AF: Small: Strategy Problems for Robots and Sensor Networks

Cuny City College, New York NY

Investigators

Abstract

This award aims at developing strategies that allow mobile robots and sensor networks to perform their tasks near-optimally, or at least competitively. The tasks considered in this proposal deal mainly with exploration, patroling, guarding and searching. Typical tasks are that a group of robots should explore an unknown structure consisting of rooms which are connected by passages, or that a mobile guard should patrol a set of corridors, so that he can look down each corridor once during his patrol round. A typical problem involving sensors is to place sensors with a limited sensing range in a region of interest in such a way that any intruder can move only a small distance without being discovered by a sensor. For each of these problems, a strategy needs to be developed, its worst-case performance needs to be analyzed, and it needs to be compared with the unknown optimal strategy, to show that the proposed strategy is nearly optimal. Some of the strategies developed in this project will be implemented and tested by a cooperating robotics lab.