This thesis examines the role of function in representations of space by robots - that is, dealing directly and explicitly with those aspects of space and objects in space that serve some purpose for the robot. It is suggested that taking function into account helps increase the generality and robustness of solutions in an unpredictable and complex world, and the suggestion is affirmed by several instantiations of functionally conceived spatial models. These include perceptual models for the "on" and "in" relations based on support and containment; context-sensitive segmentation of 2-D maps into regions distinguished by functional criteria; and, learned predictive models of the causal relationships between objects in physics simulation. Practical application of these models is also demonstrated in the context of object search on a mobile robotic platform. / QC 20111125

This dissertation presents a novel multi-robot coordination control algorithm for search and rescue (SAR) operations. Continuous and rapid coverage of the unstructured and complex disaster areas in search of possible buried survivors is a time critical operation where prior information about the environment is either not available or very limited. Human navigation of such areas is definitely dangerous due to the nature of the debris. Hence, exploration of unknown disaster environments with a team of robots is gaining importance day by day to increase the efficiency of SAR operations. Localization of possible survivors necessitates uninterrupted navigation of robotic aiding devices within the rubbles without getting trapped into dead ends. In this work, a novel goal oriented prioritized exploration and map merging methodologies are proposed to generate efficient multi-robot coordination control strategy. These two methodologies are merged to make the proposed methodology more realistic for real world applications. Prioritized exploration of an environment is the first important task of the efficient coordination control algorithm for multi-robots. A goal oriented and prioritized exploration approach based on a percolation model for victim search operation in unknown environments is presented in this work. The percolation model is used to describe the behavior of liquid in random media. In our approach robots start prioritized exploration beginning from regions of the highest likelihood of finding victims using percolation model inspired controller. A novel map merging algorithm is presented to increase the performance of the SAR operation in the sense of time and energy. The problem of merging partial occupancy grid environment maps which are extracted independently by individual robot units during search and rescue (SAR) operations is solved for complex disaster environments. Moreover, these maps are combined using intensity and area based features without knowing the initial position and orientation of the robots. The proposed approach handles the limitation of existing works in the literature such as / limited overlapped area between partial maps of robots is sufficient for good merging performance and unstructured partial environment maps can be merged efficiently. These abilities allow multi-robot teams to efficiently generate the occupancy grid map of catastrophe areas and localize buried victim in the debris efficiently.