© 2004 the mitre corporation. all rights reserved engineering a complex system: the air & space...
TRANSCRIPT
© 2004 The MITRE Corporation. All rights reserved
Engineering a Complex System:
The Air & Space Operations Center (AOC) as a
Complex Systems Exemplar
Doug Norman
Senior Technical Advisor, AOC-WS
Dept Head, AF Battle Management / Command & Control
22 Oct 2004
2004 MIT SDM Alumni Conference
“Innovative Product and
System Development”
© 2004 The MITRE Corporation. All rights reservedSDM 2004
Caveat
The author's affiliation with The MITRE Corporation is provided for identification purposes only, and is not intended to convey or imply MITRE's concurrence with, or support for, the positions, opinions or viewpoints expressed by the author.
© 2004 The MITRE Corporation. All rights reservedSDM 2004
Outline for Discussion
What is an AOC? What’s the issue? Why Rethink Systems Engineering? Complexity and Complex Systems Engineering Complex Systems
© 2004 The MITRE Corporation. All rights reservedSDM 2004
Roche cites AOC as “Top Weapon”
© 2004 The MITRE Corporation. All rights reservedSDM 2004
TBMCS: the “Engine” of the AOC
’03 Association for Enterprise Integration (AFEI) winner – Excellence in Enterprise Integration
© 2004 The MITRE Corporation. All rights reservedSDM 2004
An AOC
© 2004 The MITRE Corporation. All rights reservedSDM 2004
What an AOC Does:e.g.Global Strike CONOPS
AOC WS is the Key C2ISR node and enabler for GSTF
© 2004 The MITRE Corporation. All rights reservedSDM 2004
How it does it:Today’s ATO Process
JFCJFCCOMPONENTCOMPONENT
COORDCOORD
JOINTJOINT
ATOATODEVELOPMENTDEVELOPMENT
WEAPONEERINGWEAPONEERINGALLOCATIONALLOCATION
FORCEFORCEEXECUTIONEXECUTION
ASSESSMENTASSESSMENT STRATEGYSTRATEGYDEVELOPMENT DEVELOPMENT
FindFind
FixFix
TrackTrack
TargetTarget
EngageEngage
AssessAssess
TST
© 2004 The MITRE Corporation. All rights reservedSDM 2004
How it does it:AOC Block 10.1 System List
1. Collection Management Mission Application2. Command and Control Information Processing System3. Command and Control Personal Computer4. Generic Area Limitation Environment Lite5. Global Command and Control System6. Global Decision Support System7. Global Transportation Network8. Imagery Product Library9. Information Warfare Planning Capability10. Interim Targeting Solution11. Joint Deployable Intelligence Support System12. Operational Model Exploiting GPS Accuracy13. PC Integrated Intelligence and Imagery14. Planning and Decision Aid System15. Personnel Recovery Mission Software16. Portable Flight Planning System17. RAINDROP18. Requirement Management System19. Space Battle Management Core System20. Theater Battle Management Core System21. Theater Weather Server22. Worldwide Origin Threat System23. Weapons System Video24. Combat Survivor/Evader Locator25. Intelligence Surveillance Reconnaissance Manager26. Time Critical Targeting – F27. All Source Satellite Evaluation Tool28. Commanders Tactical Terminal29. Generic Area Limitation Environment30. Global Command and Control System – I331. PowersceneServices32. Air Operations Net33. Broadsword34. Defense Message System35. Global Broadcast System36. INTELINK and INTELINK-S37. Joint Collaboration Environment38. NSA Threat Warning Net39. Predator Video40. Purple Net41. Tactical Data Information Exchange System-Broadcast
42. Tactical Intelligence Broadcast System43. Tactical Related Applications44. CSP AUTODIN45. Internet Relay Chat46. Combat Track II47. Hummingbird Exceed48. JWARN49. Global Hawk Access50. Geospatial Product Library (Eagle Express)51. Multi-Media Message Manager (M3 AMHS)52. Outlook Web Access (OWA)53. Sky MediaInfrastructure54. Data wall55. Domain Core56. Perimeter Security System57. JICO Support System58. Radiant Mercury59. Imagery Support Server Environment Guard60. Community of Interest Network61. C2 Weapon System Part Task Trainer62. Joint Worldwide Intelligence Comm System63. Sensitive but Unclassified Internet Protocol Net64. Secure Internet Protocol Router Network65. Tactical Data Links in formats A, B, J66. Air Defense System Integrator – TSQ-21467. Deployable Transit-case System68. Joint Tactical Air Ground System69. Precision Lightweight Global Position Receiver70. Tactical Data Processing Suite71. Tactical Data Terminal72. Tactical Receive Suite 73. Air Force DoDIIS Infrastructure (aka JEDI)74. AOC Security Portal75. Access Net76. CENTRIX-S77. Gigabyte Ethernet78. GPS Timing79. Network Appliance Filer System (NAS)80. STAMPS81. Multi-Level Security
Capabilities• Improved ATO production
& dissemination • Common air picture• Initial Formal Training
Unit (FTU) infrastructure• Initial Battlespace Visual.• Initial TCT Capability • Initial Plans Automation • Initial ISR Management • Limited Coalition Interop• Transportable – limited
deployability• Improved M2M• Initial Info Services
Capability • Airspace management• Collaboration (chat)• Comm/Info management• Common weather picture• IW/Space• Air Mobility Coordination• Combat Search and
Rescue
© 2004 The MITRE Corporation. All rights reservedSDM 2004
What is Traditional Systems Engineering (TSE)?
Definition taken from the International Council on Systems Engineering (INCOSE – www.incose.org ):
“Systems Engineering is an interdisciplinary approach and means to enable the realization of
successful systems. It focuses on defining customer needs and required functionality early in the
development cycle, documenting requirements, then proceeding with design synthesis and system
validation while considering the complete problem:
– Operations
– Performance
– Test
– Manufacturing
– Cost & Schedule
– Training & Support
– Disposal
Systems Engineering integrates all the disciplines and specialty groups into a team effort forming a
structured development process that proceeds from concept to production to operation. Systems
Engineering considers both the business and the technical needs of all customers with the goal of
providing a quality product that meets the user needs.”
What are the boundary conditions for this to be successful?
© 2004 The MITRE Corporation. All rights reservedSDM 2004
Boundary conditions for TSE
Boundaries for successful applications of TSE:– The specific desired outcome must be known a priori, and it
must be clear and unambiguous (implied in this is that the edges of the system, and thus responsibility, are clear and known);
– There must be a single, common manager who is able to make decisions about allocating available resources to ensure completion;
– Change is introduced and managed centrally;
– There must be “fungible” resources (that is money, people, time, etc.) which can be applied and reallocated as needed.
© 2004 The MITRE Corporation. All rights reservedSDM 2004
The tension between users and acquirers… User-assessed value is related to:
– Immediate usefulness to the operators
– Agility of potential interconnections and interconnected elements
– Acceptance of innovation Both technical and operational
The Operator’s Problem with current Acquisition as the source of AOCs:
– Late to need
– Too bureaucratic
– Seldom delivers “what’s needed”
– Too expensive
– Too-early demand for unknown (and potentially unknowable) answers to needs
– Hard to argue with, and win – they tend to have the documentation justifying what they’ve done But often, what they do “feels wrong”
The problem is probably Systems Engineering – and that which it implies and includes
– Why Rethink Systems Engineering?
– Complexity and Complex Systems
– Engineering Complex Systems
– Complex Systems Engineering in Practice
© 2004 The MITRE Corporation. All rights reservedSDM 2004
Observation 1: No grand design possible
The AOC System of Systems is an opportunistic aggregation, not a design– Only the AOC System Program Office (SPO), which has the
acquisition responsibility, has a strong interest in an AOC design, yet it has no way to impose its design on others,
– As currently structured, since many of the elements don’t belong to the AOC SPO, and the AOC SPO has the responsibility for the integration, delivery, and operational admin of the AOC, the AOC SPO has insufficient authority to execute its responsibilities,
– Since the AOC doesn’t spend its money for many of the elements, the element-owners have little incentive to comply with an AOC design,
– The need for, and the appearance of, a specific new tool or application at the AOC is often driven by a new, immediate need.
© 2004 The MITRE Corporation. All rights reservedSDM 2004
Multiple System-ownersESC-Managed Systems in Red
Mission Applications1. Collection Management Mission Application (Navy)2. Command and Control Information Processing System (AMC)3. Command and Control Personal Computer (USMC)4. Generic Area Limitation Environment Lite (NRO)5. Global Command and Control System (ESC/DI)6. Global Decision Support System (AMC)7. Global Transportation Network (TRANSCOM)8. Imagery Product Library (NIMA)9. Information Warfare Planning Capability (ESC/SR)10. Interim Targeting Solution (AFRL)11. Joint Deployable Intelligence Support System (NMIC)12. Operational Model Exploiting GPS Accuracy (ESC/NDC)13. PC Integrated Intelligence and Imagery (ESC/ACJ)14. Planning and Decision Aid System (NSA)15. Personnel Recovery Mission Software (JPRA)16. Portable Flight Planning System (ESC/ACU)17. RAINDROP (COTS)18. Requirement Management System (DIA)19. Space Battle Management Core System (ESC/NDC)20. Theater Battle Management Core System (ESC/ACF)21. Theater Weather Server (ESC/ACW)22. Worldwide Origin Threat System (ESC/NDC)23. Weapons System Video (AF/SCM)24. Combat Survivor/Evader Locator (SMC/CZJ)25. Intelligence Surveillance Reconnaissance Manager (ESC/SR)26. Time Critical Targeting – F (ESC/ACF)27. All Source Satellite Evaluation Tool (NRO)28. Commanders Tactical Terminal (ESC/SR)29. Generic Area Limitation Environment (NRO)30. Global Command and Control System – I3 (ESC/DI)31. Powerscene (n/a)Services32. Air Operations Net (n/a)33. Broadsword (AFRL)34. Defense Message System (SSG)35. Global Broadcast System (ESC/MC)36. INTELINK and INTELINK-S (n/a)37. Joint Collaboration Environment (COTS)38. NSA Threat Warning Net (n/a)39. Predator Video (ESC/SRG)40. Purple Net (n/a)41. Tactical Data Information Exchange System-Broadcast (ESC/SR)*
42. Tactical Intelligence Broadcast System (ESC/SR)*43. Tactical Related Applications (ESC/SR)*44. CSP AUTODIN45. Internet Relay Chat (COTS)46. Combat Track II (ESC/SRK)47. Hummingbird Exceed (COTS)48. JWARN (USMC SC)49. Global Hawk Access (ASC/RAV)50. Geospatial Product Library (Eagle Express) (NIMA)51. Multi-Media Message Manager (M3 AMHS) (ESC/SR)52. Outlook Web Access (OWA) (COTS)53. Sky MediaInfrastructure54. Data wall (n/a)55. Domain Core (n/a)56. Perimeter Security System (ESC/ACF)57. JICO Support System (ESC/ACF)58. Radiant Mercury (ARFL)59. Imagery Support Server Environment Guard (ARFL)60. Community of Interest Network (n/a)61. C2 Weapon System Part Task Trainer (AFC2TIG)62. Joint Worldwide Intelligence Communications System (n/a)63. Sensitive but Unclassified Internet Protocol Network (n/a)64. Secure Internet Protocol Router Network (n/a)65. Tactical Data Links in formats A, B, J (n/a)66. Air Defense System Integrator – TSQ-214 (ESC/DI)67. Deployable Transit-case System (ESC/SRG)68. Joint Tactical Air Ground System (Army)69. Precision Lightweight Global Position Receiver (n/a)70. Tactical Data Processing Suite (ESC/SR)*71. Tactical Data Terminal (ESC/SR)*72. Tactical Receive Suite (ESC/SR)*73. Air Force DoDIIS Infrastructure (aka JEDI) (AFRL)74. AOC Security Portal (n/a)75. Access Net76. CENTRIX-S (n/a)77. Gigabyte Ethernet (n/a)78. GPS Timing (n/a)79. Network Appliance Filer System (NAS) (n/a)80. STAMPS81. Multi-Level Security (n/a)
Over 30 separate PEs are represented in the AOC
© 2004 The MITRE Corporation. All rights reservedSDM 2004
Observation 2: Issues with “Integration”
Integration-enabling technologies, and integration developments (glueware), are grafted onto the elements (systems) of the AOC after delivery– Each element in the aggregate is designed and built with its own
understanding of the world – around its own set of “conceptual atoms”
Integration among these elements requires effort to understand and bring these potentially disparate “conceptual atoms” in line so they can be composed
– Integration is a source of work and revenue – using today’s business model (employer/contractor) contractors sell engineering hours
“Big Integration” is a potential cash-cow for those who perform it Little incentive to limit the work, or find ways to be more effective
– Late integration guarantees the delivery of an integrated, operational AOC will lag behind the availability of the individual elements; however, the expectation from the field is that general availability and integrated are synonymous. This potential disappointment is further compounded by the need to expend additional funds for the integration proper.
© 2004 The MITRE Corporation. All rights reservedSDM 2004
Observation 3: Funds for Integration
Funds for integration are limited– Ability for user to wait is limited– Barriers for building automated functionality (in software) are
low, setting expectations that it’s easy and quick– Integration tends to be built around a defined work flow which
implements a specific concept of operation. Integration “glue” which implements the CONOP binds systems into rigid relationships. This is contrary to achieving “agility” and “netcentricity.”
– Testing issues
© 2004 The MITRE Corporation. All rights reservedSDM 2004
Observation 4: “Planning” and Expectations
“Planning” as a primary SoS strategy has problems– Focuses on the future – but is based in the past
Imposes expectations, and dependencies, on partially-interested participants
Design implied in the plan is based on today’s understandings. As things change in the world all the elements to be composed are subject to different pressures and decisions which likely will not align
Measures of Success based on promises – not achievements
– Plans focus on “things” rather than “collaborations about things”
© 2004 The MITRE Corporation. All rights reservedSDM 2004
Where
we s
pend our r
esource
s
What we i
gnore
Observation 5: What consumes our time and resources?
Half-life of entityShort Long
Bre
adth
of
utility
Narrow
Wide
OperatorPosition
Tools
DecisionSupportTools
SharedUnderstanding
(Language)
Storage, Access
Technologies &
Information
Distribution
Services
Loose Coupling
Tight Coupling
InformationServices
Community’sInformationArchitecture
Waters the eyes
Considered boring… mere details
Assumed Knowledge
© 2004 The MITRE Corporation. All rights reservedSDM 2004
The Issue(s), so far (condensed) The Gov’t (and industry in general – world wide, in fact) has been
unable to build large systems, or integrate systems into larger collections predictably.– This is as true for the AOC as it is for others…– How do we position ourselves and our engineering activities to avoid
this outcome? For the most part, today’s systems are not composable. The
systems:– Don’t share a common conceptual basis.– Aren’t built for the same purpose, or for use within specific work flows,
or for use exclusively at AOCs,– Share an acquisition environment which pushes them to be “stand
alone“ (regardless of any statements to the contrary),– Have no common control or management,– Don’t share common funding which can be directed to “problems” as
required,– Many of the systems have many “customers,” of which the AOC is only
one,– Evolve at different rates (as do system components).
Integrating the AOC is an unbounded, unpredictable engineering activity
© 2004 The MITRE Corporation. All rights reservedSDM 2004
Can TSE be applied to the AOC?
Boundaries for successful applications of TSE:– The specific desired outcome must be known a priori, and it
must be clear and unambiguous (implied in this is that the edges of the system, and thus responsibility, are clear and known);
Not true for the AOC
– There must be a single, common manager who is able to make decisions about allocating available resources to ensure completion;
Not true for the AOC
– Change is introduced and managed centrally; Not true for the AOC (although it attempts to be)
– There must be “fungible” resources (that is money, people, time, etc.) which can be applied and reallocated as needed.
Not True for the AOC
TSE does not fit the context of the AOC as currently structured
© 2004 The MITRE Corporation. All rights reservedSDM 2004
So… How can we characterize this new context? A metaphor: the Watchmaker and the Gardener
– A useful (set of) metaphor(s) for framing the differences in problem spaces
The watchmaker – has a specific outcome– Develops and uses an a priori design which has been well-
vetted, and well-analyzed Can “prove” the design – it is correct & complete Can price the design – all requirements/elements are known and
stable Can schedule the creation
The gardener – has an outcome space– Develops a desired outcome; generally an “architecture-like”
statement of acceptability Specific result not known a priori Sense – and – respond to conditions as they emerge Guides garden into the desired outcome space
Sounds like a “complexity” problem?
© 2004 The MITRE Corporation. All rights reservedSDM 2004
What is Complexity?
A measure of potentiality It does not mean “Difficult to understand” Contrast with “Intricacy” Mousetrap example
Mousetrap® Game by Milton Bradley
Intricate Complex
http://www.ento.csiro.au/science/ants/pests.htm
Ants at honey
Intricate Complex
Shannon numberLow High
brittle resilient
© 2004 The MITRE Corporation. All rights reservedSDM 2004
What is a Complex System?
A Complex System is a system:– Whose structure and behavior is not deducible, nor may it be
inferred, from the structure and behavior of its component parts;
– Characteristics by the presence of independent agents which introduce changes
– Whose elements can change in response to imposed “pressures” from neighboring elements (note the reciprocal and transitive implications of this);
– Has a large number of useful potential arrangements of its elements;
– That continually increases its own complexity given a steady influx of energy (raw resources);
Tend to be both Bottoms-up & Top-down
© 2004 The MITRE Corporation. All rights reservedSDM 2004
Is the AOC a Complex System? A Complex System is a system:
– Whose structure and behavior is not deducible, nor may it be inferred, from the structure and behavior of its component parts;
Partially true for the AOC
– Characteristics by the presence of independent agents which introduce changes
True for the AOC
– Whose elements can change in response to imposed “pressures” from neighboring elements (note the reciprocal and transitive implications of this);
True for the AOC
– Has a large number of useful potential arrangements of its elements;
True for the AOC
– That continually increases its own complexity given a steady influx of energy (raw resources);
True for the AOC
Yes, the AOC is a Complex System
© 2004 The MITRE Corporation. All rights reservedSDM 2004
Yet another Observation: Dissonance and Partial-connectivity between two “Ecosystems” Ecosystem 1: User/Ops
– About the flow of Information and transformations of information Ecosystem 2: Business
– About the flow of revenue How is “selective pressure” realized in each? What are the interrelations?
Flow: Information
User/Ops
Flow: Revenue
Business
Needs
Things
© 2004 The MITRE Corporation. All rights reservedSDM 2004
What would CSE involve?
Deliberate and accelerated mimicry of the processes that drive Natural Selection
– Characteristics of environments supporting natural evolution (an Ecosystem)
Co-existence and interact-ability – required to form an “environment” Evolvability – must be able to change Sensitivity to environment – change a function of “selective” pressure Variety – “competitors” within a niche Fitness – differential response to environment
– Artificial Engineering Environment for (co)Evolution Development Environment – setting up a continuous interaction space Outcome Spaces (not outcomes) Developmental Precepts – “forced” rules for interaction Continuous Characterization – continuing selective pressure Judging – choices made, with consequences Rewards based on achieved result and value Safety Regulations – sensitivity to “unsuccessful” varieties Duality – the coexistence of “build time” and “run time”
© 2004 The MITRE Corporation. All rights reservedSDM 2004
Summary
Many systems exist which have the characteristics and behaviors of Complex Systems – such as the AOC
Traditional Systems Engineering only takes one so far Something akin to Complex Systems Engineering is needed Complex Systems Engineering should support the
deliberate mimicry of evolution/co-evolution