SERC Talk: “What Lives at the Intersection of MOSA and Set-Based Design?”

TOPIC: “What Lives at the Intersection of MOSA (Modular Open System Architecture) and Set-Based Design?”

SPEAKER:  Dr. Gary Witus, Associate Director for Student Programs, Anderson Institute; Associate Professor (Research), Industrial and Systems Engineering, College of Engineering, Wayne State University

DATE: Wednesday, October 19, 2016

ABSTRACT: Dr. Witus will be discussing Modular Open System Acquisition, Set Based Design, and Model-Based Systems Engineering in reference to an Advanced Technology Demonstrator development program which attempted to employ these concepts.  The system was divided into five functional subsystems and contracted out to five vendors, with a sixth contract for system integration.  The companies did not have contractual relationships with each other, only with the Government agency.  In principle, this approach would align the structure of the engineering organization with the functional architecture of the system (rather than, as often happens, the system architecture mirroring the structure of the executing organization), enable contracting with the most highly qualified technology developer for each subsystem, and facilitate competitive bid should the technologies be transferred to a program of record.  Several issues and challenges became apparent during the engineering development.  Significantly more time and effort was required for consensus on the interface specifications.  While the contractors were required to make their subsystems compatible, there was no such requirement for the engineering models of the subsystems.  The engineering models were considered proprietary, and shared in a “software as a service” model.  Substantial translation was required to convert the outputs from one model to the inputs to another, with loss of resolution in the process.  The decomposition into subsystems was good for some engineering management decisions, but not others.   The dependencies between subsystems depended on subsystem architecture choices.  There was not sufficient information to determine the appropriate order of decisions in down-selecting set-based design option.   Guidelines for set-based design down selection were unclear, in part due to multiple goals on the part of different stakeholders.  Limitations in addressing dependencies between subsystems under different design approaches led to multiple iterations to resolve conflicts, which was aggravated by lack of an engineering workflow map.

Click here for more information on the event.