Researchers at the Hong Kong University of Science and Technology (Guangzhou) have introduced FLORES, an innovative wheel-legged robotic system that blends the high efficiency of wheeled locomotion with the adaptive agility of legged movement—ideal for navigating diverse environments including stairs, uneven ground, and smooth surfaces, reports Tech Xplore.
Mechanical Innovation
FLORES sets itself apart with a unique front-leg design, replacing the conventional hip-roll degree of freedom (DoF) with a hip-yaw joint. This reconfigured joint assembly—combining hip-yaw, hip-pitch, knee-pitch, and wheel joints—grants efficient, smooth steering on flat terrain while preserving the ability to maneuver through complex obstacles.
Adaptive Locomotion
The robot seamlessly transitions between wheeled and legged modes, selecting the optimal mode based on terrain. This hybrid approach enables both fast traversal on smooth surfaces and robust adaptability in rugged conditions.
Efficiency Gains
Lab tests comparing FLORES to earlier wheel-legged robots showed dramatic improvements in energy consumption—only 30% for straight-line travel and 35% for turning maneuvers relative to prior models.
Control System
To harness the robot’s mechanical flexibility, the team implemented a reinforcement learning (RL) controller based on a Hybrid Internal Model (HIM), with a customized reward structure tailored to FLORES’s mechanical configuration. The result is adaptive, multi-modal locomotion with smooth transitions across locomotion types.
Future Directions
Going forward, the researchers plan to enhance FLORES with a lightweight manipulator arm for object handling, and test it in more demanding environments—such as narrow plank bridges, where it can utilize a bipedal stance with aligned wheels for enhanced lateral stability.