Connectivity between models in Model Based Systems Engineering (MBSE) and Digital Engineering (DE) ecosystems is most often focused on tools and languages, with limited and ad hoc model exchange. This leads to model-based solutions that vary widely both internally and between organizations, have brittle technical architectures, and have high up-front costs due to lack of reuse. Technical architectures require frequent rework due to the pace of change in the tool industry and advancements in MBSE and DE. The approach also has a significant risk that if these connections are not updated for contextual changes, model or data corrections. While automated connections between models could be used to synchronize analytics, model changes, and updates, if not properly managed would more rapidly propagate errors and lead to systemic failures.
We believe engineers must be an integral part of the model lifecycle. How models are used, fitness for purpose, and a dynamic confidence level – improved by feedback from the digital twin – are essential. Existing automation in tooling, languages, models and workflows with data models hidden within platforms can lack context and not be easily reconfigured, changed or modified to improve fitness for use. In contrast, a reinforcing loop of a model ecosystem with openly available trust, reliability, and suitability for use will further develop and grow existing model assets. With careful thought to the social, technical, and economic value system of exchange, digital engineering will flourish with the power of connectivity.
As an illustration, a group of digital asset creators, model owners, SMEs and project team members can meet in an IPT meeting whiteboard, discuss, debate and often effectively and rapidly adapt to changes in inputs, models, or data to arrive at a decision. Confidence in model assessments, and context where the models are used is implicitly understood by their owners. The decision, however, often lacks a digital representation even through these individuals may have used models for their input. Digital engineering needs to couple the rigor of models with the speed of team communications. We need improved connectively between models, model users, model owners and contextualized model representations that are not slowed down by barriers (languages, tools, network and technology boundaries, etc.). Based on experience working with DoD, defense and aerospace companies, and commercial MBSE/DE ecosystems, we have identified several common boundaries limiting connectivity between models:
• Unintentional barriers (e.g. networks, technical domains, languages, tool environments, etc) disconnect digital assets, owners and users
• Centralized architectures, where ecosystems need to “own” rather than “use” models tend to obscure model usage
• Poor visibility of models - you cannot connect what you cannot discover or even know exists
• Digital Thread and Digital Twin are notional, but lack formal representations that can themselves be referenced and reused.
To meet these challenges, we propose the idea of a “digital model wrapper”, or model characterization profile, that can describe a model in terms of context, required parameters, and usage in the digital thread. The wrapper would respect the authoritative model source, and not require a centralization of model assets. Publicizing wrappers to an appropriate audience would create a digital asset exchange (DAX) that provides a structured and effective way to package digital assets, making them both locatable and immediately exploitable.
The model wrapper adds a unifying metadata layer to the digital engineering environment, which encapsulates lower-level complexity, increases performance, and addresses growing scalability issues, while remaining neutral to tools and hosting locations. It pulls from several industry standards and best practices such as the Dublin Core technology used by leading libraries worldwide to track and catalog hundreds of millions of diverse information artifacts, the Functional Mock-up Interface (FMI) standard, SysML, and many other standard and model types.
The combination of a digital wrapper and DAX will empower connectivity:
• Unifying metadata layer, common system representation and language
• Coupling the digital asset creator and content through a distributed network
• A digital asset exchange network provides a modern sociotechnical network to form dynamic problem-solving organizations
• Formally defined models that describe connected digital assets included in Digital Threads and Digital Twins
DAX and digital model wrappers addresses many of the Digital Engineering challenges identified in the OSD DE Strategy, and is applicable to the DoD and industry partners, and directly aligns with the USAF goal for “Implementation of a Digital Enterprise Environment framework which integrates models, data and artifacts”. These approaches are also applicable to commercial clients operating in a decentralized and distributed engineering environment. Many of the elements of DAX have been discussed and tested with academic and commercial clients.