Seamless Digital Engineering: A Grand Challenge Driven by Needs

Seamless Digital Engineering: A Grand Challenge Driven by Needs

James S. Wheaton, Daniel R. Herber

Digital Engineering currently relies on costly and often bespoke integration of disparate software products to assemble the authoritative source of truth of the system-of-interest. Tools not originally designed to work together become an acknowledged system-of-systems, with their own separate feature roadmaps, deprecation, and support timelines. The resulting brittleness and conglomeration of disparate interfaces in the Digital Engineering Ecosystem of an organization drains resources and impairs efficiency and efficacy.

If Model-Based Systems Engineering were applied to this problem, a complete system architecture model would be defined, and a purpose-built computing system-of-systems would be constructed to satisfy stakeholder needs. We have decades of research in computer science, cybersecurity, software and systems engineering, and human-computer interaction from which to draw that informs the design of a Seamless Digital Engineering tooling system, but it would require starting from a clean slate while carefully adopting existing standards.

In this paper, this problem space and solution space are characterized, defining and identifying Seamless Digital Engineering as a grand challenge in Digital Engineering research.

Comments:	17 pages, 7 figures, 4 tables, to be published in AIAA SciTech Forum 2024 conference proceedings
Subjects:	Systems and Control (eess.SY)
Cite as:	arXiv:2401.02059 [eess.SY]
	(or arXiv:2401.02059v1 [eess.SY] for this version)
	https://doi.org/10.48550/arXiv.2401.02059

Submission history

From: James Wheaton [view email]
[v1] Thu, 4 Jan 2024 04:53:46 UTC (170 KB)

 

Authors

• James S. Wheaton - Systems engineer at Colorado State University's Department of Systems Engineering. Background in model-based systems engineering.
• Daniel R. Herber - Professor in Systems Engineering at Colorado State University. Published research on model-based systems engineering, SysML, and agile systems engineering. Google Scholar
  

Previous Related Work:

• The authors have previously published on topics like engineering elegant systems, quality attributes, and SysML.
• Their work builds on prior research at universities and companies on model-based systems engineering, digital engineering, tool integration, and seamless model-driven engineering.
• Key related work cited includes research by Manfred Broy, Howard Reubenstein, and others on seamless tool integration and clean-slate secure architectures.

Institutional Associations:

• Both authors are affiliated with the Department of Systems Engineering at Colorado State University.
• The university has an active research program in systems engineering, including the CSU Systems Engineering Research Center (SERC).

Summary

The paper proposes "Seamless Digital Engineering" (SDE) as a new paradigm for digital engineering tooling. The goal is to create an integrated system of tools that guarantees model coherence and integrity through elegant interfaces and formal verification of the full stack down to the hardware level.

• Current digital engineering relies on integrating many disparate tools, which creates fragility and drains resources. There is no overarching systems architecture guiding tool integration.
• Seamless integration requires common data schemas, live object representations, and unified interfaces to remove the "seams" between tools. Formal verification of interfaces and behaviors is needed.
• SDE refines digital engineering by adding the requirement of "elegance" - the system should be robust, efficient, minimize unintended consequences, and have an intuitive interface.
• Five key architecture tenets of SDE are proposed: Seamless Models, Composable Data, Live Objects, Seamless Workflows, and Clean-slate Cybersecurity.
• Achieving SDE is a "grand challenge" requiring transdisciplinary collaboration and a clean-slate approach not bound by legacy computing constraints.
• The paper defines and characterizes SDE as a new research direction and paradigm for model-based digital engineering. Developing an integrated, formally verified system of modeling tools remains an open challenge.
  

 Five Key Architecture Tenets

Here are some more details on the five key architecture tenets proposed for Seamless Digital Engineering:

1. Seamless Models - Engineering models become interoperable through consistent and coherent meta-models based on mathematical formalisms like type theory. This enables semantic integrity across tools.
2. Composable Data - Data objects become composable by defining commensurate data schemas in a standard format. Well-typed data definitions enable seamless data sharing.
3. Live Objects - Model elements are represented as common system objects that are uniquely identifiable across tools. They are "live" in that they retain history and validity is checked against engineering models.
4. Seamless Workflows - A unified systems modeling language and activity-based computing preserves context and collaboration. This allows workflows to be defined, modeled, and transferred between tools.
5. Clean-slate Cybersecurity - New computer architectures must be designed from scratch using model-based systems engineering to guarantee security. Legacy platforms have unresolvable systemic vulnerabilities.

In summary, the tenets emphasize semantic consistency, common object modeling, verifiable interfaces, and integrated workflows as part of a ground-up secure architecture. The goal is to remove the integration "seams" that increase complexity and errors in current digital engineering ecosystems.

SysML and Seamless DE

SysML (Systems Modeling Language) can play a key role in enabling Seamless Digital Engineering:

• SysML provides a standard modeling language to define the system architecture and requirements for the Digital Engineering System (DES). This allows specifying the components, interfaces, behaviors, etc.
• The SysML model serves as the authoritative source of truth for the DES, capturing all aspects of the system design. Requirements traceability helps verify it meets stakeholder needs.
• The DES architecture can integrate SysML with other engineering models and tools through open model-based interfaces. SysML's API support facilitates this integration.
• SysML's executable modeling features allow early prototyping and simulation of DES components and workflows. This enables incremental verification.
• SysML's model analysis features like trade studies help evaluate alternative DES designs to find an optimal architecture.
• SysML models of engineering workflows and activities provide the common language for Seamless Workflow integration between tools.
• Meta-modeling in SysML can define ontology and semantics for composable data schemas and live object representations.
• SysML itself could be redesigned as part of the clean-slate approach to align with the needs of Seamless DE - for example, enhancing model composition and verification capabilities.

In summary, SysML provides a robust modeling language to architect and design the Digital Engineering System that meets the goals of Seamless DE. As an open standard, it facilitates collaborative development. SysML is a key enabler, but itself could evolve to better support this grand challenge.

Road to Seamless DE

Implementing Seamless Digital Engineering (SDE) would likely require developing new standards and controls in these areas:

• Data Schemas and Models - Common standards for engineering data representations, semantics, schemas, and meta-models are needed for composability between tools. Standards bodies like ISO, IEEE, or OMG could govern these.
• Interfaces and APIs - Standardized APIs for tool integration and live object access must be defined. These could be developed as open specifications.
• Workflows - Standard workflows, activities, lifecycles, and processes would need conformance specifications. Integrated toolchains would then implement accordingly.
• Security - New network, software, and hardware security standards would be required for the clean-slate architecture. Governance via NIST, UL, or consortium.
• Compliance Testing - Certification testing procedures would be needed to validate conformance to the standards by tool vendors. This requires a compliance authority.
• Configuration Control - Strict configuration management standards for the integrated toolchain and architecture baseline. Following ISO 10007 quality process standard.

Enforcement of these standards could be enacted through:

• Compliance Certification - Tools must pass compliance tests to integrate into the SDE ecosystem.
• Contractual Agreements - Vendors commit to standards conformance and continued certification via contracts.
• Open-Source Reference - An open source SDE reference architecture provides the standards blueprint.
• Model-Based Governance - SysML model analysis validates tools meet architecture specifications.
• Strict Change Control - Architecture baseline changes are restricted to only verified improvements.

In summary, a combination of open standards, contractual controls, rigorous certification, and model-based governance can help ensure vendors build tools that integrate seamlessly in the SDE ecosystem.