This work is focused on the problem of performing multi-robot patrolling for infrastructure security applications in order to protect a known environment at critical facilities. Thus, given a set of robots and a set of points of interest, the patrolling task consists of constantly visiting these points at irregular time intervals for security purposes. Current existing solutions for these types of applications are predictable and inflexible. Moreover, most of the previous work has tackled the patrolling problem with centralized and deterministic solutions and only few efforts have been made to integrate dynamic methods. Therefore, one of the main contributions of this work is the development of new dynamic and decentralized collaborative approaches in order to solve the aforementioned problem by implementing learning models from Game Theory. The model selected in this work that includes belief-based and reinforcement models as special cases is called Experience-Weighted Attraction. The problem has been defined using concepts of Graph Theory to represent the environment in order to work with such Game Theory techniques. Finally, the proposed methods have been evaluated experimentally by using a patrolling simulator. The results obtained have been compared with previous available approaches.
Eye-movement analysis has grown exponentially in recent decades. The reason is that abnormalities in oculomotor movements are usually symptoms of injuries in the nervous system. This paper presents a novel regulated solution named OSCANN. OSCANN aims at providing an innovative tool for the control, management and visualization of oculomotor neurological examinations. This solution utilizes an eye-tracker sensor based on video electro-oculography (VOG) technology to capture eye movements and store them in video files. Such a sensor can store images at a rate of 100 frames per second. A characterization study was performed using twenty-two volunteers (13 male, 9 female, ages 22–45 years, mean 29.3 years, SD = 6.7) to assess the accuracy and precision specifications of OSCANN during oculomotor movement analysis. The accuracy was evaluated based on the offset, whereas precision was estimated with Root Means Square (RMS). Such a study reported values lower than 0.4∘ and 0.03∘ of accuracy and precision, respectively. These results suggest that OSCANN can be considered as a powerful tool to measure oculomotor movement alterations involved in some neurological disease progression.
A great deal of work has been done in recent years on the multi-robot patrolling problem. In such problem a team of robots is engaged to supervise an infrastructure. Commonly, the patrolling tasks are performed with the objective of visiting a set of points of interest. This problem has been solved in the literature by developing deterministic and centralized solutions, which perform better than decentralized and non-deterministic approaches in almost all cases. However, deterministic methods are not suitable for security purpose due to their predictability. This work provides a new decentralized and non-deterministic approach based on the model of Game Theory called Stochastic Fictitious Play (SFP) to perform security tasks at critical facilities. Moreover, a detailed study aims at providing additional insight of this learning model into the multi-robot patrolling context is presented. Finally, the approach developed in this work is analyzed and compared with other methods proposed in the literature by utilizing a patrolling simulator.
As a result of terrorist attacks in the last years, new efforts have raised trying to solve challenges related to security task automation using robotic platforms. In this paper we present the results of a cooperative multi-robot approach for infrastructure security applications at critical facilities. We formulate our problem using a Ms. Pac-Mac like environment. In this implementation, multiple robotic agents define policies with the objective to increase the number of explored states in a grid world. This is through the application of the off-policy learning algorithm from reinforcement learning area, known as Q-learning. We validate experimentally our approach with a group of agents learning a patrol task and we present results obtained in simulated environments.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.
hi@scite.ai
10624 S. Eastern Ave., Ste. A-614
Henderson, NV 89052, USA
Copyright © 2024 scite LLC. All rights reserved.
Made with 💙 for researchers
Part of the Research Solutions Family.