Playful DoggyBot: Learning Agile and Precise Quadrupedal Locomotion

2024-09-30

arXiv.org

AI Summary

This work addresses the longstanding challenge in quadrupedal robotics of simultaneously achieving high agility and sub-centimeter end-effector positioning accuracy during dynamic locomotion—where perception and control fidelity are inherently limited. We propose a leap-and-grasp framework for legged robots that eliminates the need for external manipulators. Our approach integrates reinforcement learning–based end-to-end locomotion control, multi-modal state estimation, real-time torso-leg coordinated trajectory optimization, and sim-to-real transfer techniques, coupled with a front-mounted passive gripper for precise mid-air capture. To our knowledge, this is the first demonstration on a purely quadrupedal platform achieving both sub-2 cm end-effector positioning accuracy and highly dynamic jumping capability—resolving the fundamental agility–precision trade-off. In simulation, the system achieves grasping at 1.05 m height; on hardware—with a mere 0.3 m standing height—the robot successfully captures a suspended ball at 0.8 m.

Technology Category

• Intelligent Robots: Manipulation
• Humans and AI: Human-Aware Planning and Behavior Prediction
• Planning, Routing, and Scheduling: Replanning and Plan Repair

Application Category

• Responsible Web: Machine-in-the-loop, human agency and autonomy
• Search and Retrieval-Augmented AI: Agentic search
• Systems and Infrastructure for Web, Mobile and WoT: Applied ML and AI for Web-based mobile applications

Abstract

Quadrupedal animals have the ability to perform agile while accurate tasks: a trained dog can chase and catch a flying frisbee before it touches the ground; a cat alone at home can jump and grab the door handle accurately. However, agility and precision are usually a trade-off in robotics problems. Recent works in quadruped robots either focus on agile but not-so-accurate tasks, such as locomotion in challenging terrain, or accurate but not-so-fast tasks, such as using an additional manipulator to interact with objects. In this work, we aim at an accurate and agile task, catching a small object hanging above the robot. We mount a passive gripper in front of the robot chassis, so that the robot has to jump and catch the object with extreme precision. Our experiment shows that our system is able to jump and successfully catch the ball at 1.05m high in simulation and 0.8m high in the real world, while the robot is 0.3m high when standing.

Problem

• Achieving agile and precise quadrupedal locomotion in robots.
• Overcoming hardware limitations in robotics for object interaction.
• Developing a perception-control system for dynamic object tracking and catching.

Innovation

• Reinforcement Learning for precise control
• Perception-control decoupled system design
• Passive gripper for object interaction