Autonomous Vehicle Navigation With Dynamic Adaptation and Complete Coverage of Unknown Environments

This paper presents a Statistical Mechanics-inspired navigation algorithm with dynamic adaptation and complete coverage of unknown environments, which is built upon the concept of generalized Ising model. The algorithm enables autonomous vehicles to cover all areas in the environment, avoid unknown obstacles and adapt to target neighborhoods. Potential applications of this algorithm are humanitarian de-mining, hazard detection and floor-cleaning tasks. The algorithm has been validated on a Player/Stage simulator with an example of minesweeping.