Roi Yehoshua

Robotic adversarial coverage: Introduction and preliminary results

This paper discusses the problem of generating efficient coverage paths for a mobile robot in an adversarial environment, where threats exist that might stop the robot. First, we formally define the problem of adversarial coverage, and present optimization criteria used for evaluation of coverage algorithms in adversarial environments. We then present a coverage area planning algorithm based on a map of the probable threats. The algorithm tries to minimize the total risk involved in covering the target area while taking into account coverage time constrains. The algorithm is based on incrementally extending the coverage path to the nearest safe cells while allowing the robot to repeat its steps. By allowing the robot to visit each cell in the target area more than once, the accumulated risk can be reduced at the expense of extending the coverage time. We show the effectiveness of this algorithm in extensive experiments.

Safest path adversarial coverage

Coverage is a fundamental problem in robotics, where one or more robots are required to visit each point in a target area at least once. While most previous work concentrated on finding a solution that completes the coverage as quickly as possible, in this paper we consider a new version of the problem: adversarial coverage. Here, the robot operates in an environment that contains threats that might stop the robot. We introduce the problem of finding the safest adversarial coverage path, and present different optimization criteria for the evaluation of these paths. We show that finding an optimal solution to the safest coverage problem is NP-Complete. We therefore suggest two heuristic algorithms: STAC, a spanning-tree based coverage algorithm, and GSAC, which follows a greedy approach. These algorithms produce close to optimal solutions in polynomial time. We establish theoretical bounds on the total risk involved in the coverage paths created by these algorithms and on their lengths. Lastly, we compare the effectiveness of these two algorithms in various types of environments and settings.

Robotic adversarial coverage of known environments

Coverage is a fundamental problem in robotics, where one or more robots are required to visit each point in a target area at least once. Most previous work has concentrated on finding a coverage path that would minimize the coverage time. In this paper, we consider a new and more general version of the problem: adversarial coverage. Here, the robot operates in an environment that contains threats that might stop the robot. The objective is to cover the target area as quickly as possible, while minimizing the probability that the robot will be stopped before completing the coverage. This version of the problem has many real-world applications, from performing coverage missions in hazardous fields such as nuclear power plants, to surveillance of enemy forces in the battlefield and field demining. In this paper, we discuss the offline version of adversarial coverage, in which a map of the threats is given to the robot in advance. First, we formally define the adversarial coverage problem and present different optimization criteria used to evaluate coverage algorithms in adversarial environments. We show that finding an optimal solution to the adversarial coverage problem is N P -hard. We therefore suggest two heuristic algorithms: STAC, a spanning-tree-based coverage algorithm, and GAC, which follows a greedy approach. We establish theoretical bounds on the total risk involved in the coverage paths created by these algorithms and on their lengths. Lastly, we compare the effectiveness of these two algorithms in various environments and settings.

Online robotic adversarial coverage

In the robotic coverage problem, a robot is required to visit every point of a given area using the shortest possible path. In a recently introduced version of the problem, adversarial coverage, the covering robot operates in an environment that contains threats that might stop it. Previous studies of this problem dealt with finding optimal strategies for the coverage, that minimize both the coverage time and the probability that the robot will be stopped before completing the coverage. However, these studies assumed that a map of the environment, which includes the specific locations of the threats, is given to the robot in advance. In this paper, we deal with the online version of the problem, in which the covering robot has no a-priori knowledge of the environment, and thus has to use real-time sensor measurements in order to detect the threats. We employ a frontier-based coverage strategy that determines the best frontier to be visited by taking into account both the cost of moving to the frontier and the safety of the region that is reachable from it. We also examine the effect of the robots sensing capabilities on the expected coverage percentage. Finally, we compare the performance of the online algorithm to its offline counterparts under various environmental conditions.

Analysis of the effects of lifetime learning on population fitness using vose model

Voses dynamical systems model of the simple genetic algorithm (SGA) is an exact model that uses mathematical operations to capture the dynamical behavior of genetic algorithms. The original model was defined for a simple genetic algorithm. This paper suggests how to extend the model and incorporate two kinds of learning, Darwinian and Lamarckian, into the framework of the Vose model. The extension provides a new theoretical framework to examine the effects of lifetime learning on the fitness of a population. We analyze the asymptotic behavior of different hybrid algorithms on an infinite population vector and compare it to the behavior of the classical genetic algorithm on various population sizes. Our experiments show that Lamarckian-like inheritance - direct transfer of lifetime learning results to offsprings - allows quicker genetic adaptation. However, functions exist where the simple genetic algorithms without learning, as well as Lamarckian evolution, converge to the same local optimum, while genetic search based on Darwinian inheritance converges to the global optimum.

Capturing an area-covering robot

Area coverage is a fundamental task in robotics, where one or more robots are required to visit all points in a target area at least once. In many real-world scenarios, the need arises for protecting one’s territory from being covered by a robot, e.g., when we need to defend a building from being surveyed by an adversarial force. Therefore, this paper discusses the problem of defending a given area from being covered by a robot. In this problem, the defender needs to choose the locations of k stationary guards in the target area, each one having some probability of capturing the robot, in a way that maximizes the probability of stopping the covering robot. We consider two types of covering robots: one that has an a-priori map of the environment, including the locations of the guards; and the other has no prior knowledge of the environment, and thus has to use real-time sensor measurements in order to detect the guards and plan its path according to their discovered locations. We show that in both cases the defender can exploit the target area’s topology, and specifically the vulnerability points in the area (i.e., places that must be visited by the robot more than once), in order to increase its chances of capturing the covering robot. We also show that although in general finding an optimal strategy for a defender with zero-knowledge on the robot’s coverage strategy is

Frontier-Based RTDP: A New Approach to Solving the Robotic Adversarial Coverage Problem

{\mathcal {NP}}

Multi-Robot Adversarial Coverage.

NP-Hard, for certain values of k an optimal strategy can be found in polynomial time. For other cases we suggest heuristics that can significantly outperform the random baseline strategy. We provide both theoretical and empirical evaluation of our suggested algorithms.