⚡ electrical engineering

Rollbot: a Spherical Robot Driven by a Single Actuator

This paper introduces Rollbot, the first spherical robot capable of controlled 2D planar motion using only a single actuator by modulating trajectory curvature through acceleration and deceleration based on derived quasi-stable state dynamics.

Original authors: Jingxian Wang, Michael Rubenstein

Published 2026-05-13

📖 4 min read☕ Coffee break read

CC BY 4.0

Original authors: Jingxian Wang, Michael Rubenstein

Original paper licensed under CC BY 4.0 (http://creativecommons.org/licenses/by/4.0/). ✨ This is an AI-generated explanation of the paper below. It is not written or endorsed by the authors. For technical accuracy, refer to the original paper. Read full disclaimer

Imagine a ball that can roll around a room, but instead of having a complex engine with many moving parts, it only has one single motor inside. That is the story of Rollbot, a new invention by researchers at Northwestern University that challenges a long-held belief in robotics: that you need at least two motors to steer a ball-shaped robot in a straight line or a circle.

Here is a simple breakdown of how it works, why it's special, and what the researchers found.

The Big Idea: One Motor, Two Directions

Usually, to make a ball roll in a specific direction, you need a system that can push it left, right, forward, or backward independently. Think of it like a car: you need an engine to move forward and a steering wheel to turn. Most spherical robots use two "engines" (actuators) to do this.

Rollbot breaks this rule. It has only one motor. It doesn't have a steering wheel. Instead, it steers by changing how fast that single motor spins.

How It Works: The "Hula Hoop" Analogy

Inside the hollow ball, there is a heavy weight (a pendulum) attached to the motor.

The Setup: Imagine you are inside a giant hula hoop with a heavy bowling ball attached to a stick in your hand. If you spin the bowling ball around you, the hoop starts to roll.
The Trick: In Rollbot's case, the "bowling ball" is the internal weight. When the motor spins this weight, the whole ball rolls on the floor.
Steering: Here is the magic part. If the motor spins at a constant speed, the ball rolls in a perfect circle. But, if you speed up or slow down the motor, the size of that circle changes.
- Speed up: The circle gets bigger (like a wide turn).
- Slow down: The circle gets tighter (like a sharp turn).

By constantly adjusting the speed, the robot can trace a path that looks like a straight line or a complex shape, even though it is technically always rolling in a circle. It's like driving a car where you can't turn the steering wheel, but you can change the size of your turns by pressing the gas pedal harder or softer.

The "Quasi-Stable" State

The researchers discovered that if the motor spins at a steady pace, the robot settles into a predictable "dance." They call this a quasi-stable state.

In this state, the robot naturally finds a specific circle size based on how fast the motor is spinning.
If you bump the robot or the floor is slightly tilted, the robot naturally wobbles back into this dance pattern, much like a spinning top that rights itself after being nudged. This makes it very stable and hard to knock over.

Why Build This?

The paper highlights a few key reasons for this design:

Simplicity and Safety: Fewer motors mean fewer things to break.
The "Fail-Safe" Feature: This is a major point in the paper. Imagine a robot designed for a dangerous place (like the Moon or Mars) that usually has two motors. If one motor breaks, the robot is usually stuck. However, if that robot was built like Rollbot, it could switch to "one-motor mode" and keep working, albeit with slightly different movement. It turns a disaster into a manageable situation.
Swarm Robotics: Because it is simple and cheap to build, you could have hundreds of them working together.

What Did They Prove?

The researchers built a physical prototype (about the size of a large beach ball) and tested it in a lab.

They proved it works: They showed that by simply changing the motor speed, they could make the robot roll in circles of different sizes.
They proved it can follow a path: They programmed the robot to move in an "N" shape and stop at specific points. It successfully navigated these paths, stopping within a few inches of the target.
They proved it handles bumps: Even when the floor was slightly tilted (which would normally make the robot drift off course), the robot's control system compensated and kept it on track.

The Bottom Line

Rollbot is a "proof of concept." It's not the fastest robot (it moves about as fast as a slow walk), and it's not the most complex. But it proves that a spherical robot doesn't need a complex brain or multiple motors to move around a room. It just needs one motor, a heavy weight inside, and a clever way of changing speeds.

The researchers hope this design will inspire simpler, more robust robots for the future, especially for environments where reliability is more important than speed.

Technical Summary: Rollbot – A Spherical Robot Driven by a Single Actuator

Problem Statement
Conventional spherical robots typically require at least two actuators to achieve controlled two-dimensional (2D) planar motion. This requirement stems from the prevailing assumption that multiple degrees of freedom (DoFs) are necessary for a robot to maneuver in more than one direction. While underactuated systems offer advantages in energy efficiency, material reduction, and space savings—particularly valuable for swarm robotics and operations in hostile environments (e.g., lunar or Martian surfaces)—few mobile robots have successfully adopted "extreme underactuation" with a single actuator for continuous 2D ground locomotion. Existing single-actuator mobile robots often rely on compliant mechanisms, one-way bearings, or discrete stepping motions, but a ground robot capable of smooth, continuous 2D motion with a single motor has not been validated in the real world.

Methodology
The authors present Rollbot, a proof-of-concept spherical robot driven by a single brushless DC (BLDC) motor. The system consists of a spherical shell and an internal pendulum mass connected to the shell via the motor.

Dynamics Modeling: The authors derived the equations of motion for a general spherical robot driven by internal point masses. By applying Newton's second law and the angular momentum theorem, they formulated a complete equation of motion for the shell, accounting for the interaction between the internal mass, the shell, and the ground contact point. The model assumes no-slip contact and incorporates damping effects.
Quasi-Stable State Analysis: The core theoretical contribution is the identification and analysis of a "quasi-stable state." In this state, the robot rolls in a circular pattern where the driving speed ( $\omega_0$ ) is constant. The authors demonstrated that by numerically solving the derived dynamics, the radius of the robot's circular trajectory ( $R_0$ ) is a function of the driving speed. This allows the robot to control its path curvature simply by accelerating or decelerating the single motor.
Stability and Perturbation: The paper analyzes the stability of this quasi-stable state through first-order perturbation theory. Eigenvalue analysis of the system's linearized dynamics reveals that the state is stable (with a characteristic recovery time of approximately 7 seconds) provided the system parameters are within specific ranges. This stability allows the robot to recover from transient behaviors during acceleration/deceleration and external disturbances.
Hardware Design: The physical prototype (24 cm diameter, 1.2 kg) was designed to closely match the theoretical model. Key design choices included concentrating mass at the shell's surface to ensure a spherical moment of inertia ( $I \propto I_3$ ), positioning the motor and battery to maintain a geometric center of mass, and adding glass beads inside the shell to introduce linear damping for faster transient recovery.
Control Strategy: Instead of directly controlling the robot's position, the control algorithm manipulates the center of curvature ( $\vec{o}$ ). By modulating the driving speed to change the turning radius ( $R_0$ ), the system can steer the center of curvature toward a target. A PID controller is employed to compensate for non-idealities such as ground slope and motor limitations, ensuring the robot remains near the quasi-stable state.

Key Contributions

First Real-World Validation: Rollbot is the first spherical robot demonstrated to achieve controllable 2D planar motion using only a single actuator, challenging the assumption that two control DoFs are required.
Theoretical Framework: The paper provides a general derivation of dynamics for spherical robots with internal masses and introduces the concept of a "quasi-stable state" for single-actuator control, including a stability analysis via perturbation.
Fail-Safe Mechanism: The design serves as a potential fail-safe for conventional barycentric spherical robots. If a secondary motor fails, a standard robot could theoretically downgrade to the Rollbot configuration to continue limited operation.
Experimental Verification: The authors validated their theoretical models through open-loop motion tests, stable circular motion tracking, and waypoint following experiments.

Results

Trajectory Control: Experiments confirmed that the revolving radius ( $R_0$ ) can be modulated by changing the driving speed ( $\omega_0$ ). Theoretical predictions for $R_0$ and the tilting angle of the rotation axis ( $\xi$ ) matched experimental data with minor systematic deviations (6 mm and 1°, respectively), attributed to shell imbalance.
Stability: The robot successfully maintained stable circular motion around set points with radii ranging from 20 cm to 65 cm. Despite a ground tilt of approximately 1°, the feedback control compensated for the drift, keeping the robot within an average deviation of 3.4 cm from the expected trajectory.
Waypoint Following: Rollbot successfully navigated a complex "N" shaped path, stopping within 7 cm of target vertices.
Performance: The current prototype operates at an average speed of approximately 2 cm/s. The authors note this is slower than typical rolling robots (10 cm/s to 1 m/s) due to the conservative control strategy required to maintain the quasi-stable assumption.

Significance and Claims
The paper claims that Rollbot decisively challenges the assumption that spherical robots require multiple actuators for 2D motion. By leveraging the dynamics of a single internal mass and operating within a quasi-stable regime, the robot achieves precise, controllable locomotion. The authors emphasize that while the current implementation prioritizes concept validation over high-speed performance, the framework proves that single-actuator spherical robots are viable. The work suggests that future improvements in parameter selection and advanced control could enable these robots to match the performance of more complex conventional systems. The paper positions Rollbot not only as a novel minimalist robot but also as a robust solution for environments where actuator failure is a critical risk, offering a path for continued operation with reduced functionality.