Uncredentialed MBSE Consultants

[FWA Bulldog]

I have recently observed the presence of a number of MBSE consulting firms and consultants whose MBSE credentials are either questionable or non-existent.  Yet, they appear to be obtaining consulting contracts with organizations engaging the MBSE practice.  I'm doubtful positive outcomes are probable. Which leads me to ask the question; What are the potential negative impacts on the systems engineering profession?  I assume my familiarity with the names of a number of contributors here suggests you may have a perspective to contribute.

Cheers,

bulldog

[Editor, SysML Forum]

Dear Bulldog,

Thank you for highlighting this pertinent issue in our profession, one that bears significant implications for the quality and reliability of Model-Based Systems Engineering (MBSE) consulting. Your concern about the potential negative impacts of questionable MBSE credentials resonates with me and I would like to add some thoughts to this discussion based on my personal experience in the field.

In principle, I wholeheartedly agree that having professionals certified in MBSE + SysML through programs like OMG OCSMP and INCOSE CSEP is a laudable idea. However, it is also essential to recognize that while these certifications are useful benchmarks, they are not universally applicable nor are they always an indicator of deep, applied understanding or ability.

Drawing from my nearly two decades of experience, starting from leading the SysML Partners SysML Open Source Specification Project (https://sysml.org) in 2003, and further training and coaching medium-to-large MBSE Teams in various domains across three continents, I have come to some sobering realizations.

In practice, I've found that OCSMP and CSEP certifications often represent a relatively low bar in terms of technology proficiency. Many MBSE + SysML training programs, regrettably, "teach to the test". This results in engineers being able to pass these certifications through rote memorization of crib sheets and specialized training, but lacking in-depth understanding of the material. There have been numerous instances when OCSMP certified professionals have approached me to skip PivotPoint's ESSENTIAL (Basic) MBSE + SysML training in order to take either INTERMEDIATE or ADVANCED training, only to realize that their knowledge of SysML was superficial, outdated or both. They often lack understanding of SysML analysis, design, architecture patterns, and knowledge of Activity and Parametric ModSim – all of which we cover from day one at the ESSENTIAL level. Consequently, it often requires us to start from scratch, beginning at the ESSENTIAL level, to fill the gaps in their knowledge, regardless of their certifications.

To conclude, while I agree with your sentiments about OCSMP and CSEP certifications theoretically being a baseline for purported MBSE Experts, we must acknowledge that these certifications, as they currently stand, may not be the best indication of practical competence. It is clear that both OMG and INCOSE need to work on making their certifications more rigorous and practical. There is much room for improvement, but the path forward lies in recognizing these gaps and making concerted efforts to address them.

I hope this perspective brings some additional insight to the conversation, and I look forward to hearing the thoughts of others in this community on this critical issue.

/Cris

Editor, SysML Forum
edi...@sysmlforum.com
CTO & Founder
PivotPoint Technology Corp.
https://pivotpt.com/training/mbse-sysml/

[FWA Bulldog]

Dear Cris,

Thank you Cris for contributing your perspective to the discussion.  I extend my sincere gratitude for your continued management of this forum and to the practice.

Apologies for the use of a pseudonym here; I’m not afraid of publicly expressing my concern, but there are parties who will recognize this pseudonym should they happen to loiter here.  This topic intends to address their current state.

I too began my SysML journey during the early 2000’s while at Lockheed Martin MS2 and was influenced directly by interacting with Sandy, Michael Jesse Chonoles, Jamie Kanyok, and others of like mind.  A decade later I worked directly with John Watson, Laura Hart, and Christopher Oster; and later with Lenny Delligatti.   I add that your name is one I recognize for excellence in the practice.  I name drop to illustrate my expectations and the bar against which I measure.  Admittedly a high bar.

Cris, I am in full agreement with your conclusion and I wish I could contribute value to a resolution; if only an interim one.  I have retired from the practice, no longer able to cope with the mediocrity that seems to have engulfed it.  Perhaps the reality is that there is an insufficient number of competent practitioners to fulfill the need.  If the reality, where does that leave us?

Bulldog
Aka: Geoff

[Editor, SysML Forum]

Dear Geoff aka Bulldog,

Your perspective, though clearly steeped in sincere concern, appears to diverge significantly from the realities we've observed in the past two decades. Your recognition of an "admittedly high bar" and your commendation for the individuals who supposedly meet this standard, especially those in SysML thought leadership positions, is something I must respectfully disagree with.

Let's objectively critique these purported "MBSE Experts". Over the past two decades, their positions of leadership within SysML have resulted in a shockingly scant improvement of SysML's language syntax, semantics, usage patterns, and libraries. They've proven inept at defining authentic software/hardware interfaces and have shown a concerning lack of understanding of system architecture, analysis, and design patterns. Their evident confusion around executable ModSim semantics only adds to the growing list of shortcomings.

Let's not forget that SysML v1 essentially repackaged UML2 syntax and semantics with a few minor adjustments -- a task that any bona fide UML2 expert could easily undertake as a profile definition. Most of this work was completed during the first year of the SysML project (2003-2004). What have these 'experts' done in the intervening time? A trail of mediocre and impractical certifications, a string of minor releases, and a SysML v2 major revision that is still wrestling with basic interface and executable ModSim semantics. While you elevate these individuals, I see a missed opportunity spanning two decades!

However, let's steer clear of despair. The future of MBSE isn't chained to the SysML v2 major revision fiasco (cf. F-35 project debacle). Instead, it lies with emerging Model-Based Digital Engineering (MBDE) and hopefully a new 'System Digital Engineering Modeling Language' ('SysDEML' or whatever it eventually becomes) that benefits from the lessons and mistakes learned from SysML v1 and SysML v2. For future work we need a resilient system architectural modeling language that comes equipped with rich libraries to specify Digital Twins, complete with fine-grained Dynamic/Behavioral and Mathematical/Parametric ModSim. Moreover, it must be able to manage Smart, Cybersecure, Safe Distributed Systems that incorporate Machine Learning and Generative AI capabilities.

This is the focus of our MBDE and SysDEML work-in-progress, as well as that of our MBDE-savvy clients. We see a future that is bright for those who look ahead, and stark for those clinging to the past (for example, those applying SysML using outdated structured analysis & design techniques from the 1980s). The potential for MBDE and SysDEML progress is immense, and we must not allow the missteps of the past to hinder us.

Best regards,

/Cris

Editor, SysML Forum
edi...@sysmlforum.com
CTO & Founder
PivotPoint Technology Corp.
https://pivotpt.com/training/mbse-sysml/