Environmental Support Method for Mobile Robots Using Visual Marks with Memory Storage

According to intelligence of mobile robots, there are many researches trying to extend its application range from factory to home-office environment. This research agrees with supporting a robot by arranging its working environment. We aim at the construction of robot system (Fig.1) that can work efficiently in the home-office environment where its configuration is usually changed. The system should be useful to various tasks. Concretely speaking, this research realizes the following two steps. (1) We propose gVisual Marks with Memory Storage (Fig.2).h The mark consists of a landmark part, to support self-positioning and pose measurement of the object, and a memory part, to inform about what it is, what tasks there are and how to carry out the task. The marks are disposed to the appropriate place in the working environment. (2) We verify how it is effective to realize the task of the robots using gVisual Marks with Memory Storageh in the environment arranged under step (1) by means of an experiment; the robot is assigned to open the sliding door which is existed several meters apart from the robot (Fig.3). The progress of the experiment is shown in Fig.4. Six trials resulted in five successes. In case of success, the total running distance was approximately 10 meters and the average time to finish the task was 8 minutes and 20 seconds. The time is mainly caused by overhead of data transmission between calculation processes. This is not essential problem because the time can be reduced by the improvement of the control architecture. The result shows the effectiveness of the proposed methodology.

Keywords: service robot, environmental support,artificial landmark, mark recognition

  • Ota, J., Yamamoto,M., Ikeda,K., Aiyama,Y., Arai,T.: Environmental Support Method for Mobile Robots Using Visual Marks with Memory Storage, In Proc. the 1999 IEEE Int. Conf. on Robotics and Automation (ICRA '99), to appear 1999.
    Fig. 1 Environment with mark  Fig.2 View of mark

    (a)Mark disposition for self-positioning  (b)Mark disposition for Manipulation  (c)Self-positioning 
    (d) Pose measurement of the object  (e) Reading the memory  (f) Task realization 
    Fig. 3 An example of task realization


    (a) insert the end-effecter in the handle  (b) open the sliding door 
    Fig.4 Experimental results