Technical Report
SE M&S Integration Framework in the Digital Thread
-
Systems Engineering and Systems Management Transformation
Report Number: SERC-2016-TR-115
Publication Date: 2016-10-28
Project:
Systems Engineering Model and Simulation Integration Framework in Digital Thread
Principal Investigators:
Dr. Gary Witus
Co-Principal Investigators:
This report documents the findings of a research study addressing Model and Simulation (M&S)
integration in the transformation of Systems Engineering (SE) at a DoD Research Development and
Engineering Command (RDEC) into a model-based and digitally enabled enterprise. The study used a fieldresearch
applied case study approach to develop and demonstrate methods, procedures and tools (MPT),
and to identify issues and opportunities for additional SE research. The four project objectives were to
1. Develop methods to document the “as is” system
2. Develop a vision of the “to be” system
3. Develop an approach to develop a road-map from the “as is” to the “to be”
4. Identify SE research needs and opportunities
We found that workflows were variable depending on the particulars of the program. We found that the
SE enterprise at RDECs such as TARDEC are organized for agile SE, that can adapt and adjust to a wide
range of different types of programs - from new starts to system upgrades, Major Defense Acquisition
Programs to Technology Demonstrators and Science and Technology (S&T) projects, programs starting at
different stages of the acquisition lifecycle, different program acquisition strategies, priorities and
constraints. We found that the engineering workflow and use of M&S changed and evolved as the
program progressed. We found that different MPT and different M&S were used to produce the “same”
work product under different conditions and constraints as needed. We found that personnel changes
and re-organizations also affected workflows.
We developed methods to document the “as is” systems engineering workflow as an extension to
configuration management (CM) for the engineering work products with augmented CM meta-data. The
augmented meta-data consisted of the unique CM identifiers (IDs) of the work products that were the
sources of input models and data. This CM metadata is sufficient to reconstruct the entire workflow, to
trace the provenance of models and data, to identify potential inconsistencies (inconsistent provenance
of work products providing input to downstream processes), and to pre-empt potential for downstream
inconsistencies through mutual sign-off confirming consistent inputs as part of the CM process. This
approach is similar to the block-chain transaction authentication algorithm.
The need for an agile SE enterprise presents a challenge for digitally-enabled SE workflow and M&S
integration. The concern is that if the digital thread is rigid or brittle, it would not serve the needs of an
agile SE enterprise, could reduce effectiveness, and ultimately be discarded. The challenge was to develop
a vision for a digitally-enabled agile SE enterprise.
We considered two alternative “to be” visions. In one vision, the digitally-enabled agile SE enterprise
would have digitally-enabled engineering workflow patterns, or templates, for different types of
programs. The appropriate pattern would be instantiated for a specific program. This approach is in use
at the Jet Propulsion Laboratories (JPL), and has been the subject of some academic research. It is
extending the concept of reusable software patterns to digital engineering workflows. The other vision is
to have a grab bag of digitally-connected short segments of SE workflows, that can be used when and
where they are appropriate.
The two approaches had significantly different implications for the transformation from the “as is” system
to the “to be” vision. The vision of a grab bag of digitally-connected short segments of SE workflows can
be realized through an incremental, granular process. It can readily incorporate new M&S tools and SE
MPT. The transformation can start with high-value connections, and produce useful capabilities before full development. The workflow pattern model of digitally enabled SE has significant developmental
challenges. The first challenge is simply getting enough workflow data from which to recognize patterns.
The frequency of new programs – opportunities to collect workflow data – is small relative to the size of
the workflow state-space. Methods to discover, model, and instantiate workflow patterns will need to be
developed and refined. Methods to implement digital workflow patterns will need to be developed.
Based on these considerations, we recommend the “to be” vision of a grab bag of digitally-connected
short segments of SE workflows because the transformation process is practical, feasible, and lower risk
than the alternative.
Further research and development are needed to identify and track decisions, assumptions, and other
intangible work products, that are often not put under configuration management. MPT are needed to
reduce the CM burden, to make it more palatable to the engineering teams. Attempts to build point-to-point
digital connections along segments of the SE workflow are expected to reveal additional challenges
in the digital implementation.