Citation Context

...and time; thus defining the best sampling path based on static measurements could lead to an ill-posed optimization. Finally, there has recently been work on persistent monitoring (Smith & Rus, 2010; =-=Smith, Schwager, & Rus, 2010-=-), in which the frequency of visits to each region is adapted to the time scale on which that region changes. This study draws from the latter two references and is directed toward developing a sampli...

Citation Context

... demonstrated in numerical simulations. I. INTRODUCTION In this paper we investigate the problem of controlling a sensing robot to persistently monitor a continually changing field in its environment =-=[1]-=-, [2], [3]. Consider, for example, an Unpiloted Aerial Vehicle (UAV) with a radiation sensor that must continually fly over the site of a nuclear accident to maintain an estimate of the time changing ...

Citation Context

...n-line solution through a gradient-based algorithm. We also show that the solution is robust with respect to the uncertainty model used. This establishes the basis for extending this approach to a two-dimensional mission space. I. INTRODUCTION Enabled by recent technological advances, the deployment of autonomous agents that can cooperatively perform complex tasks is rapidly becoming a reality. In particular, there has been considerable progress reported in the literature on robotics and sensor networks regarding coverage control [1], [2], [3], surveillance [4], [5] and environmental sampling [6], [7] missions. In this paper, we are interested in generating optimal control strategies for persistent monitoring tasks; these arise when agents must monitor a dynamically changing environment which cannot be fully covered by a stationary team of available agents. Persistent monitoring differs from traditional coverage tasks due to the perpetual need to cover a changing environment, i.e., all areas of the mission space must be visited infinitely often. The main challenge in designing control strategies in this case is in balancing the presence of agents in the changing environment so that it...

Citation Context

...imalist patrolling algorithms. Notice that in these last works the focus is mainly on finding optimal trajectories and in scheduling the motion of the robots. A different setup is considered in [11], =-=[12]-=-, where the robots are constrained to move along pre-specified tours, and the This material is based upon work supported in part by NSF grant IIS0904501, NSF grant CPS-1035917 and ARO grant W911NF-11-...

Citation Context

...ain approach to persistent monitoring is to cast the problem as one in combinatorial optimization. This research typically falls under the name of patrolling or persistent surveillance. In the most basic problem, a robot seeks to minimize the time between visits to each point in space. For this problem, both centralized approaches [9, 13, 21], and distributed algorithms for multiple robots [24] have been proposed. Recently, there has been work on minimizing the weighted time between visits to each region of interest, where the weight of a region captures its priority relative to other regions [27, 23]. These works focus on controlling robots along predefined paths. In [27], the authors consider velocity controllers for persistent monitoring along fixed tours, while in [23] the authors focus on coordination issues for multiple robots on fixed tours. An optimal control formulation for persistent monitoring in one-dimensional spaces is given in [8]. In our prior work [26], we considered a specific case of the min-max latency walk problem on Euclidean graphs, where the graphs are constructed by distributing vertices in a Euclidean space according to a known probability distribution. Under thes...

Citation Context

...nly path from point A to point B may be across certain hallways. In Section VI where we furthermore have online sensing, the pre-defined path could be the path dictated due to sensing constraints. In =-=[17]-=-, the authors mention that in ocean sampling, the path is typically pre-defined. Furthermore, autonomous aircrafts may be restricted to fly along a particular trajectory to stay away from commercial a...

Citation Context

...ing features and determine policies that minimize the maximum change in features between the observations. A persistent monitoring task where robots move on a given closed path has been considered in =-=[19]-=-, and a speed controller has been designed that minimizes the penalty due to no surveillance of a region. Srivastava et al [20] present a stochastic surveillance strategy based on the Markov chain Mon...

Citation Context

...y optimizing a sequence of switching locations for the agents. The problem of generating speed profiles for robots along predetermined closed paths for keeping bounded a varying field is addressed in =-=[10]-=-. The authors characterize appropriate policies for both single and multiple robots. In [11], decentralized adaptive controllers were designed to morph the initial closed paths of robots to focus on r...

Citation Context

... sensors with finite footprints. In our prior work, we developed methods for computing closed monitoring paths for robots [1], and for computing the speed with which each robot should follow its path =-=[2]-=-. This prior work assumes the paths are non-intersecting. However, most efficient monitoring paths may intersect. Thus, a collision avoidance procedure is needed. In this paper we develop a collision ...

Citation Context

...ant positive scalar. We would like to use the stability criterion for a persistent sensing task given the speed of the robot at each point along the path, referred to as the speed profile, defined in =-=[2]-=-, and plug it into the adaptive controller, such that the control action increases the stability margin of the persistent sensing task through time. However, since the robot does not know the environm...