Probabilistic data association for simultaneous localization and mapping

This invention describes a mobile robot that can figure out its own location and build a map of its surroundings at the same time. It does this by using data from its movement sensors and pictures taken by its camera. For each picture, it identifies various objects. Then, it uses a specific mathematical technique called an Expectation-Maximization (EM) algorithm to simultaneously calculate the robot's path and determine the type and exact position of each detected object in the environment. The claims specifically focus on this EM algorithm approach.