Jama Software: What are the biggest trends you’re seeing in MBSE right now and how are they impacting product development?
Dirk Zwemer: The biggest trend we’re seeing is the growing realization that MBSE is not a function of any one software tool, not PLM, not SysML, not requirements. The goal is the Digital Thread, the complete set of domain models organized, connected and version-managed in a way that allows everyone on the development team to find the data they need to do their jobs. Each discipline and each organization have a seat at the MBSE table (Figure 1).
Implementation of the Digital Thread is still incremental in every enterprise. Early adopters look for specific integrations that enhance collaboration between team members, speeding completion of tasks and reducing errors from domain model inconsistencies. As they implement the Digital Thread more fully, they realize even greater gains in model validation and verification, which allows deeper exploration of the system design space within project schedules.
JS: What are some continuing or new trends in MBSE you expect to see over the next decade?
DZ: First, the engineering software tools and the Digital Thread infrastructure that connects them are all becoming scalable enterprise applications, sharing services and data either in the cloud or in on-premises servers. Standard interfaces such as RESTful or OSLC APIs will help support this, but the great heterogeneity of data, models and use cases will handicap approaches that restrict themselves to these technologies.
Second, no single MBSE methodology will be universally adopted, so flexibility in searching, visualizing, and documenting the Digital Thread will become important. Tools with a “one size fits all” interface will be at a disadvantage relative to open source, open standard and third-party components that can give users the reports and metrics they need with minimal effort and cost.
Third, MBSE is moving downstream through the system lifecycle. The Digital Thread will cover:
Conceptual Design → Detailed Design → Manufacturing & Logistics
With respect to tool integration, this will initially involve:
SysML → PLM → MRP
And we can expect to see increasing activity on this bridge. The advantages of devops processes in the software world will be increasingly obtainable for cyber-physical systems, as well.
JS: What sorts of process adjustments do you think development teams will need to make to accommodate these changes?
DZ: Practicing good data governance will be the biggest adjustment for most development teams building the Digital Thread.
- Cybersecurity will be an immediate concern. With the rollout of CMMC v0.6 (Cybersecurity Model Maturity Certification), DoD contractors, primes on down, will need audited cybersecurity practices to compete for government contracts starting in 2020. Information sharing will be particularly impacted.
- While cybersecurity is nominally concerned with preventing access by bad actors, information access by collaborators is also an issue. Process teams will look for ways to selectively share data, e.g. between customer and supplier, that prevent the partner from accessing deeper levels of proprietary technology.
- If the Digital Thread shares information between repositories, process teams will need to explicitly define data ownership, i.e. which repository is master, and processes for data comparison, notification, and update.
- All these adjustments will be easier if the enterprise adopts standard, well-documented authentication mechanisms. In too many organizations, these have grown ad-hoc, different for each repository, and with no central management or knowledge base. Implementing the Digital Thread in such environments will prove frustrating, so common authentication protocols will be needed.
JS: What do you think will remain the same in MBSE throughout the 2020s?
DZ: While not absolutely necessary, architecture modeling will remain an important part of MBSE and SysML will remain the dominant architecture modeling language. The SysML v2 Submission Team is working under the auspices of the Object Management Group to keep SysML relevant, with a final submission target of June 2021. Proposed changes will make SysML more precise, intuitive, and reusable, especially in the area of modeling usages and variant designs. A new underlying metamodel and a companion standard for SysML API and Services will open new opportunities for model creation, management and visualization. Together, these new features should broaden the reach of SysML across the development team while reducing the barriers of learning new tools, terminology, and notation.
JS: Anything else you’d like to add?
As new MBSE adopters flood the market and existing users refine their processes, the “Voice of the Customer” will be heard louder than ever. Buyers will want tools that are easy to use, cost-effective, and customer-configurable rather than vendor-customized. There will be plenty of opportunities for newer vendors that can meet these needs.
This article was first published in Jama Software.