Navigation Behavior Selection using Generalized Stochastic Petri Nets

People^Top

Gunhee Kim
       Woojin Chung

Description^Top

 Appropriate design and control of behaviors of mobile robots are important for their successful autonomous navigation in a real dynamic environment. This work proposes a formal selection framework of multiple navigation behaviors for a service robot. In the presented approach, modeling, analysis, and performance evaluation are carried out based on the Generalized Stochastic Petri Nets (GSPNs). By adopting a probabilistic approach, the proposed framework helps the robot to select the most desirable navigation behavior in run time according to environmental conditions. Moreover, after a mission completion, the robot evaluates its prior navigation performance from accumulated data, and automatically uses the results to improve its future operations. Also, GSPNs have several advantages over direct use of other modeling formalisms such as Finite State Automata (FSA) or Markov Processes (MPs).

We conduct experiments on real guidance tasks with visitors by implementing the framework in the guide robot Jinny at the National Science Museum of Korea. The results show that the proposed strategy is useful for a robot's selection of an appropriate navigation behavior in a dynamic environment.

Fig.1 shows the resultant GSPNs model of two navigation behaviors, AutoMove and Contour tracking. In our model, three tokens are exploited to represent the statuses of the localizer, path planner and behavior, respectively. To perform analytic evaluations of GSPNs designs, we need to obtain an embedded Markov chain (EMC). Fig.1.(c) shows the EMCs induced from the rechability graph of Fig.1.(b), which is derived from the GSPNs model of Fig.1.(a).

   

       (a) A GSPN model                       (b) The reachability graph       (c) The reduced embedded Markov chain
                                                                      Fig.1. A GSPN model of navigation behaviors

The developed framework is tested on the Jinny, the guide robot of the National Science Museum of Korea. Fig.2.(a) shows a target workspace, one of sections popular among visitors in the museum. The mission is to navigate from the start point (8.1, 6.4) to the goal (8.5, 27.5).

Fig.2.(b)-(d) shows the detailed results of the behavior transitions during the guide. The robot initially starts with AutoMove. At point A, the robot turns its motion to Contour tracking when the localization Warning is detected. At this time, many people were around the robot, and the robot was located far from the wall. Fig.2.(b) shows the resultant trajectory. Fig.7.(c) is a typical example of the localizer Success, and Fig.7.(d) shows an instance of the localizer Warning. They contain the information about the local map, laser scan data, sample distributions, and an estimated position of each calculation. As shown in these pictures, the environment is very crowded and dynamic due to visitors.

             (a) An experimental environment (b) The resultant trajectory and behavior changes
                  (c) Localization result during AutoMove (d) Localization result during Contour tracking
                                 Fig.2. Experimental results during one execution of a guide task

Fig.2 shows the resultant GSPNs model of two navigation behaviors, AutoMove and Contour tracking. In our model, three tokens are exploited to represent the statuses of the localizer, path planner and behavior, respectively. To perform analytic evaluations of GSPNs designs, we need to obtain an embedded Markov chain (EMC). Fig.2.(c) shows the EMCs induced from the rechability graph of Fig.2.(b), which is derived from the GSPNs model of Fig.2.(a).

Funding^Top

• Development of a Silver Mate Robot Platform: The Intelligent Robotics Development Program, one of the 21st Century Frontier R&D Programs (Oct. 2003 ~ Jul. 2006)
• Development of Science Museum Guide Robots (Oct. 2003 ~ Feb. 2005).

Copyright notice^Top

 The documents contained in these directories are included by the contributing authors as a means to ensure timely dissemination of scholarly and technical work on a non-commercial basis. Copyright and all rights therein are maintained by the authors or by other copyright holders, notwithstanding that they have offered their works here electronically. It is understood that all persons copying this information will adhere to the terms and constraints invoked by each author's copyright.