Towards Intelligent, Agile, Safe, Precise, and Adaptive Robots for the Real World
The Hybrid Intelligent Experimental Robotics (HIER) Lab develops next-generation humanoid and animaloid robots capable of operating autonomously in complex, dynamic, and unstructured environments. Our research integrates planning, control, learning, and physical interaction to advance robotic capabilities beyond conventional locomotion toward intelligent, autonomous loco-manipulation in the real world.
Recent advances in robotics have enabled humanoid and legged robots to walk, climb stairs, and navigate challenging terrain. However, achieving reliable autonomy in real-world environments requires substantially more than locomotion alone. Future robots must simultaneously demonstrate intelligence, agility, safety, precision, and adaptability while physically interacting with their surroundings, manipulating objects, and responding to unpredictable conditions.
These capabilities introduce fundamental challenges and trade-offs in robot design, motion planning, control, and learning. The HIER Lab addresses these challenges through an integrated research approach that combines theoretical foundations, computational algorithms, and experimental validation on real robotic platforms.
Research Directions
Intelligent and Autonomous Humanoid Loco-Manipulation
We develop autonomous loco-manipulation capabilities that enable humanoid and animaloid robots to coordinate locomotion, manipulation, and environmental interactions in complex environments. Our research explores contact-rich motion planning, multi-contact behaviors, adaptive locomotion, and the integration of whole-body movements with task-oriented manipulation.
Learning-Enabled Whole-Body Planning and Control
We investigate hierarchical planning and whole-body control frameworks that integrate model-based optimization with learning-based approaches. By combining reinforcement learning, adaptive control, and whole-body optimization, we aim to achieve coordinated, agile, and robust robotic behaviors while accounting for physical constraints and dynamic interactions.
Heterogeneous Robot Teams with Physical and Contextual Interactions
We study collaborative robotic systems composed of heterogeneous platforms, including humanoids, quadrupeds, and robotic manipulators. Our research focuses on coordinated behaviors, physical interaction, and context-aware decision-making to enable robots with complementary capabilities to accomplish complex tasks collaboratively.
Optimal, Safety-Critical, and Stochastic Control
We develop theoretical and computational methods for controlling nonlinear, constrained, and uncertain robotic systems. Our research includes nonlinear optimal control, safety-critical control, stochastic control, and scalable optimization, with an emphasis on reliable decision-making and control under dynamic uncertainty and physical constraints.
Robotic Hardware and Mechanism Innovation
We design and develop innovative robotic mechanisms and integrated experimental platforms to overcome limitations of conventional robot morphologies. Our research investigates novel locomotion and manipulation mechanisms, morphology-aware design, custom robotic hardware, and integrated actuation and sensing systems that expand robotic capabilities in challenging real-world environments.