Visible to the public File preview

Biomedical
 Cyber-­‐Physical
 Systems
 
  You
 Can
 Bet
 Your
 Life
 On
 
NSF
 CPS
 PI
 mee>ng
 2011-­‐08-­‐01
 
Patrick
 Lincoln
  Computer
 Science
 Laboratory
  SRI
 Interna>onal
 

Outline
 
•  Future-­‐present
 robo>cs
 
–  Augmen>ng
 human
 skills,
 safety,
 and
 experiences
 

•  Safety
 cases
 for
 robo>c
 systems
 
–  Evidence-­‐based
 cer>fica>on
 for
 life-­‐cri>cal
 systems
 

•  Future-­‐present
 biology
 
–  Augmen>ng
 human
 health,
 produc>vity,
 and
  environment
 

•  Safety
 cases
 for
 biological
 systems
 
–  Evidence-­‐based
 cer>fica>on
 for
 environment-­‐cri>cal
  biological
 systems
 

Future-­‐Present
 Robo>cs
 
Augmen>ng
 human
 skills,
 safety,
 and
  experiences
  Example:
 Telepresence
 Surgical
  Robo>cs
 

Origins
 of
 Remote
 Manipula>on
 Robo>cs
 
•  Leonardo
 da
 Vinci
 1464
  •  Human-­‐shape
 automaton
  •  Designed
 to
 raise
 arms,
 
  open
 visor,
 etc.
  •  Used
 cables
 and
 pullies
  to
 actuate
  •  Designs
 lost
 for
 500
 years
 
–  Rediscovered
 1950
 
 

Remote
 manipula>on
 of
 hazardous
 materials
 
 
•  Robert
 Heinlein’s
 1942
 science
 fic>on
 “Waldo”
  •  Raymond
 Goertz
 (Argonne
 Na>onal
 Lab)
 and
 others
 developed
  Master-­‐Slave
 Manipulators
 “Waldos”
 for
 radioac>ve
 handling
 in
  1950s
 

Origins
 
 

Teleopera>on
 of
 Virtual
 Systems
 
•  Brooks
 at
 University
 of
 
  North
 Carolina
 at
 
  Chapel
 Hill
 1988,
 1990
 

SRI
 Telepresence
 Surgery
 
•  Phil
 Green’s
 team
 at
 SRI
 created
 world’s
 1st
 complete
  telepresence
 surgical
 systems
 in
 1980s
 and
 early
 1990s
 
–  Primarily
 funded
 by
 DARPA
 for
 remote
 military
 surgery
  –  Built
 on
 NIH
 funded
 experiments
 at
 SRI
 and
 Stanford
 University
  –  Also
 built
 on
 NASA
 funding
 for
 remote
 teleopera>on
 in
 space
 
•  NASA
 Flight
 Telerobo>c
 Servicer
 (1980s)
 

–  Dexterous
 minimally
 invasive
 surgical
 tools –  Intui>ve
 user
 interface
 


 

•  Successful
 demonstra>ons,
 though
 no
 long-­‐range
 on-­‐ bablefield
 (let
 alone
 in-­‐space)
 deployment
  •  SRI
 has
 many
 patents
 issued
 worldwide
 for
 the
 key
  components
 (now
 licensed
 to
 Intui>ve
 Surgical)
 

The
 Basic
 Approach
 
•  Human
 operator
 puts
 hands
 on
 
  master
 controllers
 
  •  Master
 system
 uses
 forward
 
  kinema>cs
 to
 compute
 desired
 
  pose
 of
 end
 effector
  •  Master
 computer
 communicates
  to
 slave
 control
 computer
 over
 
  a
 digital
 network
  •  Slave
 computer
 applies
 inverse
 kinema>cs
  to
 compute
 required
 robot
 arm
 and
 wrist
 angles
  •  Live
 stereo
 video
 is
 fed
 back
 to
 operator
  •  (op>onal)
 Sensed
 forces
 on
 slave
 effectors
 communicated
 back
  through
 similar
 system,
 providing
 hap>c
 feedback
 

Other
 Pioneers
 in
 Robo>c
 Surgery
 

•  Russel
 Taylor
 at
 IBM
 Watson
 Research
 Center
 and
 Mark
 Talamini
  at
 Johns
 Hopkins
 developed
 the
 Laparoscopic
 Assistant
 Robot
 
  •  Hari
 Das
 at
 JPL
 NASA-­‐funded
 Robot
 Assisted
 Microsurgery
 (RAMS)
 
  •  Yulan
 Wang
 at
 UC
 Santa
 Barbara
 developed
 a
 robo>c
 system
 Zues
  NASA-­‐funded
 SBIR
 seeded
 Computer
 Mo>on
 Inc.
 
 
–  Computer
 Mo>on
 acquired
 by
 Intui>ve
 Surgical
 in
 2003
 

•  Ken
 Salisbury
 at
 MIT
 developed
 innova>ve
 hap>cs
 systems
 
–  Later
 he
 joined
 Intui>ve
 Surgical,
 now
 Stanford
 professor
 

•  Brian
 Davies
 at
 Imperial
 College
 PROBOT
  •  Plus
 several
 other
 academic
 and
 industrial
 efforts
 

Intui>ve
 Surgical
 
•  Spun
 out
 from
 SRI
 in
 1996
 
–  Large
 porjolio
 of
 SRI
 patents
 and
 prototypes
  –  Entrepreneurs
 John
 Fruend,
 Dr.
 Frederick
 Moll,
 and
  Roberge
 Younge
 
  –  Several
 SRI
 staff
 members,
 
  including
 current
 CEO
 Gary
 Guthart
  –  Venture
 funding
 from
 Mayfield,
 Sierra,
  and
 Morgan
 Stanley
 

Forming
 a
 Venture:
 
  Intui>ve
 Surgical
 
•  SRI
 spun
 out
 Intui>ve
 Surgical
  in
 1996
  •  ISRG
 Refined
 SRI
 system
 into
  “Lenny”
 1997
  •  Created
 daVinci
 robot
 1998
  •  First
 robo>c-­‐assisted
 heart
  bypass
 1998
  •  First
 bea>ng-­‐heart
 robo>c-­‐
  assisted
 heart
 bypass
 1999
  •  IPO
 in
 April
 2000
  •  FDA
 approval
 in
 2003
  •  ISRG
 market
 cap
 today:
 $15B
 

Nurses
 at
 bedside,
 surgeon
 a
 few
 steps
 away
 

Concept
 of
 Opera>ons
 

Impact
 of
 Telepresence
 Surgery
 
•  Many
 types
 of
 surgery
 improved:
 
 
–  Urology,
 Gynecology,
 Cardiothoracic,
 General
 Surgery,
  Colorectal,
 Head
 &
 Neck,
 Pediatric
 

•  ~2,000
 installed
 daVinci
 robots
 installed
  •  Nearing
 one
 million
 surgeries
 total
  •  Direct
 benefits:
 
 

#1
 treatment
 op>on
 for
 prostate
 and
 gynecological
 cancer
 

+
 Reduced
 risk
 of
 infec>on
  +
 Less
 pain
 and
 scarring
  +
 Less
 blood
 loss
 and
 less
 need
 for
 blood
 transfusions
  +
 Shorter
 hospital
 stay
 (2-­‐5
 days
 less
 for
 cardiac)
  +
 Faster
 recovery
 and
 return
 to
 normal
 ac>vi>es
  –  Note:
 Capital
 cost
 $1+M
 per
 robot,
 $1+K
 consumables
 

Example
 Impact
 on
 Cardio
 Bypass
 
•  •  •  •  •  •  •  •  •  •  No
 sternotomy
  No
 8-­‐10”
 cut
 through
 chest
  No
 cuts
 through
 sternum
  No
 cracking
 of
 ribs
  Shorter
 >me
 on
 table
  Shorter
 recovery
 >me
  Less
 blood
 loss
  Less
 pain
 and
 scarring
  Quicker
 return
 to
 normal
 ac>vi>es
  Less
 morbidity
 

Abstrac>on
 Enables
 
Purng
 a
 computer
 between
 surgeon
 and
 pa>ent
  enables
 certain
 advantages
 

•  Scaling
 up
 or
 down
  •  Virtually
 altering
 or
 stopping
 mo>on
 
 

Safety
 cases
 for
 robo>c
 systems
 
Evidence-­‐based
 cer>fica>on
 for
 life-­‐cri>cal
 systems
 

Next
 Ques>on:
 How
 Assured?
 
•  Original
 system
 used
 unreliable
 transport
 network
  •  Sotware
 and
 hardware
 originally
 constructed
  using
 standard
 engineering
 prac>ces
 
–  Not
 bad,
 but
 not
 perfect
 

•  Engineering
 cannot
 aim
 for
 perfec>on
 
99%
 yes.
 
 
 
 99.999%
 yes.
 
 
 
 
 
 100%
 no.
 

•  What
 level
 of
 assurance
 is
 appropriate
 
  for
 this
 type
 of
 system?
 

See:
 Medical
 Devices
 and
 Public
 Health,
 2011
  •  Approvals
 require
 extensive
 documenta>on,
 
  laborious
 tes>ng,
 rigorous
 science,
 expert
  review.
  This
 enables
 principled
 approval
 of
 new
 things
 
–  For
 drugs,
 not
 devices
 

FDA
 Approvals
 and
 Clearances
 

•  Medical
 devices
 are
 cleared,
 not
 approved,
  through
 the
 510(k)
 process
 
–  510(k)
 arises
 from
 1976
 congressional
 authorizing
  legisla>on
  –  Main
 topic
 of
 510(k):
 “substan>al
 equivalence”
 to
 

Defini1on
 of
 Substan1al
 Equivalence
 in
  1990
 Safe
 Medical
 Device
 Amendments
 
A.
 For
 purposes
 of
 determina>ons
 of
 substan>al
 equivalence
 .
 .
 .
 the
 term
  “substan>ally
 equivalent”
 or
 “substan>al
 equivalence”
 means,
 with
  respect
 to
 a
 device
 being
 compared
 to
 a
 predicate
 device,
 that
 the
 device
  has
 the
 same
 intended
 use
 as
 the
 predicate
 device
 and
 that
 [FDA]
 by
 order
  has
 found
 that
 the
 device
 –
 
(i)
 has
 the
 same
 technological
 characteris>cs
 as
 the
 predicate
 device,
 or
  (ii)
 has
 different
 technological
 characteris>cs
 and
 the
 informa>on
 submibed
  that
 the
 device
 is
 substan>ally
 equivalent
 to
 the
 predicate
 device
 contains
  informa>on,
 including
 clinical
 data
 if
 deemed
 necessary
 by
 FDA,
 that
  demonstrates
 that
 the
 device
 is
 as
 safe
 and
 effec>ve
 as
 a
 legally
 marketed
  device
 and
 does
 not
 raise
 different
 ques1ons
 of
 safety
 and
 efficacy
 than
 the
  predicate
 device.
 

B.
 For
 purposes
 of
 subparagraph
 (A),
 the
 term
 “different
 technological
  characteris>cs”
 means,
 with
 respect
 to
 a
 device
 being
 compared
 to
 a
  predicate
 device,
 that
 there
 is
 a
 significant
 change
 in
 the
 materials,…
 

Ensuring
 Safety
 and
 Effec>veness
 
 
 vs.
  Promo>ng
 New
 Innova>ve
 Medical
 Devices
 
•  1997
 FDA
 Moderniza>on
 act
 
 
–  Directs
 FDA
 to
 require
 “least
 burdensome”
 level
 of
  scien>fic
 evidence
 for
 manufacturers
 to
 assert
  substan>al
 equivalence
 

•  FDA
 abemp>ng
 to
 foster
 innova>on,
 but
 balance
  need
 for
 safety
 and
 evidence
 of
 effec>veness
 

Mathema>cian’s
 Issues
 With
 510(k)
 
•  Base
 case:
 no
 reason
 to
 assume
 everything
 
  used
 before
 1976
 is
 safe
 and
 effec>ve
  •  Induc1on
 case:
 broad
 defini>on
 of
 substan>ally
  equivalent
 may
 mean
 devices
 with
 really
 new,
  novel
 technology
 cleared
 without
 rigorous
  evidence
 of
 safety
 and
 effec>veness
 
 

Example
 Challenges
 in
 Verifica>on
 that
 CPS
  biomedical
 systems
 meet
 their
 requirements
 
•  •  •  •  •  •  •  •  Ethical
 tes>ng
 of
 the
 unproven
 on
 human
 subjects
  Interoperable
 devices,
 inter-­‐device
 interference
 
  Composability
 
  Lifecycle
 and
 maintenance
 issues
  Metrics
 and
 measurement
 
  Malicious
 aback
  Hybrid
 (discrete
 and
 analog)
 control
  Regulatory
 staffing
 (vs
 peer
 review)
 

Current
 FDA
 efforts
 are
 making
 progress
 on
 some
 of
  these
 challenges,
 such
 as
 assurance
 case
 frameworks
 

A
 Way
 Forward,
 How
 You
 Can
 Help
 
•  Create
 new
 approval
 procedure
 for
 de
 novo
 medical
  devices,
 and
 for
 new
 technologies
 for
 equivalents
 
 
+
 Evidence-­‐based
 medicine,
 formal
 methods
  +
 Expand
 #
 of
 applica>ons
 that
 cite
 clinical
 evidence
 

•  Like
 the
 safety
 cases
 for
 avionics
 and
 other
  industries,
 enable
 reasonable
 procedures
 and
  prac>ces
 based
 on
 rigorous
 scien>fic
 principles
  •  Enable
 post-­‐market
 monitoring
 of
 safety
 and
  effec>veness
 

+
 Many
 in
 the
 Cyber-­‐Physical-­‐Systems
 community
 could
 be
  very
 helpful
 to
 this
 process
  +
 Many
 in
 the
 HCSS
 /
 CyberTrust
 communi>es
 could
 be
  helpful
 in
 ensuring
 privacy
 and
 security
 

Looking
 to
 Other
 Industries:
  Consider
 Fly-­‐By-­‐Wire
 
•  What
 computer
 would
 you
 feel
 comfortable
  purng
 between
 the
 pilot
 and
 the
 wings
 of
 the
  aircrat
 you
 will
 fly
 home
 on?
 
–  Digital
 fly-­‐by-­‐wire
 avionics
 is
 now
 commonplace
 

•  Classic
 goal
 of
 nine-­‐nines
 in
 avionics
 
–  One
 system
 failure
 in
 a
 billion
 hours
 of
 use
  –  Prac>cally
 untestable:
 >1000
 planes
 flying
 >100
 years
  –  What
 evidence
 other
 than
 tes>ng
 should
 be
 gathered
  for
 a
 new
 aircrat
 type?
  –  Led
 by
 NASA
 and
 FAA,
 standards
 and
 prac>ces
 for
  safety
 cases
 exist
 and
 are
 in
 regular
 use
 

Can
 We
 Show
 Medical
 Robots
 Operate
 Within
  Specified
 Parameters
 Despite
 Faults?
 
•  Latency,
 Speed,
 Responsiveness,
 Accuracy,
 etc.
  Leverage
 ancient
 history
 of
 high-­‐assurance
 machines
  •  Byzan>ne
 fault-­‐tolerance
 machines:
 NASA
 and
 SRI’s
 SIFT,
  Allied’s
 MAFT,
 Draper’s
 FTP,
 Vienna
 MARS,
 AIPS
  •  Fundamental
 academic
 work
 
  in
 distributed
 systems
  And
 their
 formal
 analysis
  •  Reduc>on
 of
 ques>ons
 of
 interest
 to
  symbolic
 calcula>on:
 EHDM,
 PVS,
 ACL
  Much
 high-­‐quality
 research
 and
 
  development
 in
 academia
 and
 
  industry,
 
 including
 much
 performed
  or
 funded
 by
 speakers
 and
 abendees
  here
 at
 the
 NSF
 CPS
 mee>ng
 
 

Prac>cality
 of
 Assured
 Surgical
 Robo>cs?
 
•  Recent
 advances
 in
 formal
 methods
 make
 prac>cal
 the
  analysis
 of
 complex
 CPS
 systems
 such
 as
 medical
 robots
  •  Example
 project:
 SimCheck
 

–  Safety,
 Reliability,
 and
 Resilience
 of
 M7
 slave
 unit
  –  Matlab
 Simulink
 models
 of
 robot
 and
 control
 system
  –  PVS
 and
 Yices
 used
 to
 analyze
 proper>es
 of
 models
  –  Natarajan
 Shankar,
 John
 Rushby,
 Sam
 Owre,
 Bruno
 Dutertre
  –  Supported
 by
 NASA
 Coopera>ve
 Agreement
 NNX08AY53A
 
  and
 NSF
 Grant
 CSR-­‐EHCS(CPS)-­‐0834810
 

•  Other
 example
 projects
 at
 Berkeley,
 UPenn,
 MIT,…
 
–  Including
 speakers
 today
 

Next
 Steps
 
•  Can
 the
 lessons
 learned
 here
 and
 tools
 developed
  help
 analyze
 infusion
 pumps,
 insulin
 pumps,
 heart
  monitors,
 pacemakers,
 and
 other
 new
 medical
  devices?
  •  Can
 we
 build
 a
 tool
 bus
 to
 integrate
 many
 analysis
  engines
 for
 designing
 high-­‐assurance
 cyber-­‐physical
  biomedical
 systems?
  •  What
 kind
 of
 assurance
 case
 can
 we
 build
 for
 such
  devices?
  •  What
 kind
 of
 architecture
 (with
 sotware
 health
  management)
 yields
 the
 strongest
 assurance
 case?
 

Future-­‐Present
 Biology
 
Augmen>ng
 human
 health,
  produc>vity,
 and
 environment
 

Future
 Direc>ons
 for
 Biomedical
 CPS
 
•  Small
 assays
  •  Fast
 assays
  •  Precise
 biochemical
 actua>on
 (Synthe>c
 Biology)
 

Future
 Direc>ons
 for
 Biomedical
 CPS
  Extremely
 Small
 Assays
 
•  Today
 many
 assays
 are
 performed
 on
 large
  popula>ons
 of
 cells,
 averages
 are
 reported
  •  Move
 to
 single
 cell
 assays
 
–  Flow
 cytometry
 (Herzenbergs,
 Stanford)
 
•  15-­‐color
 cell
 sorter
 

–  Nanoliter
 PCR
 (Farris,
 SRI)
 
•  Single-­‐cell-­‐content
 PCR
 

–  Nanowire
 voltmeter
 (Lieber,
 Harvard)
 
•  30
 simultaneous
 electrical
 readings
 on
 single
 cell
 

Future
 Direc>ons
 for
 Biomedical
 CPS
  Extremely
 Fast
 Assays
 
•  Today
 assays
 are
 performed
 over
 hours
 or
 days
  •  Tomorrow
 can
 we
 move
 to
 real-­‐>me
 assays?
 
–  Real-­‐>me
 (outpa>ent
 in
 clinic)
 blood
 assays
 
•  Can
 we
 tell
 if
 pa>ent
 was
 exposed
 to
 pathogen,
 toxin,
 or
  radia>on
 from
 a
 blood
 sample,
 before
 they
 leave
 the
 clinic?
 

–  Dialysis-­‐like
 control
 systems
 
•  Can
 we
 enable
 more
 sensing
 and
 >ghter
 controls,
 enabling
  dialysis-­‐like
 treatment
 of
 sepsis,
 rapidly
 mi>gate
 shock,
 etc?
 

–  Embedded
 medical
 devices
 
 
•  Can
 we
 enable
 long-­‐term
 implantable
 medical
 devices
 to
  sense
 and
 actuate
 to
 improve
 health
 and
 wellness?
  •  Insulin
 pumps,
 pacemakers,
 and
 others
 

Example
 Enabler:
 Really
 Really
 Rapid
 PCR
  Greg
 Faris,
 SRI
 
•  Laser
 hea>ng
 of
 nanoliter
 droplets
 allows
  extremely
 fast
 polymerase
 chain
 reac>on
 (PCR)
  amplifica>on
 of
 DNA
 and
 RNA
  •  One
 of
 fastest
 PCR
 methods
  >1000
 PCR-­‐base-­‐pair
 cycles
 per
 minute
 
40
 amplifica>on
 cycles
 of
 a
 186
 base
 pair
 amplicon
 in
 370
 s
 

•  Amplifica>on
 of
 the
 contents
 of
 single
 cell
 
  demonstrated
 

PCR
 Products
 in
 Droplet
 Array
  Real
 Time
 PCR
 in
 Single
 Droplet
 

Laser
 Hea>ng
 of
 Droplet
 
H.
 Kim,
 S.
 Dixit,
 C.
 J.
 Green,
 and
 G.
 W.
 Faris,
 “Nanodroplet
 real-­‐>me
 PCR
 system
 with
 laser
 assisted
 hea>ng,”
  Opt.
 Express
 17,
 218-­‐227
 (2009).
  H.
 Kim,
 S.
 Vishniakou,
 and
 G.
 W.
 Faris,
 “Petri
 dish
 PCR:
 laser-­‐heated
 reac>ons
 in
 nanoliter
 droplet
 arrays,”
  Lab
 Chip
 9,
 1230-­‐1235
 (2009).
 

Precise
 Biological
 Actua>on:
 
  Synthe>c
 Biology
 
Defini1on
 of
 Synthe1c
 Biology:
 
  the
 design
 and
 construc1on
 of
 new
 biological
 parts,
 devices,
 and
  systems,
 and
 the
 re-­‐design
 of
 exis1ng,
 natural
 biological
 systems
 
  for
 useful
 purposes
  Synthe>c
 Biology
 is
 a
 new
 approach
 to
 engineering
 biology,
 with
 an
  emphasis
 on
 technologies
 to
 write
 DNA.
 Founda>onal
 work,
  including
 the
 standardiza>on
 of
 DNA-­‐encoded
 parts
 and
 devices,
  enables
 them
 to
 be
 combined
 to
 create
 programs
 to
 control
 cells.
 

Costs
 of
 Synthe>c
 Biology
 
•  The
 longest
 synthesized
 DNA
 sequence
 has
 been
  growing
 on
 a
 rapid
 exponen>al
 curve
 
–  It
 will
 likely
 slow
 as
 the
 u>lity
 of
 many
 megabase
  sequence
 synthesis
 is
 limited
 by
 design
 tools
 

•  More
 importantly,
 the
 cost
 of
 DNA
 sequencing
 is
  now
 low
 and
 con>nues
 to
 drop
 exponen>ally
  •  Also,
 the
 cost
 of
 DNA
 synthesis
 con>nues
 to
  drop,
 though
 somewhat
 more
 slowly
 

Moore’s
 Law
 &
 Carlson
 Curves
  The
 Cost
 of
 Fablines
 
•  The
 cost
 of
 produc>on
 for
 chips
 (especially
 the
  capital
 required
 for
 a
 fab)
 is
 rising
 
–  Though
 not
 rising
 as
 fast
 as
 in
 the
 past
  –  Astounding
 capital
 commitment
 is
 required
 (>$5B)
 

•  The
 cost
 of
 produc>on
 for
 biology
 is
 falling
 

Emerging
 Synthe>c
 Biology
 Community
 
•  Synthe>c
 Biology
 1.0,
 2.0,
 3.0,
 4.0,
 5.0
 
–  Led
 by
 Tom
 Knight,
 Drew
 Endy,
 and
 Randy
 Rhetberg
 

•  Growing
 the
 community
 from
 the
 bobom
 up
 
–  Already
 great
 interna>onal
 interest
 

Safety
 cases
 for
 biological
 systems
 
•  Evidence-­‐based
 cer>fica>on
 for
 environment-­‐ cri>cal
 biological
 systems
 

Assessing
 Risks
 of
 Synthe>c
 Biology
 
•  Presiden>al
 Commission
 for
 the
 Study
 of
 Bioethical
 Issues
  recommenda>on:
 
 
–  Risk
 Assessment
 Prior
 to
 Field
 Release
  –  See:
 “NEW
 DIRECTIONS
 The
 Ethics
 of
 Synthe>c
 Biology
 and
  Emerging
 Technologies”
 December
 2010
 

•  Risk
 Assessment
 Prior
 to
 Field
 Release
 

–  Reasonable
 risk
 assessment
 should
 be
 carried
 out,
 under
 the
  Na>onal
 Environmental
 Policy
 Act
 or
 other
 applicable
 law,
 prior
 to
  field
 release
 of
 research
 organisms
 or
 commercial
 products
  involving
 synthe>c
 biology
 technology.
 This
 assessment
 should
  include,
 as
 appropriate,
 plans
 for
 staging
 introduc>on
 or
 release
  from
 contained
 laboratory
 serngs.
 Excep>ons
 in
 limited
 cases
  could
 be
 considered,
 for
 example,
 in
 emergency
 circumstances
 or
  following
 a
 finding
 of
 substan1al
 equivalence
 to
 approved
 products
 

Risk Assessment Prior to Field Release and Substantial Equivalence Determination
How
 do
 we
 go
 about
 this?
  Living
 systems
 are
 wickedly
 complicated
  Our
 knowledge
 is
 extremely
 limited
  Our
 ability
 to
 accurately
 model
 and
 predict
  behaviors
 of
 a
 given
 organism
 is
 extremely
  limited
  •  Our
 ability
 to
 accurately
 predict
 changes
 in
  systems,
 such
 as
 DNA
 muta>on,
 is
 extremely
  limited
  •  •  •  • 

Rigorous
 Abstract
 Methods
 Are
 Needed
  To:
 
•  Accommodate
 conven>onal
 types
 of
 discrete
  reasoning
 based
 on
 experimenta>on
 
  •  Unambiguously
 define
 a
 model
 and
 allowable
  reasoning
 steps
  •  Provide
 predic>ve
 power
 for
 genera>ng
 testable
  hypotheses
 
 

A
 Way
 Forward,
 How
 You
 Can
 Help
 
•  Create
 new
 analysis
 methods
 for
 de
 novo
 biological
  devices,
 and
 for
 new
 technologies
 for
 equivalents
 
 
+
 Evidence-­‐based
 synthe>c
 biology,
 formal
 methods,
 pathway
  logic,
 pathway
 tools
 

•  Like
 the
 safety
 cases
 for
 avionics
 and
 other
 industries,
  enable
 reasonable
 procedures
 and
 prac>ces
 based
 on
  rigorous
 scien>fic
 principles
 

•  Enable
 post-­‐release
 monitoring
 of
 gene>cally
 modified
  and
 synthe>c
 organisms
 

+
 Can
 we
 close
 the
 gap
 that
 exists
 in
 design
 tools
 in
 this
 domain?
  +
 Many
 in
 the
 Cyber-­‐Physical-­‐Systems
 community
 could
 be
 very
  helpful
 to
 this
 process
  +
 Many
 in
 the
 HCSS
 /
 CyberTrust
 communi>es
 could
 be
 helpful
  in
 ensuring
 privacy
 and
 security
 

The
 End