Application of System of Systems Engineering Methodology ...faculty.nps.edu/thuynh/Conference...
Transcript of Application of System of Systems Engineering Methodology ...faculty.nps.edu/thuynh/Conference...
1
Application of System of Systems EngineeringMethodology to Study of Joint Military Systems
Interoperability
John Osmundson1, Nelson Irvine2, Gordon Schacher3, Jack Jensen4, GaryLangford5, Thomas Huynh6, and Richard Kimmel7
1Departments of Information Sciences and Systems Engineering,Naval Postgraduate School,
1 University Circle, Monterey, CA 93943-5000, [email protected]
2Department of Information Sciences Naval Postgraduate School,1 University Circle, Monterey, CA 93943-5000, [email protected]
3Professor EmeritusNaval Postgraduate School
1 University Circle, Monterey, CA 93943-5000, [email protected]
4The Monterey Technology Group10300 Saddle Road, Monterey, CA 93940, [email protected]
5Department of Systems EngineeringNaval Postgraduate School,
1 University Circle, Monterey, CA 93943-5000, [email protected]
6Department of Systems EngineeringNaval Postgraduate School,
1 University Circle, Monterey, CA 93943-5000, [email protected]
7Science & Technology AdvisorCommander Third Fleet, [email protected]
ABSTRACT
The Joint Systems Baseline Assessment (JSBA) program is currently planningoperational demonstrations of a selected set of joint and coalition intelligence andcollection management systems to be performed in early fall of 2006. The NavalPostgraduate School (NPS) is applying a system of systems engineering methodology(Osmundson and Huynh 2005) to the analysis of the interoperability of the JSBA jointand coalition intelligence and collection management systems. The JSBA systems arerepresented in the NPS approach by a unified model that captures the essential elementsof ten Department of Defense Architectural Framework (DODAF) views in a single moreeasily understood view. The NPS system of systems model is also being used to theguide the instrumentation and measurement requirements for the operational
2
demonstrations. Outputs from these measurements will be used in the near future asinputs to an executable model of the joint intelligence and collection management systemof systems, and outputs of the executable model are expected to aid the identification offuture interoperability issues, as well as help identify potential interoperability solutions.
1.0 Introduction
Because systems have been acquired over many years without strict adherence to
common specifications, the US Department of Defense (DoD) faces numerous
inoperability challenges. Significant among these is the challenge to attain
interoperability among joint and coalition intelligence collection management and
targeting systems. The Joint Systems Baseline Assessment (JSBA) program is chartered
by the US Joint Intelligence Interoperability Board (JIIB), under the direction of the Joint
Forces Command (JFCOM) with support from the Joint Systems Integration Command
(JSIC), to address selected interoperability issues in a series of annual demonstrations and
studies (Singleton and Myers 2006). The JSBA program is currently planning an
operational demonstration of a selected set of joint and coalition intelligence and
collection management systems to be performed in early fall of 2006.
The JSBA program personnel and Naval Postgraduate School personnel are
applying a system of systems engineering methodology to the analysis of the
interoperability of the JSBA joint and coalition intelligence and collection management
systems. This approach is enabling identification of architectures of operational
interoperability demonstrations to be performed in September, 2006 at China Lake, CA.
The JSBA systems are represented in this approach by a unified model that captures the
essential elements of ten Department of Defense Architectural Framework (DODAF)
views in a single view. The system of systems model is also being used to the guide the
3
instrumentation and measurement requirements for the operational demonstrations.
Outputs from these measurements will be used in the near future as inputs to an
executable model of the joint intelligence and collection management system of systems.
Outputs of this executable model are expected to aid in identifying future interoperability
issues, as well as help in identifying potential interoperability solutions.
2.0 SYSTEM OF SYSTEMS ENGINEERING METHODOLOGY
Our system of systems engineering analysis methodology consists of a sequence
of analyses, transformations, model building, and simulations. The first three steps, which
have been applied to the JSBA program, are:
1. Development of system of systems scenarios and operational architectures
2. Identification of system of systems elements and threads
3. Representation of operational architectures in a Unified Modeling Language
(UML) – like format
This approach incorporates many of the features of the DoD architectural framework
but combines approximately ten of the individual DODAF architectural views into one
comprehensive view that gives high visibility to information system of systems
interoperability issues, and further can be directly translated into executable models. An
additional benefit of this approach is that it facilitates development of system of systems
ontologies, including ontologies for interoperability (Osmundson, Huyhn and Shaw
2006).
DODAF views, and one sequence for constructing the views, are shown on figure 1
(Smith 2005). AV on figure 1 refers to all views, OV refers to operational views, SV
refers to system views and TV refers to technical views. A usual starting point is the
4
development of AV-1 which includes an overview including the scope, purpose, users
and environment of a particular system of systems operation. The AV-1 sets the scope
and the context of the system of systems architecture. Next, an OV-1 is developed which
includes the operational nodes, activities performed at each node, and connectivity and
information exchange need lines between nodes. Other OV’s provide additional detailed
information. System views show how multiple systems link and interoperate and may
also describe the internal construction and operations of particular systems within the
architecture
Figure 1. A Sequence for Constructing DODAF Views. (Smith 2005)
In our system of systems engineering methodology the first step, development of
system of systems scenarios and operational architectures, corresponds to the initial step
of the DoD C4I architectural framework, which is to determine the operational context.
In this step the organizations, systems and operational goals within specified contexts are
identified.
START
AV-1Overview/Summary
OV-1High-level Operational
Concept
OV-5
Activity Model
OV-6b
Operational State TransitionDiagram
OV-2
Operational NodeConnectivity OV-7
Logical Data Model
SV-4System Function-
Operational Activity
System InterfaceDiagram
SV-1
SV-10-c
Systems Event/TraceDiagram
OV-6c
Operational Event/TraceDiagram
OV-6aOperational Rules
Model SV-6
Report:System Data Exchange
SV-5
Operational Activity/System Function Matrix
SV-9
Technology/Tech Area
AV-2Integrated Dictionary
SA Repository
SV-10b
Systems State TransitionDiagram
TV-2
StandardsTechnology
Forecast
SV-10a
System Rules Model
AV-3Capability Maturity
ProfileSA - Future
OrganizationalRelationships
OV-4
Technical ArchitectureProfile
TV-1
SV-7 System PerformanceParameters Matrix
PhysicalData
ModelSV-11
Auto- generated Report:Information Exchange
OV-3
1.0
2.0
4.14.0
3.0
5.0
3.1.2
3.1.1
3.1
8.0
7.0
6.0
3.2.1
3.2
8.4
8.5
SV-8
System EvolutionSA Derived
6.2
SV-2
Systems Communications8.2
6.1.1
6.1.2
SV-3
Matrix; Report:System - System
8.1
6.1
9.0
8.3
10.0
START
AV-1Overview/Summary
OV-1High-level Operational
Concept
OV-5
Activity Model
OV-6b
Operational State TransitionDiagram
OV-2
Operational NodeConnectivity OV-7
Logical Data Model
SV-4System Function-
Operational Activity
System InterfaceDiagram
SV-1
SV-10-c
Systems Event/TraceDiagram
OV-6c
Operational Event/TraceDiagram
OV-6aOperational Rules
Model SV-6
Report:System Data Exchange
SV-5
Operational Activity/System Function Matrix
SV-5
Operational Activity/System Function Matrix
SV-9
Technology/Tech Area
SV-9
Technology/Tech Area
AV-2Integrated Dictionary
SA Repository
AV-2AV-2Integrated Dictionary
SA Repository
SV-10b
Systems State TransitionDiagram
TV-2
StandardsTechnology
Forecast
SV-10a
System Rules Model
AV-3Capability Maturity
ProfileSA - Future
AV-3AV-3Capability Maturity
ProfileSA - Future
OrganizationalRelationships
OV-4
Technical ArchitectureProfile
TV-1
SV-7 System PerformanceParameters MatrixSV-7 System PerformanceParameters Matrix
PhysicalData
ModelSV-11
Auto- generated Report:Information Exchange
OV-3
1.0
2.0
4.14.0
3.0
5.0
3.1.2
3.1.1
3.1
8.0
7.0
6.0
3.2.1
3.2
8.4
8.5
SV-8
System EvolutionSA Derived
6.2
SV-8
System EvolutionSA Derived
SV-8
System EvolutionSA Derived
6.2
SV-2
Systems Communications8.2
SV-2
Systems Communications
SV-2
Systems Communications8.2
6.1.1
6.1.2
SV-3
Matrix; Report:System - System
8.1
6.1
9.0
8.3
10.0
5
The second step, identification of system threads, is an extension of the first step
and corresponds to the OV-6c diagram in the Department of Defense Architectural
Framework (DODAF.) In this step initial events that trigger the initiation of a process in
one or more of the systems are identified, as well as subsequent processes in the system
of systems until a logical ending point is reached.
In the third step the system of systems is represented in a UML-like format that
corresponds to UML swim lane diagrams. Figure 2 shows an illustration of a generic
system of systems and interactions between system elements in this UML-like format.
Time increases to the right along the horizontal axis. As illustrated in Figure 2, element
A1 of system A and element C1 of system C interact with system B by passing physical
items or information to element B1 of system 2. Later element B2 of system B interacts
with element B1 of system B. Examples of elements of systems could include
organizations of people, processing systems, communication systems, production systems
or transportation systems that interact to produce information, products or to transport
things.
Figure 2. Graphical View of System of Systems Interactions in a UML Format.
Time
System A
System B
System C
Element A1
Element B1
Element C1
Element B1
Element A2
System A
System B
System C
Element A1
Element B1
Element C1
Element B1
Element A2Item exchange
6
The flow of interactions shown on Figure 2 from a starting point to a logical
ending point is similar to a thread in a software system, and this view of system-of-
system interactions is analogous to swim lane diagrams in the Unified Modeling
Language (UML) (Larman 1998). Complex system interactions can be understood by
modeling each system in terms of objects corresponding to system elements, with the
proper logical flow and timing of items of interest passed between system elements,
either within a system boundary or across system boundaries, during interactions. Passing
of items from one model object to another is analogous to passing messages between
objects in UML. The measures of performance of such system of systems could include
time to complete a thread - such as accomplishing a complex task, or the throughput of
items through the total system. Depending on how the model is constructed another
example of a measure of performance could be the quality of the final outputs.
An important goal of the JSBA architectural analyses is to determine
interoperability requirements between systems. For a system of joint and coalition
intelligence management systems the directed arrows on Figure 2 represent information
items and the graphical view of the system would indicate information exchange
requirements needed for interoperability. If the graphical view shown on Figure 2 were
converted into an executable model such as a discrete event model, the timing
requirements for interoperability could be determined.
The final four steps of our system of systems engineering methodology are:
4. Identification of system of systems design parameters and factor levels
5. Transformation of the UML-like representation into executable models
6. Application of design of experiments
7
7. Simulation runs and analysis of results
The fourth step in the methodology is the identification of system of systems
design parameters and factor levels. Before developing executable models the systems
engineer must identify the appropriate system of systems options. In many situations the
system engineer would like to examine the effects of many systems variables on a
dependent variable. The variables are often referred to as design factors and the
dependent variable is often referred to as an objective function. Design factors might
include the type of arrangement of system interconnections, methods of access, methods
of routing and controlling flow of items, and throughputs of the various links in the
network, and must be identified by the systems engineer. Typically each of the design
factors can be varied (the design factors can be said to have several different states or
levels) and combinations of the various design factors in different levels represent
potential system of systems options.
The fifth step transforms the UML-like system of systems representation into
executable models. Models of system of systems are constructed in a modular manner so that
design factors are represented by an association with modeling application objects. System
options are represented by rearranging the objects and by varying the object attributes from model
to model.
The sixth step applies design of experiments to guide the development of executable
models and the running of simulations. Once system design factors and factor states have
been identified system of systems models must be built that enable all of the design
options to be analyzed. Often the total number of options can be high, resulting in a
formidable modeling task, and design of experiments provides a means of reducing the
total number of models that need to be constructed.
8
The seventh and final step in the methodology is running the simulations as specified
by the design of experiments and analyzing the results. It should be noted that some high
level system of systems’ issues can identified and resolved merely by constructing the
paper model in UML-like format. Questions such as whether two systems have the
required connectivity and meet basic item exchange requirements can often be resolved
based on an inspection of a well constructed paper model. Obtaining qualitative measures
of system of systems’ performance, however, requires the analytical basis provided by
executable model results.
3.0 APPLICATION OF SoS ENGINEERING METHODOLGY TOINTEROPERABILITY OF COLLECTION MANAGEMENT SYSTEMS
There are a number of issues associated with interoperability of collection
management systems such as: interfaces between intelligence collection management
tools and ground stations; the capability to monitor health, status, and location of
intelligence surveillance reconnaissance (ISR) assets; tasking for collection platforms
capabilities; integrating human intelligence data into command and control systems;
intelligence, surveillance and reconnaissance support to detect and defeat improvised
explosive device (IED); and enhanced workflow management. Interoperability issues
that relate to assuring that there is proper inter-system connectivity and common data
formats and structures are typically addressed in specific system developments and then
tested during JSBA technical demonstrations. In this work the focus is on interoperability
issues associated with the time flow of collection management processes.
9
In JSBA, we identify scenarios of interest and corresponding operational
architectures, beginning with a high-level operational view as shown on the operational
view (AV-1) shown on figure 3.
Figure 3. DODAF All View 1 (AV-1) of the Joint and Coalition Intelligence andCollection Management System of Systems.
Next, a UML-like representation of the joint intelligence and collection
management SoSs was developed, a portion of which is shown on figure 4.
10
Figure 4. A portion of the joint intelligence and collection management systemof systems represented in a UML-like format.
Specific threads of interest that met key JSBA objectives were identified for the
planned operational demonstration. These threads are: intelligence preparation for seizure
of a facility, a naval aviation strike against a costal target, re-tasking of sensors to collect
against high value and time sensitive targets, surveillance of border crossings,
surveillance of border crossing cued by human intelligence, and reconnaissance in
11
support of convoy operations. UML-like representations of each of the six threads were
developed by extracting the relevant organizations and functions from the total
intelligence and collection management system of systems representation, a portion of
which is shown on figure 4. An example ULM-like representation of the naval aviation
strike against a costal target thread, corresponding to a DODAF OV-6c diagram, is
shown on figure 5.
Development of system of systems scenarios and operational architectures
Identification of system of systems threads
Representation of operational architectures in a Unified Modeling Language (UML) –
like format
Figure 5. UML-like representation of the thread: Naval strike against a costal target.
Next, systems to be deployed in the operational demonstrations were mapped to
each of the UML-like thread representations. Example systems overlain on the naval
strike thread are shown on figure 6.
12
Figure 6. Example systems mapped to the thread: Naval strike against a costal target.
A useful technique to aid in understanding the system of systems interoperability
problems due to systems’ time behavior is to construct executable models of threads of
interest (Osmundson 2000). Time delays are associated with the performance of each
function and interface, such as manually inputting data, accessing communications links
and transporting data, and waiting at queues, for example. One approach to improving
systems interoperability is to minimize the total time delay associated with each
important thread of functions through the systems.
In our approach, executable models are built, instrumented, and simulations run in
order to identify bottle-necks and to give insight into potential system of systems
architecture improvements through application of design of experiments (Osmundson
GCCS -M
GlobalHawk
DGCS
13
2004.) Figure 7 shows a portion of an executable model of the mission thread for a naval
aviation strike against a costal target.
Figure 7. Executable model of the thread: Naval strike against a costal target.
JSBA technical demonstrations and operational demonstrations will be
instrumented wherever possible to provide values for executable model parameters. The
models can then be run to provide a number of important outputs including:
Measurements of thread timelines
Measurements of system throughputs
Identification of bottlenecks and choke points
Assessment of technology improvements
Assessment of potential process improvements
Identification of best architectures
In addition the models can be run under much more stressing conditions,
including high tempo of battle, that are not possible in technical and operational
count
event
CFMCC HQ
CJTF HQ
JCWG
JCMB
JTF J2
JTCB
COCOM JIC
NTM(inactive)
CFACC HQ
CAOC
ISR TacticalUnits
PED Node
Strike TacticalUnits
T0
startV
F
L W
0
Available platforms,sensors, orbits
D
T U
Receive platforms,etc.
T1
startV
Nominate
F
L W
0
D
T U
Approvenominations
T2
D
T U
Draft CPCL
T3
ApproveCPCL
D
T U
T4
DraftJPCL
T5
ApproveJPCL
D
T U
T6
Forward approvedJPCL
Create collectiondeck
T7
D
T U
D
T U
D
T U
Create expoitationregmts
Createexploitation plan
D
T U
T8
D
T Ua
b
T9 T10
D
T U
DraftJCMP
ApproveJCMP
D
T U
D
T U
D
T U
D
T U
Naval Aviation Thread Model 1.0 John Osmundson
Forward approvedJCMP
Forward approvedJCMP
D
T U
Forward approvedJCMP
D
T Ua
b
T11
D
T U
Draft MAAP
D
T UApprove MAAP
DevelopATORSTA Annex
D
T U
ForwardATORSTA Annex
D
T Ua
b
D
T U
Collect Image
D
T Ua
b
Exploit Image
D
T Ua
b
Receive Exploited Image
F
L W
0 Exit#
0
F
L W
0 Exit#
1
F
L W
0 Exit#
1
a
b
T12 T13 T14 T15 T16 T17 T18 T19
a
b
F
L W
0Exit
#
0
a
b
ndemand
Var
#
n
demand
Var
#
n
demand
Var
#
ndemand
Var
#
ndemand
Var
#
ndemand
Var
#
14
demonstrations, therefore allowing the robustness of interoperability solutions to be
tested.
4.0 CONCLUSIONS
The JSBA program is following a disciplined system of systems engineering
methodology to address joint collection management systems interoperability issues. By
combining traditional DODAF views into a single comprehensive view, all of aspects of
interoperability issues, both systems issues and organizational issues, are illuminated.
Executable models can be developed based on the same single comprehensive view, and
the executable models can be exercised to test robustness of interoperability solutions and
explore further improvements.
5.0 ACKNOWLEDGEMENTS
The authors would like to acknowledge the following sponsoring and supporting
organizations: the Joint Staff – J2 (JS-J2), Joint Forces Command – J2 (JFCOM-J2),
National Geospatial-Intelligence Agency (NGA), Joint Systems Integration Command
(JISC), and the Systems of Systems Engineering Center of Excellence (SOSECE); and
personnel including LT COL Tony L. Barker, USAF, JS-J2; CDR Joseph A. Smith, USN,
NGA; LT COL Jill E. Singleton, USAF, JISC; LT COL Anthony S. Lombardo, USAF,
JS-J2, and Russell Myers, General Dynamics contractor to JISC.
References
Larman, Craig, Applying UML and Patterns, Prentice-Hall, Upper Saddle River, NJ,1998.
Osmundson, John, and Thomas Huynh, “A Systems of Systems (SoS) SystemsEngineering Methodology.” 1st Annual System of Systems Engineering ConferenceProceedings, Johnstown, PA, June 13-14, 2005.
15
Osmundson, John S.;Russell Gottfried; Chee Yang Kum: Lau, Hui Boon; Lim,Wei Lian;Poh, Seng Wee Patrick; and Tan, Choo Thye, Process Modeling: A Systems EngineeringTool for Analyzing Complex Systems, Systems Engineering, Vol. 7, No. 4, 2004.
Osmundson, John, Thomas Huynh and Paul Shaw, “Developing Ontologies forInteroperability of Systems of Systems,” Proceedings of the Conference on SystemsEngineering Research, Los Angeles, CA.April 7-8, 2006.
Osmundson, John, A Systems Engineering Methodology for Information Systems,Systems Engineering, Vol. 3, No. 2, July, 2000.
Singleton, Jill and Russ Myers, Joint Systems Baseline Assessment 2006 (JSBA-06)Interoperability Assessment Plan V 4.0, Joint Systems Integration Command, 5 July,2006.
Smith, Joseph A., DOD Architecture Framework (DODAF) 101, Joint Staff J2P-3 March4, 2005.