Date post: | 27-Dec-2015 |
Category: |
Documents |
Upload: | theresa-stevenson |
View: | 219 times |
Download: | 1 times |
page 1
Software Engineering InstituteCarnegie Mellon University
Defense Acquisition University
Sponsored by the U.S. Department of Defense© 2004 by Carnegie Mellon University
School of GovernmentUniversity of North Carolina
OSD PA&E CAIG
System of Systems Integration Cost Driver Research
March 2004
Jeannine M. Siviy: SEPG 0545Jeannine M. Siviy: SEPG 0545
page 2
Carnegie MellonSoftware Engineering Institute
Agenda
Research Effort Background 1230-1300• Initial Model and Hypotheses • Collaborators• ScheduleDiscussion 1300-1420• Session Goals• Ground rules• Q&A• Wrap up Next Steps 1420-1430
page 3
Carnegie MellonSoftware Engineering Institute
Research Goals
To identify leading indicators of SoS cost and schedule• Gain insight into drivers of size, complexity, cost, and schedule • Gain insight into drivers of process efficiency and effectiveness
Output: • Analytical tools and methodologies to support Joint Capabilities
investment decisions• Practical management methodologies to assist SoS/FoS
implementers and operators
page 4
Carnegie MellonSoftware Engineering Institute
Research StatusFunded in FY04 as part of broad software research initiativeCurrent research partnership
• Software Engineering Institute (SEI)- Dr. David Zubrow, SEI Software Engineering Measurement &
Analysis (SEMA)- Mr. Bill Anderson, Integration of Software Intensive Systems
(ISIS)• University of North Carolina (UNC)
- Dr. Mary Maureen Brown, UNC School of Government• Defense Acquisition University (DAU)
- Ms. Martha Spurlock, Director, Cost Analysis Curriculum- Mr. Robert Skertic, Software, Technical & Engineering Dept.
CAIG Study POC’s • Rob Flowe (703) 692-8052 [email protected]• Tom Coonce (703) 697-3845 [email protected]
Seeking collaboration opportunities, ideas, data!
page 5
Carnegie MellonSoftware Engineering Institute
Focus Group GoalsWhat key [issues, drivers, factors] differentiate the programmatic issues associated with SoS acquisition from single system development?
To what extent do the stakeholder, end-user, and developer relationships influence the technical architecture design and implementation?
To what extent is the strategic intent of the SoS effort accompanied by the necessary funding, operational, and personnel incentives to promote successful accomplishment?
page 6
Carnegie MellonSoftware Engineering Institute
Some General Precepts
“Interoperability” is a multi-dimensional aspect of system engineering.• Scope is far greater than simply interoperability of data –
requires integration of enterprise rules• Scope includes degrees of operational, system, and technical
coupling, ownership, …• Scope includes integration at the programmatic level
We can never anticipate fully the boundaries within which a given system will be expected to operate.• One man’s system-of-systems is another’s system.• There will always be new things to integrate into the system. • Integrating systems in a network can affect all other systems
in the network in unintended ways.
page 7
Carnegie MellonSoftware Engineering Institute
More General Precepts
Size matters:• As integration in the small gets larger, new problems
creep in: management, organizational.• As systems get more complex, interoperability issues
increase.• As operators become more agile, systems must
become dynamically reconfigurable.
Interoperability must be quantifiable to be achievable.
Interoperability must be sustainable and sustained.
page 8
Carnegie MellonSoftware Engineering Institute
A View of the Problem SpaceIncomplete understanding of scope and nature of the engineering to be accomplished• There are weaknesses in translating high level
concepts of operations into a set of system requirements that lead to interoperability.
Ongoing inertia toward separate programs executed independently
Technologies do not currently exist that permit quantification of interoperability.
page 9
Carnegie MellonSoftware Engineering Institute
System “C”
We know quite a lot about constructing systems from components (over which we may have little or no control).
We know very little about constructing an interoperable network of systems…the key distinction being that the network is unbounded (or very loosely bounded) and has no single controlling authority.
An Instance of the Problem
We know something about composing systems of systems from individual systems (over which we may have little or no control).
System “B”
System “A”
Network “X”
Unplanned, unexpected, emergent behavior here…
page 10
Carnegie MellonSoftware Engineering Institute
Interoperability Activities Model
Activities performed to manage the acquisition of a system.
Activities performed to create and maintain a system. Focus is on architecture, standards, components.
Activities performed to operate a system. Focus is on interactions with other systems and data distribution.
OperationalSystem
Sy
ste
mC
on
str
uct
ion
ProgramManagement
page 11
Carnegie MellonSoftware Engineering Institute
SoSI Interoperability Model
System
Co
nst
ruct
ion
PM
SystemC
on
stru
ctio
n
PM
System
Co
nst
ruct
ion
PM
Programmatic
Interoperability
Operational Interoperability
Constructive
Interoperability
page 12
Carnegie MellonSoftware Engineering Institute
Operational Interoperability
Fixing interoperability problems at the Operational level is key: Operational problems reduce the capabilities of the warfighter.
• insufficient sharing of data among systems- sometimes operators make up for the deficiencies
manually- sometimes unique interfaces obscure sharing that is
possible- users uncover the needs
• fixing problems in the field costs scarce resources- problems may not be found until time of greatest
need
page 13
Carnegie MellonSoftware Engineering Institute
Constructive Interoperability
Developers may not know how their system is to interoperate with others; frequently they:• are given “just a piece” of the overall system of systems• SoS architectures are often insufficiently specified• have no model indicating the meaning of data (semantics)• are unaware of timing and sequencing of interactions
Technologies exist, but not clear how or when to use them.• Exceptions: technologies for real-time, high-speed
applications and multilevel security are not sufficient• Information Exchange Requirements make sense in
point-to-point communication, less so in a net-centric environment
page 14
Carnegie MellonSoftware Engineering Institute
Relationship among DoD AF Views
page 15
Carnegie MellonSoftware Engineering Institute
Programmatic Interoperability
Programmatic problems dominate constructive ones.
SoSs often lack an ultimate authority for overall system (of systems) engineering.• Requirements for interoperability either are not given or are
vague.• Dependence on other systems can lead to local sub-
optimization: no universal enforcement of policy. • Program offices are stove-piped, doing their own thing
- even if they wanted to, it is hard to share information
“Unfunded mandates”• interoperability may not be sufficiently funded• interoperability is the first quality to be lost when schedule or
money are tight• IAM is often not funded and therefore not developed
page 16
Carnegie MellonSoftware Engineering Institute
Characterizing InteroperabilityDimensions apply variously to three system aspects:• the whole system of systems• the components that make up the system of systems
(“nodes”)• the relationships between the components (“arrows”)
Characteristics cover such dimensions as:• Evolution• Ownership• Technology availability• Communication• Planning and management• Usage patterns• Dynamism• Semantic consistency• Data management
page 17
Carnegie MellonSoftware Engineering Institute
Concept Map of SoSI Attributes as Related to Acquisition Outcomes
CostSchedule
Performance(Outcomes)
DrivesTotalEffort
InherentEffort
PartOf
InducedEffort
PartOf
Product-RelatedEffort Drivers:e.g., size, &complexity
Process-RelatedEffort Drivers:
e.g., efficiency &effectiveness
EnterpriseRules
InterfaceAttributes
ElementAttributes
ProcessAttributes
EnvironmentalAttributes
page 19
Carnegie MellonSoftware Engineering Institute
Discussion
page 20
Carnegie MellonSoftware Engineering Institute
• Keep focused
• Maintain momentum
• Get closure on questions
• Only one person speaks at a time
• Be concise, get to the point
• Stay on the topic and minimize diversions
Ground Rules
page 21
Carnegie MellonSoftware Engineering Institute
Questions for Discussion
What key [issues, drivers, factors] differentiate the programmatic issues associated with SoS acquisition from single system development?
• What differentiating activities in your experience are key to the success of SoS acquisition?
• How do the differences impact cost, schedule, performance and functionality?
page 22
Carnegie MellonSoftware Engineering Institute
Questions for Discussion
To what extent do the stakeholder, end-user, and developer relationships influence the technical architecture design and implementation? Are the DoDAF views useful for translating operational capabilities into a roadmap for SoS integration? What impact do these relationships have on program cost, schedule, performance and functionality?
page 23
Carnegie MellonSoftware Engineering Institute
Questions for Discussion
To what extent is the strategic intent of the SoS effort accompanied by the necessary funding, operational, and personnel incentives to promote successful accomplishment? How do the requisite funding, operational, and personnel incentives impact program cost, schedule, performance and functionality?
page 24
Carnegie MellonSoftware Engineering Institute
Next Steps
Field StudiesSuccessful programsChallenged programs
Ongoing Interviews with Subject Matter ExpertsPoints of Contact
Appreciate your comments on draft results and opportunity to continue collaboration
THANK – YOU!