robomechanics/quad-sdk

Software tools for agile quadrupeds, developed by the Robomechanics Lab at Carnegie Mellon University.

What it solves

Quad-SDK provides a unified software stack for agile quadrupedal locomotion, reducing the need for custom software changes when moving between different robot platforms or switching between simulation and hardware. It addresses the complexity of integrating planning, control, and estimation in legged robotics research.

How it works

Built on ROS (Robot Operating System), the framework uses a modular architecture that allows researchers to swap out components. It integrates several key capabilities:

  • Planning and Control: Supports both twist-based input for movement and a global planner for dynamic motions.
  • Estimation: Uses an Extended Kalman Filter (EKF) for reliable onboard state estimation.
  • Specialized Locomotion: Includes a controller specifically designed for walking through entanglements (like vines) to help the robot disentangle itself.
  • Simulation: Provides Gazebo simulation support and a suite of visualization and data-processing tools.

Who it’s for

Robotics researchers and developers working with quadrupedal robots who need a full-stack framework for experimenting with planning, control, and estimation algorithms.

Highlights

  • Full-stack integration: Combines planning, control, estimation, and communication in one package.
  • Modular design: Allows for easy experimentation with custom implementations of different components.
  • Modular simulation: Integrated Gazebo support for rapid development and testing.
  • Outdoor reliability: Includes EKF state estimation and a specialized entanglement-handling controller for outdoor environments.

Related

  • Project
  • Project
  • Project
  • Project
  • Project