A unified framework for developing planning and control algorithms for active sensing, with examples of applications for modern sensor technologies.
Active sensor systems, increasingly vital to such applications as unmanned vehicles, mobile robots, and environmental monitoring, are characterized by a high degree of autonomy, reconfigurability, and redundancy. This book is the first to offer a unified framework for the development of planning and control algorithms for active sensing with multiple agents, with application examples including cameras and acoustic and gas sensors. The methods presented are characterized as information-driven because their goal is to optimize the value of information, rather than to optimize traditional guidance and navigation objectives.
The book explains relevant background in systems and control, graph, probability, and information theories; develops an integrated mathematical representation, or model, of system components and their interactions; and shows how motion planning, network, and control theoretic algorithms can be used to manage agent mode, position, and motion. It describes information-driven placement, navigation, and control methods that can be used to allocate limited resources so that sensing objectives, including coverage, detection, classification, and tracking, are optimized. These systems are able to process and learn from data, adapt autonomously to unexpected situations, self-organize to meet multiple objectives, and evolve over time to exhibit greater functionality in changing and complex environments. The book's unified notation and treatment allows direct comparison and parallel implementations of methods and algorithms drawn from disparate communities and disciplines.