1. Description
2. References
3. Manual
4. Play
5. New mobot 3D version


polski english


Description

        For years guarding and investigating real environment has been a domain of human activity. However, nowadays more and more often mobile robots are used to this aim. In case of static environment, planning a minimal robot's tour to see all what is needed to see is well known problem. The art gallery problem and the watchman problem concern this case and they have already found numerical solutions [1, 2]. However, if the environment changes rapidly then a mobile robot is, at least partly, driven by an operator. Human factor is then crucial as far as quality of detecting changes is concerned. Therefore,
to achieve the highest investigating quality robot operators can offer, they should be appropriately trained and ranked. In [5] it is showed how playing
a specific game running on a computer in virtual reality can help in this case.

        The game is similar to a flight simulator. Virtual robot tours a designed scene driven by the player (robot operator). The scene describes
an environment where objects are placed. The player should detect all scene changes as fast and as precise as possible [3, 4]. The detecting quality
is calculated and rated.

        The operator is equipped with a map of initial scene state. All the objects are marked on the map and described. The player does not know how often the changes occur although he knows that the objects may change position or color, appear or disappear. The player knows punishment function.
It is based on the number of detected changes, the number of mistakes and the number of commands send to the virtual robot. He knows that his goal
is to minimize this function.

Research team (in alphabetical order):
  • mgr in¿. M. Bedla
  • mgr in¿. A. Chrobot
  • mgr in¿. J. Jantura
  • mgr in¿. K. Kurczyna
  • mgr in¿. B. £ukawska
  • dr in¿. G. £ukawski
  • mgr in¿. P. Paduch
  • prof. dr hab. in¿. K. Sapiecha - team leader


References


[1]

Thierry Dagnino, "The Art Gallery Problem", http://cgm.cs.mcgill.ca/~godfried/teaching/cg-projects/97/Thierry/thierry507webprj/artgallery.html


[2]

Jorge Urrutia, "Watchman’s problem", http://www.site.uottawa.ca/~jorge/openprob/Watchman/


[3]

Krzysztof Sapiecha, Barbara £ukawska, Pawe³ Paduch: "Experimental Data Driven Robot for Pattern Classification",
Annales Universitatis Mariae Curie-Sk³odowska, Informatica, 2005,


[4]

Krzysztof Sapiecha, Barbara £ukawska, Pawe³ Paduch: "System Wspomagaj±cy Profilowanie Operatorów Mobilnego Robota",
Zeszyty naukowe Politechniki ¦wiêtokrzyskiej, nr 43, 2005 (in polish)


[5]

Krzysztof Sapiecha, Barbara £ukawska, Pawe³ Paduch, "An application of virtual reality for training and ranking operators of mobile robot",


[6]

Sándor P. Fekete, Rolf Klein, and Andreas Nüchter, "Online Searching with an Autonomous Robot", http://arxiv.org/PS_cache/cs/pdf/0404/0404036.pdf


[7]

Nii Saka Addo, Suor Kim, and Eli M Silk, "Navigation Using Behavior-based and Path-planning Strategies", http://palantir.swarthmore.edu/maxwell/classes/e28/S00/reports/addo-kim-silk-lab2/


[8]

Volkan Isler, "Theoretical Robot Exploration", http://www.cis.upenn.edu/~isleri/research/wpe/wpe.pdf


[9]

"The Polygon Exploration Problem I: A Competitive Strategy", http://www.pi6.fernuni-hagen.de/publ/tr241.pdf


[10]

"The Polygon Exploration Problem II: The Angle Hull", http://www.pi6.fernuni-hagen.de/publ/tr245.pdf


Manual

        Before the game starts every player have to register entering a login and a password (picture below). When the player connects the game for the first time it is required to mark the box " Create new player".


        If the logging is successful, the name of the next board will be presented to the player. The game starts after pressing the button "Start" (picture below).



        When the game starts time begin to be counted. The player moving in the given path (red line on the picture below) has to find every change of the environment before a deadline. Time, position and frequency of changes are not known to the player.

        The player can move pressing arrows shown on the picture above or arrows on a computer keyboard:
  • "Arrow - Up" - move forward,
  • "Arrow - Down" - move backward,
  • "Arrow - Left" - turn left 45°,
  • "Arrow - Right" - turn right 45°.

        In order to take a picture the player should press "camera button" on the picture above or a space bar. When the player wants to report a change he should point at a position on a map (picture above) and press a left mouse button or press a button "Object changed" (picture above) or press a return key. Then the player should enter position and type of the change (picture below).

        When a button "Show grading scale" is pressed current grading scale is shown (picture below). The grading scale may change during play.

        Initial number of points: 100.