1. Home
2. Faculty of Engineering

EE585 PROBABILISTIC ROBOTICS

Course Objectives

The purpose of this course is to teach the use of advanced probabilistic techniques to solve canonical problems of mobile robotics and autonomous systems. These problems include localization, map building, navigation and exploration. To this end we also need to investigate noisy sensor and actuator models. By the end of this course, you will have gained theoretical and practical application knowledge about the use of the Bayes Filter for estimation and its various implementations to solve these problems. Hopefully, you will be at a point to use this knowledge to advance your research goals or to find more competitive engineering solutions in your work.

Course Content

Introduction to probabilistic methods for mobile robots and autonomous systems. Recursive state and parameter estimation using Bayes filter. Probabilistic robot motion, actuator and motion models. Probabilistic robot sensing and perception, sensor models. Gaussian and non-parametric filters for estimation. Canonical problems of localization and mapping. Simultaneous localization and mapping (SLAM). Introduction to probabilistic planning and control.

Course Learning Outcomes

By the end of this course, the students will be able to:

• Apply recursive state estimation through the Bayes Filter framework,
• Build probabilistic motion models for mobile robotic systems such as differential drive wheeled robots,
• Build probabilistic measurement (sensor) models for mobile robotic sensors such as laser scanners, cameras, proximity sensors etc,
• Design and implement estimation algorithms using the Kalman Filter, Extended Kalman Filter, Unscented Kalman Filter and Particle Filter approaches,
• Solve Localization, Mapping and SLAM canonical robot perception problems using the aforementioned filter structures,
• State assumptions, limitations and practical issues and problems with probabilistic robotics algorithms.