o-SE Conceptual Model

We have been inquiring into the relationship between systems, knowledge domains and engineering, triggered by the challenges of integrating domain-specific modelling tools with overall systems models. A simple initial realization was that a distinctive characteristic of software and systems engineering, as compared to other engineering disciplines, is that correctness of solutions in these disciplines depends to a large extent on the body of knowledge in other fields. Perhaps this is why it has proven harder to create theoretical foundations that explain engineering practice in these disciplines, resulting in their being viewed widely as empirical disciplines.

Our inquiry led us to realize that knowledge domains can be categorized into knowledge domain that focus on wholes (e.g. network routers), and knowledge domains that focus on aspects (e.g. network routing). Wholes knowledge include knowledge about how wholes relate to aspects, as well as how they relate to other wholes. They carry the knowledge that enables multiple knowledge disciplines to come together in systems. This insight suggests a modular approach to knowledge representation (modular ontologies) with bridges that establish linkages among them based on their relative roles in the system of interest.

The result of our inquiry is a strawman conceptual model of systems engineering (attachment below). The focus is on a model of how knowledge domains relate to each other, how knowledge comes together in systems, and the set of concerns that need to be addressed by systems engineering practice, in particular the need to detect and close gaps between the model world and the real world. Our ultimate goal is to create a framework for reasoning about systems engineering practice, so that the discipline can go beyond being seen as empirical.

One of our other objectives is to help create semantic foundations for system modelling (SysML). To enable this, we are attempting to express the conceptual model mathematically. This work is ongoing.

While we have some seed ideas in this space, we are also keenly aware that systems engineering and systems science are vast areas, of which we have only a necessarily limited grasp. Building conceptual foundations for systems engineering based on systems science, and formulating it mathematically, will require synthesis of ideas and understanding from people with diverse perspectives and expertise. This is why we are making it a project in the systems science working group.

We would very much welcome inputs and participation. Please contact Swami at <swami.n@tcs.com>.