The SAVI Program began with a review of current architectural description languages.
The effort sought to identify a language in the embedded systems domain that was extensible to other development domains, such as hydraulic and mechanical systems. A comprehensive language is the best way to keep systems designs integrated from inception through deployment.
These projects concluded that no languages, as they existed, had the necessary capabilities. But, that the Architecture Analysis and Design Language (AADL) appeared to be a suitable language for experimentation.
The SAVI Program started with a scope estimation effort that led to the first Proof-of-Concept project.
| Project | Year | Description |
|---|---|---|
| Improved Software Verification Methods and Support Tools | 2003-2004 | Identify the software verification problems to be addressed, and the plan for developing the solutions to those problems. |
| Improved Software Verification Methods and Support Tools – Supplement | 2004-2005 | Survey Existing Architectural Specification Languages (ASL’s), Identify ASL Requirements, Define a Strawman ASL, and Develop an Example Problem to demonstrate the use of the strawman ASL. |
| Systems and Software Integration Virtualization (SSIV) | 2006-2008 | Develop WBS for full development of SAVI capabilities. |
| SSIV Proof-of-Concept Project | 2008 | Exercise the key concepts of a mutli-aspect, single-truth data repository and a model data translation layer. |
| SAVI Proof-of-Concept II | 2010-2011 | Develop SAVI use cases, derive requirements, extend sample problem demonstration. |
| SAVI Proof-of-Concept II – Supplement | 2011-2012 | Continue development of SAVI requirements. Introduce “shadow projects” to accelerate maturation of SAVI concept. |
Accomplishments
- Surveyed existing architectural description languages ADL’s
- Demonstrated extensibility of an architectural description language to include electronic, functional, and mechanical design parameters in a consistent, quantitatively analyzable model set.
- Demonstrated interoperability of diverse tools to perform system analyses. This included an integrated analyses of mechanical components controlled by embedded computing systems.
- Demonstrated multi-level OEM/supplier model integration and analysis, including iterative model verification and refinement, in a distributed, heterogenous development environment.
- Developed a return on investment model (ROI) for development and implementation of SAVI.
- Developed an extensive list of use cases defining potential SAVI implementations and leading to high-level SAVI requirements.
- Refined and demonstrated several use cases including perform behavioral analysis, demonstrate system safety, and verify mechanical fit.
- Continually reached out to related research efforts to identify best-in-class technologies to support SAVI and minimize duplication of effort throughout the aerospace community.
- Forged a collaboration between commercial and military aerospace system integrators, subsystem suppliers, regulators, academics, and software tool producers in the U.S., Europe, and South America.
