Experiences in applying Model Based Systems Engineering (MBSE) techniques on a large-scale, distributed, real-time system are presented. Challenges and applied approaches with associated lessons learned are described for both technical and socialogical issues. Examples of technical issues addressed include: defining timing requirements and analyzing the ramifications of design choices on these requirements; early execution of the design for behavior analysis to include the use of realistic publish and subscribe mechanisms; managing the rapidly evolving design model so that a large number of engineers can make modifications to it (including the need to incorporate practices of managing software builds); and incorporating legacy software into an MBSE approach. Socialogical issues addressed include: using MBSE to work as a team when members span multiple organizations and technical/business domains; the need for a variety of views of the model based on team members backgrounds (i.e., one type of view such as UML will not meet all needs); information hiding so that there is one logically consistent model, yet parts of it can be hidden from some members of the team (for example, when individuals may not have a security clearence that allows them to see certain aspects of the design); and aspects of MBSE that need to be accounted for when creating project schedules (for example, there may need to be model integration periods scheduled similar to software integration periods). A major source of material for this paper was the authors collaboration on the requirements analysis and preliminary design of the U.S. Air Force’s E- 10A Multi-Sensor Command and Control system……………………….Experiences in applying Model Based Systems Engineering
Sep 28