PhD Oral Exam - Rupinder Kaur, Information and Systems Engineering

When studying for a doctoral degree (PhD), candidates submit a thesis that provides a critical review of the current state of knowledge of the thesis subject as well as the student’s own contributions to the subject. The distinguishing criterion of doctoral graduate research is a significant and original contribution to knowledge.

Once accepted, the candidate presents the thesis orally. This oral exam is open to the public.

Abstract

Medical device development relies on complex and evolving regulatory frameworks, such as FDA 21 CFR 820 and ISO 13485, whose interdependent clauses must be interpreted and transformed into actionable engineering requirements. This process remains largely manual, expertise-dependent, and is difficult to verify for completeness and traceability.

This thesis investigates the use of Environment-Based Design (EBD) as a foundation for systematically extracting engineering requirements from medical device regulations and enabling their automated analysis.

First, a bibliometric analysis of 124 publications identifies major research challenges in regulatory compliance, including regulatory complexity, compliance cost, requirement elicitation, traceability and collaboration. Second, a manual EBD-based methodology supported by the Recursive Object Model is developed to extract traceable requirements from regulatory clauses. Its application demonstrates the ability to identify compliance-critical requirements while revealing the limitations of manual iterative analysis. Third, an automated framework, EBDREx (Environment-Based Design-Guided Requirement Extraction), is introduced by integrating EBD with large language models and retrieval-augmented generation

The framework incorporates a grounding-based convergence mechanism that restricts extracted objects to those supported by regulatory text, enabling controlled, traceable, and reproducible requirement extraction. Applied to the design and development planning subclause of the FDA’s Quality Management System Regulation (QMSR), which covers essentially the same foundational principles as ISO 13485, the framework converges after six iterations and generates traceable compliance checklist items, identifies regulatory gaps, and establishes clause dependency relationships.

The proposed approach demonstrates that a design methodology can provide both the reasoning structure and convergence criterion required for automated regulatory analysis. By bridging regulatory text and engineering requirements, this thesis establishes an automatic and traceable framework to support compliance-oriented decision-making in medical device development.