In this Phase I, we will develop system-level architectures, protocols, methodologies, and optimal cooperative planning techniques to effectively implement Cooperative Behavior Controls, methods for enhancing cooperation through learning, and demonstrate their performance by simulating distributed heterogeneous unmanned agents on a set of mission scenarios with changing dynamic environments and mission parameters. The proposed system provides a comprehensive solution for facilitating and managing cooperative behaviors amongst heterogeneous systems, with a focus on enhanced behavior through learning. Our solution addresses multiple groups and echelons of agents, with mixed levels of autonomy, on mixed--initiative mission sets. The construct proposed is robust, scalable and portable. To demonstrate the implementation of the proposed architecture and paradigm, we will leverage our existing software ingredients for Intelligent--Autonomous Mobile Systems (I--AMS) that can be quickly integrated into various platform types. These decentralized individual intelligent systems are then simulated together to evaluate their overall cooperative behaviors with respect to the applied learning methodologies. We treat cooperative behaviors merely as an extension of the decision making process, taking into consideration the probable decisions that neighboring agents would make, allowing for cooperation with minimal communication requirements; similar to how humans and other higher cognitive forms of life cooperate.
Benefit: There are many military and civilian applications of the proposed cooperative behavior control system to automatically readjust mission objectives and tasks based on perceived changes in environments and mission requirements for cooperative unmanned systems. Such applications range from search and rescue, inspection and surveillance, reconnaissance, anti-terror surveillance, fire monitoring, pipeline inspection, traffic monitoring, disaster management and response to name a few.
Keywords: Cooperative Control, Learning, Uav, Ugv, Distributed Processing, Decentralized, Communication Bandwidth Constraint, Scalable Architecture
