12. 20th Century DOD/IC Dominance
• We owned & deployed the most advanced technology
• R&D & contractors had the smartest domain experts
• We could design and manufacture the best systems
18. The Offset Strategy
• 1950’s Eisenhower
– mass, mechanized warfare to nuclear deterrence
• 1970’s Perry: stealth, precision strike,
– Technology as a “force multiplier”
– battlefield information and communications systems,
intelligence systems, positioning and navigation
capabilities
• 2014 Hagel
– robotics and system autonomy, miniaturization, big
data, and advanced manufacturing and innovation
19. The Offset Strategy
• 1950’s Eisenhower
– mass, mechanized warfare to nuclear deterrence
• 1970’s Perry: stealth, precision strike,
– Technology as a “force multiplier”
– battlefield information and communications systems,
intelligence systems, positioning and navigation
capabilities
• 2014 Hagel
– robotics and system autonomy, miniaturization, big
data, and advanced manufacturing and innovation
Adversaries aren’t waiting for our
3rd offset strategy
27. Three Horizons of Innovation
Source: Baghai, Coley, White
Mature Business
our established capabilities
Rapidly Growing Business Emerging Business
28. Three Horizons of Innovation
Source: modified Baghai, Coley, White
programs
capabilities
29. Three Horizons of Innovation
Known
Unknown
Partially Known
Level of innovation is defined by whether
the Business model is being executed, extended or explored!
Execute
Explore
Extend
30. Three Horizons of Innovation
Existing Mission Model:
Process Innovation
Execute Core Mission
Known
31. Three Horizons of Innovation
Existing Business Model:
Process Innovation
Execute
Known
Partially Known
New Opportunities via
Mission Model Innovation
Extends Core Business
32. Three Horizons of Innovation
Existing Business Model:
Process Innovation
Execute
New Opportunities via
Business Model Innovation
Execute/Search
Known
Unknown
Partially known
New/Disruptive
Mission Model
Explores
35. Return on Investment by Horizon
Known
Unknown
Partially known
ROI 1-3 years
• Improve
• Partner
• Acquire
ROI 4-6 years
• Extend
• Invest
• Partner
• Acquire
ROI 4-10 years
• Incubate
• Invent
• Invest
• Acquire
Evangelos Simoudis/Steve Blank
36. Horizon 1: Process and Procedure =
Profit
KPIs
HR
Legal
Finance
Bonus
Known
Unknown
Partially known
Process and
Procedures
37. Horizon 3: Process and Procedure =
Death
KPIs
HR
Legal
Finance
Bonus
Product/Market Fit
Pivots
Agile
Risk
Large Reward
Known
Unknown
Partially known
Risk
Process and
Procedures
38. New Ventures/Startups are Horizon 3
Known
Unknown
Partially known
• Improve
• Partner
• Acquire
• Extend
• Invest
• Partner
• Acquire
ROI 1-7 yearsEvangelos Simoudis/Steve Blank
39. Startups are Horizon 3
Not burdened by:
• Existing Customers, Channels, Inventory
• Process, Procedures, KPI’s
Everyone focused on a single mission and survival
40. Four Startup Principles
• Money matters
• Performance counts
• Time is the enemy
• Your investors eventually want liquidity
43. Lean = 3 parts
Business Model Canvas
Part 1
Customers
Channels
Customer
Relationships
Revenue Model
Value
Proposition
Activities
Resources
Partners
Costs
Source: Alexander Osterwalder- Business Model Generation
44. Business Model Canvas = hypotheses of
how you create and deliver value for the company
and its customers
Part 1
Source: Alexander Osterwalder- Business Model Generation
Beneficiaries
Deployment
Buy-
in/Support
Mission Achievement
Value
Proposition
Activities
Resources
Partners
Costs
47. Mission Model Canvas
Source: Alexander Osterwalder- Business Model Generation
Customer
Segments
Channel
Customer
Relationships
Revenue Streams
Value
Proposition
Activities
Resources
Partners
Costs
how does the
team get “Buy-
In” from all the
beneficiaries?
How will we deploy
the product to
widespread use?
What constitutes a
successful
deployment?
Who are our
most important
customers?
Stakeholders?
What are their
pains/gains?
What job do
they want us to
get done for
them>
How are we
solving each
customers
pains/gains?
How?
What
product/servic
e features
match their
needs?
What key activities
do we need to be
expert in?
What key resources
do we need to own
or acquire?
Financial? Human?
Who are our key
partners?
Suppliers?
What are we
getting from
them?
Giving them?
What is the Budget/Cost? What is the revenue model?
What are the pricing tactics?
48. 2. Test Hypotheses
• Frame Hypotheses
• Test Hypotheses
Business Model
Customer Development
Customer Development is how you search for the model
52. 3. Build Incrementally & Iteratively
• Frame Hypotheses
• Test Hypotheses
• Build the product
incrementally &
Iteratively
Business Model
Customer Development
Agile Engineering
53. The Minimum Viable Product (MVP)
• Smallest feature set that gets you the most …
- learning, feedback, failure, orders, …
- incremental and iterative
• It is not a prototype
• It is not a deployable version with the fewest features
• It is what enables a test of a hypothesis
• It may be a drawing, a slide, a wireframe, clickable
workflow, etc…
54. The Pivot
• Definition: A substantive change to one or more of the
business model canvas components
• Iteration without crisis
• Fast, agile and opportunistic
• Weeks and $100K
55. Pivot Cycle Time Matters
• Speed of cycle minimizes cash needs
• Minimum feature set speeds up cycle time
• Near instantaneous customer feedback drives feature set
Customer
Discovery
Customer
Validation
Company
Building
Customer
Creation
ExecutionSearch
Pivot
57. Lean Gets Theory
Customer Development
2003
Blank
Agile Engineering
2011
Ries
Business Model Canvas
2010
Osterwalder
HBR Cover
2013
Blank
58. Lean Gets Practice
MS&E 297: “Hacking for Defense”: Solving National Security
issues with the Lean Launchpad
In a crisis, national security initiatives move at the speed of a startup yet in
peacetime they default to decades-long acquisition and procurement cycles. Startups
operate with continual speed and urgency 24/7. Over the last few years they’ve
learned how to be not only fast, but extremely efficient with resources and time using
lean startup methodologies.
In this class student teams will take actual national security problems and learn how
to apply “Lean Startup” principles, ("business model canvas," "customer
development," and "agile engineering”) to discover and validate customer needs and
to continually build iterative prototypes to test whether they understood the problem
and solution. Teams take a hands-on approach requiring close engagement with actual
military, Department of Defense and other government agency end-users.
Team applications required in February. Limited enrollment. Course builds on
concepts introduced in MS&E 477.
Terms: Spr | Units: 3-4 | Grading: Letter (ABCD/NP)
Instructors: Blank, S. (PI) ; Byers, T. (PI) ; Felter, J. (PI)
2015-2016 Spring
• MS&E 297 | 4 units | Class # 47395 | Section 01 | Grading: Letter (ABCD/NP) | LEC
• 03/28/2016 - 06/01/2016 - with Blank, S. (PI); Byers, T. (PI); Felter, J. (PI)
Lean LaunchPad
For Students
2011
1250+ teams
Taught in 75
Universities
800+ teams
Taught by 53
Universities
I-Corps
For SBIR/STTR
2012
I-Corps
For Life Sciences
2014
I-Corps
For NSA
2015
~250,000 on-line
students
Udacity.com
Hacking for
Defense
2016
62. Live Tactical Threat Toolkit (LTTT)
Surgical-grade EOD, remotely
We seek to create a stream of accurate information and insights for real-time defeat of improvised
explosive devices by partner and coalition forces. The device we are building will be network-
agnostic and designed for adverse and low-connectivity environments, and will be scalable across
levels of information fidelity.
Alex Zaheer
CS
Cyberpolicy
Andreas Pavlou
Physics/International Security
Nitish Kulkarni
Hardware
Engineering
Alex Richard
CS
Gaurav Sharma
GSB
Main Sponsor: OSD // JIDA
Customers spoken to this week: 14
Support team/Sponsors:
Booz Allen Hamilton
US Army AWG
AGUA Holdings
Wolters Kluwer
Crossdeck, Inc.
63. Live Tactical Threat Toolkit (LTTT)
surgical-grade EOD, remotely
Team Member 1 Nitish Kulkarni Alex Zaheer Alex Richard Gaurav Sharma
Academic Program
MA IPS 2016 ME 2016 CS 2019 CS 2016 GSB MSx
LinkedIn https://www.linkedin.c
om/in/andreas-pavlou-
02676753
linkedin.com/in/nitishk
2
linkedin.com/in/alex-
zaheer-ba6548bb
https://www.linkedin.co
m/in/arichard4
https://www.linkedin.co
m/in/gauravsharmastanf
ord
Subject Matter
Expert?
No
(Some Software/Proj
Mgmt)
No
(Sensor Fusion, Systems
Integration)
Yes
(Software
Engineering)
Yes
(Software Engineering)
Yes
(Military Information
Logistics)
Role
Jack of All Trades Hardware Engineering SWE SWE PM
Experience Solving a
Problem that
Seemed Impossible
Developing a curriculum
for a course on
Cybersecurity &
International Relations in
less than two days
Developed and designed
early-stage concepts for
multi-million dollar
DARPA aerospace and
ISR vehicle programs,
hypersonics.
Educating mostly
elderly workforce as a
youth regarding
“good” security
practices -- many blank
stares
Senior in CS at Stanford,
specializing in Logic
Led a team that built a
system to ensure seamless
flow of equipment to build
rockets for SpaceX in 3
months (when everyone
said it would take atleast
twice as long)
64. Current MVP
Key Pain Points:
- Real-time translation (text, voice, video)
- Persistent Communication Links
- Operation in RF-denied or RF-unavailable environments
- Language/Culture-agnostic UI/UX
65. Customer Discovery
Hypothesis:
• Build a Live Tactical
Toolkit digital
notebook for
differently located
coalition & US forces
to address diffusing
remote IED
(Improvised Explosive
Device)
Experiments:
• Interviewed 14
stakeholders including
end users, planners,
solution providers, and
procurement teams
• Ascertained
organization structure
and its relation to JIDA
Results:
Challenges include
• Information deficit, data reliability
• Coalition forces are illiterate, semi-literate
• Language barrier, no minimum training standards for
coalition forces
• Incentivizing coalition forces to use technology
• Bandwidth and network capability
• Possibility of RF-denied environments
• Solution should be as easy as for kindergarten children
• Hardware is not an issue, software is
• Data encryption
• Limited subject matter experts on IED & language
translators available
Action:
• Study existing solutions within JIDA & private solution
providers
• Interview more end users and private solution
providers
• Possibly interview coalition forces
66. DARPA PCAS Team -
modifying existing
systems
Joint Improvised
Threat Defeat Agency -
work with technicians
who deal with problem
day-in day-out
General Atomics,
Raytheon - contractors
already working in this
space may be able to
provide key insight
ARMY Research Lab
NAVEODTECHDIV
ARMY Asymmetric
Warfare Group
Localization -
ensuring system is
applicable to every
forward deployment
Minimization -
ensuring product fits
every tactical situation
Bridge - bridging the
gap between forward
deployed technicians
and US remote
technicians
Personnel-centered -
creating a system
designed around those
who need it most
Integrated - creating a
tool that is an aid, not a
hinderance, to time-
critical operations
Engineers -
contractor/DARPA
employees may provide
technical insight
Technicians -
connections to those who
need this product most
Embedded operations
need support from those
who need product
Translation experts -
those who do job
already
Iterate - Multiple simulated
field test scenarios with
technicians to fit product to
need
Incorporation - product
has many possible futures -
can potentially be
incorporated into other DoD
product needs
Primary - JIDA, EOD
specialists in recurring
forward deployment
operations across
systems
Secondary - Military
remote technical
communications
wings. Software can
be co-opted for many
different
communications
missions
Fixed: Variable:
Software development optimization
IT Management/maintenance
Deployment training
Mission Achievement Criteria
Enable clear, safe, common-sense
communication of time-sensitive high-
stress technical information
Real-time IED Defeat Assistance for Coalition
and Remote Forces
68. Live Tactical Threat Toolkit (LTTT)
Surgical-grade EOD, remotely
We seek to advance JIDA’s mission of countering improvised threats by developing
systems and workflows for real-time remote advise and assist for allied foreign militaries, with the
short-term product goal of real-time defeat of IEDs.
This is of critical importance to JIDA in the context of reduced presence of US forces and the
consequently mitigated ability of allied militaries to leverage US EOD expertise.
Alex Zaheer
CS
Cyberpolicy
Andreas Pavlou
Physics/International Security
Nitish Kulkarni
Hardware
Engineering
Alex Richard
CS
Gaurav Sharma
GSB
Main Sponsor: OSD // JIDA
Total Partners spoken to: 31
Partners spoken to this week: 17
Support team/Sponsors:
Booz Allen Hamilton
US Army AWG
DIUX
US Army Explosive Ordnance Disposal
AGUA Holdings
Wolters Kluwer
Crossdeck, Inc.
CENTCOM
69. Customer Discovery
Hypothesis:
• Build a Live Tactical
Toolkit for coalition
forces to do and gain
assistance with:
• Initial situation
assessment
• Blast damage
mitigation
procedures
• Post-incident
reporting and
information
exploitation
Experiments:
• Interviewed 17
stakeholders including
end users, planners,
solution providers, and
procurement teams
• Ascertained organization
structure and its relation
to JIDA
Results:
Challenges include
• Incentivization to cover 50 km radius
• Information deficit, data reliability
• Coalition forces are illiterate, semi-literate
• Language barrier, no minimum training standards
• Security is important yet solution should be simple
• US Forces also not incentivized to act as R&D testbeds
• Bandwidth and network capability
• No awareness of current information environment
• Possibility of RF-denied environments
• US-side EOD users adamant on not using toolkit during EOD
operations
• Data encryption for report dissemination on JIDA end
• Too many solutions already available
• Limited subject matter experts on IED & language translators
available
Action:
• Continue prototype development and testing at EOD centers
• Working on a connection to local training centers for
iterative product development
• Continue to study existing solutions and developers
• Work with JIDA to explain white/black box approach
• Possibly interview partner nation forces (?)
70. DARPA PCAS Team
Joint Improvised
Threat Defeat Agency
General Atomics,
Raytheon - contractors
ARMY Research Lab
NAVEODTECHDIV
ARMY Asymmetric
Warfare Group
US Central Command
US Army EOD
Alaska
Airlines/American
Airlines
Localization -
ensuring system is
applicable to every
forward deployment
Minimization -
ensuring product fits
every tactical situation
even in A2/AD areas
Bridge - bridging the
gap between forward
deployed technicians
and US remote
technicians
Personnel-centered -
creating a system
designed around those
who need it most
Integrated - creating a
tool that is an aid, not a
hindrance, to time-
critical operations
Engineers -
contractor/DARPA
employees may provide
technical insight
EOD Technicians –
iterative design process
needs more user input
Comm Technicians
Civilian Transportation
Embedded operations
need support from those
who need product
Translation experts -
those who do job
already
EOD Operators
Partner Militaries (?)
Iterate - Multiple simulated
field test scenarios with
technicians to fit product to
need
Incorporation - product
has many possible futures -
can potentially be
incorporated into other DoD
product needs
Primary - JIDA, EOD
specialists in recurring
forward deployment
operations across
systems
Secondary - Military
remote technical
communications
wings. Software can
be co-opted for many
different
communications
missions
Transportation
Industry – remote
medical care (airlines)
Fixed: Variable:
Software development Device Cost
IT Management/maintenance
Deployment training
Mission Achievement Criteria
Enable clear, safe, common-sense
communication of time-sensitive high-stress
technical information and post-incident
reporting and analysis dissemination
Real-time IED Defeat Assistance for Coalition
and Remote Forces
71. Virtual Threat Toolkit:
Value Proposition Canvas for Public
Sector
Products
& Services
Virtual Toolkit
Linking Iraqi
and
American
Forces
- Real time consultation with Iraqi counterparts on how
to disarm IED’s
- Gain and document intelligence from Iraqi security
forces as they encounter IED’s
Customer
Jobs
Assess and defeat
IEDs
Restart report
workflow
Lack of information on
evolution of IED’s
utilized by insurgents
and their tactics
Difficulty advising
operators in the field on
how to defuse IEDs
Explosive Ordnance Disposal
Teams
Gains
Pains
Gain
Creators
Pain
Relievers
Ability to advise coalition
counterparts on how to dismantle
IEDs
Development of predictive models
for insurgent use of IEDs
Accurate reporting data on IED
usage
- Reduction of communication barrier
between base-deployed US forces and
Iraqi security forces in the field
72. Virtual Threat Toolkit:
Value Proposition Canvas for
Private Sector
Products
& Services
Virtual Toolkit
linking
information
sources
(emergency
services) and
remote semi-
trained
operators
- Real time access to expert-level lifesaving medical
knowledge in situations where it is otherwise
unavailable
- Dynamic provision of necessary information in
situations with communications obstacles
Customer
Jobs
Assess and solve
EMS problems
remotely
Lack of information
on the specific
problem set that
commercial partners
want to address
Current needs are
vague and need
client-side definition
Commercial Partners interested
in VAA (e.g. Airlines, Transport
Providers)
Gains
Pains
Gain
Creators
Pain
Relievers
Ability to provide expert-level care
without needing expensive
resources tasking
Remote assistance in areas where
EMS expertise is sparse or
temporally difficult
Reduction of the communications barrier
between semi-trained EMS operators and
knowledge base
Development of non-connectivity solutions
to access a dynamic knowledge base
Reducing communication latency and
improving information flow bandwidth
73. Customer Workflow
Partner military unit
encounters IED
Link to EOD expert using
LTTT initiated: video
feed/or minimal
workflow depending on
connectivity
IED data logged
Report sent to
JIDA
IED diagnosis
and EOD
process
leveraging US
expertise
Resolution
JIDA analyzes
data and
distributes
intelligence as
incentive for
use
74. Current MVP
Key Pain Points:
- Real-time translation (text, voice, video)
- Persistent Communication Links
- Product adoption
- Non-disruptive UI/UX during ops
JIDA Database
+
Analysis
76. Right of Boom
Surgical-grade EOD, before and after
Alex Zaheer
CS
Cyberpolicy
Andreas Pavlou
Physics/International
Security
Nitish Kulkarni
Hardware
Engineering
Alex Richard
CS
Gaurav Sharma
GSB
Main Sponsor: OSD // JIDA
US Army Explosive Ordnance Disposal
US Air Force Explosive Ordnance Disposal
FBI Bomb Defusal – San Francisco
US Army Asymmetric Warfare Group
US Central Command
DIUx
Wolters Kluwer Romania
Booz Allen Hamilton
Crossdeck, Inc.
Iraqi Embassy
We seek to advance JIDA’s mission of
countering improvised threats by
developing systems, workflows, and
incentives for real-time remote
advise and assist for allied foreign
militaries with the goal of improved
intelligence fidelity.
12
Interviews
1
Buyer
6
Experts
5
Users
77.
78. Customer Discovery
Hypothesis: JIDA’s previous unwillingness to share
finished intel reduces Iraqi participation incentives
Hypothesis: JIDA is willing grant partner forces limited
access to IED reporting information via PiX
Hypothesis: EOD units will want a LTTT to facilitate
their ability to defuse IEDs.
12
Interviews
1
Bomb Suit
3
Prototypes
4
EOD Techs
Solution:
Incentives for EOD squadron
commanders to generate data
Next Week:
Contacting Iraqi Military via Ambassador
Field test of form+function(?) prototype of VAA
Contacting JET
79. Joint Improvised
Threat Defeat Agency
ARMY Research Lab
NAVEODTECHDIV
ARMY Asymmetric
Warfare Group
US Central Command
US Army EOD
US Air Force EOD
FBI Bomb Squad – SF
Alaska Airlines/American
Airlines
Post-Incident
Reporting
- Ensuring high
volumes of detailed
IED reporting data
Saving lives
- Using VAA/PAA to
mitigate allied
casualties
Bridge - bridging the
gap between forward
deployed technicians
and US remote
technicians
Incentivization –
gamification to increase
likelihood of use by
partner militaries
Integrated - creating a
tool that is an aid, not a
hindrance, to time-
critical operations
Reporting – a richer
data set for JIDA’s
mission
Engineers -
contractor/DARPA
employees may provide
technical insight
EOD Technicians –
iterative design process
needs more user input
JIDA SETAs
Coalition Forces
- EOD + teams
- Coalition Intel
JIDA
- Accurate reporting
- Incentivized Intel
Collection
Iterate - Multiple simulated
field test scenarios with
technicians to fit product to
need
Incorporation - product
has many possible futures -
can potentially be
incorporated into other DoD
product needs
Primary
- JIDA
- Coalition EOD
- People in IED-
impacted zones
Secondary
- Military remote
technical
communications.
- Software can be co-
opted for many
different
communications
missions
Transportation
Industry – remote
medical care
Fixed:
Software development
IT Management/maintenance
Deployment training
Mission Achievement Criteria
Enable clear, safe, common-sense
communication of time-sensitive high-stress
technical information and post-incident
reporting and analysis dissemination
Accurate high-volume post-incident IED
reporting
Variable:
Maintenance
Equipment retrofit
Equipment refurb
80. Virtual Threat Toolkit:
Value Proposition Canvas for
Coalition EOD
Products
& Services
-Virtual Toolkit
Linking Iraqi
and
American
Forces
-Gamified EOD
analytics and
reporting
- - Real time consultation with US EODs on how to
dismantle IEDs
- - EOD commander analytics and transparency
Customer
Jobs
-Assess and defeat
IEDs
-Restart report
workflow
-Lack of information on
evolution of IEDs utilized
by insurgents and their
tactics
-Insufficient EOD
knowledge
Explosive Ordnance Disposal
Teams
Gains
Pains
Gain
Creators
Pain
Relievers
-Less Iraqis dying
-Advice on how to better defuse IEDs
-Accurate reporting data on IED usage
(?)
- - Reduction of communication barrier with
base-deployed US forces
- - Less likely to catastrophically fail while
defusing an IED
81. Virtual Threat Toolkit:
Value Proposition Canvas for
JIDA/Intelligence Community
Products
& Services
-Higher-fidelity
post-incident
EOD analytics
and reporting
-Higher-
accuracy
device
classification
and
intelligence
- Gain and document intelligence from Iraqi security
forces as they encounter IEDs
- Alignment of incentives between foreign EODs
and JIDA such that blast data is generated
- Real time consultation provides environmental
and circumstantial intelligence (?)
Customer
Jobs
-Analyze IED
patterns and
provide actionable
intel
-Lack of information on
evolution of IED’s
utilized by insurgents
and their tactics
-Wasting money on
apps that nobody
uses
Explosive Ordnance Disposal
Teams
Gains
Pains
Gain
Creators
Pain
Relievers
-Accurate reporting data on IED
usage
-Development of predictive
models for insurgent use of IEDs
- Reduction of communication barrier
between base-deployed US forces and
Iraqi security forces in the field
85. Right of Boom
Surgical-grade EOD, after
Alex Zaheer
CS
Cyberpolicy
Andreas Pavlou
Physics/International
Security
Nitish Kulkarni
Hardware
Engineering
Alex Richard
CS
Gaurav Sharma
GSB
Main Sponsor: OSD // JIDA
US Army Explosive Ordnance Disposal
US Air Force Explosive Ordnance Disposal
FBI Bomb Squad – San Francisco
US Army Asymmetric Warfare Group
US Central Command
DIUx
Wolters Kluwer Romania
Booz Allen Hamilton
Crossdeck, Inc.
We seek to advance JIDA’s mission of
countering improvised threats by
developing systems, workflows, and
incentives for allied foreign militaries
with the goal of improved
intelligence fidelity.
19
Interviews
3
In MENA
61Total
Interviews
10
EOD Techs
86. Customer Discovery
Product Managers
• J8 Program Co-
ordinator, JIDA/OSD
• J6 Operations,
JIDA/OSD
Concept Developers
• [Stanford]
• PiX SETA, CENTCOM-OM
Users
• Iraqi / ANA acq
chiefs
• J2 Analysist,
JIDA/OSD
Recommenders
• Knowledge
Manager,
CENTCOM
• EOD Team Lead
Sgt., 65th
Ordnance
Company, US
Army
• JET
• Coalition EOD
Technician [PLS
HELP]
Saboteurs
• Organizational
Saboteurs
• JIDA
Classification
Office
Buyers
• Mr. James P.
Craft, Deputy
Director,
Information
Enterprise
Management,
JIDA
[Full Names Redacted For Privacy]
87. Hypotheses + Future Plans
Hypothesis: EOD units will want a LTTT to facilitate
their ability to defuse IEDs.
Hypothesis: No infrastructure for a unified data
system for IED incident reporting [PiX + JIDA SQL]
Hypothesis: Material benefits are sustainable long-
term for an information exchange relationship with
coalition forces
19
Interviews
3
In MENA
1
Prototypes
6
EOD Techs
Next Week:
- Obtaining test/demo devices
- Deep dive into PiX + API access
- Accessing + using exisiting apps (HYBRID,
PiX MENA, INDURE)
- Talk to Iraqi liaison at Defense Language
Institute
- Speak with FBI, JET, HR/softer skills experts
- Understand JIDA/Coalition workflows
better
88. Coalition EOD
Technicians
(US Point of Contact)
JET
Coalition EOD
Commanders
US FOB
[EOD]
Post-incident
reporting
Some Intel
Assistance (?)
Some Incentive/Assistance (?)
JIDA
[CONUS]
89. Coalition EOD
Technicians
US POC
[JET]
[ONSITE]
FBI
[CONUS]
DOD EOD
DATABASE
[CONUS]
JIDA DATABASE
[CONUS]
US AIR FORCE EOD
[CONUS]
US ARMY EOD
[CONUS]
US NAVY EOD
[CONUS]
[UNCLASS]
[UNCLASS]
[UNCLASS
FOUO
S/TS]
Information flow, in an ideal world
[UNCLASS
FOUO
S/TS]
[UNCLASS
FOUO
S/TS]
[UNCLASS
FOUO
S/TS]
90. INTERNET
AFGHAN NETS
SIPRnet
NIPRnet
DOD
(DISA)
APAN
DOD
(NGA)
PiX
DOD (DISA)
APAN
AF RONNA
SharePoint 2010
BASEER
NIMS
CIDNE
RED DOT
DAIIS
PORTALS
HERMES
Open Source
Program
CDC
Dissemination
Program
Dept. of Army Intelligence Information
System
Smartphone
Web Browser
PiX Android/iOS
Apps
INDURE Android
View/Report App
INDURE Android
View/Report App
JIDA Hybrid
Reporting App
SOCCENT ATAK
Tactical Ops App
PiX SMS
PiX WIKI
Mediated WIKI
INDURE API
IRAQ/Other Nets
Collaboration Portal
Zimbra/Sharepoint 2007
FOUO/SBU
INDURE
Reporting/DB
GPS
Mapping
Current Information Map
91. Joint Improvised
Threat Defeat Agency
ARMY Research Lab
NAVEODTECHDIV
ARMY Asymmetric
Warfare Group
US Central Command
US Army EOD
US Air Force EOD
FBI Bomb Squad – SF
Alaska Airlines/American
Airlines
Post-Incident
Reporting
- Ensuring high
volumes of detailed
IED reporting data
Saving lives
- Using VAA/PAA to
mitigate allied
casualties
Bridge - bridging the
gap between forward
deployed technicians
and remote technicians
IED collection
capabilities and analysis
Alignment of
Incentives
High-speed
intelligence allows for
same-day incident
analysis
Improved Reporting –
a richer data set for
JIDA’s analysis
Engineers -
contractor/DARPA may
provide technical insight
EOD Technicians –
iterative design process
needs more user input
JIDA SETAs – provide
details on existing
solutions and flaws
Coalition Forces
- NEED: intel, supplies
- HOW: Incentivize
- SABOTEURS: Bureaucracy
JIDA
- NEED: increased reporting
- HOW: incentivizing
coalition forces
- SABOTEURS: Over-
classification
Iterate - Multiple simulated
field test scenarios with
technicians to fit product to
need
Incorporation - product
has many possible futures -
can potentially be
incorporated into other DoD
product needs
Primary
- JIDA (Matt, Carie,
James)
- EoD (Ike, Ethan,
Christine, Whitner,
especially J2
(Intelligence) and J5
(Strategic Plans and
Policy)
- Coalition EOD (Sgt.
Bashir)
- Other coalition
forces (Romania,
Georgia, Bulgaria:
no contacts)
- Other entities such
as FBI (Sean)
Fixed:
Software development
IT Management/maintenance
Deployment training
Mission Achievement Criteria
Enable clear, safe, common-sense
communication of time-sensitive high-stress
technical information accurate actionable intel
and post-incident reporting and analysis
dissemination.
Accurate high-volume post-incident IED
reporting
Variable:
Maintenance
Equipment retrofit
Equipment refurb
[Full Names Redacted For Privacy]
92. Products
& Services
-Higher-fidelity
post-incident
EOD analytics
and
reporting
-Higher-
accuracy device
classification
and intelligence
- Gain and document intelligence from Iraqi
security forces as they encounter IEDs
-Alignment of incentives between foreign EODs
and JIDA such that blast data is generated
-Bringing coalition forces into intelligence
umbrella
Customer
Jobs
-Analyze IED patterns
and provide
actionable intel
-Lack of information on
evolution of IED’s utilized
by insurgents and their
tactics
-Reliance on material
incentives to coerce
Iraqis into giving intel
Gains
Pains
Gain
Creators
Pain
Relievers
-Accurate reporting data on IED usage
-Development of predictive models for
insurgent use of IEDs
- -Reduction of communication
barrier between base-deployed US
forces and Iraqi security forces in
the field
- -Transition from material
incentives to informational
incentives, allowing for more
sustainable cooperation
Virtual Threat Toolkit:
Value Proposition Canvas for
JIDA J2 and J5
Explosive Ordnance Disposal
Teams
93. Virtual Threat Toolkit:
Value Proposition Canvas for
Coalition EOD
Products
& Services
-Integrated
platform for
both reporting
raw intel and
receiving
analysis
-Gamified EOD
analytics and
reporting
- -Timely and refined intelligence on IEDs and
the tactics of insurgents using them
- - EOD commander analytics and transparency
Customer
Jobs
-Assess and defeat
IEDs
-Restart report
workflow
-Lack of information on
evolution of IEDs utilized
by insurgents and their
tactics
Explosive Ordnance Disposal
Teams
Gains
Pains
Gain
Creators
Pain
Relievers
-Less Iraqis dying
-Improved counter-IED capability
- - Reduction of communication
barrier with base-deployed US
forces
- -Less likely to encounter IEDs
without being forewarned
96. Narrative Mind
Michael May, Zach Maurer, Shazad Mohamed, Will Kim
Customers spoken to this week: 14
Total customers spoken to: 14
Sponsor: Army Cyber Command (ARCYBER) - MAJ Kirk Kudrna, LTC Matthew Zais
Liaisons: DIUx - Lt. Col. Ernie Bio
The Narrative Mind team contains experts in software engineering, social media design, and
web-based information operations (IO). We seek to develop tools that will optimize discovery
and investigation of adversary communication trends on social media, allowing ARCYBER and
others to more efficiently respond and mitigate threats posed by enemy messaging.
97. Michael May
M.A. IPS 2017
Former PM of
counterterrorism
communications
at USCENTCOM
Zach Maurer
B.S. CS 2016
Software engineer,
int’l development, UI
research
William Kim
M.S. CS 2017
Software engineer,
private equity and
community finance
Shazad Mohamed
B.A. Political Science
2016
Former PM Facebook,
co-founder of security
startup.
Team
Michael May
M.A. IPS 2017
Former PM of
counterterrorism
Zach Maurer
B.S. CS 2016
Software engineer,
int’l development, UI
98.
99.
100.
101. Value Proposition Canvas
Products
& Services
Desktop tracking
and analysis
platform.
Customer
Jobs
Develop potential
counter narratives
to propaganda
No way to easily track all
the known extremists
and real-time action of
tweets/hashtags
Gains
Pains
Gain
Creators
Pain
Relievers
Platform for automatically
categorizing tweets,
escalating potentially viral
threats early on
- Better detection ability and
improve response time to
potentially viral narratives, shut
down or respond before it gains
momentum
- Semi-automation of tweet
categorization and virality
detection
- Filing tools for user &
hashtag histories
104. Team Info
Sponsor: US Army Cyber Command (ARCYBER)
Liaisons: DIUx - Lt. Col. Ernie Bio
The Narrative Mind team contains experts in software engineering, social
media design, and web-based information operations (IO).
We seek to develop tools that will optimize discovery and investigation of
adversary communication trends on social media, allowing ARCYBER
and others to more efficiently respond and mitigate threats posed by
enemy messaging.
Weekly Total: 10 interviews
Users: 11
Experts: 31
Buyers: 4
Cumulative Total: 46 interviews
105. All the Hypotheses
Global
Tweet-level
Awareness Response
Adversary IO Org Chart
No baseline for monitoring/aggregating
adversary use of tech; cyber targeting
Automatic Narrative Detection
Language/culture experts aren’t able
to work at scale
Virality Predictor
Understand which narratives
are the most salient
Important Event Predictor
Preempt terrorist attacks
Counter-Narrative Generator
DoD/Gvt. social media
presence is weak
Persistent ID-Alias tracker
Account bans and multiple aliases across
different networks make IDs hard to track
Poison Link Generator
Monitor who views
adversary dark-web
content
File sharing Site Scraper
Need more cached information
access
Expedited Content Categorization
Scale of social media makes manual
efforts painful
106. Organizational
Chart
(updated)
Analyst “Fish”
(CW3 or NCO)
MAJ “Kramer”
Reports info to...
COL “Tom”
LTC “Zeke”
Director & Deputy
Assesses threats 1-3yrs out
Proposes CONOP?
Gets approval?
BG Frost
Deputy Commander for Ops
ARCYBER
(service component)
NSA
ADM Rogers
Internal ARCYBER
office -
Uses OTA option
LTG Cardon
BG (O8-level)
JOINT FORCES HQ-CYBER
USCYBERCOM
ADM Rogers
USSTRATCOM
COCOMs
CENTCOM,
AFRICOM,
NORTHCOM
Proposes
concepts
Blurred lines
between the two
Secretary Carter
CYBER-only
Acquisition
process
NSA
107. Hypotheses ❏ Last week’s hypotheses were not relevant to updated problem
definition.
❏ Too MVP specific.
Experiments ❏ Weren't able to consistently test any previous hypotheses.
Results ❏ ARCYBER confidently re-focused the problem on “identifying
narratives in real time”.
❏ Full-circle evolution.
❏ The problem is not content categorization nor org chart-level
awareness nor persistent ID tracking nor event prediction.
Actions 1. Increased frequency of communication with sponsor.
2. Help ARCYBER define what a “narrative” is.
3. Research how commercial companies are visualizing
“narratives”.
4. Continue expanding network to beneficiaries who can answer
the above questions.
- PACOM, CENTCOM, NPS CORE Lab
Customer Discovery
108. Mission Model Canvas- Categorize social media posts
and users by content for
monitoring and tactical purposes.
- Understand viral potential of
extremist social media posts in
real-time.
- Track how extremist
groups use technology
over time.
- Gnip/Twitter/Facebook
- CrowdFlower,
Samasource, or
Mechanical Turk
- Pre-existing social
media service and micro-
labor aggregators
- Third-party access
platforms for social
media, e.g. Hootsuite
which allow users to
circumvent barriers, but
allows us to track IP
addresses.
-Data visualization (Zoic)
-Content analysis
platforms (Sens.ai,
Leidos)
ARCYBER
-Bg. General (decision
maker)
-MAJ/LTC/COL
(operational plan)
-Analysts/Operators
(actionable insights)
COCOMs
-Generals/Analysts
-Tactical decision
makers
- Optimize workflow for
social media analysts
- Effectively visualize
how extremist groups
use technology as they
grow.
- Aggregate known
research?
- Expedite categorization of
social media content.
- Use MechanicalTurk to
crowdsource categorization of
content and users.
- Algorithmic virality predictor to
create alerts for important, time-
sensitive threats.
- Use design of now-defunct
Palantir Torch as inspiration for
how to present content in a
streamlined manner.
- Force multiplier for intelligence analysts: receive cleaner, pre-categorized data, target the
most urgent threats to national security in real-time.
- Increase throughput to quantify and flag viral extremist content.
- Improve the categorization of unstructured social media data points using crowdsourced
micro-task labor.
??????????
- Architecture that can
support massive
concurrent data
aggregation and
analysis.
- Customized UI
- Testing with analysts
- MechanicalTurk or crowdsourcing labor (microtasks)
- UI Development/Testing with ARCYBER analysts.
- Software Development
- Research aggregation? (e.g. Import and reformat existing
knowledge about extremist organizational structure)
- Access to Twitter
firehose or API
- Local language
speaking crowdsourcing
staff.
- Accurate testing for
intercoder reliability
- Individual
Analysts/Operators
- ARCYBER (Bg. General,
LTC, Strategic Intiatives
Group)
- Continued partnership
with crowdsourcing firms,
Beneficiaries
Mission AchievementMission Budget/Costs
Buy-In/Support
Deployment
Value
Proposition
Key Activities
Key Resources
Key Partners
109. MVP - 3 Types of “Narratives” Outputted by Our System
E.g.
Madison, AL: Pro-Trump
(Positive: 61% ± 5%)
Millbrae, CA: Anti-Trump
(Positive: 11% ± 5%)
A ranked list of
extremists’ most
popular content
over time
A list of hashtags
that appear with
the same
message with
extreme
frequency.
Positive or
negative opinion of
adversaries
among geographic
or demographic
population.
Narrative Type #1 Narrative Type #2 Narrative Type #3
E.g.
Adversary: Trump Campaign
#NY4Trump
#Trump2016
Could be classified by an
analyst as “New York
Primary Push” or similar.
E.g
Could be classified by an
analyst as “Political
Outsider” or similar.
110. Value Proposition Canvas - ARCYBER
Products
& Services
Web/Desktop
Application Provide high-level
overview of terrorists
and groups,update
research
Customer
Jobs
- New mission priorities.
- Little familiarity with overall
environment.
- No repository for high-level
learnings
Gains
Pains
Gain
Creators
Pain
Relievers
- New opportunities for cyber-
ops against adversary groups
- Social network graph of all
terrorist groups?
-Illustrates evolution of
enemy info dissemination
networks
- Provides insight for
mapping enemy “online
terrain”
Visualize information about
how adversary
organizations are using
social media
ARCYBER - Operator/Analyst (Actionable Insights)
111. Value Proposition Canvas - ARCYBER
Products
& Services
Web/Desktop
Application
Aggregate insights
from analysts,
compile into plan
passed to decision
maker
- New mission priorities.
- Proliferating/viral social
media platforms
- Urgency of threat clouds
ability to prioritize
overabundance of threats
Customer
Jobs
Gains
Pains
Gain
Creators
Pain
Relievers
-Faster turnaround time with
better data gained from
information awareness software
and technology
- Clear directive for
neutralizing cyber threats
- Clear hierarchy of cyber
threat levels
ARCYBER - MAJ/LTC/COL (Operational Plans)
-Better data in informing
operational plans provided
to decision makers
-Reducing uncertainty in
operational planning
112. Value Proposition Canvas - ARCYBER
Products
& Services
Web/Desktop
Application
Act on reports and
plans generated
about various cyber
threats across globe
- Little understanding of
ground-level nuances of
threats
- Little understanding of
proliferating platforms
Customer
Jobs
Gains
Pains
Gain
Creators
Pain
Relievers
- Organizational chart of
technology broken down by
functional purpose
- Timeline-based viewer
- Social network graph of all
terrorist groups?
-Reduce uncertainty of
decision making
-Add methodological rigor
to ARCYBER’s
offensive/defensive
operations
- Neutralize terrorist threats
on cyber
ARCYBER - Brigadier General (Decision Maker)
113. Value Proposition Canvas - COCOM
Products
& Services
Web/Desktop
Application
Target audience
analysis, finding
opportunities to
optimize influence
operations
Customer
Jobs
-Loss of visibility on key
communicators (when
accounts shut down/opened)
- No repository for high-level
learnings
Gains
Pains
Gain
Creators
Pain
Relievers
-Auto-populated list & real-time
visualization of key accounts,
dissemination networks
-Allows more accurate targeting
-Illustrates evolution of enemy
info dissemination networks
- Provides insight for mapping
enemy “online terrain”
Closure of visibility gap on
key adversary accounts
COCOM - Operator/Analyst (Actionable Insights)
114. Value Proposition Canvas - COCOM
Products
& Services
Web/Desktop
Application
-Long term
operational planning
- Measuring
effectiveness
Customer
Jobs
- Unclear picture of how
some adversary networks
evolve, adapt over time
Gains
Pains
Gain
Creators
Pain
Relievers
- Organizational chart of
technology broken down by
functional purpose
- Visualization of key accounts,
dissemination networks over
time
-Illustrates evolution of
enemy info dissemination
networks
- Provides insight for
measuring operations,
mapping enemy “online
terrain”
Greater awareness of
network evolution improves
planning, indicators of
operational effectiveness
COCOM - Maj/LTC/Col. (O4-O6) (Operational Plans)
115. Value Proposition Canvas - COCOM
Products
& Services
Web/Desktop
Application
- Judge potential
effect and risk of
operations
- Approve/reject
plans
Customer
Jobs
- Less familiarity with
internet as an operational
environment
- Needs high level summary
of how adversaries use
technology
Gains
Pains
Gain
Creators
Pain
Relievers
- Organizational chart of
technology broken down by
functional purpose
- Timeline-based viewer
-Illustrates evolution of
enemy info dissemination
networks
- Provides insight for
mapping enemy “online
terrain”
Visualize high-level
information about how
regional adversaries
use technology.
COCOM - Brigadier General (O7) (Decision Maker)
117. Hypotheses ❏ “C2”, “Recruiting”, “Propaganda” are sufficiently granular content
categories.
❏ Advanced warning virality prediction is valuable for countering threats early.
❏ Content categorization is central to analyst workflow.
❏ Provide higher-level information around adversaries’ organizational
structure
Experiments ❏ Focused on getting sponsor to re-articulate original problem.
❏ Explicitly asked for thoughts on crowdsourcing mechanism.
❏ Interviewed end users about their tasks and current tools, as well as our
own hypotheses
Results ❏ Understanding terrorist groups is difficult as they are dynamic
organizations, need some way of tracking how these organizations shift
over time, and how they interact with each other
❏ Geospatial information is valuable, more so that social network analysis
❏ Government agencies are OK with crowdsourced tools for OSINT
Actions Testing our two main MVPs:
1. Organizational chart of the various terrorist groups over time
2. Social network map of how various terrorist groups interact with each
other
Customer Discovery
118. ARCYBER Org Chart●Talk of
CYBERCOM
becoming its
own COCOM
●Both
CYBERCOM
and NSA are
technically led
by ADM Rogers
119. Organizational chart of various terrorist groups
over time and their social media platforms
- Organizational structure of terrorist groups
and affiliates
- Sub-nodes display main channels for
propaganda and platforms used
- Time slider shows temporal change
MVP
ISIL Nusra Front
C2 Propaganda Recruitment
Google
Maps
Logistics Battle
Whats
App
Public Private
Twitter Telegram
Twitter
Time Slider
APR 2015 MAY 2015
121. Disruptive
Innovation
Continuous
Innovation
Lean Means Getting Out of Your Office
Horizon 2
Horizon 3
Speed &
Urgency
Lean
Steve Blank
• If you’re not talking to 100’s of customers, it’s not lean
• If you’re not building iterative and incremental minimum
viable products, it’s not lean
122.
123. Help DOD/IC Tap
University Talent
Pool
Create Tech
National Service
Help Gov’t
Innovate at Speed
Connect the DOD/IC
Islands of Innovation
124. Thanks!
Steve Blank: https://steveblank.com/about/ sblank@kandsranch.com
Hacking for Defense Class: http://hacking4defense.stanford.edu
Hacking for Defense Blogs: https://steveblank.com/category/hacking-for-defense/
Lean Open Source: https://steveblank.com/slides/
Lean in the HBR: https://hbr.org/2013/05/why-the-lean-start-up-changes-everything/ar/1