ICT Role in 21st Century Education & its Challenges.pptx
Skynet Lessons Learned H4D Stanford 2016
1. Skynet
Increasing Situational Awareness
Through Computer VisionSOCOM
Week 10
Increase situational
awareness by
automatically detecting
humans on
autonomous drone
feeds
Week 0
Increase situational
awareness while
minimizing cognitive
load with a new
human-drone
interaction system
107
Interviews
Many Interviews Later…
Companies
8
Users
52
Experts
26
Buyers
14
2. The Skynet Team
Kevin Mott Alvin Goh Sam Gussman Olga Musayev
International Policy
Studies 2017
Mgmt Sci & Engr
2016
Symbolic Systems
2016
JD 2017
AD Infantry Officer
Embedded
Systems
Engineer
UI/UX Design
Software Engineer
Data Scientist and
Lawyer
3. The Problem
Don’t want to take Operator out of
the fight to watch a drone feed
NAVY SEALs Fighting
ISIS, 2016
5. Mission Model Canvas V1- Week 0
SkyNetUI: Mission Model Canvas
- SOF, Rangers, AWG
- Law Enforcement
Agencies
- Drone manufacturers (ex:
DJI)
- Haptic startups
- Augmented Reality
Visualization Companies
(ex: Google Glass)
- Threat/Friendly detection
- using
light/sound/motion/indicat
or/marking
-HCI - feedback
mechanism
-Making system easy to
use/learn/maintain
- Primary: SOF elements
and Ground Force
Commanders
- Secondary: Command
elements at TOC/TAC
-Tertiary: Search and
Rescue teams
-Enhance situational
awareness - Drone
becomes another set of
eyes, frees up an operator.
-Reduced risk to force-
Drone goes so humans do
not have to
- video recording to
facilitate AARs in training
-Low cost alternative to
costly assets- frees up
other assets in a resource
constrained environment
- Reduces dependence on
human capital
-Improved situational awareness for small units in combat
-Demand across SOF elements and requests for fielding from
conventional units
- Field test with one unit.
Evaluate training method.
- Expand field-tests to
multiple units in same
deployment context
- With successful field
tests, contract COTS
vendors to mass produce.
Fixed:
- Equipment - drones, development toolkits,
- Hapkit: Haptic Starter kit ($50 x 4)
- Software design & engineering
Variable:
- Costs of maintenance, updates, and training
- Drones w/ camera
(provided by SOCOM)
- Augmented Reality
Visualizer with developer
tools
- Haptic Specialists to act
as advisors
- Hapkit
- Drone pilots
- Soldiers / team leaders
-Need current methods
used to mark friendly
forces
-Support of an automated
threat detection system
Beneficiaries
Mission AchievementMission Budget/Costs
Buy-In/Support
Deployment
Value
Proposition
Key Activities
Key Resources
Key Partners Value
proposition:
Novel Human-
drone interface
to allow
operators to
control UAV
drones easier
Beneficiaries:
SOF elements,
Commanders,
Search &
Rescue
Key partners:
Law
enforcement
Augmented
Reality
Companies
6. Weeks 0-3 Customer Discovery
(Moffett field, Week 3)
Experiment: audio vs visual cues
What we did:
- Interviews, Moffett Field visit
- Cognitive Obstacle Course
- Gear Assessment
What we heard:
“I don’t want another system…”
What we thought:
Operators need a new way of
interacting with drones while
remaining alert on the
battlefield.
8. Mission Model Canvas- Week 3
SkyNetUI: Mission Model Canvas
- SOF, Rangers, AWG
- Law Enforcement
Agencies
- Drone manufacturers (ex:
DJI)
- Haptic startups
- Augmented Reality
Visualization Companies
(ex: Google Glass)
- Threat/Friendly detection
- using
light/sound/motion/indicat
or/marking
-HCI - feedback
mechanism
-Making system easy to
use/learn/maintain
- Primary: SOF elements
and Ground Force
Commanders
- Secondary: Command
elements at TOC/TAC
-Tertiary: Search and
Rescue teams
-Enhance situational
awareness - Drone
becomes another set of
eyes, frees up an operator.
-Reduced risk to force-
Drone goes so humans do
not have to
- video recording to
facilitate AARs in training
-Low cost alternative to
costly assets- frees up
other assets in a resource
constrained environment
- Reduces dependence on
human capital
-Improved situational awareness for small units in combat
-Demand across SOF elements and requests for fielding from
conventional units
- Field test with one unit.
Evaluate training method.
- Expand field-tests to
multiple units in same
deployment context
- With successful field
tests, contract COTS
vendors to mass produce.
Fixed:
- Equipment - drones, development toolkits,
- Hapkit: Haptic Starter kit ($50 x 4)
- Software design & engineering
Variable:
- Costs of maintenance, updates, and training
- Drones w/ camera
(provided by SOCOM)
- Augmented Reality
Visualizer with developer
tools
- Haptic Specialists to act
as advisors
- Hapkit
- Drone pilots
- Soldiers / team leaders
-Need current methods
used to mark friendly
forces
-Support of an automated
threat detection system
Beneficiaries
Mission AchievementMission Budget/Costs
Buy-In/Support
Deployment
Value
Proposition
Key Activities
Key Resources
Key PartnersKey partners:
Law
enforcement
Augmented
Reality
Companies
??
Value
proposition:
Something that
isn’t another
new system
A way to spot
enemies
keep track of
each other
Beneficiaries:
SOF elements,
Commanders,
Search &
Rescue
9. Weeks 3-4
What We Heard:
“Object detection in drone feeds
would be huge.”
What we thought:
Operators need an extremely wide
range of capabilities- drone
swarms, facial recognition, etc.
10. - Knowing the locations of
friends/enemies/civilians
- Reduced ambiguity
- autonomous assessment of
aerial perspectives and video
feed Customer
Jobs
Shoot bad
guys without
getting shot
- Cognitive overload from
existing tools
-Manpower losses
- Hassle of systems not
communicating
Gains
Pains
Gain
Creators
Pain
Relievers
-Increased situational awareness on
the battlefield
- Reconnaissance of
inaccessible places
- Integrate with current
systems that soldiers are
familiar with
- Frees up another soldier
Products
& Services
Situationally aware
drone with visual
interface
Value Proposition Canvas
Pains
New system to
learn
Gains
Increase situation
awareness
Pains
Integrate with
ATAK
Gains
Detection
ClassificationProduct
ATAK
plugin
Customer
jobs
shoot bad guys
without getting
shot
11. Mission Model Canvas- Week 4
SkyNetUI: Mission Model Canvas
- SOF, Rangers, AWG
- Law Enforcement
Agencies
- Drone manufacturers (ex:
DJI)
- Haptic startups
- Augmented Reality
Visualization Companies
(ex: Google Glass)
- Threat/Friendly detection
- using
light/sound/motion/indicat
or/marking
-HCI - feedback
mechanism
-Making system easy to
use/learn/maintain
- Primary: SOF elements
and Ground Force
Commanders
- Secondary: Command
elements at TOC/TAC
-Tertiary: Search and
Rescue teams
-Enhance situational
awareness - Drone
becomes another set of
eyes, frees up an operator.
-Reduced risk to force-
Drone goes so humans do
not have to
- video recording to
facilitate AARs in training
-Low cost alternative to
costly assets- frees up
other assets in a resource
constrained environment
- Reduces dependence on
human capital
-Improved situational awareness for small units in combat
-Demand across SOF elements and requests for fielding from
conventional units
- Field test with one unit.
Evaluate training method.
- Expand field-tests to
multiple units in same
deployment context
- With successful field
tests, contract COTS
vendors to mass produce.
Fixed:
- Equipment - drones, development toolkits,
- Hapkit: Haptic Starter kit ($50 x 4)
- Software design & engineering
Variable:
- Costs of maintenance, updates, and training
- Drones w/ camera
(provided by SOCOM)
- Augmented Reality
Visualizer with developer
tools
- Haptic Specialists to act
as advisors
- Hapkit
- Drone pilots
- Soldiers / team leaders
-Need current methods
used to mark friendly
forces
-Support of an automated
threat detection system
Beneficiaries
Mission AchievementMission Budget/Costs
Buy-In/Support
Deployment
Value
Proposition
Key Activities
Key Resources
Key PartnersKey partners:
ATAK
programmers
Computer
vision experts
Value
proposition:
Detection
Classification
Integration with
ATAK
Beneficiaries:
SOCOM Green
Berets,
Rangers,
SEALs
JTACS, Drone
Operators
12. Weeks 5-6
What we thought:
Difficult/slow process to get
funding for development
Channels
What we heard:
“There’s lots of efficiencies in the
acquisition process.”
S&T, SBIR, BAAs, RRTO and CTTO
offices…
Field testing
Before
After
14. Mission Model Canvas-- Weeks 5-6
SkyNetUI: Mission Model Canvas
- SOF, Rangers, AWG
- Law Enforcement
Agencies
- Drone manufacturers (ex:
DJI)
- Haptic startups
- Augmented Reality
Visualization Companies
(ex: Google Glass)
- Threat/Friendly detection
- using
light/sound/motion/indicat
or/marking
-HCI - feedback
mechanism
-Making system easy to
use/learn/maintain
- Primary: SOF elements
and Ground Force
Commanders
- Secondary: Command
elements at TOC/TAC
-Tertiary: Search and
Rescue teams
-Enhance situational
awareness - Drone
becomes another set of
eyes, frees up an operator.
-Reduced risk to force-
Drone goes so humans do
not have to
- video recording to
facilitate AARs in training
-Low cost alternative to
costly assets- frees up
other assets in a resource
constrained environment
- Reduces dependence on
human capital
-Improved situational awareness for small units in combat
-Demand across SOF elements and requests for fielding from
conventional units
- Field test with one unit.
Evaluate training method.
- Expand field-tests to
multiple units in same
deployment context
- With successful field
tests, contract COTS
vendors to mass produce.
Fixed:
- Equipment - drones, development toolkits,
- Hapkit: Haptic Starter kit ($50 x 4)
- Software design & engineering
Variable:
- Costs of maintenance, updates, and training
- Drones w/ camera
(provided by SOCOM)
- Augmented Reality
Visualizer with developer
tools
- Haptic Specialists to act
as advisors
- Hapkit
- Drone pilots
- Soldiers / team leaders
-Need current methods
used to mark friendly
forces
-Support of an automated
threat detection system
Beneficiaries
Mission AchievementMission Budget/Costs
Buy-In/Support
Deployment
Value
Proposition
Key Activities
Key Resources
Key Partners Beneficiaries:
Operators
Buyers (S&T,
acquisitions)
SOCOM
Channels
S&T SOCOM
SBIR SOCOM
or DoD
DARPA BAAs
Buy-In
Low-cost
Replaceable
COTS
15. What we thought:
Operators want to have
automatic detection AND
classification of humans in
video feed
What we did:
- Even more interviews!
- Explore current technology
(is classification feasible? At
what confidence levels?)
Weeks 7-9
Lesson:
Soldiers need to trust the UAV to make
accurate classifications.
Object Detection is a good starting point
16. Mission Achievement
Organization Beneficiary Mission Achievement
SOCOM Tactical operators - Human detection
- Location of humans detected
- Autonomous flight
PMs Get the right equipment to troops at a reasonable cost.
DARPA ATAK PM Get ATAK utilized in multiple military organizations, create a
plugin ecosystem to add value over time
TAK sUAS plugin team Fully implemented plugin features
DHS Border Patrol Improved identification and tracking of suspected illegal border
crossers.
Potential commercial
partners
Kespry/OceanIT/motionDSP Decreased R&D Costs, new revenue streams, relationships with
US Government customers. Improvements to existing technology
Skynet Team Skynet (Us) Awarded SBIR phase 1 & 2
OR
Transition to S&T acquisition process (via SOCOM) or corporate
17. Partners and Benefits
Computer Vision Experts (Stanford
Computer Vision Lab)
AFRL
J. (independent engineer)
DARPA
Activities Resources
Partners
People Recognition
Geolocation of person
ATAK Integration
Easy UI/UX
Pre-plan Routes
Autonomous Flight
Web Servers / Dataset Tagging,
Training Data,
Algorithm / Model
Camera Metadata (DJI API),
Drone Location data
ATAK Source Code,
ATAK Testing
User Feedback
Litchi,
Open Source Navigation Code,
Drones
Trained Deep Learning Model,
Drones
ATAK Team
- DARPA
- AFRL
SOCOM Operators
Litchi (?)
Open source drone software community
TBD
18. Mission Model Canvas- Week 7-9
SkyNetUI: Mission Model Canvas
- SOF, Rangers, AWG
- Law Enforcement
Agencies
- Drone manufacturers (ex:
DJI)
- Haptic startups
- Augmented Reality
Visualization Companies
(ex: Google Glass)
- Threat/Friendly detection
- using
light/sound/motion/indicat
or/marking
-HCI - feedback
mechanism
-Making system easy to
use/learn/maintain
- Primary: SOF elements
and Ground Force
Commanders
- Secondary: Command
elements at TOC/TAC
-Tertiary: Search and
Rescue teams
-Enhance situational
awareness - Drone
becomes another set of
eyes, frees up an operator.
-Reduced risk to force-
Drone goes so humans do
not have to
- video recording to
facilitate AARs in training
-Low cost alternative to
costly assets- frees up
other assets in a resource
constrained environment
- Reduces dependence on
human capital
-Improved situational awareness for small units in combat
-Demand across SOF elements and requests for fielding from
conventional units
- Field test with one unit.
Evaluate training method.
- Expand field-tests to
multiple units in same
deployment context
- With successful field
tests, contract COTS
vendors to mass produce.
Fixed:
- Equipment - drones, development toolkits,
- Hapkit: Haptic Starter kit ($50 x 4)
- Software design & engineering
Variable:
- Costs of maintenance, updates, and training
- Drones w/ camera
(provided by SOCOM)
- Augmented Reality
Visualizer with developer
tools
- Haptic Specialists to act
as advisors
- Hapkit
- Drone pilots
- Soldiers / team leaders
-Need current methods
used to mark friendly
forces
-Support of an automated
threat detection system
Beneficiaries
Mission AchievementMission Budget/Costs
Buy-In/Support
Deployment
Value
Proposition
Key Activities
Key Resources
Key PartnersPartners:
DARPA
AFRL
Computer
Vision Experts
SOCOM
Resources
Data,
algorithms
ATAK source
code
Drone
Key Activities
Computer
Vision
ATAK
Integration
Drone Hacking
Mission Achievement
Low-cost COTS drones that increase
situational awareness
Help DARPA start a program on drones
19. Week 9 MVP
“How am I supposed to feel vibration when I’m
running around being shot at when I can’t feel my
phone vibrate in my pocket walking to the
bathroom?”
“How does it know that someone is bad? I don’t
want to go to jail…”
“That’s badass”
“Can you demo this in July”
“I oddly feel more comfortable knowing there is a skynet
drone watching out for me”
Week 2 MVP
From this... ...to this
20. No Money
20192017 2018 2020 20212016
Q1 Q2 Q1 Q2 Q1 Q2 Q1 Q2 Q1 Q2Q1 Q2 Q3 Q4 Q3 Q4 Q3 Q4 Q3 Q4 Q3 Q4
$500k
$1m
$5m
Product
Milestones
MVP
Businesses
Milestones
Beta 2.0 Release
Class ends
Demo
events
Apply SOCOM S&T, DARPA BAAs, SBIRs
SOCOM S&T Funding or DARPA BAA funding
1.0 Release
DIUx
Demo
events
Demo
events
Sell product to large DoD
contractor
Find other Agency S&T Funding
Try to become program of record
while sustaining individual sales
Apply Other Agency S&T Funding
Finance and Operations Timeline
Q1 Q2Q1 Q2 Q3 Q4
21. Who Helped us Get Here
Wayne Chen
Michael Hard
Brian CS
Jooyong Lee
Andrew Smallwood
Mentors:
Army Rangers, Green Berets, Navy SEALs, Air Force
Pararescuemen, and Norwegian SOF
Sponsor:
Tactical Operators:
Thanks Brendon!
23. Appendix
Final mission model canvas
Pictures
Tech stack
Key partners, resources, activities
Progression of beneficiaries
Progression of value proposition
Acquisitions & deployment (timeline, get keep grow diagram, S&T process)
24. Hardware:
DIY custom drone kit
Camera (color / thermal)
Software:
Open-sourced firmware
Drone
ATAK Cursor On Target*
*Schema much like JSON
Androidplugin
Skynet
Computer Vision (openCV) Drone navigation
Object
detection
Object
recognition DJI sdk / open-sourced sdk
Android device
What we have now
25. What we learned
About the product
Refrain from building a new
system
Integrate with prefered existing
systems (ATAK)
COTS system - flexible, less
expensive
About the process
Get out of the building!
Distill the common pains and
gains
Understand the customer
Dig deep but look broad
26. Mission Model Canvas
-Software development
computer vision on drone
feeds
-People detection
-Geolocation of person
populated on map
-Integration with ATAK and
existing DoD tech
-Autonomous drone
hacking-not manufacturing
- Hack COT drone like DJI
- Partner with drone maker
- Become sensor neutral
DoD
- SOCOM: S&T and
acquisition PMs
- DARPA PMs
ATAK Integration
- ATAK Programmers in
DARPA and AFRL
Computer Vision
- Computer vision experts
- Increase situational
awareness through computer
vision, detection of people
and objects
- Provide low-cost method of
reconnaissance
-Easy UI and use through
integration w/ ATAK
-Added value to current
program/technology
-Provide demo to pitch new
program to Director of
DARPA
- Low-cost detection of
movement and people at
border
- Increase ability to respond
to sensors
- Fulfill S&T Topic of Interest requirements
- Establish and meet Key Performance Parameters (KPPs)
- Achieve 80% accuracy rate on detecting people, with few false positives
- Start new program for small UAVs in DARPA
- Fulfill technical wishlist for ATAK
- Improved identification and tracking of suspected illegal border crossers.
- Achieve sustainable business selling in S&T phase to multiple agencies
Q3, 2016 (May – June): $5,925 ($407 / Week)
Q4, 2016 (July – Sept): $16,847 ($1,296 / Week)
Q1, 2017 (Oct – Dec): $129,000 ($9,923 / Week)
Q2, 2017 (Jan – March): $249,600 ($19,200 / Week)
-Demonstrate utility + low-
cost- gain support from
senior personnel
-Add utility to current program
of record to increase adoption
-Start a company or get
acquired by another company
Beneficiaries
Mission AchievementMission Budget/Costs
Get awareness w/ demo
days, technical experiment.
conferences, etc.
-S&T UAV Topic of Area--
submit white paper
-Exchange expertise and
collaborate
-Talk with PMs, submit to
BAA-16-31
-Demo TE, TILO, and Ranger
training events in July
Buy-In/Support
Deployment
Value PropositionKey Activities
Funding/Money, Access
ATAK Integration
-ATAK source code
-DJI SDK or open-source
Computer Vision
-Algorithms (CV lab)
- AMT / AWS
-Training data: blimp or
drone footage (AFRL)
Autonomous Drone (DJI)
Key Resources
Key Partners
SOCOM
- Tactical operators - SOF
(SF, Ranger, SEALS,
JSOC, MARSOC/Recon,
drone operators)
- SOCOM PMs- Small UAV
S&T PM, Acquisition PM,
DARPA
Tactical Technology
Office-SquadX and future
Small UAV program
ATAK Program Manager
AFRL
DHS
Border Patrol
Skynet
Because you asked for more color coding . . .
27. What we did
Customer Discovery
(Moffett field, Week 3)
ATAK MVP (Week 4)
Processing a drone video feed (Week 4/5)
28. ATAK mockup (Week 7)
Product planning (Week 8)
Week 9 MVP
What we did
30. The MVP Evolution
“How am I supposed to feel
vibration when I’m running
around being shot at when
I can’t feel my phone
vibrate in my pocket
walking to the bathroom?”
“I need to know without
a doubt that they’re
holding a gun and not a
baby”
“As far as
recognition,
that’s huge.”
“Are you sure it’s not
a baby? I really don’t
want to go to jail.”
V0.01 V0.1 Current
31. - Primary: SOF elements
and Ground Force
Commanders
- Secondary: Command
elements at TOC/TAC
-Tertiary: Search and
Rescue teams
Our Beneficiaries
COMBAT: SOF in a combat
environment (TIC)
--JTACs
--Team leader
NON-COMBAT: SOF in a
non-combat environment
DUAL USE: Search and
Rescue teams
COMBAT: SOF and Rangers
in a combat environment
(TIC)
--JTACs
--Team leader
NON-COMBAT: SOF in a
non-combat environment,
interacting and training with
allied troops
DUAL USE: Search and
Rescue teams
Tactical operators - SOF
(SF, Ranger, SEALS,
MARSOC/Recon, drone
operators)
Private sector and academic
partners (Neurala, drone
manufacturer, Stanford lab)
ATAK Program Manager
SOCOM Small UAV
S&T/Acquisition Program
Managers
Dual Use: Border Patrol
Week 0 Week 2 Week 3 Week 6
32. -Enhance situational
awareness - Drone becomes
another set of eyes, frees up
an operator.
-Reduced risk to force-
Drone goes so humans do
not have to
-Live video feed
-Low cost alternative to
costly assets- frees up other
assets in a resource
constrained environment
Minimize new load that
needs to be carried
- Reduces dependence on
human capital
Value proposition
COMBAT: Drone uses visual
recognition to separate
enemies, friendlies, and
noncombatants-- prevents
collateral damage and
fratricide
NON-COMBAT: Drone flies
route autonomously,
increases awareness of
surroundings, and could do
patrols
- Low cost alternative to
costly assets
Minimize new load that needs to be
carried
- Reduces human capital
needs-- frees up man to
fight
- Identify and locate hostile
and neutral actors
- Enhanced ability to
conduct real time recon (esp
at night).
- Enable extra soldier in fight
- Autonomously fly routes
and identify potential
trouble.
- Extending communications
(e.g. inter-unit, with allied
troops)
- Conduct perimeter
surveillance
- Gather intelligence on
nearby residents and
distinguish between regular
and irregular behavior of
civilians in close proximity to
military installations
- Keep track of friendly,
potential hostile and neutral
actors/equipment
automatically
- New contracts and revenue
streams
- Opportunities for tech
development
-Added value to current
program/technology
-Cost effective, agile and
quick method of filling
capability gaps
- Low-cost detection of
movement and people at
border
Week 0 Week 2 Week 3 Week 6
33. Week 9
SOCOM
- Tactical operators - SOF
(SF, Ranger, SEALS,
MARSOC/Recon, drone
operators)
- SOCOM PMs- Small UAV
S&T PM, Acquisition PM,
Kespry, OceanIT,
motionDSP, Stanford CV
lab, drone manufacturers
DARPA
Tactical Technology Office-
Persistent Air Support
Project
ATAK Program Manager
DHS
Border Patrol
Skynet
- Increase situational
awareness through computer
vision, recognition of people
and objects
- Provide low-cost method of
reconnaissance
-Easy UI and use through
integration w/ ATAK
- New contracts and revenue
streams through the DoD
- New use cases for product
- Reputation and awareness
-Added value to current
program/technology
- Low-cost detection of
movement and people at
border
- Increase ability to respond
to sensors
product - market fit
Value Proposition Beneficiaries
35. S&T Process - Prototype to Product
BAA for
existing gap
White
Paper
White
Paper
White
Paper
S&T PM
(TBD)
S&T PM’s Group of
Operators
Other operators
who talk to
operators in PM’s
group
Bureaucracy line
Other operators
who talk to
operators in PM’s
group
S&T “Council”:
S&T PMs +
Director
Engineer
Proposal
Contract Negotiations
between business
and PM: Statement of
Work (COW);
Contracts Data
Requirements List
(CDROLS)
Step 1: find gap on
fedbizops + submit
whitepaper
Step 2: All whitepapers get
evaluated by operators
and engineer
Step 3: S&T PM brings selected
white paper to “council” of PMs.
They choose which projects to
prioritize and fund.
Step 4+5: Submit proposal
and negotiate deliverables
(CDROLS) and what you
will do (COW)
Step 0: “Free flowing” communication with operators
and PM to show off product and get selected and
prioritized
36. Mission Model Canvas- Week 1
- Threat / Friendly detection
using light / sound / motion /
indicator / or marking
-HCI - feedback mechanism
Haptic = bad
familiar = really important
-Making system easy to
use/learn/maintain
- SOF, Rangers, AWG
- Law Enforcement Agencies
- Drone manufacturers (ex:
DJI)
- Haptic startups
- Augmented Reality
Visualization Companies (ex:
Google Glass)
- Primary: SOF elements
and Ground Force
Commanders
- Secondary: Command
elements at TOC/TAC
-Tertiary: Search and
Rescue teams
-Enhance situational
awareness - Drone becomes
another set of eyes, frees up
an operator.
-Reduced risk to force-
Drone goes so humans do
not have to
- video recording to facilitate
AARs in training Live video
feed
-Low cost alternative to
costly assets- frees up other
assets in a resource
constrained environment
Minimize new load that
needs to be carried
- Reduces dependence on
human capital
-Improved situational awareness for small units in combat
-Demand across SOF elements and requests for fielding from conventional
units
- Field test with one unit.
Evaluate training method.
- Expand field-tests to
multiple units in same
deployment context
- With successful field tests,
contract COTS vendors to
mass produce.
Fixed:
- Equipment - drones, development toolkits,
- Hapkit: Haptic Starter kit ($50 x 4)
- Software design & engineering
- Augmented reality hardware & software (equivalent of Google Glasses and
relevant developer toolkits)
Variable:
- Costs of maintenance, updates, and training
-Need current methods used to mark
friendly forces
- Rank and file soldiers, unit
leaders, and commanders.
- Support of an automated
threat detection system
- Military engineers to
facilitate integration with
systems
Beneficiaries
Mission AchievementMission Budget/Costs
Buy-In/Support
Deployment
Value PropositionKey Activities
Key Resources
Key Partners
- Drones w/ camera
(provided by SOCOM)
- Augmented Reality
Visualizer with developer
tools
- Access to existing gear and
standard equipment
- Haptic Specialists to act as advisors
- Hapkit
- Drone pilots
- Soldiers / team leaders
37. Mission Model Canvas- Week 2
- Threat/friendly/noncombat.
detection using visual
recognition light/sound/
motion/indicator/or marking
-HCI - feedback mechanism
-Making system easy to
use/learn/maintain
- Integration with prefered
sit. awareness platforms,
including smart phones (As a
Service)
- SOF, Rangers, AWG
- Law Enforcement Agencies
- Drone manufacturers (ex:
DJI)
- Sparrow (Stanford CS210
Autonomous Drone project)
- Haptic startups
- Augmented Reality
Visualization Companies (ex:
Google Glass)
-DARPA
COMBAT: SOF in a combat
environment (TIC)
--JTACs
--Team leader
NON-COMBAT: SOF in a
non-combat environment
DUAL USE: Search and
Rescue teams
COMBAT: Drone uses visual
recognition to separate
enemies, friendlies, and
noncombatants-- prevents
collateral damage and
fratricide
NON-COMBAT: Drone flies
route autonomously,
increases awareness of
surroundings, and could do
patrols
- Low cost alternative to
costly assets
Minimize new load that needs to be
carried
- Reduces human capital
needs-- frees up man to fight
-Improved situational awareness for small units in combat
-perform manpower intensive tasks in non-combat roles
-Reduce fratricide and collateral damage
- Identify program manager
- Secure funding + contract
from sponsor
- Field test small-scale
- Contract a manufacturer for
large scale production
- Deploy iteratively, for each
integrate with a new system
- Continue relationship via
maintenance + integration
service
Fixed:
- Equipment - drones, development toolkits,
- Hapkit: Haptic Starter kit ($50 x 4)
- Software design & engineering
- Augmented reality hardware & software (equivalent of Google Glasses and
relevant developer toolkits)
Variable:
- Costs of maintenance, updates, and training
-Need current methods used to mark
friendly forces
- Rank and file soldiers, unit leaders,
and commanders.
- Support of an automated threat
detection system
- Military engineers to facilitate
integration with systems
??????
Beneficiaries
Mission AchievementMission Budget/Costs
Buy-In/Support
Deployment
Value PropositionKey Activities
Key Resources
Key Partners
- Drones w/ camera
(provided by SOCOM)
- Augmented Reality
Visualizer with developer
tools
- Access to existing gear and
standard equipment
- Haptic Specialists to act as advisors
- Hapkit
- Drone pilots
- Soldiers / team leaders
38. Mission Model Canvas- Week 3
- Combatant classification
and threat detection
- Integration with ATAK &
other military systems
- Operable without GPS
- SOF, Rangers, AWG
- Law Enforcement Agencies
- Autonomous Drone
manufacturer (eg: Kespry)
- ATAK Design group
- Sparrow (Stanford CS210
Autonomous Drone project)
- Augmented Reality
Visualization Companies (ex:
Osterhout Design Group)
-DARPA
COMBAT: SOF and Rangers
in a combat environment
(TIC)
--JTACs
--Team leader
NON-COMBAT: SOF in a
non-combat environment,
interacting and training with
allied troops
DUAL USE: Search and
Rescue teams
- Identify and locate hostile
and neutral actors
- Enhanced ability to conduct
real time recon (esp at night).
- Enable extra soldier in fight
- Autonomously fly routes
and identify potential trouble.
- Extending communications
(e.g. inter-unit, with allied
troops)
- Conduct perimeter
surveillance
- Gather intelligence on
nearby residents and
distinguish between regular
and irregular behavior of
civilians in close proximity to
military installations
-Improved situational awareness for small units in combat
-perform manpower intensive tasks in non-combat roles
-Reduce fratricide and collateral damage
- Secure funding + contract
- Small-scale testing
- Contract a manufacturer for
large scale production
- Deploy iteratively, integrate
with a new systems and
provide lessons on use
- Continue relationship via
maintenance, integration
service, and training
Fixed:
- Equipment - drones, development toolkits,
- Software design & engineering
- Augmented reality hardware & software (equivalent of Google Glasses and
relevant developer toolkits)
Variable:
- Costs of maintenance, updates, and training
25k - relatively quick to
acquire
-for entirely new technology,
writeup to SOCOM
Saboteur - competition with
other groups
Beneficiaries
Mission AchievementMission Budget/Costs
Buy-In/Support
Deployment
Value PropositionKey Activities
Key Resources
Key Partners
- DJI Drones w/ Dev. tools
- AR Device w/ Dev. tools
- ATAK API
- Access to existing gear and
standard equipment
39. Mission Model Canvas- Week 4
- Identifying and tracking
friendlies
- Object detection
- Object recognition,
classification and threat
detection
- Integration with ATAK
- SOF (green beret, ranger,
SEALS)
- SOCOM
- Programmable drone
manufacturers (eg: Solo)
- ATAK Program
- Neurala (or other computer
vision firm)
Tactical operators - SOF
(SF, Ranger, SEALS,
MARSOC/Recon, drone
operators)
Private sector partners
(Neurala, drone
manufacturer)
ATAK Program Manager
SOCOM Small UAV
S&T/Acquisition Program
Managers
- Keep track of friendly,
potential hostile and neutral
actors/equipment
automatically
- New contracts and revenue
stream
- Opportunities for tech
development
-Added value to current
program/technology
-Cost effective, agile and
quick method of filling
capability gaps
-Improved situational awareness for small units in combat
-reduce workload for tactical operators and feed-monitoring operators
-decrease strain on other traditional aerial assets
-Deploy to select teams
under $25k level. Hardware
purchased COTS
-Incorporate commercial
partner software in app dev
-Work with ATAK to include
future app update with our
capabilities
-Work to make program of
record after buy in
Fixed:
- Equipment - drones, development toolkits
- Software design & engineering
Variable:
- Costs of maintenance, updates, and training
-Demonstrate utility- gain
support from senior personnel
-Show opportunities for new
revenue streams
-Add utility to current program
of record to increase adoption
-Provide low cost product
with new capability filling gap
Saboteur - competition within
SOCOM PMs
Beneficiaries
Mission AchievementMission Budget/Costs
Buy-In/Support
Deployment
Value PropositionKey Activities
Key Resources
Key Partners
- DJI phantom 3
programmable drone
- Computer vision SDK
(preferably neurala)
- ATAK API + source code
40. Mission Model Canvas- Week 5
- Fly autonomously on route
- Object detection,
recognition, classification
- Identifying and tracking
friendlies, enemies, civilians
- Integration with ATAK
- SOF (green beret, ranger,
SEALS)
- SOCOM, RRTO, CTTSO,
JCTD, CTO
- Programmable drone
manufacturers (eg: Solo)
- ATAK Program
- Neurala (or other computer
vision firm)
-Stanford Computer Vision
Lab
Tactical operators - SOF
(SF, Ranger, SEALS,
MARSOC/Recon, drone
operators)
Private sector and academic
partners (Neurala, drone
manufacturer, Stanford lab)
ATAK Program Manager
SOCOM Small UAV
S&T/Acquisition Program
Managers
Dual Use: Border Patrol
- Keep track of friendly,
potential hostile and neutral
actors/equipment
automatically
- New contracts and revenue
streams
- Opportunities for tech
development
-Added value to current
program/technology
-Cost effective, agile and
quick method of filling
capability gaps
- Low-cost detection of
movement and people at
border
-Improved situational awareness for small units in combat
-reduce workload for tactical operators and feed-monitoring operators
-decrease strain on other traditional aerial assets
Options:
1). S&T Process
2). SBIR Process
3). Find existing contract
(short term funding)
Fixed:
- Equipment - drones, development toolkits
- Software design & engineering
Variable:
- Costs of maintenance, updates, and training
-Demonstrate utility- gain
support from senior personnel
-Show opportunities for new
revenue streams
-Add utility to current program
of record to increase adoption
-Provide low cost product
with new capability filling gap
Saboteur - competition within
SOCOM PMs
Beneficiaries
Mission AchievementMission Budget/Costs
Buy-In/Support
Deployment
Value PropositionKey Activities
Key Resources
Key Partners
- Programmable Drone
- Computer vision SDK
(preferably neurala)
- ATAK API + source code
__= New
41. Mission Model Canvas- Week 6
- Fly autonomously on route
- Object detection,
recognition, classification
- Identifying and tracking
friendlies, enemies, civilians
- Integration with ATAK
- SOF (green beret, ranger,
SEALS)
- SOCOM, RRTO, CTTSO,
JCTD, CTO
- Programmable drone
manufacturers (eg: Solo)
- ATAK Program
- Neurala (or other computer
vision firm)
-Kespry
-Stanford Computer Vision
Lab
Tactical operators - SOF
(SF, Ranger, SEALS,
MARSOC/Recon, drone
operators)
Private sector and academic
partners (Neurala, drone
manufacturer, Stanford lab)
ATAK Program Manager
SOCOM Small UAV
S&T/Acquisition Program
Managers
Dual Use: Border Patrol
- Keep track of friendly,
potential hostile and neutral
actors/equipment
automatically
- New contracts and revenue
streams
- Opportunities for tech
development
-Added value to current
program/technology
-Cost effective, agile and
quick method of filling
capability gaps
- Low-cost detection of
movement and people at
border
-Improved situational awareness for small units in combat
-reduce workload for tactical operators and feed-monitoring operators
-decrease strain on other traditional aerial assets
Demo days, technical
experimentation
conferences, etc.
S&T or SBIR Process,
depending on level of
prototype
“How to” youtube videos,
customer support
Fixed:
- Equipment - drones, development toolkits
- Software design & engineering
Variable:
- Costs of maintenance, updates, and training
-Demonstrate utility- gain
support from senior personnel
-Show opportunities for new
revenue streams
-Add utility to current program
of record to increase adoption
-Provide low cost product
with new capability filling gap
Saboteur - competition within
SOCOM PMs
Beneficiaries
Mission AchievementMission Budget/Costs
Buy-In/Support
Deployment
Value PropositionKey Activities
Key Resources
Key Partners
- Programmable Drone
- Computer vision SDK
- ATAK API + source code
__= New
42. Mission Model Canvas- Week 7
- Fly autonomously on route
Phase 1: Computer vision:
recognize people, vehicles
- Integration with ATAK
Phase 2: Computer vision:
classify people into
friendlies, enemies, civilians
- Stand alone from ATAK
- Augmented Reality
- SOF (green beret, ranger,
SEALS)
- SOCOM, RRTO, CTTSO,
JCTD, CTO
- ATAK Program
-Kespry, OceanIT,
motionDSP
-Stanford Computer Vision
Lab
- Programmable drone
manufacturers (eg: Solo)
-DARPA
- Increase situational
awareness of potential
threats and collateral
damage
- Provide low-cost method of
reconnaissance
- New contracts and revenue
streams
- Opportunities for tech
development
-Added value to current
program/technology
- Low-cost detection of
movement and people at
border
- Increase ability to respond
to sensors
- A job, money
- Fulfill S&T Topic of Interest requirements
- Establish and meet Key Performance Parameters (KPPs)
- Discover what’s possible under SBIR OSD-162-003X
- Achieve 80% accuracy rate on detecting people, with few false positives
- Fulfill the DARPA wish list of technical features
- Improved identification and tracking of suspected illegal border crossers.
Get awareness w/ demo
days, technical experiment.
conferences, etc.
-S&T UAV Topic of Area--
submit white paper
-SBIR Phase 1 Grant
“How to” youtube videos,
customer support
-Exchange expertise and
collaborate
Fixed:
- Equipment - drones, development toolkits
- Software design & engineering
Variable:
- Costs of maintenance, updates, and training
-Demonstrate utility + low-
cost- gain support from
senior personnel
-Show opportunities for new
revenue streams
-Add utility to current program
of record to increase adoption
-Start a company or get
acquired by another company
Beneficiaries
Mission AchievementMission Budget/Costs
Buy-In/Support
Deployment
Value PropositionKey Activities
Key Resources
Key Partners
- Programmable Drone
- Computer vision SDK
- ATAK API + source code
__= New
SOCOM
- Tactical operators - SOF
(SF, Ranger, SEALS,
MARSOC/Recon, drone
operators)
- SOCOM PMs- Small UAV
S&T PM, Acquisition PM,
Kespry, OceanIT,
motionDSP, Stanford CV
lab, drone manufacturers
DARPA
Tactical Technology Office-
Persistent Air Support
Project
ATAK Program Manager
DHS
Border Patrol
Skynet
43. Skynet Origins
Original Problem Statement: Traditional UAV video platforms require a tactical operator to control the UAV and
watch the video, which takes them out of the fight and has a high cognitive load. Our goal is to develop an -
autonomous drone that flies routes autonomously and gives reliable threat information without forcing an
operator to take his hands off the trigger or eyes off the scope.
Our Initial Idea: Since the root of the problem is cognitive overload, which stems from poorly designed UI and UX
systems, we aimed to create an improved Human-Drone Interaction system that could be scaled across
platforms.
90 In Person interviews (30 people surveyed)
Our product would target the Small Unmanned Aerial System (SUAS) market
TAM SAM (SUAS)
SOCOM 3,000 units 800-1000 units
DOD 18,210 units 17,296 units
44. Mission Model Canvas- Week 8
-Software development
computer vision on drone
feeds
-People recognition
-Geolocation of person
populated on map
-Integration with existing
DoD tech
-Autonomous drone
hacking-not manufacturing
- Hack COT drone like DJI
- Partner with drone maker
- Become sensor neutral
Funding from DoD
- SOCOM
- S&T and acquisition PMs
- Contracting PMs
Computer Vision
- Amazon--AWS
- Movidius
-MotionDSP, OceanIT
-Stanford CV Lab
-Independent computer
vision experts/contractors?
ATAK Integration
- ATAK Programmers
- DARPA
- “JL” for geolocation
Autonomous Drones
-DJI
-Kespry
- Programmable drone
manufacturers (eg: Solo)
- Increase situational
awareness through computer
vision, recognition of people
and objects
- Provide low-cost method of
reconnaissance
-Easy UI and use through
integration w/ ATAK
- New contracts and revenue
streams through the DoD
- New use cases for product
- Reputation and awareness
-Added value to current
program/technology
- Low-cost detection of
movement and people at
border
- Increase ability to respond
to sensors
- Fulfill S&T Topic of Interest requirements
- Establish and meet Key Performance Parameters (KPPs)
- Discover what’s possible under SBIR OSD-162-003X
- Achieve 80% accuracy rate on detecting people, with few false positives
- Find long-term customers in DoD
- Fulfill the DARPA wish list of technical features
- Improved identification and tracking of suspected illegal border crossers.
Get awareness w/ demo
days, technical experiment.
conferences, etc.
-S&T UAV Topic of Area--
submit white paper
-SBIR Phase 1 Grant
“How to” youtube videos,
customer support
-Commercial partnership
-Exchange expertise and
collaborate
Computer Vision
- Computer Vision expert: over $100 per hour
-MotionDSP- free
- Data- ??
ATAK- free
Drones- $1k and up
Other- costs of maintenance, updates, and training
-Demonstrate utility + low-
cost- gain support from
senior personnel
-Show opportunities for new
revenue streams
-Add utility to current program
of record to increase adoption
-Start a company or get
acquired by another company
Beneficiaries
Mission AchievementMission Budget/Costs
Buy-In/Support
Deployment
Value PropositionKey Activities
Key Resources
Key Partners
Funding/Money
Computer Vision
-Computing power
-Algorithms
-Training data: blimp or
drone footage
ATAK Integration
-ATAK source code
-DJI SDK or open-source
Autonomous Drone
__= New
SOCOM
- Tactical operators - SOF
(SF, Ranger, SEALS,
MARSOC/Recon, drone
operators)
- SOCOM PMs- Small UAV
S&T PM, Acquisition PM,
Kespry, OceanIT,
motionDSP, Stanford CV
lab, drone manufacturers
DARPA
Tactical Technology Office-
Persistent Air Support
Project
ATAK Program Manager
DHS
Border Patrol
Skynet