IMPLEMENTATION OF BEHAVIOUR BASED NAVIGATION IN A PHYSICALLY CONFINED SITE

Behaviour-based architecture is one of the most effective autonomous
navigation techniques, second only to machine learning. However, specific algorithm
may only be effective for a specific environment. When the environment is changed
the algorithm may fail to adapt itself to the new configuration. This study looks at
developing a behavior-based algorithm that adaptable to several different
environment configurations. Selected basic behaviour-based algorithms such as wallfollower,
obstacle avoidance, escape route and target detection are to be combined
and its efficiency is measured. Experiments conducted in different mazes show that
robot is able to navigate itself from the starting point to the target with 4% to 10%
extra turns. The efficiency of the mobile robot increases by 77% compared to
manually controlled robot. The behaviour based implementation on the mobile robot
managed to find the shortest path with the rate of failure of less than 15%. However,
the time taken for the robot to reach the target can be variable depending on the
complexity of the maze.