Robotic Patrolling Systems based on Hidden Markov Model for Moving Visitors

This paper presents a robotic patrolling system for moving visitors. A mobile robot equipped with a sensor for human detection is used in this system. The patrol mission for the robot is to detect as many visitors as possible. The visitors move toward their destinations. Thus the robot is required to monitor what they are doing at their destinations rather than the moving paths. For this challenge, we propose an optimal patrolling strategy considering the destinations. However, since the detectable range of the sensor is limited, the robot can only observe the environment locally. In order to correctly estimate the moving destinations of visitors from the local information, we use a Hidden Markov Model, HMM. The estimated destinations are used as a reward function in the value iteration method. Finally, the patrolling simulations show that the robot is enabled to detect more visitors by monitoring their destinations. Through the results, we discuss the effectiveness of the robotic system for moving visitors.

[1]  Satoshi Hoshino,et al.  Patrolling robot based on Bayesian learning for multiple intruders , 2015, 2015 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS).

[2]  Satoshi Hoshino,et al.  Probabilistic surveillance by mobile robot for unknown intruders , 2015, 2015 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS).

[3]  Satoshi Hoshino,et al.  Robot Vision System for Real-Time Human Detection and Action Recognition , 2018, IAS.

[4]  H. Asama,et al.  A Method to Estimate Destination of a Walking Person with Hidden Markov Model for Safety of Human Friendly Robots , 2008, 2008 IEEE International Workshop on Safety, Security and Rescue Robotics.

[5]  Noa Agmon,et al.  Multi-robot area patrol under frequency constraints , 2007, Proceedings 2007 IEEE International Conference on Robotics and Automation.

[6]  Sarit Kraus,et al.  The impact of adversarial knowledge on adversarial planning in perimeter patrol , 2008, AAMAS.

[7]  Van Nostrand,et al.  Error Bounds for Convolutional Codes and an Asymptotically Optimum Decoding Algorithm , 1967 .

[8]  R Bellman,et al.  On the Theory of Dynamic Programming. , 1952, Proceedings of the National Academy of Sciences of the United States of America.

[9]  Satoshi Hoshino,et al.  Optimal patrolling methodology of mobile robot for unknown visitors , 2016, Adv. Robotics.

[10]  Yoichi Tomioka,et al.  Generation of an Optimum Patrol Course for Mobile Surveillance Camera , 2012, IEEE Transactions on Circuits and Systems for Video Technology.

[11]  Lynne E. Parker,et al.  Behavioral control for multi-robot perimeter patrol: A Finite State Automata approach , 2009, 2009 IEEE International Conference on Robotics and Automation.

[12]  Jr. G. Forney,et al.  The viterbi algorithm , 1973 .

[13]  Sarit Kraus,et al.  Multi-robot perimeter patrol in adversarial settings , 2008, 2008 IEEE International Conference on Robotics and Automation.

[14]  D. Rubin,et al.  Maximum likelihood from incomplete data via the EM - algorithm plus discussions on the paper , 1977 .

[15]  Alex Bateman,et al.  An introduction to hidden Markov models. , 2007, Current protocols in bioinformatics.

[16]  Lawrence R. Rabiner,et al.  A tutorial on hidden Markov models and selected applications in speech recognition , 1989, Proc. IEEE.