Autonomous Mobile Robots and Multi-Robot Systems

1Motion‐Planning Schemes in Global Coordinates

Oded Medina and Nir Shvalb

1.1 Motivation

During the robot's tasks performance, it may change its configuration frequently and gather information from its environment using its sensors. Changing the robot's configuration from one to another can turn out to be not an easy task. The robot might perform an intricate maneuver from its initial configuration to its goal configuration rather than proceeding in a straight line in order to avoid obstacles. Such maneuvers may require detailed information regarding the robot's environment. For example, in a robot competition that involves several mobile robots playing soccer autonomously (RoboCup competition, small size league), the information is collected by a camera that captures the entire game field from above the playing ground. Another example where motion is a complicated task is the case of hyper‐redundant robots, especially when motion takes place in the vicinity of obstacles.

The term “motion planning” indicates the algorithm that calculates the maneuver from a start configuration c_s to a goal configuration c_g.

Many examples of motion‐planning algorithms are known (see, e.g., [Canny 1988; Laumond, Sekhavat, and Lamiraux 1998] for robotic arm and mobile robots motion planning designed to avoid obstacles collisions). Some of these algorithms are deterministic (i.e., they assume all information is accurately known or calculable) and some are stochastic (i.e., known to some extent). ...