Engineering Decisions in MBSE: Insights for a Decision Capture Framework Development

Decision-making is a core engineering design activity that conveys the engineer’s knowledge and translates it into courses of action. Capturing this form of knowledge can reap potential benefits for the engineering teams and enhance development efficiency. Despite its clear value, traditional decision capture often requires a significant amount of effort and still falls short of capturing the necessary context for reuse. Model-based systems engineering (MBSE) can be a promising solution to address these challenges by embedding decisions directly within system models, which can reduce the capture workload while maintaining explicit links to requirements, behaviors, and architectural elements. This article discusses a lightweight framework for integrating decision capture into MBSE workflows by representing decision alternatives as system model slices. Using a simplified industry example from aircraft architecture, we discuss the main challenges associated with decision capture and propose preliminary solutions to address these challenges.

💡 Research Summary

The paper addresses the persistent problem of capturing engineering design decisions in a way that preserves their rationale and context while minimizing the effort required from engineers. Traditional approaches rely on scattered documents, meeting minutes, and personal files, which makes it difficult to retrieve, understand, or reuse decisions later—especially in large, long‑lived projects such as commercial aircraft. To overcome these shortcomings, the authors propose a lightweight decision‑capture framework that is tightly integrated with Model‑Based Systems Engineering (MBSE).

The core idea is to represent each design alternative as a “model slice,” i.e., a subset of the system’s structural and behavioral model that reflects the impact of that alternative. When a decision is made, the chosen slice is marked (e.g., with a “preferred” label) and the unchosen alternatives are retained for traceability. Evaluation artifacts—simulation results, test data, performance analyses—are linked to the corresponding slices through the digital thread, ensuring that the rationale is co‑located with the model elements it influences. By keeping the capture activity inside the same modeling environment used for system description, the framework reduces the perceived intrusiveness of documentation and makes access to decision information intuitive: a user can navigate from a model element to its associated decision rationale and, if needed, explore alternative slices.

The authors illustrate the approach with a simplified aircraft example: allocating a cabin depressurization function either to the door system or to an extended control system. Both alternatives are modeled as separate slices, linked to performance evaluations performed by a specialist, and the final allocation is recorded by labeling the door‑system slice as preferred. This example demonstrates how the framework can provide immediate traceability from the “what” (the allocation link) to the “why” (performance trade‑off) without requiring separate documents.

Beyond basic traceability, the framework adopts relationship types defined in ISO 42010 (constraints, influences, triggers, enables, etc.) to model dependencies among decisions. For instance, a decision to use door vent flaps may trigger a new design problem concerning vent placement. By explicitly modeling such relationships, the impact of changing one decision on other linked decisions can be analyzed automatically, supporting impact assessment and design evolution.

The paper also identifies several practical challenges: (1) the effort required to capture decisions must not impede engineers’ primary tasks; (2) decision information must be easy to locate and understand; (3) sufficient context (system architecture, requirements, analysis artifacts) must be preserved; (4) only “key” decisions that are likely to be revisited should be captured, to avoid information overload. Additional concerns such as scalability, confidentiality, granularity, and complexity are acknowledged as topics for future work.

In conclusion, the authors argue that embedding decision capture directly into MBSE models via slices offers a promising path to reduce documentation overhead, improve knowledge consolidation, and enable systematic reuse of design rationale. The approach aligns well with emerging digital‑thread initiatives and could become a foundational element of future systems‑engineering knowledge‑management practices, provided that the identified challenges are addressed through tooling, automation, and refined selection criteria for “critical” decisions.