|M.Sc Student||Batan Inbar|
|Subject||Enhancing Model- Based System Engineering With Model|
|Department||Department of Industrial Engineering and Management||Supervisors||Professor Dov Dori|
|Dr. A. Zonnenshain|
|Full Thesis text|
Model Based System Engineering (MBSE) is the application of formal qualitative and quantitative modeling to support the system definition, design, development, deployment, and operation, i.e., the entire system’s lifecycle. MBSE relies on modeling languages, such as Object-Process Methodology (OPM) or Systems Modeling Language (SysML). Modeling is done using software tools, such as OPM's free modeling tool, OPCAT, IBM's Rational Rhapsody, or MathWorks’ Matlab-Simulink. The tools help system analysts model the system, its function, structure, and behavior. However, most organizations developing complex systems still do not use MBSE; rather, they base their systems engineering process on textual documentation and occasional diagrams.
The model of a complex engineered system captures and provides information and knowledge about the system and its components, and may therefore become very big and complex in itself. It is difficult to measure manually the amount and quality of information in the model, to identify information gaps, to assess the model’s descriptiveness, or the sufficiency of the information about the system. There is currently no tool that measures, monitors, and analyzes the different informational aspects of the model, and allows the systems engineer to determine that a model is sufficiently descriptive, contributing to the understanding of the system, or missing critical information. Hence, it is not possible to assess the value of system models in terms of the information they convey and their contribution to designing, developing, testing, deploying, or operating the system.
In order to address the problem at hand, we propose a tool that provides structure of metrics, Model Informativity Level (MIL), that measure and aggregate the amount, quality, and utility of the various information aspects of the system, as represented by its model. MIL provides an estimate for the model’s expressive power, descriptiveness, and usability. It also helps identify aspects with insufficient informativity and consequently allocate engineering and modeling efforts.. We explore the evolution of MIL along the modeling stages for OPM models constructed by individuals and teams in both academia and industry.
Our findings show that when modeling a system in the MBSE approach using OPM, the model contains the system environment, requirements, main flow and alternative flows of the system. The MIL evaluation also shows that the model informativity increases along the modeling processes when almost no model parts are remodeled. In addition, by using MIL outcomes, the system engineer receives feedback of the system informativity, and can see where there are gaps in information.
This research was motivated by the assertion that quantitative assessment of model utility and model-based system development trend analysis will help drive the adoption of a MBSE approach along the lifecycle of the system, in order to provide a better model, which leads to better system understanding. As part of this research, we strive to promote and support the adoption of MBSE by establishing an open-access repository of MBSE case studies, including OPM modeling projects by MBSE course attendees, as well as MBSE applications on actual systems in commercial companies, including large Fortune 500 enterprises.