Model-Based Systems Engineering Implementation

Considerations
Warren K. Vaneman, Ph.D. Ronald Carlson
Systems Engineering Department Systems Engineering Department
Naval Postgraduate School Naval Postgraduate School
Monterey, CA, USA Monterey, CA, USA

Abstract— As organizations strive to implement Model-Based Systems Engineering” (MBSE) was defined and popularized by
Systems Engineering (MBSE), there are significant discussions the International Council on Systems Engineering (INCOSE)
about model-based languages, tools, and the potential benefits of MBSE Initiative in 2007 to transform systems engineering to a
MBSE. However, discussions not occurring, but should be, are the model-based discipline [2].
implementation challenges. Issues such as: (i) a comprehensive
understanding or the organization’s vision and goals for the In spite of the emphasis being placed on MBSE, systems
MBSE environment; (ii) establishing a comprehensive model – the engineering methods have failed to keep pace with the ever
proverbial “single source of technical truth;” (iii) the governance increasing complexities of systems. As a result, many system
and management of the MBSE environment; (iv) establishing an engineering activities have been relegated to the beginning of
infrastructure and support environment that allows for the systems engineering lifecycle [3].
collaboration and communication across stakeholders; and, (v)
realizing the need, and establishing a plan, for organizational Model-Based Systems Engineering is the formalized
culture change. This paper discusses implementation challenges, application of modeling (static and dynamic) to support system
associated with those five issues, that must be overcome to design and analysis, throughout all phases of the system
transition to a true MBSE environment. This narrative is lifecycle, through the collection of modeling languages,
important because the successful realization of a MBSE structures, model-based processes, and presentation frameworks
environment requires a well-planned implementation to used to support the discipline of systems engineering in a model-
fundamentally change the way organizations perform systems based or model-driven context [3]. The four tenets of this
engineering. definition are shown in Fig. 1.
Keywords— Model-Based Systems Engineering (MBSE) x Modeling Languages – Serve as the basis of tools, and enable
the development of system models. Modeling languages are
I. INTRODUCTION based on a logical construct (visual representation) and/or a
meta-model1.
“Advancements in computing, modeling, data management,
and analytical capabilities offer great opportunities for the
engineering practice. Applying these tools and methods, we
are shifting toward a dynamic digital engineering
ecosystem. This digital engineering transformation is
necessary to meet new threats, maintain overmatch, and
leverage technology advancements.”
- Ms. Kristin Baldwin [1]

The Systems Engineering Community has long recognized

the need to evolve from the traditional document-based
approach to a model-based approach. The challenge driving this
change is how to deal with this increased systems complexity,
while ensuring a comprehensive and high quality design.
Model-Based Systems Engineering (MBSE) was conceived to
address the increasing complexity of systems by replacing a
traditional document-based approach with a model-based
approach where the models can be easily tailored to changing
conditions and needs, re-used, and able to observe the system
architecture from a static perspective, or explore system
Fig. 1. Four Tenets of MBSE.
behavior from a dynamic perspective. The term “Model-Based

1
A meta-model is a collection of standardized, defined terms or concepts and
the relationships among the terms and concepts.

978-1-5386-8396-5/19/$31.00 ©2019 IEEE

x Structure – Defines the relationships between the system provides insight across the entire system lifecycle from
entities, establishes concordance 2 within the model, and definition of need to disposal. The system has “height” which
allows for the emergence of system behaviors and provides for the decomposition from the highest to the lowest
performance characterizations within the model. levels of the system. The system also has “depth,” which
provides the complex relationships between systems, functions,
x Model-Based Processes – Provides the analytical framework requirements, analysis, risk, costs, schedule, etc. [5].
to conduct the analysis of the system virtually defined in the
model. The model-based processes may be traditional II. MBSE IMPLEMENTATION CONSIDERATIONS
systems engineering processes such as requirements
The transformation from traditional systems engineering to
management, risk management, or analytical methods such
a MBSE environment requires deliberate and fundamental
as discrete event simulation, systems dynamics modeling,
changes within the organization. However, most organizations
and dynamic programming.
begin the transformation process by considering model-based
x Presentation Frameworks - Provides the framework for the languages, tools, and the envisioned potential benefits of MBSE,
logical constructs of the system data in visualization models while ignoring their vision and goals for MBSE. A vision for
that are appropriate for the given stakeholders. These MBSE is a comprehensive conceptualization of how the MBSE
visualization models take the form of traditional systems environment will be employed in the future. The vision is further
engineering models. These individual models are often decomposed into specific goals.
grouped into frameworks that provide the standard views The fours tenants of MBSE serve as a good starting point for
and descriptions of the models, and the standard data establishing the MBSE goals. Fig.1 shows that maximum MBSE
structure of architecture models. effectiveness occurs at the intersection of modeling languages,
One element that may appear to be missing from this model-based processes, structure, and presentation frameworks.
definition is MBSE tools. Model-Based Systems Engineering Conducting an assessment to determine the organization’s
tools are general purpose software applications that use perspective on each of the four tenants will enable better
modeling languages, and support the specification, design, formalization of MBSE goals, and therefore will yield a better
analysis, validation and verification of complex system implantation strategy.
representations. Given that MBSE applies to both static and A.Considerations for Modeling Language Goals
dynamic models, a wider range of models are included in the
MBSE environment to meet the needs of the systems Modeling languages serve as the basis of tools, and enable
engineering effort. the development of system models. The following questions
could provide insight into the goals for modeling languages.
Transitioning to a MBSE environment means more than x Does the organization have a preferred modeling
using model-based tools and processes to create hard-copy text-
language? This may be the case if the organization
based documents, drawings, and diagrams. To successfully
progress to MBSE, the model must serve as a virtual engaged in preliminary MBSE efforts. Ideally a MBSE
representation of the system, thus the “dimensions” of a system language would be selected based on the MBSE vision,
project must be considered. To illustrate, assume that the cube but realistically a modeling language may have been
in Fig. 2 [4] is a system. The system has “width” that selected, and significant investment in tools and
training may have been expended.
x What is the technical knowledge of systems
engineering and MBSE among the staff? If the
systems engineering competency is new to the
organization, introducing a MBSE environment from
the outset may be feasible. However, selecting a
difficult to learn language may not be appropriate as the
team is learning the fundamentals of systems
engineering.
x What impact will the modeling language have on
productivity? Moving to a MBSE environment will
have a learning curve. Depending on the pace of
systems engineering efforts, the time to learn a
modeling language may be a limiting factor, this may
dictate the complexity of the language selected.
Fig. 2. Dimensions of a Systems Project [4]. x What level of corroboration is required for the MBSE
environment? Corroboration demonstrates that the
model corresponds to the real-world system problem

2
Concordance is the ability to represent a single entity such that data in one
view, or level of abstraction, matches the data in another view, or level of
abstraction, when talking about the exact same thing [3].

978-1-5386-8396-5/19/$31.00 ©2019 IEEE

 with sufficient detail for the fidelity of the intended be proactive in identifying generic data needs to
engineering and analysis [6]. To answer this question, support these issues.
significant knowledge into the real-world system
x How is data with a similar purpose and attributes
engineering issues must be understood. For example, if discovered and shared, given that each data set is stored
the systems engineering problem requires a focus on and maintained in its native format? A significant
the entire system lifecycle, the modeling language challenge is how to use the data in a collaborative
goals should reflect that need. manner.
B. Considerations for Model-Based Process Goals x How is quality assurance of the model-based processes
The formalization of the development, integration, and use evaluated? Quality assurance ensures the adherence of
of models to inform decision-making is essential to realize the the processes and that the resulting models adhere to
benefits of MBSE. This includes the formalized planning for applicable standards.
modeling and supporting engineering activities, the use of
models for decision-making across the systems lifecycle, and the x What processes are needed for verification and
policies that must be in place for MBSE to replace traditional validation of the model? Verification asks the
systems engineering. Model-based processes provide the question, “Was the model built correctly?” Often the
analytical framework to conduct the analysis of the system static system data is compared against the actual or
virtually defined in the model. The following questions could envisioned real-world system. However, this
provide insights into the goals for model-based processes. comparison often overlooks the relationships that may
not be apparent in the real-world system, but are
x What are the engineering and analysis objectives for essential to representing system complexity in a model.
the problems? Will the analysis being conducted There is no easy solution to model verification in the
include both static and dynamic problems? Static and static realm. However, when exercising the model
dynamic problems are often addressed in separate dynamically, through simulation, there are several
environments, and by engineers. MBSE can serve as a verification techniques that can be used [7]. Simulation
bridge that brings the static and dynamic analysis will identify areas where the model is logically
together in one environment. inconsistent.
x Does the organization have existing processes in a non- x What processes are needed for validation of the model?
MBSE, or preliminary MBSE environment that Validation asks the question, “Was the right model
produces trusted results to engineering and analysis built?” Validation increases the confidence that the
issues? This is an important question since model adequately represents the real-world system.
organizations may be hesitant to adopt new processes, Model credibility is essential since the model will serve
even when those new processes are more compatible as the foundation for many decisions about the real-
with a MBSE environment, if the current results are world system.
trusted.
x Are model-based processes meant to be used by
x Will the model-based processes be used represent the engineers only, or will stakeholders at various levels of
entire lifecycle, or just portions of it? Many the organization use the model?
applications of MBSE focus on early systems
architecting and requirements development. However, C. Considerations for Structure Goals
MBSE is well-suited for applications across the entire In traditional document-based systems engineering, a
systems lifecycle. change to an “system entity” is required for every instance where
that “system entity” is represented. For example, the same
x How compatible are existing model-based processes requirement is often represented in several documents within the
with using data from several different sources? The requirements hierarchy. When a requirement changes in a
single source of technical truth will be discussed in document-based environment, it must be changed everywhere it
Section III, however, strength of a MBSE environment is depicted or referenced. Structure defines the relationships
is the ability to represent system data as few times as between the system entities, establishes concordance within the
possible - ideally once. model, and allows for the emergence of system behaviors and
x What is the timeliness required to address engineering performance characterizations within the model. Systems
and analysis issues? It may not be possible to conduct consists of “building blocks” and the relationships between them
a disciplined modeling approach due to the decision that form a complete and functional entity. Complexity in the
horizon. Is this the norm, or the exception, for the model-based environment is significantly reduced by
organization? MBSE can shorten the time to address separating, and characterizing, systems issues into various data
engineering and analysis issues if the data is at least types that can be corroborated with the real-world system. This
partially developed at the outset of the issue [6]. approach increases emphasis on the model, specifically the
However, MBSE is not a “magic environment” that can entities and relationships it contains, rather than the diagram to
quickly address previously unimagined or ill- encourage better model development, usage, and decision-
structured problems. This requires the organization to making.

978-1-5386-8396-5/19/$31.00 ©2019 IEEE

 The following questions will provide insights into the goals x What additional viewpoints, and views, are required to
for the model structure. provide the stakeholders with the requisite information
to make decisions? Frameworks often focus only on
x How willing is the organization to migrate to a true systems architecture viewpoints and views, thereby
MBSE environment where a virtual representation of focusing on the early systems engineering activities.
the system replaces the traditional, document-based These frameworks are very good at representing
view of the system? This may appear to be an odd capabilities, functions, and conceptual and physical
question given that the decision to transition to a MBSE systems views, but don’t address other systems
environment has already been made. While, the engineering views such as requirements, and risks, and
decision to migrate to a MBSE environment may have do not capture the results from numerical analysis. The
been easy due to the direction of the system engineering purpose of MBSE is to serve an enabler for systems
community, grasping with representing the system engineering across the entire lifecycle. As previously
virtually, where data entities are ideally represented addressed, what portions of the systems lifecycle will
only once is more difficult. This requires a fundamental be modeled? This answer will drive the additional
change in the thought process of the engineering views of the data needed.
workforce.
III. MBSE UNIVERSAL ENABLERS
x Does a meta-model of existing data related to system
entities exist? If a meta-model does exist, will it be Developing a MBSE implementation strategy, based on a
sufficient to represent the data types and sources in the sound understanding of the organization’s vision and goals is
single source of technical truth? essential. Regardless of the strategy, each MBSE environment
must have three universal enablers: MBSE authoritative data in
D. Considerations for Presentation Frameworks context; MBSE management and governance standards; and,
Organizations often focus on a single presentation staffing and workforce development.
framework (e.g. the Department of Defense Architecture A. MBSE Authoritative Data in Context
Framework [DoDAF]), but fail to realize that no existing
framework includes the full spectrum of informational needs Data in a MBSE environment is ideally maintained within a
across the system lifecycle. For example, DoDAF represents single environment – the proverbial single source of technical
architectural data, but fails to capture other important systems truth. While a single repository, and tool, is desirable, it may
engineering data such as analysis from simulation, risks, and not be feasible due to the size and scope of the system or system
requirements. For MBSE to be effective, there is a strong need of systems being modeled. Regardless, if the environment is a
to ensure that the systems engineers and stakeholders understand single data repository, or composed of multiple data sources
the different model types and what information can be gleaned into an integrated data repository, the four tenets apply, and
from them. must be considered for the MBSE environment to be fully
Presentation frameworks provide the logical construct of the effective. Fig. 3 [9] depicts a notional MBSE environment.
system data in visualization models that are appropriate for the The MBSE environment should have the following
given stakeholders. Data types are grouped into frameworks that characteristics:
consists of viewpoints3, and provide the standard views4. The x Interoperable - The ability to exchange and transform
following questions could provide insights into the goals for the data with other models and structured data to include
presentation frameworks. both numeric and non-numeric analysis tasks in a
manner which each model understands, and without
x What system perspectives (i.e. viewpoints) do the data loss for the underlying repository;
system stakeholders represent? What information
x Model Constructability/Manageability - Ability to
format (i.e. views) do the stakeholders need to make
efficiently and intuitively construct and manage model
decisions? Viewpoints and views foster the ability to
effectively communicate with diverse stakeholders. construction and model extensions for special or
This includes standard systems engineering and fit-for- domain-specific concepts and terminology;
purpose views, answering the questions who, what, x Usability – The ability for stakeholders to efficiently
when, where, why, and how. and intuitively create, maintain, and use the model and
data repository.
x What level of detail will be required to represent The single environment, has a singular definition for any
system data precisely? System representation should model element, and allows for the static and dynamic
be as unambiguous and concise as needed at various representations of a system from several different perspectives
levels of data abstraction.
and levels of decomposition. The relationships formed between
x Does the organization have a preferred, or mandated, the principal data entities define structure, address complexity,
presentation framework? Some organizations (e.g. and ensures system traceability across the model.
U.S. Department of Defense) dictate a framework.

3 4
A viewpoint describes data drawn from one or more perspectives and A view is a related set of information using models for the representation of
organized in a particular way useful to decision-making [8]. data in any understandable format [8].

978-1-5386-8396-5/19/$31.00 ©2019 IEEE

 x Defines the communication path among key
stakeholders and enables a means of conflict resolution
among the constituent models;
x Maintains the overall MBSE structure that allows data
from the constituent models to be combined to achieve
the systems engineering goals.
One way organizations can account for the technical
management activities, as defined by model curation, and the
governance responsibilities, is by establishing a lead curator for
the MBSE environment. The role of the curator is to employ a
series of methods, practices, and principles to optimize the
MBSE environment, to ensure that the environment provides
Fig. 3. Notional MBSE Environment [5]. accurate and relevant system analysis. The role of the curator,
and the model curation process, cannot be over-stated because
To address the single source of technical truth, organizations the data in the single source of technical truth will be the
must consider implementing a common meta-model that can be foundation for making system decisions.
used across the full spectrum of data and applications with the
MBSE environment. The meta-model will characterize data into C. Establishing an Infrastucture and Support Environment
thematic types and allow for discovery, and collaboration among A MBSE environment requires the establishment of an
disparate data sets. For this collaboration to occur, new policies infrastructure and support environment that allows for
must be established to ensure governance of the authoritative collaboration and communication across stakeholders. The
source of technical truth, and define data standards and establishment of this infrastructure requires that new standards
processes that lead to search and discovery of data while it and guidance be established for new tool development, and
resides in its native format. provides a secure environment that protects intellectual
property.
B. MBSE Management and Governance
Recently, the concept of “model curation,” as means to Establishment of a MBSE environment may eventually
manage the MBSE environment, was introduced. Model support diverse toolsets and controlled data exchanges. Initially,
curation is defined as [10]: the data being exchanged will be produced from model-based
tools that were developed asynchronously to perform stand-
“the lifecycle management, control, preservation and alone systems analysis. Data standards and guidance should be
active enhancement of models and associated information to established to aid current tools to best integrate into the MBSE
ensure value for current and future use, as well as environment, and guide the development of new tools and data
repurposing beyond initial purpose and context. Curation that emerge over time into the MBSE environment. A data
activities include model identification, acquisition, standard will also be required to ensure the exchange of data
accession, composition, evaluation, preservation, and within the appropriate model-based tools in the MBSE
archiving. Curation practices promote formalism and environment.
provide for the management and control of models and
associated digital artifacts, particularly when managed as a Data cyber-security is a challenge that must be addressed in
collection, at the program and/or enterprise levels. Data the MBSE infrastructure and support environment. One of the
associated with a model includes model technical data, benefits of a robust MBSE environment is the discovery and
model metadata, and model pedigree.” reuse of system data from disparate model-based tools so that a
more complete analysis of the system can be achieved.
Model curation is essential to the management of the MBSE However, the strength of openness subjects the MBSE
environment. To be successful in establishing a complex MBSE environment to a cyber-security. Policies must be established to
environment, organizations must also implement a governance ensure that the data is protected from cyber criminals, and from
structure. Governance is the set of rules, policies, and decision- inadvertent spillage.
making criteria defines and guides the structure and
relationships among key stakeholders, and models, that D. Staffing and Workforce Development
determines the direction of the MBSE environment. Organizations must make focused efforts to transform their
culture to focus on the virtual model of the system, and away
MBSE governance structure:
from technical documentation. Workforce skill composition
x Provides the set of policies and decision-making criteria must change through the education of current engineers and the
which guides the MBSE environment; hiring practices for new engineers. This cultural change must
also be adopted by organizations responsible for all constituent
x Defines the distribution of authority within the models as well, to be truly effective. Organizations often discuss
overarching MBSE environment and the the importance of change, but the magnitude of the cultural
responsibilities of the constituent models; change cannot be overstated.
Progress towards MBSE implementation can be slow, and
short-term benefits may be unclear to leadership. Meanwhile,

978-1-5386-8396-5/19/$31.00 ©2019 IEEE

resource investment can be high with apparently no end in sight. REFERENCES
These issues often make funding issues difficult. The [1] Office of the Deputy Assistant Secretary of Defense for Systems
establishment of clear goals to measure incremental success is Engineering, “Department of Defense Digital Engineering Strategy
Guide,” Department of Defense, Washington, DC, June 2018.
essential. These measures can be qualitative and quantitative.
[2] S. Friedental, R. Gregigo, and M. Sampson, “INCOSE Model Based
Two questions that should be addressed are: (i) How do we Systems Engineering (MBSE) Workshop Outbrief,” INCOSE
know that progress is being made towards the envisioned MBSE International Workshop, Alburque, NM, January 2008.
end-state; and, (ii) Does the roadmap meet the goals of the [3] W.K. Vaneman, “Enhanced Model-Baseed Systems Engineering with the
MBSE environment? Lifecycle Modeling Language”, Proceedings of the 10th Annual IEEE
Systems Conference, Orlando, FL, April 18-21, 2016.
IV. ONWARD TOWARDS A MBSE ENVIRONMENT [4] W. Larson, D. Kirkpatrick, D. Verma, T. Dale, and J.J. Sellers, “Applied
For MBSE to be truly successful, the model-based Space Systems Engineering” in Space Systems Engineering, McGraw
Hill, Boston, MA, 2013.
environment must replace traditional systems engineering. This
[5] W.K. Vaneman, J.J. Sellers, and S.H. Dam, “Essential LML,” SPEC
requires the organization to examine, and understand, their Innovations, Manassas, VA, 2018.
vision and goals for this new environment. Failure to have an [6] R.W. Covey and D.J. Hixon, “The Creation and Use of an Analysis
understanding of the vision and goals, will potentially lead to Capability Maturity Model (ACMM),” Technical Report No. TR-
consequences that are counter to or fall short of successful 2005(3909)-1, The Aerospace Corporation, El Segundo, CA, 2005.
MBSE implementation. The consequences may include: [7] A.M. Law and W. D. Kelton, “Simulation Modeling and Analysis, 3rd ed,”
McGraw Hill, Boston, MA, 2000.
x The MBSE implementation may be delayed while the [8] S.H. Dam, “DoD Architectuew Framework 2.0: SA Guide to Applying
organization struggles to identify the boundary of the Systems Engineering to Develop Integratedm Executable Architectures,”
MBSE problems; SPEC Innovations, Manassas, VA, 2015.
[9] W.K. Vaneman. “Model-Based Systems Engineering De-Mystified,”
x The MBSE environment may fail to meet the data and unpublished white paper, Naval Postgraduate School, Monterey, CA,
information needs for the required decisions; 2019.
[10] D.H. Rhodes, “Model Curation,” OMG MBSE Wiki, November 19,2018,
x Additional costs and time may be required for rework; http://www.omgwiki.org/MBSE/doku.php?id=mbse:curating_digital_art
ifacts
x The model may lack credibility due to incorrect data, or
flawed engineering and analysis;
x The MBSE environment may be costlier than expected due
to “false starts” as a result of poor planning.
MBSE requires an organizational mindset change in systems
engineering processes, and a change in expectations of the data
required to make decisions. A well-developed system model
will be a virtual representation of the system, with relationships
made between the various system entities, and exhibit a high
level of concordance.
MBSE data is required to be exchanged, combined, and used
by various MBSE tools, therefore standards need to be defined
and enforced to support these activities. To achieve the standards
for an authoritative database, the associated infrastructure and
support environment need to be established. This includes data
being maintained in a single environment to foster discovery,
reuse, and collaboration. Evolving from a single to multiple
models is tantamount to the evolution of systems engineering to
system of systems engineering.
The challenges of implementing a MBSE environment
extends past the technical challenges. To succeed at
implementing MBSE, organizations also need a deliberate effort
to transform the workforce and promote cultural change.
Organizations must also realize that implementing MBSE often
takes a leap of faith as it deviates from for the proven analog
processes.

978-1-5386-8396-5/19/$31.00 ©2019 IEEE