Fast visibility evaluation for covert robotics path planning

Abstract

In previous publications we have presented a promising new idea on covert path planning. The aim is to allow a robot to plan a path with minimum exposure to known or unknown hostile sentries within the same environment. An important element of our approach is to create what we have called a visibility risk map. In case of unknown sentries' locations, the visibility risk map is produced by measuring the exposure value of each free space cell. The early method for this process suffers from high computation time when dealing with high resolution maps. This drawback appears clearly in the case of unknown environments where the visibility risk map needs to be updated continuously as the environment's map is updated online. In this paper, a faster method is presented that based on triangulation geometry to find an approximated exposure value of each cell. The new method has been tested on a simulated environment which has shown its speed improvement over the previous method, as well as generating a smoother covert path.