Principal Investigator: Dr. Michael zur Muehlen, Stevens Institute of Technology (Michael.zurMuehlen@stevens.edu)
Timeframe: September 2010 to December 2011
Category: Systems Engineering and Systems Management Transformation
Objectives: Major DoD/IC acquisitions are required to produce Department of Defense Architecture Framework (DoDAF) artifacts to enable assessment by those external to the program. However, these artifacts are often not leveraged during the follow-on design and development processes and therefore, much of their potential value is not realized in most cases. In addition, it is difficult to keep the DoDAF artifacts synchronized with rapidly changing architectures and capabilities. The net result is that the system being delivered does not meet the expectations of the acquiring party. Increased integration of Architecture Development, Modeling and Simulation (M&S), and software design could significantly improve the Systems Engineering (SE) process. The research objective is to develop ways to facilitate closer integration of DoDAF into the architecture, M&S, and software design activities, that would result in reduced re-work, reduced development cycle time, and improved resource efficiency.
Approach: The project proceeded in three phases.
- Phase I: Mappings between different modeling languages (viz., BPMN 2.0, SysML, and Arena) were created.
- Phase II: In order to determine the usefulness of DoDAF for software engineering, architecture documentation for a medical training simulator starting with the full set of DoDAF 2.0 architecture views. In parallel, different model types of SysML were mapped to the DoDAF 2.0 views to determine the extent of coverage.
- Phase III: Prototypes were implemented to illustrate the feasibility of the findings from phase 1 and 2. In the simulation area, a model converted from BPMN 2.0 XML to Arena was developed in order to demonstrate the reuse of conceptual models for simulation purposes. In the software engineering area, requirements models were transferred from a news use case into an executable workflow model.
Significant Research Findings & Products:
- A clear definition of underlying terms and conditions. The first step of any architecture and development project should be the definition of key vocabulary. This includes capabilities, activities, resources, and performers.
- Not all perspectives of the DoDAF 2.0 framework are equally valuable for software-intensive design projects.
- Process models integrate concepts from many disparate architecture views and should be developed early in the system development life cycle, both to check the completeness of the architecture scope and the dynamic behavior of the intended system.
- Simulation models are a valuable design aid that is increasingly affordable to integrate in the software design process.
Publications: none to date
► Research Team
- Dr. Michael zur Muehlen, Stevens Institute of Technology
- Dr. Drew Hamilton, Auburn University
- Dr. Russell Peak, Georgia Tech University