Odometry
Odometry is the estimation of a robot's change in position and orientation over time using motion sensors, such as wheel encoders, cameras, or inertial units. It provides the short-term motion tracking that navigation and SLAM build on.
In this article
Building visual or physical AI?
Let's talk.
What is odometry?
Odometry estimates how far and in what direction a robot has moved by integrating readings from its own motion sensors. Wheel odometry counts wheel rotations, visual odometry tracks how the scene shifts between camera frames, and inertial odometry integrates acceleration and rotation from an IMU. Adding up these increments gives a running estimate of the robot's trajectory.
Its weakness is drift: small per-step errors accumulate, so odometry alone slowly diverges from the true path.
Key takeaways
- Odometry estimates motion by integrating sensor readings over time.
- It comes in wheel, visual, and inertial forms, often combined.
- It drifts over time, so it is usually corrected by other methods.
How it works
Each sensor provides an increment of motion between time steps, and those increments are summed to track pose. Because every measurement carries a little noise, the errors compound, which is why odometry is typically fused with other sensors and corrected by SLAM's loop closure or absolute references. Fusing complementary sensors reduces drift compared to any one alone.
Why it matters
Odometry is the fast, local motion estimate that higher-level navigation depends on, and it is a key input to SLAM. Understanding its drift explains why physical AI systems fuse many sensors rather than trusting a single stream.
Frequently asked questions
Why does odometry drift?
Each step's small measurement errors accumulate as they are summed, so the estimate gradually diverges from the true position.
How is odometry different from SLAM?
Odometry estimates incremental motion, while SLAM also builds a map and uses it to correct the accumulated drift in that motion estimate.
Related terms
Jesse Mostipak
SEO & Content
Jesse Mostipak is the SEO and Content Manager at Voxel51, where the work is helping humans find and trust what the brand knows, and teaching the Google knowledge graph and the LLMs answering on their behalf to do the same. That question, how knowledge gets built inside a system, is one Jesse has been chasing for years. Earlier versions of it ran through a New York City high school science classroom, data science and machine learning, and developer relations at Kaggle, Posit (formerly RStudio), and Baseten. The answer doesn't change much depending on whether the learner is a teenager, a software engineer, or a knowledge graph. Jesse holds a Master's in Education from CUNY Hunter College. LinkedIn
See all articles by Jesse Mostipak
Last updated July 9, 2026
Building visual or physical AI?
Let's talk.
© 2026 Voxel51 All Rights Reserved