Bipedal Robot
Bipedal robots aim to replicate human locomotion, focusing on stable and efficient walking, running, and manipulation tasks in diverse environments. Current research emphasizes developing robust control algorithms, often employing reinforcement learning, model predictive control, and various neural network architectures (e.g., diffusion models, convolutional neural networks) to achieve agile and adaptable locomotion, including navigation in crowded spaces and on uneven terrain. These advancements are significant for improving robot dexterity and reliability in challenging real-world scenarios, impacting fields such as search and rescue, manufacturing, and assistive technologies.
Papers
Torque-based Deep Reinforcement Learning for Task-and-Robot Agnostic Learning on Bipedal Robots Using Sim-to-Real Transfer
Donghyeon Kim, Glen Berseth, Mathew Schwartz, Jaeheung Park
Integrating Reconfigurable Foot Design, Multi-modal Contact Sensing, and Terrain Classification for Bipedal Locomotion
Ted Tyler, Vaibhav Malhotra, Adam Montague, Zhigen Zhao, Frank L. Hammond, Ye Zhao