Welcome to the Robotics and State Estimation Lab!

The UW Robotics and State Estimation Lab was established by Dieter Fox in 2001. We are interested in the development of embedded computing systems that interact with their environment in an intelligent way. To investigate such systems, we focus on probabilistic methods and their application to mobile robotics and activity recognition. We develop rich yet efficient methods for representing uncertainty and for reasoning under uncertainty. These methods are evaluated using problems in mobile robot localization, map building, and multi-robot collaboration. Furthermore, we investigate state estimation and machine learning in areas such as people tracking and high-level activity recognition.

News

Jonathan's paper GP-UKF: Unscented Kalman Filters with Gaussian Process Prediction and Observation Models won the Best Student Paper Award at IROS-07!

The book on Probabilistic Robotics is out.

Take a look at the MCL in action web site.

WiFi-based tracking of a person using a Gaussian process sensor model (true position in yellow).
Click here for more information on our activity recognition research.