Jeffrey R. Peters

Camera Network Coordination for Intruder Detection

This paper proposes surveillance trajectories for a network of autonomous cameras to detect intruders. We consider smart intruders, which appear at arbitrary times and locations, are aware of the cameras configuration, and move to avoid detection for as long as possible. As performance criteria, we consider the worst case detection time (WDT) and the average detection time (ADT). We focus on the case of a chain of cameras, and we obtain the following results. First, we characterize a lower bound on the WDT and on the ADT of smart intruders. Second, we propose a team trajectory for the cameras, namely equal-waiting trajectory, with minimum WDT and with guarantees on the ADT. Third, we design a distributed algorithm to coordinate the cameras along an equal-waiting trajectory. Fourth, we design a distributed algorithm for cameras reconfiguration in the case of failure or network change. Finally, we illustrate the effectiveness and robustness of our algorithms via numerical studies and experiments.

Human Supervisory Control of Robotic Teams: Integrating Cognitive Modeling with Engineering Design

This article focuses on the design of systems in which a human operator is responsible for overseeing autonomous agents and providing feedback based on sensor data. In the control systems community, the term human supervisory control (or simply supervisory control) is often used as a shorthand reference for systems with this type of architecture [5]-[7]. In a typical human supervisory control application, the operator does not directly manipulate autonomous agents but rather indirectly interacts with these components via a central data-processing station. As such, system designers have the opportunity to easily incorporate automated functionalities to control how information is presented to the operator and how the input provided by the operator is used by automated systems. The goal of these functionalities is to take advantage of the inherent robustness and adaptability of human operators, while mitigating adverse effects such as unpredictability and performance variability. In some contexts, to meet the goal of single-operator supervision of multiple automated sensor systems, such facilitating mechanisms are not only useful but necessary for practical use [8], [9]. A successful system design must carefully consider the goals of each part of the system as a whole and seamlessly stitch components together using facilitating functionalities.

Sensor Network Localization on the Group of Three-Dimensional Displacements

We consider the problem of estimating relative configurations of nodes in a sensor network based on noisy measurements. By exploiting the cyclic constraints induced by the sensing topology to the network, we derive a constrained optimization on

Attention Allocation Aid for Visual Search

SE(3)^n

Robust Task Scheduling for Multi-Operator Supervisory Control Missions

. For the case of a network with a single cyclic constraint, we present a closed-form solution. We show that, in certain cases, namely restriction to pure rotation and pure translation, this solution is independent of the particular representation of the constraint function and is the unique, constrained minimizer of an appropriate cost. For sensing topologies represented by a general, weakly connected digraph, we present a solution method which is based on the limits of the solution curves of a continuous-time ordinary differential equation. We show that solutions obtained by our method satisfy all semicycle (generalized cycle) constraints induced by the sensing topology of the network. Further, we show through numerical simulation that for “Gaussian-like...

Coverage control with anytime updates for persistent surveillance missions

This paper outlines the development and testing of a novel, feedback-enabled attention allocation aid (AAAD), which uses real-time physiological data to improve human performance in a realistic sequential visual search task. Indeed, by optimizing over search duration, the aid improves efficiency, while preserving decision accuracy, as the operator identifies and classifies targets within simulated aerial imagery. Specifically, using experimental eye-tracking data and measurements about target detectability across the human visual field, we develop functional models of detection accuracy as a function of search time, number of eye movements, scan path, and image clutter. These models are then used by the AAAD in conjunction with real time eye position data to make probabilistic estimations of attained search accuracy and to recommend that the observer either move on to the next image or continue exploring the present image. An experimental evaluation in a scenario motivated from human supervisory control in surveillance missions confirms the benefits of the AAAD.

Cloud-Supported Coverage Control for Persistent Surveillance Missions

In future collaborative missions involving humans and unmanned aerial vehicles, especially those involving multiple human operators, the issue of resource allocation will be a crucial component to system success. Traditional deterministic strategies for task allocation and scheduling, such as those designed for job-shop applications, can often lead to poor performance in human-centered systems because these strategies fail to account for operator cognitive requirements or for the large amounts of uncertainty in human behavior. In light of this, a flexible framework that can potentially address both of these issues in finite horizon scheduling applications involving multiple operators is presented. Specifically, operator task load constraints are included as a part of a mixed integer program-based scheduling framework, which also incorporates robustness to uncertain processing times through the use of scenarios. The utility and modularity of this framework is then explored through the introduction of adapt...

Robust scheduling strategies for collaborative human-UAV missions

Coordination schemes for multi-agent surveillance missions typically require ideal data transfer between spatially separated agents, an assumption that is too restrictive for many realistic missions. This paper develops dynamic coverage control algorithms that only require unplanned and sporadic exchanges between mobile agents and a central base station. In particular, the proposed schemes are designed to operate within a decomposition-based multi-agent surveillance framework, which pairs dynamic partitioning with single-agent routing. The present work extends our previous work [13] by introducing two alternative coverage update rules that operate in anytime. We show that these variations add robustness to premature algorithmic termination, while preserving many desirable properties, namely, production of connected coverage regions, assurance of persistent coverage, and convergence to a Pareto optimal configuration in certain conditions.

Thinking Robotics: Teaching Robots to Make Decisions

A cloud-supported coverage control scheme is proposed for multi-agent, persistent surveillance missions. This approach decouples assignment from motion planning operations in a modular framework. Coverage assignments and surveillance parameters are managed on the cloud, and transmitted to mobile agents via unplanned and asynchronous exchanges. These updates promote load-balancing, while also allowing effective pairing with typical path planners. Namely, when paired with a planner satisfying mild assumptions, the scheme ensures that (i) coverage regions remain connected and collectively cover the environment, (ii) regions may go uncovered only over bounded intervals, (iii) collisions (sensing overlaps) are avoided, and (iv) for time-invariant event likelihoods, a Pareto optimal configuration is produced in finite time. The scheme is illustrated in simulated missions.

Multi-Layer Framework for Synthesis and Evaluation of Heterogeneous System-of-Systems Composed of Manned and Unmanned Vehicles

Resource allocation in collaborative human-UAV missions has become an important research area in recent years. Traditional deterministic strategies for task scheduling, such as job-shop schemes, can lead to poor performance, since these strategies fail to account for human cognitive requirements or behavioral uncertainty. In response, we present a flexible mixed-integer linear programming framework that can potentially address both of these issues in finite horizon scheduling applications. Specifically, we illustrate how cognitive workload constraints can be formulated as a mixed-integer linear program, and introduce robustness to uncertain processing times through the use of scenarios. We explore the modularity and utility of this simple framework by introducing additional layers of complexity, including receding horizon planning and adaptive estimation. Throughout the discussion, we use simulation studies to discuss the functionality of these algorithms, as well as various issues regarding practical implementation.