Motion-planning schemes in global coordinates

This chapter considers the basic concepts of motion-planning of mobile robots acting in global coordinates. It addresses the widely accepted concept of configuration space and its corresponding notation. Motion-planning schemes are presented for two settings: completely known configuration space and partially known configuration space. There are many examples of motion-planning algorithms in the completely known configuration space. Informally, they can be categorized into three main groups: potential-field algorithms, grid-based algorithms, and sampling-based algorithms. For motion planning with partially known configurations space, it is assumed that when it starts moving, the robot does not know about the obstacles and partially observes them only when it is close enough. For such problems, the chapter also considers three basic algorithms of the same family - BUG0, BUG1, and BUG2. The considered algorithms form the basis of motion-planning methods widely used for both mobile robots and also other tasks that deal with the activity of autonomous agents.