XXooptRobotics
โ† Roadmap/State Estimation

Kalman / EKF / UKF

Optimal recursive estimation for (nearly) linear-Gaussian systems.

hardState Estimation

Why it matters in robotics

The Kalman filter is THE workhorse of robotic state estimation โ€” localization, IMU fusion, tracking. Interviewers love it because it tests probability, linear algebra, and systems thinking at once. You'll be asked why it's optimal, when it breaks, and how the EKF/UKF cope with nonlinearity.

Application focus

The same topic, tailored to the robot you're building. Your choice is remembered across the roadmap and every topic.

Select an application above.

At a glance

Belief(ฮผ, ฮฃ)Predict(motion model)Update(measurement)uโ‚œzโ‚œnew (ฮผ, ฮฃ)

The Kalman filter is a recursive predict โ†’ update loop over a Gaussian belief.

What to study

  • โœ“Linear KF: predict/update equations & the Kalman gain
  • โœ“Why it's the optimal linear-Gaussian estimator (MMSE)
  • โœ“EKF: linearization via Jacobians, and its failure modes
  • โœ“UKF: the unscented transform vs. EKF
  • โœ“Tuning Q and R; observability & filter divergence

Study by time budget

Pick the path that fits the time you have before your interview.

  1. โ–ถKalman Filter โ€” 5-minute intuitionโ†—VideoCyrill Stachnissยท ~6 min
  2. โœŽHow a Kalman filter works, in picturesโ†—ArticleTim Babbยท ~25 min

Prerequisites

Practice questions (3)