Learning Bipedal Locomotion on Gear-Driven Humanoid Robot Using Foot-Mounted IMUs

Learning Bipedal Locomotion on Gear-Driven Humanoid Robot Using Foot-Mounted IMUs Sotaro Katayama 1, Y uta Koda 2, Norio Nagatsuka 2, and Masaya Kinoshita 1 Abstract -- Sim-to-real reinforcement learning (RL) for humanoid robots with high-gear ratio actuators remains challenging due to complex actuator dynamics and the absence of torque sensors. T o address this, we propose a novel RL framework leveraging foot-mounted inertial measurement units (IMUs). Instead of pursuing detailed actuator modeling and system identification, we utilize foot-mounted IMU measurements to enhance rapid stabilization capabilities over challenging terrains. Additionally, we propose symmetric data augmentation dedicated to the proposed observation space and random network distillation to enhance bipedal locomotion learning over rough terrain. The experimental results demonstrate that our method improves rapid stabilization capabilities over non-rigid surfaces and sudden environmental transitions. I. INTRODUCTION Bipedal and humanoid robots have fascinated people for decades. One of their anticipated roles is to replace human workers. Humanoid robots, which have morphologies similar to humans, are expected to navigate environments accessible to humans and perform tasks that humans can accomplish.