Projects

Humanoid Robot AI

Development of Reinforcement Learning–Based Free-Floating and Contact-Based Locomotion Control in Microgravity for Humanoid Robots Operating Inside Spacecraft

Research Problem

Stable humanoid mobility and reliable recontact with structures inside spacecraft under microgravity.

Research Approach

• Learn push-off, attitude stabilization, and handrail grasping policies.
• Integrate floating and contact-based motion with hierarchical reinforcement learning.

Advanced Air Vehicle AI

Development of Reinforcement Learning-based Flight Control for AAV under Propulsion Performance Degradation

Research Problem

Maintaining AAV flight stability, position and attitude tracking, and recovery when propulsion performance degrades.

Research Approach

• Augment baseline flight control with reinforcement learning–based compensation.
• Evaluate tracking and recovery under varied propulsion degradation.

Defense AI

Drone Detection and Interception Using a Quadruped Robot

Research Problem

Detection of threatening drones and interception from a mobile quadruped platform in counter-drone defense scenarios.

Research Approach

• Study threatening-drone detection from a quadruped platform.
• Explore drone interception using a mounted gun.