Research

Research at the intersection of reinforcement learning and optimal control for autonomous systems. Focused on combining learning-based methods with classical control theory to achieve safe, reliable robot behavior.

Key work includes Safe Autonomous Vehicle Navigation using Reinforcement Learning and Optimal Control — developing navigation policies that satisfy safety constraints through a hybrid approach combining RL exploration with control-theoretic guarantees.

Designed and tested physical lung phantoms for medical imaging calibration and validation. These anatomically representative models simulate lung tissue properties, enabling researchers to benchmark imaging algorithms and diagnostic tools without requiring patient data.

Work involved material characterization, fabrication of multi-layer tissue-mimicking structures, and systematic testing against known imaging standards.

Capstone mechatronics project integrating mechanical design, electronics, embedded systems, and control software into a complete autonomous system. Covered sensor integration, actuator control, real-time embedded programming, and system-level design for robustness.