Online strategies for coverage of uncertain terrains using an autonomous mobile robot

Abstract

Coverage path planning (CPP) algorithms customize an autonomous robot’s trajectory for various applications. In surveillance and exploration of unknown environments, random CPP can be very effective in searching and finding the objects of interest. Robots with random-like search algorithms need to be unpredictable in their motion and simultaneously scan an uncertain environment, avoiding intruders and obstacles in their path. Inducing chaos into the robot’s controller system makes its navigation unpredictable, accounts for better scanning coverage, and avoids any hurdles (obstacles and intruders) without the need for a map of the environment. The unpredictability, however, will come at the cost of increased coverage time. Due to the associated challenges, previous studies have ignored the coverage time and focused instead on the coverage rate only. This paper establishes novel methods that address the coverage time challenge of chaotic path planners. Initially, four chaotic state-of-the-art dynamic systems were selected for their fastest coverage time in an unknown environment of size 50 m × 50 m and greater. These systems were then coupled together and manipulated using 3 chaos control techniques and 2 obstacle avoidance techniques to achieve a fast coverage of the environment and simultaneously avoid obstacles. Considering the average coverage time for the manipulated systems, the outcome has been a technique that can cover 90% of an area, 53.4% less than the state-of-the-art systems.