mission model, mission model canvas, customer development, Hacking for Defense, lean startup, stanford, startup, steve blank, Pete Newell, Joe Felter, minimum viable product
AMERICAN LANGUAGE HUB_Level2_Student'sBook_Answerkey.pdf
Xplomo Hacking for Defense 2017
1. Standoff IED Detection Using UAVs
Week 1: Original Focus
Hardware - A small, portable drone:
● that uses different sensors to;
● replace human capabilities in
detecting IEDs
Week 10: Final Focus
Software - A sense-making system:
● that uses image processing
techniques to;
● augment human capabilities in
detecting IEDs; and
● scales across platforms.
103
Interviews
PLOMOJIDO
Sponsor
3. Yicheng An Weihan Zhang Robert André
Borochok
Marko Jakovljevic
M.S. Computer
Vision & Machine
Learning
M.S. Business M.S. Management
Science and
Engineering
Postdoc Imaging
School of Medicine
Software
Engineering
Law Enforcement /
Strategy &
Operations
Industrial
Engineering /
Operations
Radiology / Image
Processing
3
PLOMO
10. “Infantrymen are the most
vulnerable to IED
attacks...they are trained to
detect IEDs, but rely mainly
on visual cues.”
- Operational Commander
PLOMO
11. Focusing on Our Primary Beneficiary
PRIMARY:
DISMOUNTED
INFANTRY
Most vulnerable to
attack
Relies on visual cues
(potholes, disturbed
earth)
Needs standoff detection
capability
Load is a major constraint
Needs near real-time
detection
11
PLOMO
12. Week 3: Mission Model Canvas
Value Proposition
Replace infantrymens’
capabilities to detect IEDs
A drone system
System to analyze drone
feed and indicate risk
areas to war fighter
Beneficiaries
Primary: Dismounted
infantry patrolling known
area
12
PLOMO
13. Let’s build a counter-
IED drone for ground
infantry...
PLOMO
14. ...Or not.
“We already use surveillance
drones…why would we need
another one?”
- Operational Commander
PLOMO
15. Let’s Modify Existing Hardware!
Build a Drone Add Sensors to Existing
Drones
PIVOT
PLOMO
16. Pivot to Sensor Improvements
PLOMO
Hypothesis:
Sensors added to
drones like Raven will
increase situational
awareness and ease
of IED detection.
Reality:
Sensors on their own
do not give analytical
insights
17. Pivot to Sensor Improvements
PLOMO
Hypothesis:
Sensors added to
drones like Raven will
increase situational
awareness and ease
of IED detection.
Reality:
Sensors aren’t
perfect
“There is no vapor for
chem sensors to sniff
in an open
environment.”
- Explosive Signature
Specialist
18. Pivot to Sensor Improvements
PLOMO
Hypothesis:
Sensors added to
drones like Raven will
increase situational
awareness and ease
of IED detection.
Reality:
Adding sensors will
require long
deployment time
“JIDO can’t … add a
sensor to a program
of record.”
- JIDO Tech Chief
29. Week 6: Mission Model Canvas
Key Partners
JIDO J8/J6
Image Processing
Experts
Mission Achievement
- Algorithm that successfully detects potholes
with a false alarm rate <5 per frame
- At least 80% accuracy
Buy-in/Support
JIDO
Troops in field
Military Leadership
Mission Achievement
- Algorithm that successfully detects potholes with a
false alarm rate <5 per frame
- At least 80% accuracy
30. In the rush to develop a
working product,
you’ve got to
Fail Fast, Move Quick
PLOMO
38. Thank You
JIDO
Wayne A. Stanbery
W. Richards Thissell
James McGuyer
Industry Mentor
Kevin Ray
Robert Medve
Special Contributors
Robert Best
Caitlin Cima
Todd Forsman
Andrea Gilli
George Hasseltine
Rafi Holtzman
Michael Leone
David Zinn
Special Thanks
Camp Pendleton Commanders and Staff
& to all our 103 Interviewees
38
Teaching Team
Steve Blank
Joseph Felter
Peter Newell
Steve Weinstein
Teaching Assistants
Darren Hau
Isaac Matthews
Melisa Tomak
PLOMO