Mobile robot navigation using a behavioural strategy
Panus Nattharith, Robert Bicker · 2009
This paper addresses the issue of mobile robot navigation in an indoor environment. In the system, an obstacle is defined as any object which prevents the mobile robot from achieving its goal. The robot needs to detect and avoid these objects and make its way to the desired location. Hence the robot’s behavioural ability in avoiding obstacles becomes a major issue, along with its ability to plan a path to its goal. The collision avoidance behaviour is a strategy which the mobile robot will adopt to prevent it from hitting obstacles as well as maintaining its track to its goal. In the proposed work, a hybrid architecture is adopted. It combines two components: a deliberative architecture and a reactive architecture which is based upon the motor schema. Several perceptual and motor schemas are employed in order to facilitate both the aspects of global and local navigation. The deliberative planning creates the safe path using a wavefront algorithm. It is then passed to the reactive module in terms of a series of waypoints. These are treated as immediate subgoals and are used to form behaviours leading the mobile robot to achieve its goal without experiencing any collision within an optimum travel distance and reasonable amount of time. High level processing of the data will be undertaken as part of the overall behavioural and path planning strategies, which will require significant consideration and design in order to implement them as part of a Player/Stage architecture.