In this thesis, we explore the topic of designing an architecture and processing algorithms for a multi-agent system, where agents need to address potential unexpected events in the environment, operating under soft real-time constraints. We first develop a classification of unexpected events into Opportunities, Barriers and Potential Causes of Failure, and outline the interaction required to support the allocation of tasks for these events. We then propose a hybrid architecture to provide for agent autonomy in the system, employing a central coordinating agent. Certain agents in the community operate autonomously, while others remain under the control of the coordinating agent. The coordinator is able to determine which agents should form teams to address unexpected events in a timely manner, and to oversee those agents as they perform their tasks. The proposed architecture avoids the overhead of negotiation amongst agent teams for the assignment of tasks, a benefit when operating under limited time and resource constraints. It also avoids the bottleneck of having one coordinating agent making all decisions before work can proceed in the community, by allowing some agents to work independently. We illustrate the potential usefulness of the framework by describing an implementation of a simulator loosely based on that used for the RoboCup Rescue Simulation League contest. The implementation provides a set of simulated computers, each running a simple soft real-time operating system. On top of this basic simulation we implement the model described above and test it against two different search-and-rescue scenarios. From our experiments, we observe that our architecture is able to operate in dynamic and real-time environments, and can handle, in an appropriate and timely manner, any unexpected events that occur. We also comment on the value of our proposed approach for designing adjustable autonomy multi-agent systems and for specific environments such as robotics, where reducing the overall level of communication within the system is crucial.

In this thesis, we explore the topic of designing an architecture and processing algorithms for a multi-agent system, where agents need to address potential unexpected events in the environment, operating under soft real-time constraints. We first develop a classification of unexpected events into Opportunities, Barriers and Potential Causes of Failure, and outline the interaction required to support the allocation of tasks for these events. We then propose a hybrid architecture to provide for agent autonomy in the system, employing a central coordinating agent. Certain agents in the community operate autonomously, while others remain under the control of the coordinating agent. The coordinator is able to determine which agents should form teams to address unexpected events in a timely manner, and to oversee those agents as they perform their tasks. The proposed architecture avoids the overhead of negotiation amongst agent teams for the assignment of tasks, a benefit when operating under limited time and resource constraints. It also avoids the bottleneck of having one coordinating agent making all decisions before work can proceed in the community, by allowing some agents to work independently. We illustrate the potential usefulness of the framework by describing an implementation of a simulator loosely based on that used for the RoboCup Rescue Simulation League contest. The implementation provides a set of simulated computers, each running a simple soft real-time operating system. On top of this basic simulation we implement the model described above and test it against two different search-and-rescue scenarios. From our experiments, we observe that our architecture is able to operate in dynamic and real-time environments, and can handle, in an appropriate and timely manner, any unexpected events that occur. We also comment on the value of our proposed approach for designing adjustable autonomy multi-agent systems and for specific environments such as robotics, where reducing the overall level of communication within the system is crucial.

When working with a complex AI or robotics system in a specific application, users often need to incorporate their special domain knowledge into the autonomous system. Such needs call for the ability to manage autonomy. However, managing autonomy can be a difficult task because the internal mechanisms and algorithms of the autonomous components may be beyond the users' understanding. We propose an approach where users manage autonomy indirectly by managing information provided to the intelligent system hierarchically at three different temporal scales: strategic, between-episodes, and within-episode. Information management tools at multiple temporal scales allow users to influence the autonomous behaviors of the system without the need for tedious direct/manual control. Information fed to the system can be in the forms of areas of focus, representations of task difficulty, and the amount of autonomy desired. We apply this approach to using an Unmanned Aerial Vehicle (UAV) to support Wilderness Search and Rescue (WiSAR). This dissertation presents autonomous algorithms/components and autonomy management tools/interfaces we designed at different temporal scales, and provides evidence that the approach improves the performance of the human-robot team and the experience of the human partner.

Les robots vont être de plus en plus utilisés dans les domaines civils, comme dans le domaine militaire. Ces robots, opérant en flottes, peuvent accompagner des soldats au combat, ou accomplir une mission en étant supervisés par un poste de contrôle. Du fait des exigences d'une opération militaire, il est difficile de laisser les robots décider de leurs actions sans accord d'un opérateur ou surveillance, en fonction de la situation. Dans cette thèse, nous nous attardons sur deux problématiques:D'une part, nous cherchons à exploiter l'autonomie ajustable de sorte à ce qu'un robot puisse accomplir sa mission de la manière la plus efficace possible, tout en respectant des restrictions assignées par un opérateur sur son niveau d'autonomie. Pour cela, celui-ci est en mesure de définir pour un ensemble d'états et d'actions donné un niveau de restriction. Ce niveau peut par exemple imposer au robot la télé-opération pour accéder à une zone à risque.D'autre part, comme nous envisageons la possibilité que plusieurs robots soient déployés en même temps, ces robots doivent se coordonner pour accomplir leurs objectifs. Seulement, comme les opérateurs peuvent prendre le contrôle de certains d'entre eux, la question de la coordination se pose. En effet, l'opérateur ayant ses propres préférences, perception de l'environnement, connaissances et étant sujet aux stress, hésitations, il est difficile de prévoir les actions que celui-ci va effectuer, et donc de s'y coordonner. Nous proposerons dans cette thèse une approche visant à estimer la politique exécutée par un robot télé-opéré à partir d'apprentissage basé sur les actions observés de ce robot.La notion de planification est très présente dans ces travaux. Ceux-ci se baseront sur des modèles de planifications comme les Processus Décisionnels de Markov. / Robots will be more and more used in both civil and military fields. These robots, operating in fleet, can accompany soldiers in fight, or accomplish a mission while being supervised by a control center. Considering the requirement of a military operation, it is complicated to let robots decide their action without an operator agreement or watch, in function of the situation.In this thesis, we focus on two problematics:First, we try to exploit adjustable autonomy to make a robot accomplishes is mission as efficiency as possible, while he respects restrictions, assigned by an operator, on his autonomy level. For this, it is able to define for given sets of states and actions a restriction level. This restriction can force, for example, the need of being tele-operated to access a dangerous zone.Secondly, we consider that several robots can be deployed at the same time. These robots have to coordinate to accomplish their objectives. However, since operators can take the control of some robots, the coordination is harder. In fact, the operator has preferences, perception, hesitation, stress that are not modeled by the agent. It is then hard to estimate his next actions, so to coordinate with him. We propose in this thesis an approach to estimate the policy executed by a tele-operated robot from learning methods, based on observed actions from this robot.The notion of planning his important in these works. These are based on planning models, such as Markov Decision Processes.

Searching for lost people in a Wilderness Search and Rescue (WiSAR) scenario is a task that can benefit from large numbers of agents, some of whom may be robotic. These agents may have differing levels of autonomy, determined by the set of tasks they are performing. In addition, the level of autonomy that results in the best performance may change due to varying workload or other factors. Allowing a supervisor and a searcher to jointly decide the correct level of autonomy for a given situation (“mixed initiative”) results in better overall performance than giving an agent absolute control over their level of autonomy (“adaptive autonomy”) or giving a supervisor absolute control over the agent's level of autonomy (“adjustable autonomy”).

Human-controlled robots are utilized in many situations and such use is becoming widespread. This thesis details research that allows a single human to interact with a team of robots performing tasks that require cooperation. The research provides insight into effective interaction design methods and appropriate interface techniques. The use of team-level autonomy is shown to decrease human workload while simultaneously improving individual robot efficiency and robot-team cooperation. An indoor human-robot interaction testbed was developed at the BYU MAGICC Lab to facilitate experimentation. The testbed consists of eight robots equipped with wireless modems, a field on which the robots move, an overhead camera and image processing software which tracks robot position and heading, a simulator which allows development and testing without hardware utilization and a graphical user interface which enables human control of either simulated or hardware robots. The image processing system was essential for effective robot hardware operation and is described in detail. The system produced accurate robot position and heading information 30 times per second for a maximum of 12 robots, was relatively insensitive to lighting conditions and was easily reconfigurable. The completed testbed was utilized to create a game for testing human-robot interaction schemes. The game required a human controlling three robots to find and tag three robot opponents in a maze. Finding an opponent could be accomplished by individual robots, but tagging an opponent required cooperation between at least two robots. The game was played by 11 subjects in five different autonomy modes ranging from limited robot autonomy to advanced individual autonomy with basic team-level autonomy. Participants were interrupted during the game by a secondary spatial reasoning task which prevented them from interacting with the robots for short periods of time. Robot performance during that interruption provided a measure of both individual and team neglect tolerance. Individual robot neglect tolerance and performance did not directly correspond to those quantities at the team level. The interaction mode with the highest levels of individual and team autonomy was most effective; it minimized game time and human workload and maximized team neglect tolerance.

This thesis presents a human interface which helps users efficiently allocate multiple unmanned ground vehicles (UGVs) cooperating to accomplish timing-sensitive missions in an urban environment. The urban environment consists of obstacles and a hazardous region. The obstacles represent a "no-go zone" while the hazardous region represents a high-risk area. The main object of this problem is to minimize the team operational cost while satisfying timing constraints. Operational costs for individual vehicles are based on risk and power consumption, and are calculated using path length and vehicle velocity. In this thesis, three types of timing constraints are considered: simultaneous arrival, tight sequential arrival, and loose sequential arrival. Coordination variables and functions are the strategy by which both temporal and spatial information is used to achieve cooperative timing at a minimum cost. Specifically, coordination variables and functions are used to plan trajectories for a team of UGVs that satisfy timing constraints. The importance of properly representing information to users, allowing them to make efficient decisions, is also discussed. Four different control interfaces (temporal, spatial, cost, and coordination variable/function control) were tested. A full factorial design of experiments was performed with response time, workload, and quality of decision as metrics used to evaluate the quality and effectiveness of each interface. Based on the results of this experiment, a final graphical user interface (GUI) was designed and is described. It incorporates a combination of coordination variable/function control and cost control. This GUI is capable of planning paths for vehicles based on cooperative timing constraints and enables users to make high quality decisions in deploying a group of vehicles.