10 Best Practices for a Successful Transition to MBSE

Shifting from traditional, document-based engineering to model-based systems engineering (MBSE) is no easy path. It takes time and a lot of adjustments. Misunderstanding what MBSE is and considering that adopting MBSE requires an additional workload also make the task of a smooth and efficient transition much more daunting. In nearly all cases, it involves a genuine change of mindset within teams and habits. That's why we believe it's best to go gradually, step by step, without imposing any drastic changes. Yet, adopting an MBSE approach doesn’t have to be that daunting. Organizations of all sizes and sectors can adopt simple strategies with a long-term focus on transitioning from traditional engineering to MBSE. In this article, we'll share ten best practices to ensure an efficient transition to MBSE. You'll see that these initiatives involve both cultural and technical transformations because they will entail a profound change in terms of tools and workloads.

So, without further ado, let's get started!

TABLE OF CONTENTS

1. Define the scope of your MBSE project implementation

2. Establish leadership support

3. Start easy with a pilot project

4. Engage and train stakeholders early in the process

5. Choose the right MBSE methodology

6. Build a model-driven approach to requirements and architecture with the help of traceability 

7. Implement Incremental and Iterative Development

8. Choose your MBSE tools carefully

9. Improve Processes and Tools Based on Feedback

10. Rely on metrics for concrete proof of MBSE's value

1. Define the scope of your MBSE project implementation

Defining the scope and the goals of an MBSE project implementation with the different stakeholders involved is essential. In fact, it is the very first step for a successful transition. Ensure each expert understands the goals and key benefits of adopting an MBSE approach for their daily tasks. Emphasize the benefits it brings:

• Better communication between teams
• Reduced development time
• A more optimized way of creating traceability between requirements
• Improved documentation and enhanced system integration
• And many more!

During this definition phase, be sure to clarify any misconceptions that may still be associated with MBSE. A common misconception is that “since I'm using models, I'm already using the MBSE approach” or that implementing MBSE means that all project data must be converted. This is, of course, not the case, but you should say it explicitly. Make sure that everyone really understands the strength and scope of this discipline and the tools and methodologies it implies.

2. Establish leadership support

A key factor for a successful MBSE project is the commitment and active involvement of top management. MBSE requires a thorough overhaul of processes, habits, and culture that may have been in place for decades. So, you need to get their support as early as possible. Management should allocate the necessary resources and lead the way with their teams. And we can't emphasize it enough: Communication is key, and leaders must communicate and share their vision at every level, always!

3. Start easy with a pilot project

Resist the urge to rush the whole organization into this significant change. Instead, take a step-by-step approach and begin implementing MBSE with a pilot project. This is a common practice in many businesses and one that has proved successful.

Select a standard project, one without too much complexity, where MBSE can be introduced in a well-controlled and monitored way. Alternatively, choosing a project where it's clear that a traditional engineering approach is no longer suitable and prone to major errors may also be an option. Whatever the case, this will help you demonstrate to your teams the benefits of MBSE.

Pilot projects will also let you identify problems and challenges so that you can tackle them confidently in future projects. They are also an excellent way of testing new tools and methods without disrupting the rest of the organization.

4. Engage and train stakeholders early in the process

Involving key players early is paramount. Indeed, providing them with quality training and follow-up helps ensure a smooth transition. Getting involved early will align expectations, address concerns, and create a strong teamwork environment. Now, who should be trained? Engineers, systems architects, project managers, and all other staff who have to handle MBSE and associated processes. Teams need to be trained in the key disciplines of MBSE. It includes:

• Modeling languages, especially SysML or SysML v2
• MBSE methodology (See section 5)
• System Architecture Design
• Lifecycle Integration

These courses should cover tool usage, change management, verification, and validation. In short, they must provide your teams with all the skills needed for successful MBSE implementation.

➡️ You can consult our workshops, tailored coaching, and training sessions:

• SodiusWillert Workshops & Coaching for Systems & Software Engineering
• IBM ELM Training Courses and Workshops
• Engineering Efficiency Empowerment Workshop
  

A few years ago, a Caltech survey defined the concept of methodology in an MBSE context as follows: “a collection of related processes, methods, and tools used to support the discipline of systems engineering in a model-based or model-driven context”. MBSE methodologies intend to support and orient systems engineering toward a model-driven approach. Think of it as a series of instructions enabling you to move from theory to practice. Whatever the methodology you select, it's a fundamental ingredient in the success of the MBSE implementation. It's what will enable you to streamline system model development and establish a common framework, methods, and language for each team and stakeholder.

Some examples of MBSE methodologies

This methodology was introduced by IBM Rational and created with the support of IBM Rhapsody. It focuses on integrating systems engineering practices with model-based techniques using UML and SysML. Its strengths include requirements analysis, system functional analysis, and design synthesis. Harmony is particularly well suited to industries requiring rigorous systems engineering practices and managing complex and safety-critical systems, such as automotive, A&D, and medical.

MagicGrid, another major MBSE methodology, was developed by NoMagic prior to its acquisition by Dassault Systèmes. It aims to provide a structured approach to modeling systems using SysML and CATIA Magic technologies. MagicGrid was created with the support of MagicDraw and has been designed to be flexible and adaptable, making it particularly suitable for a wide range of industries.

➡️ Object-Oriented Systems Engineering Methods (OOSEM)

This methodology was developed by Sandy Friedenthal and was later adopted by INCOSE as a reference MBSE practice. OOSEM tends to combine object-oriented techniques with traditional systems engineering practices. The object orientation approach derives from the nature of SysML, based on UML, which itself was created to describe object-oriented software. This emphasis on object-oriented techniques makes it a versatile methodology that can be adapted to a variety of sectors. Unlike MagicGrid and Harmony, OOSEM was not created with the support of a specific modeling tool.

Various MBSE methodologies are available. Some are SysML-oriented, and others are not. Most methodologies follow an “RFLP” process: 

• Requirements Analysis
• Functional Analysis
• Logical Architecture
• Physical Architecture

MagicGrid adds the parametric aspect. Harmony emphasizes verification and validation (V&V) through model simulation and model-based testing. Another noteworthy factor about the methodologies we have just discussed is the focus of analysis on “use cases”.  

6. Build a model-driven approach to requirements and architecture with the help of traceability 

Adopting an MBSE approach will significantly change the way you manage requirements and system architecture design. You'll be shifting from document-based requirements to model-driven specifications. These requirements will be linked to system components and behaviors. Models will be used to capture, analyze, and manage both requirements and architectural elements. In a model-centric approach, you'll create detailed SysML diagrams, such as requirements diagrams and block definition diagrams. This will enable teams to represent requirements and link them to specific design components visually.

This demonstrates that traceability will be of major importance in the implementation of an MBSE strategy.  You need to ensure that models are traceable and that each design decision is verified against the requirements. MBSE facilitates the creation of traceability, making it easier to design and verify the system. When something changes, it's easier to see the impact if the modeling includes traceability.

7. Implement Incremental and Iterative Development

Our long experience in this area has proven the point time and time again. The MBSE methodology is more likely to work in an iterative and incremental environment, where models are intended to evolve throughout the project lifecycle. In such an approach, models are continually refined and updated rather than waiting for final designs to be approved. This approach implies dividing the system development process into smaller but more manageable increments.

In a document-centric workflow, each phase of a process must wait until the previous phase has been fully completed and approved, which leads to frustration and many bottlenecks. Incremental and iterative development environments help keep models aligned with system requirements. They also allow quick changes to be added to the model.

8. Choose your MBSE tools carefully

Selecting the right tools is a decisive challenge in the transition to an MBSE approach. However, before diving into the tools themselves, you will first need to identify the different criteria for choosing your MBSE solutions. Your choice will heavily depend on the needs of the team and the organization, the type and stage of the project, the structure of the team for that project, etc. Some of the key factors to consider include:

• Integration with existing tools – Look for tools that integrate well with existing systems and support models and workflows.
• Ease of use – Are you looking for simplicity and intuitiveness? Or are you comfortable with a more complex platform? Does the tool require extensive training?
• Scalability of the tool – Can it support projects that grow in complexity?
• Compliance – Does it comply with your industry standards specifically?
• Customization – Does it cover the basic features you need? Does it offer additional customization features?
• Method automation – Do you require a tool capable of reducing repetitive manual tasks and processes?
• Artifact Generation – Can your tool generate essential documentation, diagrams, or reports?

Really think about your non-negotiables and your nice-to-haves, and use these criteria to drive your choice toward an MBSE solution.

➡️ Explore SodiusWillert solutions for MBSE

9. Improve Processes and Tools Based on Feedback

Adopting the MBSE methodology is not a one-time event but rather a long-term process of continuous refinement and improvement. As your teams use MBSE more intensively, you'll need to gather information and feedback on what's working and what still needs to be improved. Once you are actively seeking observations and incorporating them in a structured way, you will help your teams refine their MBSE practices. If the results are conclusive, you'll know that it's time to expand the MBSE approach to other areas, such as testing, validation, and others you consider relevant.

First of all, you'll need to establish a continuous feedback loop involving all stakeholders. Feedback can take many shapes and forms:

• Reviews
• Retrospectives
• User testing, etc.

Reviews and feedback should cover methodologies, team workflows but also tools. Examine this feedback closely. Make changes step by step. This way, you can fix the issues effectively.

10. Rely on metrics for concrete proof of MBSE's value

 Nothing irrefutably demonstrates how MBSE improves development processes better than metrics. You can measure elements such as model consistency, traceability, avoided error rates, time-to-market, or cost savings. Keep an eye on performance indicators related to model usage, quality, and component reuse. Finally, monitor the speed with which changes are made to the system.

CONCLUSION 

Transitioning to an MBSE approach for your systems development can significantly benefit your projects. By following best practices and understanding the functionalities and capabilities of each tool and how it fits into your objective, you can efficiently adopt MBSE into your workflow and boost productivity, quality, and communication. Please note that there are certainly other best practices for transitioning to MBSE, but we have focused on our practical experience and fields of expertise.

To wrap this article up, we’d say that an optimal MBSE methodology needs:

• Continuous learning
• Strong adaptability
• Knowledge-sharing within the organization

These elements are essential for success!

Implement MBSE best practices with SodiusWillert

SodiusWillert has been assisting organizations to successfully implement MBSE practices for over 20 years.  We have deep expertise in MBSE and especially in helping customers implement IBM Rhapsody into their development projects. More specifically, we assist our customers by:  

• Handling the licensing and implementation of IBM Rhapsody with technical demonstration, Proof of Concept, flexible licensing options, procurement and renewals, and license deployment.
• Coaching, workshops, and training SysML and UML courses for beginners and advanced learners looking to learn or expand their knowledge of IBM Rhapsody.
• Providing tailored solutions for IBM Rhapsody (e.g. add-ons) to make working with IBM Rhapsody easier, smoother, and more productive than ever. 
• Providing a Harmony toolkit for IBM Rhapsody to help you automate this MBSE methodology
  
  

Some of our SodiusWillert solutions and technologies can also help you implement an MBSE approach:

• SECollab facilitates the review and verification of models by many stakeholders, as well as traceability to other lifecycle artifacts.
• OSLC Connect enables traceability for planning activities in Jira (in the context of agile/iterative processes)
  
  

➡️ Do you have a question on how to successfully implement MBSE within your organization? We'd love to hear from you!