What is Localization and Mapping

Localization and mapping are similar, yet opposite, problems.

Localization is the process of determining the location and orientation of a robot relative to its environment. Solving the localization problem requires that the robot has a map of the entire environment based on which it can determine it's position.

Mapping is the process of generating a map of the robot's environment. In order to generate a map of the environment, the robot's exact position must be known (so that the map can be generated relative to the robot's position and orientation).

These two problems are inherently coupled and often must be solve simultaneously (determine the robot's location in an unknown environment). This is where simultaneous localization and mapping algorithms come in. These algorithms are capable of building a map of the environment while also localizing the robot relative to that map using probabilistic techniques.

Resources

These are a collection of resources that I have found to be helpful in understanding Localization, Mapping, and SLAM.

• Probabilistic Robotics, by Sebastian Thrun, Wolfram Burgard and Dieter Fox
  • The authors of this textbook are the leaders in the field of SLAM and probabilistic robotic techniques. This covers all of the fundamentals of the field with good details an explanation. It can be hard to translate the math into actual code, however.
• SLAM course video lectures, by Cyrill Stachniss @ University of Freiburg
  • This is a great lecture series on SLAM. He starts at the very basics to explain the motivation for SLAM and the basic mathematics principles driving it, but also goes in depth into actual implementation details. Complements the Probabilistic Robotics textbook very well
• Kalman Filter Video Series, by Michel van Biezen
  • This is a 55 video series on Kalman filters. It really dives deep into explaining the math behind them with plenty of good examples.
• Simultaneous Localisation and Mapping (SLAM): Part I The Essential Algorithms, by Hugh Durrant-Whyte and Tim Bailey
  • Another good paper explaining basic algorithms and implementation - recommended by Dr. Peters