Mazes have intrigued the human mind for thousands of years, and have been used to measure cognitive abilities of laboratory animals. In recent years, mazes have been used to examine the artificial intelligence of robots by observing their ability to traverse mazes using algorithm for maze exploration and exploitation.A simulation of a multi-agent system is used to demonstrate the benefits of utilizing a group of several robots in maze exploration. Using a behavioral algorithm based on Tarry’s algorithm, it is shown that the group performance improves and becomes more robust as the number of robots increases. In addition, the amount of data transfer required for group coordination can be minimized to a small set of data items, which is independent of either the number of robots in the group or the maze size.As a result, the above multi-agent approach can be scaled up to mazes or groups of any size, as indicated by the results of the MATLAB-based simulation.