Research Area:

Explainability in Validation & Verification of Decision-Making Systems:

My research focuses on enhancing the validation and verification (V&V) of autonomous systems through explainable testing methodologies. Autonomous decision-making models process multimodal sensory inputs, making safety, reliability, and transparency critical challenges. My work integrates explainability, interpretability, and transparency into V&V processes by leveraging test oracles, scenario generators, and interactive diagnostic tools. The ultimate goal is improving failure analysis, reducing debugging time, and enhancing trust by providing clear, structured insights into decision-making processes. Through empirical studies and real-time validation in simulation environments, my research contributes to the development of robust, user-centric testing frameworks for autonomous technologies across diverse domains, including automotive, robotics, and industrial automation.

Automotive Software Testing:

In the rapidly evolving landscape of software-defined vehicles (SDVs), ensuring the reliability and safety of automotive software is a critical challenge. My work in automotive software testing focuses on integrating fuzzing-based testing methodologies to uncover vulnerabilities in autonomous driving software. As part of my research, I contributed to the design of a mutation-based scenario fuzzing framework to enhance the testing of highly configurable automotive systems. This approach systematically generates and mutates driving scenarios to expose potential system failures, improving fault detection, robustness, and test coverage. By integrating software-in-the-loop (SiL) testing and CI/CD pipelines, my work supports the continuous validation of variant-rich automotive functions while optimizing testing efficiency. This research plays a crucial role in advancing safe and transparent software deployment for next-generation autonomous vehicles.