks thrive in research facilities from
superpower countries like the United States and China can cross-pollinate for research to
advance research in AI software. Top popular AI emerging technologies topics are
Machine Learning (ML), Computer Vision (CV) and Pattern Recognition (PR). ML
enables machines to learn from a task from a previous experience without new
programming by a human; CV uses AI to detect and identify images; and, PR combines
the two in which a computer recognizes patterns based on ML and CV results. 282 The
researchers writing AI papers at academic research facilities can be recruited for their
skills.

The data is from a repository arena within Cornell University’s arXiv database of
research submissions worldwide. The purpose is to provide a potential network analysis of
how AI can be adopted and researched in academia (in terms of computer vision, machine
learning, pattern recognition, etc.), by who and paper topics. This network is a two-mode
data set (author to paper).

The ML, CV, and PR data used in this researched is derived from Cornell
University’s arXiv website. Cornell’s arXiv is a central database and electronic repository
for approved scientific papers with over 10,000 submissions per month and not fully peer
reviewed. research, recognition, and possible funding for future work. 283 The benefit of
such a worldwide, open-source website with AI authors and AI papers is access to trending
or innovative research. A second benefit is to be able to disrupt their network and directly
recruit individuals or a pool of people talented and driven in AI research. Relational ties
and organizations included in this network will encompass the approved papers submitted

282 “8 Best Topics for Research and Thesis in Artificial Intelligence,” GeeksforGeeks: A Computer
Science Portal for Geeks, accessed February 20, 2020, https://www.geeksforgeeks.org/8-best-topics-for-
research-and-thesis-in-artificial-intelligence/.
283 “Artificial Intelligence.”

111
within arXiv database from 2009 to 2019 (Figure 21). The dataset will be bounded by the
following topics: AI bulk, computer vision and pattern recognition, statistical machine
learning, and database.

Figure 21. Cornell University’s ArXiv Researchers: Computer Vision and

Pattern Recognition: Paper to Person 1,999 Nodes, Density,
Clustering Coefficient

Figure 22 represents the group size statistics on the number of groups, isolated groups,
two per group (dyad), three per group (triad), and larger groups of researchers working
together. The algorithm in a computer software called ORA, an analytical tool used for
social network analysis, automatically generates visual representations of interconnected
nodes from a specific dataset input into the software, which in this case was derived from
arXiv. 284

284 “ORA,” Center for Computational Analysis of Social and Organizational Systems (CASOS),
accessed November 11, 2019, http://www.casos.cs.cmu.edu/index.php.
112
 Figure 22. Computer Vision and Pattern Recognition: Density, Clustering
Coefficient

The aggregated sociography for CV and is moderately connected network. The

metrics shows a total of 1,220 groups with 5.4 percent dyad and 11.5 percent triad co-
authorship and 55 percent single authorship, which indicates the majority work
independently on papers. Figure 23 displays the top 10 actors in the network within 10
diverse types of groups and his or her value through the Girvan-Newman algorithm.

Figure 23. Girvan-Newman Clustering Group (Nodes with Highest Internal

Degree)

113
 K-core is a “maximal group of actors, all of whom are connected to some number
(k) of other group members.” 285 According to Naval Postgraduate School Professor Sean
Everton, “Girvan-Newman begins with a connected network and then strategically
removes ties, a process that partitions the network into an increasing number of
clusters.” 286 The key to this algorithm is the notion of edge betweenness, in which the
algorithm estimates the ties, or betweenness centrality of the edges. In other words, it
measures the shortest path between the authors.

I have pulled this group and searched each of their affiliated schools, organizations,
and particular fields of research and each of them are from organizations around the world.
Guillermo Sapiro has over 64,000 citations on Google Scholar and one of his most recent
studies in 2020 on deep neural networks relates to detecting adversaries using influence
and nearest neighbors. Group 1 reveals the top 10 authors or nodes and separate from the
other Girvan-Newman groups. In Table 7, the top 10 AI experts who have written papers
are shown, including associated schools, organizations, and research topics.

Table 7. Group 1: Top 10 AI Experts with an Internal Node Count of 158

285 Sean Everton, Disrupting Dark Networks (Cambridge, MA: Cambridge University Press, 2012),
182.
286 Everton, 195.

114
 By using ORA combined with researching on Google Scholar, this social network
analysis algorithm illuminates hidden talent that the DOD and United States Navy can
leverage to maintain its competitive advantage over other adversaries. In particular, the
DOD can visualize and seek well-connected authors and co-authors who are researching
today on deep neural networks that can or will be used in the future for counter-AI tools or
future work that can be further researched and connected with relevant DOD components.
Figure 24 shows social network analysis using the software ORA and data set derived from
arXiv.

Figure 24. K-Core Subgroup of CV and PR

Figure 25 shows the group statistics and large group statistics of AI researchers
studying topics on computer vision and pattern recognition. The group statistics helps
visualize how many groups and how many groups of two, three or more are collaborating.

Figure 25. K-Core Group Size and Large Group Statistics

115
 However, this data set is limited to mostly one author to one author on a research
paper in which 83 percent of the nodes are tied to one other node.

RESULTS: ANALYSIS OF SPECIFIC NETWORKS

In the Machine Learning Group 1, four main nodes were taken out of the whole
network for further analysis. This group showed the most connections, as depicted in light
blue, blue, lime green, and gold, as shown in Figure 26.

Figure 26. Top Connected AI and Machine Learning Researchers Derived

ArXiv and Visualized through ORA

Guillermo Sapiro (ML, CV; over 64,000 scholarly citations), Yann LeCun (ML,
CV, image compression; over 140,000 citations), Michael Bronstein (geometric deep
learning; over 14,000 citations), and Jonathan Masci (AI, ML, CV; over 5,000 citations)
appear to be the center nodes with the most connections. French mathematician and École
Normale Supérieure Professor Stephane Mallat (over 100,000 citations) is represented by
a center gold node and appears to be a potential bridge between Sapiro’s and LeCun’s
networks, as show in Figure 27. If severed, these two would not be connected any longer.
However, if one wanted to be connected to both of their networks, ORA’s algorithm has
revealed Mallat is the “go-to” person regarding this field of research in AI. The farthest in
116
distance of connected nodes appears to be Rama Chellappa and her network. Mallat has
researched heavily on facial and pattern recognition, specifically mathematics on high
degree approximations and deep convolutional networks for image identification and
classification (see Figure 27).

Figure 27. Social Network Analysis Shows Stephane Mallat as a Bridge

Between LeCun’s and Sapiro’s Networks

In CV and PR, a more well-connected and smaller part of the whole network was
separated out to visually display a few central nodes. Depicted by the pink arrows, the AI
authors Zhouchen Lin, Allen Yang, and Yi Ma are limited in centrality, betweenness,
brokerage potential, clustering coefficient, and so on and so forth. Clearly, this network is
not very dense. Zhouchen Lin is the Vice President of Samsung Research and works at
Peking University in China. Allen Y. Yang is a research expertise in the fields of CV,
robotics, and pattern recognition at Berkeley University as the Chief Scientist, Fung
Institute for Engineering Leadership and Executive Director for the Center of Augmented
Cognition. Yi Ma is a Berkeley professor with an expertise in Artificial Intelligence (AI)

117
Control, Intelligent Systems, and Robotics (CIR) Signal Processing (SP) Computer Vision
Compressive Sensing. These top three nodes show a direct tie between the United States
and China both heavily researching in AI, but with very low out-degree.

However, this specific network reveals heavy international connections. For

instances, Lin is connected to Shuicheng Yan (center left, central node) who is a professor
at National University of Singapore. Yan is connected to a U.S. Johns Hopkins and
biomedical engineer Professor Rene Vidal (Figure 28).

Figure 28. Low Density of CV and PR Researchers, But Well Connected

Internationally

118
 LIST OF REFERENCES

Aerospace Technology. “BeiDou-3 Navigation Satellite System.” Accessed March 7,

2020. https://www.aerospace-technology.com/projects/beidou-3-navigation-
satellite-system/.

Albus, James S. “Outline for a Theory of Intelligence.” IEEE Transactions on Systems,

Man and Cybernetics 21, no. 3 (June 1991): 473–509. https://doi.org/10.1109/
21.97471.

Ali, S. Mahmud. China’s Belt and Road Vision: Geoeconomics and Geopolitics. Cham,
Switzerland: Springer International Publishing, 2020.

Allard, Kenneth. Command, and Control, and the Common Defense. Fort McNair,
Washington, DC: National Defense University, Institute for National Strategic
Studies, 1996.

Allen, Gregory C. Understanding China’s AI Strategy: Clues to Chinese Strategic

Thinking on Artificial Intelligence and National Security. Washington, DC:
Center for New American Security, 2019. https://www.cnas.org/publications/
reports/understanding-chinas-ai-strategy.

Allison, Graham. “Is China Beating America to AI Supremacy?” National Interest,

December 22, 2019. https://nationalinterest.org/feature/china-beating-america-ai-
supremacy-106861.

Amazon Web Services. “What Is Artificial Intelligent?” Amazon Web Services.

Accessed June 5, 2020. https://aws.amazon.com/machine-learning/what-is-ai/.

Anton, Philip S., Megan McKernan, Ken Munson, James G. Kallimani, Alexis Levedahl,
Irv Blickstein, Jeffrey A. Drezner, and Sydne Newberry. Assessing the Use of
Data Analytics in Department of Defense Acquisition. Santa Monica, CA: RAND
Corporation, 2019. https://www.rand.org/pubs/research_briefs/RB10085.html.

Anyoha, Rockwell. “The History of Artificial Intelligence.” SITN Blog (blog), August 28,
2017. http://sitn.hms.harvard.edu/flash/2017/history-artificial-intelligence/.

Arquilla, John. In Athena’s Camp: Preparing for Conflict in the Information Age. Santa
Monica, CA: RAND Corporation, 1997.

———. “Uncle Spam Wants You!” Foreign Policy, February 5, 2013.

https://foreignpolicy.com/2013/02/05/uncle-spam-wants-you/.

Arquilla, John, and Peter Denning. “Automation Will Change Sea Power.” Naval
Institute Proceedings 145, no. 6 (June 2019): 32–36.

119
“Artificial Intelligence.” In Oxford Reference. Accessed June 8, 2020.
https://www.oxfordreference.com/view/10.1093/oi/
authority.20110803095426960.

Assistant Secretary of the Navy for Research, Development, and Acquisition (ASN
RDA). “Welcome to NavalX.” NavalX. Accessed March 14, 2020.
https://www.secnav.navy.mil/agility/Pages/default.aspx.

Atwood, Chandler. “Activity-Based Intelligence: Revolutionizing Military Intelligence

Analysis.” Joint Force Quarterly 77 (April 2015): 24–33.
https://ndupress.ndu.edu/JFQ/Joint-Force-Quarterly-77/.

Axe, David. “How China Is Militarizing the South China Sea with a Ton of Missiles.”
National Interest, March 23, 2020. https://nationalinterest.org/blog/buzz/how-
china-militarizing-south-china-sea-ton-missiles-136297.

———. “Stealth Refuel Drone: Meet the Navy’s New MQ-25 Stingray.” National
Interest, March 12, 2020. https://nationalinterest.org/blog/buzz/stealth-refuel-
drone-meet-navys-new-mq-25-stingray-132302.

Barnaby, Frank. The Automated Battlefield. New York: The Free Press, 1986.

Barreveld, Dirk Jan, ed. Air Crash Investigations: Killing 290 Civilians - The Downing of
Iran Air Flight 655 By the USS Vincennes Kindle Edition. Morrisville, NC: Lulu
Enterprises, Inc., 2016.

Biltgen, Patrick, and Stephen Ryan. Activity-Based Intelligence: Principles and

Applications. Boston: Artech House, 2016.

Blanken, Leo, Philip Swintek, and Justin Davis. “Special Operations as an Innovation
Laboratory.” War on the Rocks, February 25, 2020. https://warontherocks.com/
2020/02/special-operations-as-an-innovation-laboratory/.

Blasko, Dennis. The Chinese Army Today. 2nd ed. Thousand Oaks, CA: Routledge, 2012.

Bottoms, Michael. “SOFWERX: A Smart Factory of Innovation Helping the

Warfighter.” USSOCOM - United States Special Operations Command, February
2, 2018. https://www.socom.mil/Pages/SOFWERX--A-smart-factory-of-
innovation-helping-the-warfighter.aspx.

Bourbakis, Nikolas G., ed. Artificial Intelligence and Automation. Hackensack, NJ:
World Scientific Publishing, 1998. https://doi.org/10.1142/3079.

Burgess, Richard R. “Four Navy Ships Set for Delivery of Newest SSDS Configuration.”
Seapower, January 14, 2020. https://seapowermagazine.org/four-navy-ships-set-
for-delivery-of-newest-ssds-configuration/.

120
Bussert, James C., and Bruce A. Elleman. People’s Liberation Army Navy: Combat
System Technology, 1949–2010. Annapolis, MD: Naval Institute Press, 2011.

Cain, Mary, and Robert Mittman. Diffusion of Innovation in Health Care. Oakland, CA:
California Health Care Foundation, 2002. http://kpworkforce.org/projects/include/
DiffusionofInnovation.pdf.

Campbell, Thomas A. Artificial Intelligence: An Overview of State Initiatives. Evergreen,

CO: FutureGrasp, LLC, 2019. http://www.unicri.it/in_focus/files/Report_AI-
An_Overview_of_State_Initiatives_FutureGrasp_7-23-19.pdf.

Carnegie Mellon University, School of Computer Science. “ORA.” Center for

Computational Analysis of Social and Organizational Systems (CASOS).
Accessed November 11, 2019. http://www.casos.cs.cmu.edu/index.php.

Chandrashekar, S., and N. Ramani. China’s Space Power & Military Strategy – the Role
of the Yaogan Satellites. ISSSP Report No. 02–2018. Bangalore, India:
International Strategic and Security Studies Programme, National Institute of
Advanced Studies, 2018. http://isssp.in/wp-content/uploads/2018/07/Chinas-
Space-Policy_July2018.pdf.

Chatzky, Andrew, and James McBride. “China’s Massive Belt and Road Initiative.”
Council on Foreign Relations, January 28, 2020. https://www.cfr.org/
backgrounder/chinas-massive-belt-and-road-initiative.

Chen, Stephen. “PLA Navy Eyes China’s Deep-Sea Underwater Glider After Successful
Test Shows It Rivals U.S. Vessel.” South China Morning Post, September 1,
2016. https://www.scmp.com/news/china/article/2012137/pla-navy-eyes-chinas-
deep-sea-underwater-glider-after-successful-test.

Chief of Naval Operations Public Affairs. “CNO Holds Video Teleconference with
Chinese Counterpart Vice Adm. Shen Jinlong.” Navy News Service, July 9, 2019.
https://www.navy.mil/submit/display.asp?story_id=110159.

China Institute for Science and Technology Policy. China AI Development Report 2018.
Beijing: China Institute for Science and Technology Policy at Tsinghua
University, 2018. http://www.sppm.tsinghua.edu.cn/eWebEditor/UploadFile/
China_AI_development_report_2018.pdf.

China Power Team. “How Is China Modernizing Its Navy?” China Power, December 17,
2018. https://chinapower.csis.org/china-naval-modernization/.

Choudhary, Mahashreveta. “What Are Various GNSS Systems?” Geospatial World

(blog), November 20, 2019. https://www.geospatialworld.net/blogs/what-are-the-
various-gnss-systems/.

121
Clarivate Analytics. “Web of Science.” Web of Science Group, 2020.
https://clarivate.com/webofsciencegroup/solutions/web-of-science/.

Clausewitz, Carl von. On War. Edited by Michael Howard and Peter Paret. Reprint.
Princeton, NJ: Princeton University Press, 1989.

Coenn, Daniel. Abraham Lincoln: His Words. New York: Barnes & Noble, 1994.

Cohen, Eliot A. Commandos and Politicians: Elite Military Units in Modern

Democracies. Cambridge, MA: Center for International Affairs, Harvard
University, 1978.

Copeland, B.J. “Artificial Intelligence.” In Encyclopedia Britannica, March 24, 2020.

https://www.britannica.com/technology/artificial-intelligence.

Cordesman, Anthony H. China’s New 2019 Defense White Paper: An Open Strategic
Challenge to the United States, But One Which Does Not Have to Lead to
Conflict. Washington, DC: Center for Strategic & International Studies, 2019.
https://csis-prod.s3.amazonaws.com/s3fs-public/publication/
190724_China_2019_Defense.pdf.

Craymer, Lucy. “Virus Grounds a U.S. Aircraft Carrier as Crew Quarantined in Guam.”
Wall Street Journal, April 1, 2020. https://www.wsj.com/articles/virus-grounds-a-
u-s-aircraft-carrier-as-crew-quarantined-in-guam-11585736476.

Daggett, Stephen. Quadrennial Defense Review 2010: Overview and Implications for
National Security Planning. CRS Report No. R41250. Washington, DC:
Congressional Research Service, 2010. https://fas.org/sgp/crs/natsec/R41250.pdf.

Damanpour, Fariborz, and Marguerite Schneider. “Phases of the Adoption of Innovation

in Organizations: Effects of Environment, Organization and Top Managers.”
British Journal of Management 17, no. 3 (September 2006): 215–36.
https://doi.org/10.1111/j.1467-8551.2006.00498.x.

Davis, Zachary, Michael Nacht, and Ronald Lehman, eds. Strategic Latency and World
Power: How Technology Is Changing Our Concepts of Security. Livermore, CA:
Center for Global Security Research, Lawrence Livermore National Laboratory,
2014. https://cgsr.llnl.gov/content/assets/docs/Strategic_Latency.pdf.

Davis, Zachary S. Artificial Intelligence on the Battlefield: An Initial Survey of Potential

Implications for Deterrence Stability and Strategic Surprise. Livermore, CA:
Center for Global Security Research, Lawrence Livermore National Laboratory,
2019. https://cgsr.llnl.gov/content/assets/docs/CGSR-AI_BattlefieldWEB.pdf.

Defense Advanced Research Projects Agency. “AI Next Campaign.” DARPA - Defense
Advanced Research Projects Agency. Accessed March 14, 2020.
https://www.darpa.mil/work-with-us/ai-next-campaign.
122
Defense Innovation Unit. “Accelerating Commercial Technology for National Security.”
Defense Innovation Unit. Accessed March 14, 2020. https://www.diu.mil/.

———. “Defense Innovation Unit: Who We Are/Our Mission.” Defense Innovation

Unit. Accessed April 22, 2020. https://www.diu.mil/about.

Defense Intelligence Agency. China’s Military Power. Washington, DC: Defense

Intelligence Agency, 2019. https://www.dia.mil/Portals/27/Documents/News/
Military%20Power%20Publications/
China_Military_Power_FINAL_5MB_20190103.pdf.

Department of Defense. Autonomy in Weapon Systems. DOD Directive 3000.9.

Washington, DC: Department of Defense, 2012.
https://www.hsdl.org/?abstract&did=726163.

———. DOD Cloud Strategy. Washington, DC: Department of Defense, 2018.

https://media.defense.gov/2019/Feb/04/2002085866/-1/-1/1/DOD-CLOUD-
STRATEGY.PDF.

———. Summary of the 2018 Department of Defense Artificial Intelligence Strategy:

Harnessing AI to Advance Our Security and Prosperity. Washington, DC:
Department of Defense, 2018. https://media.defense.gov/2019/Feb/12/
2002088963/-1/-1/1/SUMMARY-OF-DOD-AI-STRATEGY.PDF.

Department of the Navy. The Commander’s Handbook on the Law of Naval Operations.
Washington, DC: Office of the Chief of Naval Operations, 2017.
https://www.hsdl.org/?view&did=806860.

DIA Public Affairs. “DIA’s Vision of MARS: Decision Advantage for the 21st Century.”
DIA News, May 23, 2019. https://www.dia.mil/News/Articles/Article-View/
Article/1855910/dias-vision-of-mars-decision-advantage-for-the-21st-century/.

Dobb, Maurice. “Economic Backwardness in Historical Perspective: A Book of Essays.

by A. Gerschenkron.” Economic Journal 74, no. 296 (December 1964): 1001–
1003. https://doi.org/10.2307/2228863.

Doornbos, Caitlin. “Freedom-of-Navigation Ops Will Not Dent Beijing’s South China
Sea Claims, Experts Say.” Stars and Stripes, April 4, 2019.
https://www.stripes.com/news/pacific/freedom-of-navigation-ops-will-not-dent-
beijing-s-south-china-sea-claims-experts-say-1.575609.

Drun, Jessica. “China’s Maritime Ambitions: A Sinister String of Pearls or a Benevolent

Silk Road (or Both)?” Center for Advanced China Research Blog (blog),
December 5, 2017. https://www.ccpwatch.org/single-post/2017/12/05/
China%E2%80%99s-Maritime-Ambitions-a-Sinister-String-of-Pearls-or-a-
Benevolent-Silk-Road-or-Both.

123
Engineers Australia. Jindalee: Over-the-Horizon Radar. Barton ACT: Engineers
Australia Engineering Heritage Victoria, 2016.
https://portal.engineersaustralia.org.au/system/files/engineering-heritage-australia/
nomination-title/Jindalee.Nomination.V14.June%202016.pdf.

Erickson, Andrew, and Gabe Collins. “China’s Real Blue Navy.” The Diplomat, August
30, 2012. https://thediplomat.com/2012/08/chinas-not-so-scary-navy/.

Everton, Sean. Disrupting Dark Networks. Cambridge, MA: Cambridge University Press,
2012.

Faggella, Daniel. “What Is Artificial Intelligence? An Informed Definition.” Emerj,

December 21, 2018. https://emerj.com/ai-glossary-terms/what-is-artificial-
intelligence-an-informed-definition/.

FAS Strategic Air Defense Systems. “AN/FPS-118 Over-The-Horizon-Backscatter

(OTH-B) Radar,” June 29, 1999. https://fas.org/nuke/guide/usa/airdef/an-fps-
118.htm.

Fathom5, and Booz Allen Hamilton. “HACKtheMACHINE 2019.”

HACKtheMACHINE, June 2019. https://www.hackthemachine.ai/.

Feigenbaum, Edward A. “Expert Systems in the 1980s.” Stanford, CA: Stanford

University, 1980. https://stacks.stanford.edu/file/druid:vf069sz9374/
vf069sz9374.pdf.

FireEye Mandiant. M-Trends 2019. Milpitas, CA: FireEye Inc., 2019.

https://content.fireeye.com/m-trends/rpt-m-trends-2019.

Frost & Sullivan. Extended Reality Applications in U.S. Defense Training, 2020. San
Antonio, TX: Frost & Sullivan, 2020. https://www.researchandmarkets.com/
reports/5007728/extended-reality-applications-in-us-defense.

Fuller, John Frederick. Armament and History: The Influence of Armament on History
from the Dawn of Classical Warfare to the End of the Second World War.
Cambridge, MA: Da Capo Press, 1998.

Gallivan, Michael J. “Organizational Adoption and Assimilation of Complex

Technological Innovations: Development and Application of a New Framework.”
The DATABASE for Advances in Information Systems 32, no. 3 (Summer 2001):
51–85. https://doi.org/10.1145/506724.506729.

Garner, Robert. “Early Popular Computers, 1950 - 1970.” Engineering and Technology
History Wiki, January 9, 2015. https://ethw.org/
Early_Popular_Computers,_1950_-_1970.

124
Garside, M. “Major Countries in Silicon Production from 2014 to 2019 (in 1,000 Metric
Tons)*.” Statista, February 12, 2020. https://www.statista.com/statistics/268108/
world-silicon-production-by-country/.

GeeksforGeeks. “8 Best Topics for Research and Thesis in Artificial Intelligence.”

GeeksforGeeks: A Computer Science Portal for Geeks. Accessed February 20,
2020. https://www.geeksforgeeks.org/8-best-topics-for-research-and-thesis-in-
artificial-intelligence/.

Gerschenkron, Alexander. Economic Backwardness in Historical Perspective.

Cambridge, MA: Harvard University Press, 1962.

Girasa, Rosario. Artificial Intelligence as a Disruptive Technology: Economic

Transformation and Government Regulation. New York: Springer International
Publishing, 2020.

Goertzel, Ben, and Cassio Pennachin. Artificial General Intelligence. New York:
Springer, 2007.

Goldstein, Lyle J. “How China Sees the U.S. Navy’s Sea Hunter Drone.” National
Interest, January 31, 2017. https://nationalinterest.org/feature/how-china-sees-the-
us-navys-sea-hunter-drone-19264.

———. “Robot War in the South China Sea?” National Interest, October 11, 2018.
https://nationalinterest.org/feature/robot-war-south-china-sea-33146.

Habib, Maki K. Revolutionizing Education in the Age of AI and Machine Learning.

Hershey, PA: IGI Global, 2019.

International Society of Automation. “What Is Automation?” International Society of

Automation. Accessed January 28, 2020. https://www.isa.org/about-isa/what-is-
automation/.

Jing, Meng, and Sarah Dai. “China Recruits Baidu, Alibaba and Tencent to AI ‘National
Team.’” South China Morning Post, November 21, 2017. https://www.scmp.com/
tech/china-tech/article/2120913/china-recruits-baidu-alibaba-and-tencent-ai-
national-team.

Johns Hopkins University. “COVID-19 Dashboard by the Center for Systems Science
and Engineering (CSSE) at Johns Hopkins University (JHU).” Johns Hopkins
Coronavirus Resource Center. Accessed May 26, 2020.
https://coronavirus.jhu.edu/map.html.

Johnson, Jesse. “U.S. Warship Challenges Chinese Claims in Disputed South China Sea.”
Japan Times, September 14, 2019. https://www.japantimes.co.jp/news/2019/09/
14/asia-pacific/u-s-warship-challenges-chinese-claims-disputed-south-china-
sea/#.Xhd7aMhKj_M.
125
Joint Chiefs of Staff. Cyberspace Operations. Joint Pub 3–12. Washington, DC: Joint
Chiefs of Staff, 2018. https://www.jcs.mil/Portals/36/Documents/Doctrine/pubs/
jp3_12.pdf.

———. Department of Defense Dictionary of Military and Associated Terms. Joint Pub.
1–02. Washington, DC: Joint Chiefs of Staff, 2016. https://fas.org/irp/doddir/dod/
jp1_02.pdf.

———. Joint Intelligence Support to Military Operations. Joint Pub 2–01. Washington,
DC: Joint Chiefs of Staff, 1996. http:// www.hsdl.org/?view&did=3737.

Kania, Elsa B. “The AI Titans’ Security Dilemmas.” Governance in an Emerging New

World, no. 218 (October 29, 2018). https://www.hoover.org/research/ai-titans.

Kauchak, Marty. “Naval Channel SNA 2020: ‘A Whole New Heat and Light.’” Mönch
Publishing Group, January 15, 2020. https://monch.com/mpg/news/naval-channel/
6606-sna-2020-lockheed-ssds-acb-20.html.

Koch, Gary. “NORAD-USNORTHCOM C2 Systems Overview for AFCEA JC2C:

Evolving the GCCS FoS.” Presented at the AFCEA Course, Fairfax, VA, March
25, 2010. https://www.slideshare.net/gdkochjr/afcea-jc2c-gccs-presentation.

Koleski, Katherine. The 13th Five-Year Plan. Washington, DC: U.S.-China Economic
and Security Review Commission, 2017. https://www.uscc.gov/sites/default/files/
Research/The%2013th%20Five-Year%20Plan_Final_2.14.17_Updated%20%
28002%29.pdf.

Kroll, Joshua. “Classifying Machines by Kinds of Learning: Automation.” Presented at

the Harnessing AI Course, September 30, 2019. https://nps.edu/web/ai-
consortium/harnessing-ai-course.

LaCamera, Jr., J. Jerry. “Rapid Acquisition - The Challenge to Accelerate.” Herndon,

VA: National Defense Industrial Association, 2019. https://www.ndia.org/-
/media/sites/ndia/divisions/ipmd/2019-04-meeting/207-middle-tier-of-acquisition-
lacamera-190508215620.ashx?la=en.

Larter, David B. “With $294 Million in Contracts, the U.S. Navy Keeps Its Promise to
Upgrade Cruisers.” Defense News, August 24, 2018.

Lazar, Seth. “War.” In Stanford Encyclopedia of Philosophy, edited by Edward N. Zalta.

Stanford, CA: Metaphysics Research Lab, Stanford University, 2020.
https://plato.stanford.edu/entries/war/.

Lee, Kai-fu. AI Superpowers: China, Silicon Valley, and the New World Order. Boston:
Houghton Mifflin Harcourt, 2018.

126
Lei, Zhao. “CNS Liaoning Leads Live-Fire Drill.” China Daily, December 16, 2016.
https://www.chinadaily.com.cn/china/2016-12/16/content_27685144.htm.

LeoLabs. “Three Ways to Use LeoLabs.” LeoLabs Platform for Operators and
Developers. Accessed March 8, 2020. https://platform.leolabs.space/.

Liedtka, Jeanne. “In Defense of Strategy as Design.” California Management Review 42,
no. 3 (Spring 2000): 8–30. https://doi.org/10.2307/41166040.

Lilley, James, and David L. Shambaugh. China’s Military Faces the Future. Abingdon,
UK: Routledge, 1999.

Lindgren, David. “Global Remittances and Space-Based Cryptocurrencies: A

Transformational Opportunity for the Post-2030 Agenda.” In Post 2030-Agenda
and the Role of Space, edited by Annette Froehlich, 11–25. Cham, Switzerland:
Springer International Publishing, 2018. https://www.springer.com/gp/book/
9783319789538.

Lockheed Martin. “Future Warfighting Capabilities Secure with Lockheed Martin

Engineering Experience.” Lockheed Martin Newsroom, August 20, 2019.
https://news.lockheedmartin.com/future-warfighting-capabilities-secure-with-
lockheed-martin-engineering-experience.

———. “Lockheed Martin Receives $150 Million Contract to Deliver Integrated High
Energy Laser Weapon Systems to U.S. Navy.” Lockheed Martin Newsroom,
March 1, 2018. https://news.lockheedmartin.com/2018-03-01-Lockheed-Martin-
Receives-150-Million-Contract-to-Deliver-Integrated-High-Energy-Laser-
Weapon-Systems-to-U-S-Navy.

Lopez, C. Todd. “DOD Adopts 5 Principles of Artificial Intelligence Ethics.” DOD

News, February 25, 2020. https://www.defense.gov/Explore/News/Article/
Article/2094085/dod-adopts-5-principles-of-artificial-intelligence-ethics/.

Lowenthal, Mark M., and Robert M. Clark, eds. The Five Disciplines of Intelligence
Collection. Thousand Oaks, CA: CQ Press, 2015.

Lye, Harry. “US Navy Deploys ODIN Counter-UAS Laser.” Naval Technology, February
20, 2020. https://www.naval-technology.com/news/us-navy-deploys-odin-
counter-uas-laser/.

Lynch, Shana. “Andrew Ng: Why Ai Is the New Electricity.” Stanford News, March 14,
2017. https://news.stanford.edu/thedish/2017/03/14/andrew-ng-why-ai-is-the-
new-electricity/.

MacFarquhar, Roderick. The Origins of the Cultural Revolution: Volume II, the Great
Leap Forward 1958–1960. 2nd ed. New York: Columbia University Press, 1983.

127
Mahan, Alfred Thayer. The Influence of Sea Power Upon History, 1660–1783. New
York: Dover Publications, 1987.

Marine Traffic. “Marine Traffic Live Map.” Accessed March 5, 2020.

https://www.marinetraffic.com.

Marr, Bernard. “The Key Definitions of Artificial Intelligence (AI) That Explain Its
Importance.” Forbes, February 14, 2018. https://www.forbes.com/sites/
bernardmarr/2018/02/14/the-key-definitions-of-artificial-intelligence-ai-that-
explain-its-importance/#532664514f5d.

McAfee, Heather. “Aery Aviation, LLC Wins Naval Special Warfare Contract.” PR
Newswire, February 20, 2020. https://www.prnewswire.com/news-releases/aery-
aviation-llc-wins-naval-special-warfare-contract-301008746.html.

McCamley, Nick. Cold War Secret Nuclear Bunkers: The Passive Defence of the Western
World During the Cold War. Barnsley, South Yorkshire: Pen and Sword Military,
2007.

McCarthy, John, Marvin L. Minsky, Nathaniel Rochester, and Claude E. Shannon. “A

Proposal for the Dartmouth Summer Research Project on Artificial Intelligence.”
Hanover, NH: Dartmouth College, August 31, 1955. http://jmc.stanford.edu/
articles/dartmouth.html.

McGregor, Richard. “5 Myths about the Chinese Communist Party.” Foreign Policy,
January 3, 2011. https://foreignpolicy.com/2011/01/03/5-myths-about-the-
chinese-communist-party/.

McRaven, William H. Spec Ops: Case Studies in Special Operations Warfare: Theory &
Practice. Novato, CA: Presidio Press, 1996.

Mendel, Peter, Lisa S. Meredith, Michael Schoenbaum, Cathy D. Sherbourne, and

Kenneth B. Wells. “Interventions in Organizational and Community Context: A
Framework for Building Evidence on Dissemination and Implementation in
Health Services Research.” Administration and Policy in Mental Health 35, no.
1–2 (March 2008): 21–37. https://doi.org/10.1007/s10488-007-0144-9.

“Metcalfe’s Law.” In Cambridge Dictionary. Accessed April 2, 2020.

https://dictionary.cambridge.org/us/dictionary/english/metcalfe-s-law.

Mish, Frederick. “Adoption.” In Merriam-Webster’s Collegiate Dictionary. Merriam-

Webster Inc., 2004.

Missile Defense Advocacy Alliance. “What Is Anti-Access/Area Denial (A2/AD)?”

Missile Defense Advocacy Alliance. Accessed February 27, 2020.
https://missiledefenseadvocacy.org/missile-threat-and-proliferation/todays-
missile-threat/china-anti-access-area-denial-coming-soon/.
128
Modley, Thomas B. “SECNAV Vector 11.” Official memorandum. Washington, DC:
Secretary of the Navy, February 14, 2020.

“Moore’s Law.” In Encyclopaedia Britannica, July 21, 2011.

https://www.britannica.com/technology/Moores-law.

Morrison, Elting. Men, Machines, and Modern Times. Cambridge, MA: MIT Press, 1966.

Muro, Mark, Robert Maxim, and Jacob Whiton. Automation and Artificial Intelligence:
How Machines Are Changing People and Places? Washington, DC: Brookings
Institute: Metropolitan Policy Program, 2019. https://www.brookings.edu/wp-
content/uploads/2019/01/2019.01_BrookingsMetro_Automation-
AI_Report_Muro-Maxim-Whiton-FINAL-version.pdf.

National Security & Technology Council. The Networking & Information Technology
Research and Development Program: Supplement to the President’s FY2020
Budget. Washington, DC: National Security & Technology Council, 2019.
https://www.whitehouse.gov/wp-content/uploads/2019/09/FY2020-NITRD-AI-
RD-Budget-September-2019.pdf.

Naval Reserve Intelligence Program. “Module 8—Intelligence Automated Data

Processing (ADP) Systems: L. Gale Lite.” In Ready-for-Sea: Modular Course &
Handbook. Naval Air Station North Island, CA: Naval Reserve Intelligence
Program, 1999. https://fas.org/irp/doddir/navy/rfs/part08.htm.

Naval War College Library. “Intelligence Studies: Types of Intelligence Collection.”

Intelligence Studies Research Guide, March 5, 2020. https://usnwc.libguides.com/
c.php?g=494120&p=3381426.

NAVSEA Office of Corporate Communication. “Navy Leverages Workforce; Delivers

C-ISR Capability Rapidly to Surface Fleet.” Naval Sea Systems Command,
February 20, 2020. https://www.navsea.navy.mil/Media/News/
SavedNewsModule/Article/2089079/navy-leverages-workforce-delivers-c-isr-
capability-rapidly-to-surface-fleet/.

Newell, Allen, and Herb Simon. “Computer Science as Empirical Inquiry: Symbols and
Search.” Communications of the ACM 19, no. 3 (March 1976): 113–26.
https://doi.org/10.1145/360018.360022.

Nezda, Christine. “Machine Learning for Patterns of Life.” Technology Today, no. 1
(2018): 26–31. https://www.raytheon.com/sites/default/files/technology-today/
2018/issue1/wp-content/uploads/2018/08/
Raytheon_TechnologyToday_Issue1_2018.pdf.

Nilsson, Nils J. “Edward A (‘Ed’) Feigenbaum: United States - 1994.” Association for
Computing Machinery, A.M. Turing Award, 2019. https://amturing.acm.org/
award_winners/feigenbaum_4167235.cfm.
129
Office of Naval Intelligence. China’s Navy 2007. Washington, DC: Office of Naval
Intelligence, 2007. https://fas.org/irp/agency/oni/chinanavy2007.pdf.

———. The PLA Navy: New Capabilities and Missions for the 21st Century.
Washington, DC: Office of Naval Intelligence, 2015. https://www.oni.navy.mil/
Portals/12/Intel%20agencies/China_Media/
2015_PLA_NAVY_PUB_Print.pdf?ver=2015-12-02-081247-687.

Office of Science and Technology Policy. Summary of the 2018 White House Summit on
Artificial Intelligence for American Industry. Washington, DC: White House,
2018. https://www.whitehouse.gov/wp-content/uploads/2018/05/Summary-
Report-of-White-House-AI-Summit.pdf.

Office of the Secretary of Defense. Quadrennial Defense Review. Washington, DC:

Department of Defense, 2010.

Office of Undersecretary of Defense (Comptroller)/Chief Financial Officer. Defense

Budget Overview: United States Department of Defense Fiscal Year 2020 Budget
Request. Washington, DC: Department of Defense, 2019.
https://comptroller.defense.gov/Portals/45/Documents/defbudget/fy2020/
fy2020_Budget_Request_Overview_Book.pdf.

O’Rourke, Ronald. China Naval Modernization: Implications for U.S. Navy

Capabilities—Background and Issues for Congress. CRS Report No. RL33153.
Washington, DC: Congressional Research Service, 2020.
https://crsreports.congress.gov/product/pdf/RL/RL33153/233.

———. Navy Aegis Ballistic Missile Defense (BMD) Program: Background and Issues
for Congress. CRS Report No. RL33745. Washington, DC: Congressional
Research Service, 2018. https://crsreports.congress.gov/product/pdf/RL/RL33745/
189.

Owens, William A. Lifting the Fog of War. Baltimore: Johns Hopkins University Press,
2001.

Page, Jeremy. “China Sees Itself at Center of New Asian Order: Beijing Builds Roads,
Pipelines, Railways and Ports to Bind Itself to Region.” Wall Street Journal,
November 9, 2014. https://www.wsj.com/articles/chinas-new-trade-routes-center-
it-on-geopolitical-map-1415559290.

Pickrell, Ryan. “A Chinese Warship Is Believed to Have Been Sent to Spy on War
Games Between the U.S., Australia, and Japan.” Business Insider, July 7, 2019.
https://www.businessinsider.com/australia-tracks-chinese-spy-ship-watching-
talisman-saber-war-games-2019-7.

130
“Protocol Additional to the Geneva Convention of August 12, 1949.” International
Committee of the Red Cross, May 2010. https://www.icrc.org/en/doc/assets/files/
other/icrc_002_0321.pdf.

Quindazzi, Mike. “Artificial Intelligence and Robotics 2017: Leveraging Artificial

Intelligence and Robotics for Sustainable Growth.” ASSOCHAM India, April 4,
2017, 8. https://www.slideshare.net/MikeQuindazzi/artificial-intelligence-and-
robotics-in-2017.

———. “Human Decisions Supplemented by #ArtificialIntelligence along the #AI

Continuum; 1-#augmented, 2-#assisted, 3-#autonomous. #pwc.” Twitter, March
22, 2017. https://twitter.com/mikequindazzi/status/844764424917286912.

Richardson, John. A Design for Maritime Superiority. Version 2.0. Washington, DC:
Department of the Navy, 2018. https://www.navy. mil/navydata/people/cno/
Richardson/Resource/Design_2.0.pdf.

Rielage, Dale C. “Chinese Navy Trains and Takes Risks.” Naval Institute Proceedings
142, no. 5 (May 2016). https://www.usni.org/magazines/proceedings/2016/may/
chinese-navy-trains-and-takes-risks.

Robinson, Lori. “The Evolution of Artificial Intelligence and Future of National

Security.” National Interest, March 13, 2020. https://nationalinterest.org/feature/
evolution-artificial-intelligence-and-future-national-security-133032.

Roff, Heather M., and David Danks. “‘Trust but Verify’: The Difficulty of Trusting
Autonomous Weapons Systems.” Journal of Military Ethics 17, no. 1 (2018): 2–
20. https://doi.org/10.1080/15027570.2018.1481907.

RT World News. “‘Whoever Leads in AI Will Rule the World’: Putin to Russian
Children on Knowledge Day,” September 1, 2017. https://www.rt.com/news/
401731-ai-rule-world-putin/.

Russell, Stuart, and Peter Norvig. Artificial Intelligence: A Modern Approach.

Englewood Cliffs, NJ: Prentice Hall, 1995.

Sapio, Flora, Weiming Chen, and Adrian Lo, trans. A New Generation of Artificial
Intelligence Development Plan. State Council Document [2017] No. 35. Beijing:
Chinese State Council, 2017. https://flia.org/notice-state-council-issuing-new-
generation-artificial-intelligence-development-plan/.

Scharre, Paul, Anthony Cho, and Eric Schmidt. “Eric Schmidt Keynote Address at the
Center for a New American Security Artificial Intelligence and Global Security
Summit.” Washington, DC: Center for a New American Security, November 17,
2017.

131
Schmidt, Eric. Promoting DOD’s Culture of Innovation: Testimony before the
Committee on Armed Services (2018). https://armedservices.house.gov/2018/4/
promoting-dod-s-culture-of-innovation.

Scholoer Consulting Group. “What Is Artificial Intelligence?” Understanding Artificial

Intelligence. Accessed January 27, 2020. http://www.schloerconsulting.com/
understanding-artificial-intelligence.

Schwieterman, Joseph P., and Lauren A. Fischer. “The S-Curve of Technological

Adoption: Mobile Communication Devices on Commuter Trains in the Chicago
Region, 2010–2015.” Journal of Public Transportation 20, no. 2 (2017): 1–18.
https://doi.org/10.5038/2375-0901.20.2.1.

Seidel, Jamie. “Beijing Releases 12 ‘Haiya’ Sea-Wing Surveillance Drones into the
Disputed South China Sea.” News Corp Australia Network, July 28, 2017.
https://www.news.com.au/world/beijing-releases-12-haiya-seawing-surveillance-
drones-into-the-disputed-south-china-sea/news-story/
16bd0faf5195f31fdc7c3afeb5b826c2.

Sheridan, Thomas, and William Verplank. Human and Computer Control of Undersea
Teleoperators. Cambridge, MA: Massachusetts Institute of Technology, 1978.

Shore, Zachary. Blunder: Why Smart People Make Bad Decisions. New York:
Bloomsbury USA, 2015.

Silver, David, Thomas Hubert, Julian Schrittwieser, Ioannis Antonoglou, Matthew Lai,
Arthur Guez, Marc Lanctot et al. “A General Reinforcement Learning Algorithm
That Masters Chess, Shogi, and Go Through Self-Play.” Science 362, no. 6419
(December 2018): 140–44. https://doi.org/10.1126/science.aar6404.

Smith, Phillip. “Oceanalpha Debuts Maritime Drone Swarm.” Drone Below, June 6,
2018. https://dronebelow.com/2018/06/06/oceanalphas-debuts-maritime-drone-
swarm/.

Snowden, Dave, and Cynthia F Kurtz. “The New Dynamics of Strategy: Sense-Making in
a Complex and Complicated World.” IBM Systems Journal 42, no. 3 (2003): 462–
83. http://alumni.media.mit.edu/~brooks/storybiz/kurtz.pdf.

State Council. “New Generation Artificial Intelligence Development Plan [国务院关于

印发新一代人工智能发展规划的通知],” July 8, 2017. ttp://www.gov.cn/
zhengce/content/2017- 07/20/content_5211996.htm.

Strawser, Bradley J., and David Whetham. Killing by Remote Control: The Ethics of an
Unmanned Military. Oxford, UK: Oxford University Press, 2013.

Tate, Andrew. “China Integrates Long-Range Surveillance Capabilities.” Jane’s

Intelligence Review, November 1, 2017. Janes.
132
Threston, Joseph T. “The AEGIS Weapon System.” Naval Engineers Journal 121, no. 3
(October 2009): 85–108. https://doi.org/10.1111/j.1559-3584.2009.00205.x.

Till, Geoffrey. Seapower: A Guide for the Twenty-First Century. 2nd ed. Abingdon, UK:
Routledge, 2009.

Trump, Donald J. “Maintaining American Leadership in Artificial Intelligence: Executive

Order 13859.” Federal Register 84, no. 31 (February 14, 2019): 3967–72.
https://www.federalregister.gov/documents/2019/02/14/2019-02544/maintaining-
american-leadership-in-artificial-intelligence.

———. National Cyber Strategy of the United States. Washington, DC: White House,
2018. https://www.whitehouse.gov/wp-content/uploads/2018/09/National-Cyber-
Strategy.pdf.

Tse-tung, Mao. Problems of Strategy in China’s Revolutionary War. Selected Works of

Mao Tse-Tung. Peking: Foreign Language Press, 1936. https://www.marxists.org/
reference/archive/mao/selected-works/volume-1/mswv1_12.htm.

Tucker, Patrick. “Pentagon to Adopt Detailed Principles for Using AI.” Defense One,
February 18, 2020. https://www.defenseone.com/technology/2020/02/pentagon-
adopt-detailed-principles-using-ai/163185/.

Tzu, Sun. Art of War. Translated by Samuel B. Griffith. Oxford, UK: Clarendon Press,
1964. https://suntzusaid.com/book/3/18.

———. Sun Tzu on the Art of War: XIII. The Use of Spies. Translated by Lionel Giles,
1994. https://fas.org/man/artofwar.htm#13.

U.S. Census Bureau. “U.S. and World Population Clock.” United States Census Bureau,
April 3, 2020. https://www.census.gov/popclock/.

U.S. Navy. “Status of the Navy,” April 3, 2020. https://www.navy.mil/navydata/

nav_legacy.asp?id=146.

U.S. Navy. “US Navy Maritime ISR Roadmap: Industry Day.” Presented at the 2010
Navy Information Dominance Industry Day, Heritage Conference Center,
Chantilly, VA, June 22, 2010. https://www.afcea.org/mission/intel/documents/
IndustryDay-Kraft.pdf.

U.S. Navy Information Dominance Corps. U.S. Navy Information Dominance Roadmap
2013–2028. Washington, DC: Department of the Navy, 2013.
https://defenseinnovationmarketplace.dtic.mil/wp-content/uploads/2018/02/
Information_Dominance_Roadmap_March_2013.pdf.

Wang, Pei. “On Defining Artificial Intelligence.” Journal of Artificial General

Intelligence 10, no. 2 (2019): 1–37. https://doi.org/10.2478/jagi-2019-0002.
133
Wang, Pei, and Ben Goertzel. “Introduction: Aspects of Artificial General Intelligence.”
In Proceedings of the 2007 Conference on Advances in Artificial General
Intelligence: Concepts, Architectures and Algorithms, edited by Ben Goertzel and
Pei Wang, 1–16. Amsterdam: IOS Press, 2007.

Wang, Pei, Kai Liu, and Quinn Dougherty. “Conceptions of Artificial Intelligence and
Singularity.” Information 9, no. 4 (2018): 1–15. https://doi.org/10.3390/
info9040079.

Werner, Ben. “Aegis Combat System Baseline 10 Set to IOC in 2023.” USNI News,
January 15, 2019. https://news.usni.org/2019/01/15/40397.

Wilson, James M., Marat G. Polyak, Jane W. Blake, and Jeff Collmann. “A Heuristic
Indication and Warning Staging Model for Detection and Assessment of
Biological Events.” Journal of the American Medical Informatics Association 15,
no. 2 (April 2008): 158–71. https://doi.org/10.1197/jamia.M2558.

Winstanley, Graham, ed. Artificial Intelligence in Engineering. Tiptree, Essex, UK:

Courier International Ltd, 1991.

Wisdom, Jennifer P., Ka Ho Brian Chor, Kimberly E. Hoagwood, and Sarah M.

Horowitz. “Innovation Adoption: A Review of Theories and Constructs.”
Administration and Policy in Mental Health 41, no. 4 (2014): 480–502.
https://doi.org/10.1007/s10488-013-0486-4.

Worldometer. “Current Population Growth.” Accessed February 6, 2020.

https://www.worldometers.info/world-population/.

Wright, Nicholas. “How Artificial Intelligence Will Reshape the Global Orde.” Foreign
Affairs, July 10, 2018. https://www.foreignaffairs.com/articles/world/2018-07-10/
how-artificial-intelligence-will-reshape-global-order.

134
 INITIAL DISTRIBUTION LIST

1. Defense Technical Information Center

Ft. Belvoir, Virginia

2. Dudley Knox Library

Naval Postgraduate School
Monterey, California

135