By: Heidi Davidz, Intelligent Systems Engineering Subject Matter Expert, ManTech
In systems engineering, model governance is the process an organization uses to control access to models, implement policies, track the inventory of diverse models, and improve efficiency and effectiveness. But implementing model governance is not always easy. As a result, the International Council on Systems Engineering (INCOSE) Model-Based Capabilities Matrix finds that most organizations score poorly on achieving it.
Among the top challenges:
- Managing large teams that may develop and deploy model governance inconsistently and without awareness across a program.
- Heterogenous models that don’t lend themselves to easy integration or tracking.
- Synchronizing changes, navigating contractual boundaries, and accessing hidden data.
To resolve these issues, ManTech has developed a holistic model governance approach that informs the work we do for customers and the work we do internally, as well. We combined our systems integration expertise and cybersecurity capabilities to develop a model governance guide that allows agencies to understand who their stakeholders are for modeling activities and what their unique needs and objectives are for model governance.
The ManTech Model Governance Guide includes:
- Guidance: Specifically, model-based guidance with in-model process and work instructions.
- Integration: An integration framework of the overall model governance system, the digital engineering environment, individual models and composite models.
- Scope: Articulates why you are modeling and how you will trace each model’s purpose and the resolution of its technical debt.
- Validation: Leverages automated validation for insight on compliance.
- Elasticity: Customization and tailoring for lifecycle scope, SE scale, technical management approach, formality and digital maturity, which are essential tenets of model governance.
As our holistic ManTech Model Governance Guide shows, we understand that not every model can be governed the same way. But an overall structure can ensure that the models can do what they need to do while still being trackable and traceable in the context of an overall program or objective.
As an example, systems engineering models must be able to talk to one another. A piece of equipment on a space vehicle may need to behave differently in different temperatures. The models that predict what temperatures will be present at a specific point in time need to interact with the models that control equipment functions so that the appropriate actions can be taken.
Of course, whenever models are communicating and automating behaviors, additional risks may arise. Because the threat landscape is larger, the likelihood of an interference – cyber or operational – is greater. An integrated, end-to-end model governance system ensures that we are accounting for risks and baking-in security into the model development process.
Holistic model governance can improve overall program effectiveness, security, and performance. It enhances usability and demonstrates the efficiency of model-based methods while improving integration with the introduction of referenceable, linkable, and checkable elements. Effective model governance can provide flexibility and tailoring for context, and it can institute standards and best practices that create clear and consistent guidelines for teams while driving transparency into the development process. Finally, it can better synchronize data structuring for analytics applications and ensure the veracity of authoritative sources of truth.
To learn more about ManTech’s Model Governance Guide, visit: www.mantech.com