A Comparison between Rapidly Randomized Tree and Efficient Frontier Methods for Autonomous Mobile Robot Exploration

Traditionally, an available off-line map was used and more recently, by using novel techniques, robots could automatically generate a map of their environment. These techniques provided major improvements in the navigation process. Exploration of an environment consists of leading a mobile to some intermediate points where the robot expands its gathered map until the whole environment map becomes known. There are many environment exploration methods, two of which are more common; Rapidly Randomized Tree and Efficient Frontier (based on boundary search) Exploration Methods. This article aims to investigate the performance of these two methods in real environments by implementing them first within the Gazebo simulation environment and then on a commercial TurtleBot® mobile robot. The simulation and experimental results of both methods are demonstrated in terms of some crucial parameters, including explored area dimensions and traveling time and distance to determine the efficiency of each exploration technique. While the output parameter of time indicates whether the related method is quick enough for the desired application or not, the distance traveled by the robot is an indicator of energy consumption. Lastly, it is shown that the second method improves the robot’s performance in exploration for all two independent parameters.