What Is Wheel Odometry? How Robots Estimate Distance From Wheel Rotation.

What Is Wheel Odometry?

Wheel odometry is a method of estimating a mobile robot's movement using measurements of how its wheels rotate.

It is commonly used in wheeled robots, autonomous mobile robots, and automated guided vehicles.

For example, a warehouse robot may use wheel encoders to estimate how far it has traveled along an aisle.

Wheel odometry is relatively straightforward and provides useful information for navigation, although its accuracy can deteriorate as errors accumulate over time and distance.

Key Takeaways

• Wheel odometry estimates movement from wheel rotation and robot geometry.
• It commonly uses encoders to measure wheel movement.
• Wheel slip, incorrect wheel dimensions, and other mechanical effects can cause accumulated position errors.

How Does Wheel Odometry Work?

Imagine a robot moving on two driven wheels.

Each wheel has an encoder that measures rotation.

The robot's software knows the wheels' nominal radius.

By combining the wheel rotation with the radius, it estimates how far each wheel traveled.

For example, a wheel with a radius of 0.1 meters has a circumference of approximately:

2 × π × 0.1 = 0.628 meters

If the wheel completes ten revolutions without slipping, the corresponding travel distance is approximately:

10 × 0.628 = 6.28 meters

This is an idealized calculation.

Actual travel may differ because of wheel slip, deformation, or other effects.

How Does Differential-Drive Odometry Work?

A differential-drive robot uses independently controlled wheels on opposite sides.

When both wheels rotate at similar speeds, the robot generally travels straight under suitable conditions.

When the wheel speeds differ, the robot turns.

Odometry software uses the wheel measurements and the distance between the wheels to estimate the robot's movement.

For example, if the left wheel travels farther than the right, the robot has likely turned toward the right.

The software updates the robot's estimated position and orientation according to its kinematic model.

Wheel Odometry vs. Visual Odometry

Wheel odometry: Estimates movement from wheel rotation.

Visual odometry: Estimates movement from changes across camera images.

For example, a robot's wheel encoders may report movement even when its wheels are spinning on a slippery floor.

A camera-based system may provide information that helps identify the difference between wheel rotation and actual movement.

However, visual odometry can fail when lighting or visible features are insufficient.

The methods can complement each other.

Wheel Odometry vs. Localization

Wheel odometry estimates relative movement from a starting state.

Localization estimates a robot's pose relative to a reference map or coordinate system.

For example, wheel odometry may estimate that a robot has moved five meters forward.

However, it cannot independently determine the robot's absolute location in the warehouse without a reference.

Localization software may use LiDAR or other observations to correct accumulated odometry errors.

What Causes Wheel Odometry Drift?

Several factors can create errors.

Wheel slip: Rotation does not correspond exactly to ground travel.

Incorrect wheel diameter: The software converts rotations using the wrong distance.

Uneven surfaces: Changes in contact affect movement.

Encoder errors: Measurements may miss or misinterpret movement.

Incorrect wheel separation: Turning calculations become inaccurate.

Small errors can accumulate as the robot travels.

This is why wheel odometry is commonly combined with other sensor information.

What Is the Odom Coordinate Frame?

In ROS navigation, the odom frame provides a continuous local reference for estimated movement.

Odometry information is often designed to change smoothly as the robot travels.

However, the estimate may drift over time.

A separate map frame can provide a broader localization reference that corrects accumulated position errors.

The distinction helps navigation software use continuous local movement estimates while maintaining map-relative positioning.

Frequently Asked Questions

Can wheel odometry work without a camera?

Yes. It can use wheel encoders and the robot's mechanical geometry.

Does wheel odometry measure actual ground movement directly?

No. It estimates movement from wheel rotation, which can differ from actual travel when slipping occurs.

The Bottom Line

Wheel odometry provides a simple way to estimate robot movement from wheel rotation. It is useful for local navigation, but its accumulating errors make complementary positioning methods important.