Obstacle Avoidance

Given the a priori knowledge of the environment and the goal position, mobile robot navigation refers to the robot’s ability to safely move towards the goal using its knowledge and the sensorial information of the surrounding environment. Even though there are many different ways to approach navigation, most of them share a set of common components or blocks, among which path planning and obstacle avoidance (may) play a key role. Given a map and a goal location, path planning involves finding a geometric path from the robot actual location to the goal. This is a global procedure whose execution performance is strongly dependent on a set of assumptions that are seldom observed in nowadays robots. In fact, in mobile robots operating in unstructured environments, or in service and companion robots, the a priori knowledge of the environment is usually absent or partial, the environment is not static, i.e., during the robot motion it can be faced with other robots, humans or pets, and execution is often associated with uncertainty. Therefore, for a collision free motion to the goal, the global path planning has to be associated with a local obstacle handling that involves obstacle detection and obstacle avoidance. Obstacle avoidance refers to the methodologies of shaping the robot’s path to overcome unexpected obstacles. The resulting motion depends on the robot actual location and on the sensor readings. There are a rich variety of algorithms for obstacle avoidance from basic re-planning to reactive changes in the control strategy. Proposed techniques differ on the use of sensorial data and on the motion control strategies to overcome obstacles. The Bug’s algorithms (section 2:1), [1], [2], follow the easiest common sense approach of moving directly towards the goal, unless an obstacle is found, in which case the obstacle is contoured until motion to goal is again possible. In these algorithms only the most recent values of sensorial data are used. Path planning using artificial potential fields, [3], (section 2:2) is based on a simple and powerful principle that has an embedded obstacle avoidance capabil-