Multi-robot collaborative transportation is a critical capability that has attracted significant attention over recent years. To reliably transport a kinematically constrained payload, a team of robots must closely collaborate and coordinate their individual velocities to achieve the desired payload motion. For quadruped robots, a key challenge is caused by their anisotropic velocity limits, where forward and backward movement is faster and more stable than lateral motion.
In order to enable dual-quadruped collaborative transportation and address the above challenges, we propose a novel Bilevel Learning for Collaborative Transportation (BLCT) approach. In the upper-level, BLCT learns a team collaboration policy for the two quadruped robots to move the payload to the goal position, while accounting for the kinematic constraints imposed by their connection to the payload. In the lower-level, BLCT optimizes velocity controls of each individual robot to closely follow the collaboration policy while satisfying the anisotropic velocity constraints and avoiding obstacles. Experiments demonstrate that our BLCT approach well enables collaborative transportation in challenging scenarios and outperforms baseline approaches.
A motivating scenario for dual-quadruped collaborative transportation in a warehouse environment: the robots must not only make team decisions for collaboration but also generate individual navigational controls, while satisfying multiple constraints due to quadruped and payload kinematics, boundary collisions, and anisotropic velocity limits.
Overview of our BLCT approach: By formulating collaborative transportation as a bilevel learning problem, BLCT uses an upper-level optimization to learn collaborations and a lower-level optimization to generate robot controls, while considering the kinematic, collision, and anisotropic velocity constraints.
We utilize two Unitree Go1 quadruped robots to deploy and validate BLCT in a real-world scenario. Two ball joints are 3D printed and attached at both ends of a 2kg wooden payload, and are then mounted on top of the two quadruped robots. The quadruped robots are velocity-controlled using the Go1 SDK and ROS. We use an OptiTrack Motion Capture system to determine the ground truth locations of the robot team and the boundaries, which are published over the ROS network. All videos are running in real-time.
We employ two Unitree Go1 quadruped robots to run and validate BLCT in simulation. Two ball joints and a 45-inch payload are modeled in 3D and imported to the Gazebo simulator. The quadruped robots are velocity-controlled via ROS using the Champ locomotion controller. We obtain robot and boundary poses from Gazebo and ROS directly. The videos are running at around 0.6x Gazebo Real Time Factor.
@article{jose2024blct,
author = {Jose, Williard Joshua and Zhang, Hao},
title = {Bilevel Learning for Dual-Quadruped Collaborative Transportation under Kinematic and Anisotropic Velocity Constraints},
year = {2024}
}