Design and implementation of an efficient path planning algorithm for networked robots

Networked robots refer to multiple robots operating together in coordination. This work focuses on the problem of path planning of these robots in an unknown environment with obstacles. In the existing techniques, the robots collaboratively find the location of obstacles and available path and exchange the entire map of the environment among themselves. This creates computational and communication overhead in complex environments. We propose the Directed Ant Colony Optimization Algorithm(D-ACO) in which the environment detection task is shared between the robot and the static sensor network. The location of the target and obstacles is stored in the static sensor nodes instead of storing it in the robot's memory, thereby reducing the computational overhead. The information about the neighboring nodes and the obstacles within the communication range is relayed in the form of packets to the robots. Thus the proposed scheme reduces the communication overhead. The D-ACO algorithm is analyzed for both grid and random deployment of sensors. On comparison with the existing ACO techniques, simulation results show that the rate of convergence of the D-ACO algorithm is increased by two times in grid deployment, and by four times in random deployment of static sensors.