Transcript Document

Integrating MBSE into a Multi-Disciplinary
Engineering Environment
A Software Engineering Perspective
Mark Hoffman
20 June 2011
Copyright © 2011 by Lockheed Martin Corporation. Published and used by INCOSE with permission.
Background
• Large Systems Integration that includes large
scale distributed S/W systems
– 100’s of system and software engineers
– Over a 100 Components
• Hardware components and software components (CSCIs)
– Well over 1 million lines of code
– 100’s of Hw / Sw Interfaces
– 1000’s of Requirements
• Engineering disciplines use multiple languages
and tools whose results are not always easily
integrated
– The potential for MBSE is that it provides a means to
integrate multi-disciplinary engineering including
systems, hardware, software, analysis, and test
throughout the development life cycle
Current Approach
• Software Engineering is mostly Model Based
Software Development
– SysML models – requirement analysis
– UML models and Matlab/Simulink models – design
& implementation
• Systems Engineering flow down of System
Design, Interfaces and Requirements to
Software Engineering typically consists of
Textual Documentation
– Word Documents, PowerPoint slides, Excel
spreadsheets, email, etc.
Problem
• Lack of integration is a source of design
discrepancies and errors
• As design, interfaces or requirements change
integration issues are introduced
– Time lag for information
– Coordination of changes showing up at the same
time
– Often manually intensive to gather up the changes
and get distributed to the stake holders
Considerations For Incorporating
MBSE Approach
• Incorporation of MBSE into a broader multidisciplinary engineering environment could
provide
–
–
–
–
More timely information
Similar semantics using standards like SysML
Easier integration with Software Modeling
Tighter integration to improved traceability
• Requirements
• System model artifacts such as interfaces
• Test Cases – System scenarios (Activity Diagrams)
– Impact Analysis
Questions to be addressed with an
MBSE Approach – (1)
• What should other engineering disciplines
expect from MBSE?
– Software engineering should expect the high level
conops of the system (mission scenarios)
• How the system will be used
– Software engineering should expect interface
definitions including component channelization
– Software engineering should expect change impact
due to requirement changes that effect software
Questions to be addressed with an
MBSE Approach – (2)
• What can MBSE learn from model-based
approaches used in other engineering
disciplines?
– MBSE can learn model organization from software
engineering
• For large systems, model organization is key for achieving
– Reusable model components
– Minimizing merging of modeling artifacts
– MBSE can learn effective approaches for configuration
management of model components
• To facilitate good collaboration, configuration management
of these model components is critical
– MBSE can learn effective design patterns / frameworks
that allows for effective collaboration and reuse
Questions to be addressed with an
MBSE Approach – (3)
• How should the practices and tools be
integrated/coupled across disciplines?
– For large scale, complex designs the following
areas need to be considered:
• Model Organization
• Number of Team Members
– Geographical distribution
• Configuration Management of model components
• Integration of tools
–
–
–
–
Requirements Management tool
Modeling tool
Configuration Management tool
Databases (that contain program specific data such as
parameter values)
Questions to be addressed with an
MBSE Approach – (4)
• How are the system, hardware, and software
models managed to ensure an integrated
technical baseline?
– To manage an MBSE effort of this scale the
following topics would need to be addressed
• Program Management – Manage overall modeling
activities across all disciplines
• Configuration Management – Manage the CM
architecture to facilitate technical baselines, reuse
components, distributed CM for collaboration
• Model Architecture – Manage the modeling structure and
model organization to enable consistency and seamless
integration between disciplines