Haluk Bayram

Coalition formation games for dynamic multirobot tasks

In this paper we study the problem of forming coalitions for dynamic tasks in multirobot systems. As robots, either individually or in groups, encounter new tasks for which individual or group resources do not suffice, robot coalitions that are collectively capable of meeting these requirements need to be formed. We propose a hybrid approach to this problem where coalitions proceed with the task if they have sufficient resources after liberating redundant members while they report it to a task coordinator in cases where their resources do not suffice. In turn, the task coordinator forms capable coalitions based on coalition formation games in which groups of robots are evaluated together in regards to each task’s required resources and cost of forming a coalition. The resulting coalitions are such that no group of robots has a viable alternative to staying within their assigned coalition. Thus, as new tasks are confronted, coalitions merge and split so that the resulting coalitions are capable of the newly encountered tasks. Simulations and experiments performed on groups of heterogeneous mobile robots demonstrate the effectiveness of the proposed approach.

Decentralized network topologies in multirobot systems

This paper studies the problem of forming and updating the network topology in a multirobot system that is simultaneously engaged in a given task. The contribution of this paper is to propose a decentralized model of how the network may evolve based on network-related payoff functions and pairwise games. As such, pairwise games provide a practical and general scheme for contacting other robots and revising the network topology. A network is deemed acceptable by all the robots using pairwise stability and pairwise Nash equilibrium. As an application, we consider networks that are generated with mutual link-based payoff functions and show that – under some assumptions regarding changes in the configuration states – each game is ensured of converging to a pairwise stable network. This approach is then integrated with a common robotic task where the network is critical to successful task completion. The resulting performance is evaluated with respect to a variety of measures including task completion, network density, and the average payoff along with comparative results with all-to-all communication. Graphical Abstract

Pairwise vs coalition game networks for multi-robot systems

Abstract This paper considers the bidirectional interaction between states and communication network in a multi-robot system. It proposes two alternative games that are extremes of each other for network formation. In pairwise games, each robot decides to establish or break off a link based on the improvement this offers to it individually. In turn, its state evolves based on state information received from its immediate neighbors only. In coalition games, all the robots in each coalition decide collectively and new links are added or removed based on the improvement they offer to the respective coalition. This time, the state of each robot evolves based on state information from all the other robots in its coalition, albeit with delays that depend on their geodesic proximity. Simulation results provide insights on comparative performance with respect to task completion and connectivity under varying move periods.

Tracking wildlife with multiple UAVs: System design, safety and field experiments

We present a multi-UAV system capable of localizing radio tagged animals. Each UAV carries a Yagi antenna and is capable of obtaining bearing-measurements by using the directionality of the signal emitted from the tag. In this paper, we address the following question: given a set of measurement locations, should the UAVs move in coordination and obtain synchronized measurements, or should they move independently? For each scenario, we present a path planning algorithm. The algorithms are compared in simulations. We also present a complete system implementation with three UAVs and results from field experiments.

Multirobot communication network topology via centralized pairwise games

This paper studies the problem of communication network topology in a multirobot system while the agents are engaged in a given task. The contribution of this paper is to propose a centralized approach to network evolution. In this approach, a communication coordinator is responsible for determining the network topology. The robots periodically send their state information to the communication coordinator. In turn, the communication coordinator considers the individual communication payoff functions of all the robots, their current states and the current network simultaneously and finds a network topology acceptable to all the robots. It models the network topology formation as a pairwise game where it forms or severs pairwise links based on the improvement the resulting network offers the robot pairs relative to the current network. We show that with the assumed form of communication payoff functions, each pairwise game is ensured of convergence to a pairwise stable network. Furthermore, simulation results provide statistical results on the resulting network topology, the number of game moves and the processing time as well as comparative results with the proximity based approach.

Active localization of VHF collared animals with aerial robots

Tagging animals with radio devices is a commonly used technique in wildlife and conservation biology. We study the problem of localizing a tagged animal using the audio output of an off-the-shelf receiver mounted on an autonomous Unmanned Aerial Vehicle (UAV). This problem is motivated by the application of taking pictures of bears to document their interactions with the environment after contact with the radio signal has been established. We make two contributions. (1) Received signal characteristics depend on the environment, device and driver characteristics as well as target/receiver geometry. Due to complex interactions among these factors, directly estimating range or bearing from the signal may not be possible. Instead, we show that given a sufficient step-size, we can use signal characteristics to compare the proximity of two sensing locations with respect to the source. We present a novel model from data collected in field experiments. (2) Based on this model, we also present a local-search algorithm which drives the UAV toward the target. We study the performance of the algorithm in extensive simulations using field measurements. The algorithm has also been implemented on the UAV and tested in preliminary experiments.

Assistance networks for dynamic multirobot tasks

In this paper, we consider dynamic multirobot tasks that can be done by any of the robots, but only with the assistance of any other robot. We propose a novel approach based on the concept of ‘assistance networks’ with two complementary aspects, namely assistant finding and network topology update. Each robot, encountering a new task, seeks an assisting robot among its immediate neighbors in the assistance network in a decentralized manner. The network topology is defined based on pairwise stability via payoff functions that consider general task-related guidelines. As such, the number of potential assisting robots can be ensured a priori depending on tasks’ requirements. As robots move around, the topology is updated via pairwise games. If the games are conducted by a network coordinator, each game is shown to result in a pairwise stable network. A series of simulation and experimental results in a variety of different scenarios demonstrate that the robots are able to get assistance or give assistance flexibly.