As ongoing work, I am developing reinforcement learning-based autonomous reconnaissance planning algorithms for the OSIRIS-APEX mission to asteroid Apophis, focusing on uncertainty-aware decision making and autonomous spacecraft operations.

Developed deep learning-based state estimation algorithms for vehicle tracking and autonomous systems. Current neural process-based estimation framework achieves approximately 32% lower estimation error than state-of-the-art baseline approaches across multiple tracking scenarios. Research has resulted in an accepted paper at SPIE Defense + Security, poster at NAML, and other machine learning-focused AFRL conferences.

Constructed a 6-DOF NASA HL-20 spacecraft orbital re-entry simulation and implemented a State-Dependent Riccati Equation (SDRE) guidance law. Achieved approximately 19% improved trajectory tracking performance and 50% reduction in thrust energy consumption relative to baseline iLQR guidance.

Designed and implemented an autonomous navigation system for a wheeled mobile robot using a single HC-SR04 ultrasonic sensor and Arduino-based feedback control. Demonstrated navigation performance within 7% of human teleoperation metrics while utilizing significantly reduced sensing requirements.

Developed a Unity-based simulation environment for manipulator-equipped soft robotic systems performing shared-autonomy object manipulation. Validated control algorithms using real-time physics simulation to support future deployment on physical robotic platforms.