Table of Contents
Fetching ...

Steerable rolling of a 1-DoF robot using an internal pendulum

Christopher Y. Xu, Jack Yan, Kathleen Lum, Justin K. Yim

TL;DR

ROCK (Rolling One-motor Controlled rocK), a 1 degree-of-freedom robot consisting of a round shell and an internal pendulum, is presented, allowing for mechanically simple designs that may be feasible to scale to large quantities or small sizes.

Abstract

We present ROCK (Rolling One-motor Controlled rocK), a 1 degree-of-freedom robot consisting of a round shell and an internal pendulum. An uneven shell surface enables steering by using only the movement of the pendulum, allowing for mechanically simple designs that may be feasible to scale to large quantities or small sizes. We train a control policy using reinforcement learning in simulation and deploy it onto the robot to complete a rectangular trajectory.

Steerable rolling of a 1-DoF robot using an internal pendulum

TL;DR

ROCK (Rolling One-motor Controlled rocK), a 1 degree-of-freedom robot consisting of a round shell and an internal pendulum, is presented, allowing for mechanically simple designs that may be feasible to scale to large quantities or small sizes.

Abstract

We present ROCK (Rolling One-motor Controlled rocK), a 1 degree-of-freedom robot consisting of a round shell and an internal pendulum. An uneven shell surface enables steering by using only the movement of the pendulum, allowing for mechanically simple designs that may be feasible to scale to large quantities or small sizes. We train a control policy using reinforcement learning in simulation and deploy it onto the robot to complete a rectangular trajectory.

Paper Structure

This paper contains 13 sections, 2 equations, 4 figures, 1 table.

Figures (4)

  • Figure 1: Overlaid snapshots of rock turning and jumping.
  • Figure 2: Left: Internals of ROCK, showing the electronics and pendulum. Right: Uneven exterior plastic shell of ROCK.
  • Figure 3: Left: The projection controller computes pendulum angle $\theta$ such that its projection onto the horizontal plane is in commanded direction $\textbf{d}$. Right: A policy is trained in simulation to track a randomly sampled commanded direction on rough terrain.
  • Figure 4: Overhead view: A user controls the robot to follow a rectangular trajectory marked by blue tape on the ground. The actual path of the robot is shown in yellow. The trajectory is completed in 80 seconds, resulting in an average speed of about 0.13 m/s.