Experimental mobile robots.
Mobile robots have become very popular on universities and even in public. There are many robotic competitions with many categories. First presented project is aimed at development and implementation of small mobile robots for competition. Results of this project are two functional robots. These robots are mostly used for education. Second presented project is aimed at development of bigger robots called UNIBOT. These robots were developed to be modular and universal and will be used for validation and testing of algorithms mostly for rescue and reconnaissance robots. Especially multi-robot cooperative area exploration algorithms will be tested. Rescue and reconnaissance robots have become important part of robotics. These robots are used by rescue teams, firemen and army in dangerous and hardly accessible areas. Using these robots can avoid human injuries and rescuers do not need go in to dangerous areas. In addition performance of the rescue team is increased by using these robots and their sensors. Especially robots equipped with thermal camera perform very well exploration of ruins. Nowadays there are many sources of information about and navigation of mobile robots (Siegwart & Nourbakhsh, 2004; Borenstein et al., 1996), robotic hardware, software and sensors (Solc & Zalud, 2002). An extensive amount of research has been carried out in the area of localization and mapping (Thrun, 2002).
2. SMALL ROBOTS
Micromouse is a category where autonomous robots are solving the maze. The goal is to find the fastest route to the centre of the maze within time limit. The maze for the contest contains 16x16 cells. Each cell has 18x18 cm dimension. There is more than one way to the centre. Maximum time for one robot to perform is 10 minutes. The time taken to travel from the start cell to the centre cell is called the "run" time. This time considers penalties for touching the robot and for time spent in the maze. The robot with the shortest run time wins.
Path-Follower is a category where autonomous robots must travel the whole defined path in the shortest time. The path is a black line on white background and it includes obstacles. The obstacles are mostly: a brick, an interruption of the path and a tunnel. This task is a model of the real situation for example in automatic storage manipulators and other. Exact rules of categories are described in (***, 2009). Two small mobile robots have been developed. Construction considers maximum dimensions and has ability to manoeuvre within the maze as well as follow the line. The robots are built on differential chassis.
These robots can compete in both Micromouse and PathFollower category. Sensory subsystem acquires desired data about distances of walls in the maze and in addition data about black path on the ground. Sensory subsystem is made of three analogue distance GP2D120 SHARP sensors and of six reflective CNY70 sensors. Microcontroller Atmel with 8 bit RISC architecture is used as a control unit. Drivers provide power signals for actuators. Actuators are two stepper motors with resolution of 200 steps per round. This resolution allows good speed regulation with no need of using encoders and it also makes possible to use odometry. The robot is powered by 4 cell Li-Pol battery. Communication subsystem provides data from the robot and achieves basic user interface. It includes LCD display and RF communication module. Electronic system diagram of the robot is shown in the Figure 1.
[FIGURE 1 OMITTED]
Control program for Micromouse competition must be able to control robot along whole competition. Program process data and locates the robot in the maze. It is also responsible for mapping, route planning and motion regulation. The robot has no information about walls in the maze at the start of the run. Program explores the maze with the robot and acquired information about walls is stored in the map. The Flood-Fill algorithm is used in our robots. The algorithm uses map, controls exploration of the maze and finds the shortest way to the centre of the maze. The Flood-Fill assigns value to each cell of the maze. The value is the distance from actual cell to the centre of the maze. The robot moves following cell's values in decreasing order towards the centre.
Control program for Path-Following competition must be able to control the robot along the whole defined path in the shortest time and must be able to deal with obstacles on the path. The program should include some searching algorithm for finding lost path.
Robot behaviour during the task must be effective and reliable. The PD regulator algorithm is used for motion regulation in both competition categories.
Most of present rescue and reconnaissance robots are telepresence robots. That means that robot is remotely controlled by operator. Information from robot sensors and video camera are transmitted to the operator.
Performance of rescue and reconnaissance systems can be increased by autonomous and cooperative elements (Birk & Caprin, 2006). Cooperation of robots lowers time needed for mission tasks. Typical example of such task is exploring an area with multi-robot system. Autonomous elements in the system allow the operator to control whole group of robots.
In the present the operator is still irreplaceable. Hence semi-autonomous systems are developed in which robots autonomously perform individual tasks and human operator is responsible for important decisions. Proposal of such system follows. A group of autonomous cooperative robots is incorporated in rescue system. The robots autonomously explore selected area and create map. The map contains important information like positions of obstacles, dangers and positions of victims. Autonomous activity of robots will be interrupted when important event happen for example victim is found. In that case the operator is noticed and he can make a decision about mission's proceeding. The operator can also assign tasks to individual robot or he can control one of the robots telepresently. The robots have collective communication system and share information. Behaviour of the system is based on state and position of each robot and on state of the mission. The basic element of this multi-robot system is a single autonomous robot.
Autonomous robots are robots which can perform desired tasks in unstructured environments without continuous human guidance. Autonomous behaviour is particularly desirable in fields such as space exploration. Autonomous robot gathers information about environment with various types of sensors. Important data are extracted and passed in to environment model and local map. Navigation system localizes robot in the map and updates this map with new information. Cognition system controls behaviour of the robot and plans robot's path according to mission. The path is executed by power drivers connected to motors and manipulators. UNIBOT robots were developed to be testing platform for these autonomous robot algorithms.
Computation system of the robot can be divided in to three layers (see Figure 2). All decision making, path planning and algorithms responsible for robot's behavior are executed in the first layer. This layer is equipped with mATX PC motherboard with Intel Celeron processor and wifi communication with other robots. Hardware parts are common and can be easily updated from the actual market. The second layer contains sensors, power drivers and electronics needed for gathering data from sensors. The last hardware part of the robot is mechanic construction, motors and manipulators.
[FIGURE 2 OMITTED]
UNIBOT is constructed for indoor environment. It is made of aluminium construction profile and aluminium plate. Construction is developed to be reliable simple and cheap. Robot can be easily demounted or modified. Most of joints are detachable. All of mentioned robots are built on differential chassis. It has two active wheels. Each active wheel has one degree of freedom. The third wheel is freely revolving and acts as a supporting point. This solution is not suitable for hard terrain but it can be controlled easily. Velocity and turning of the robot can be controlled by velocity of each active wheel.
The results of small robots project are two functional mobile robots for the competitions. The robots have successfully participated in competitions and are as well used for education. Students are involved in this project developing SW and HW. Performance of rescue and reconnaissance robotic systems can be increased by autonomous and cooperative elements. Project presented in the second part of the paper is aimed at development and implementation of bigger mobile testing robots UNIBOT. The robots will be used for testing autonomous and cooperative algorithms. The results of the project are three prototypes of the testing robots. Next steps are development of the control software for single autonomous robot and then development of the cooperative robotic system and incorporation of individual robots.
This work has been supported in part by Ministry of Education, Youth and Sports of the Czech Republic (Research Intent MSM0021630529 Intelligent systems in automation), Grant Agency of the Czech Republic (102/09/H081 SYNERGY --Mobile Sensoric Systems and Network) and by Brno University of Technology.
Birk, A. & Caprin, S. (2006). Rescue Robotics--a crucial milestone on the road to autonomous systems. Advanced Robotics Journal, Vol. 20, No. 5, p. 11
Borenstein, J.; Everett, B. & Feng, L. (1996). Navigating Mobile Robots: Systems and Techniques, A. K. Peters, Ltd., MA, ISBN 1-56881-058-X, Wellesley
Siegwart, R. & Nourbakhsh, I. R. (2004). Introduction to Autonomous Mobile Robots, A Bradford Book, The MIT Press, ISBN 0-262-19502-X, London, England
Solc, F. & Zalud, L. (2002). Robotics, lecture notes, Brno University of Technology (BUT), Brno
Thrun, S. (2002). Robotic mapping: A survey, In: Exploring Artifical Intelligence in the New Millenium, p. 1-35, Morgan Kaufmann, ISBN: 1-55860-811-7, San Francisco
*** (Istrobot), Slovak robotic website (2009). Available from: www.robotika.sk, Accessed: 2009-07-05
|Printer friendly Cite/link Email Feedback|
|Author:||Vesely, Milos; Sembera, Jaroslav|
|Publication:||Annals of DAAAM & Proceedings|
|Date:||Jan 1, 2009|
|Previous Article:||Economy development with introduction of clusters.|
|Next Article:||New approach in evaluating virtual labs safety into the frame of real-time DAQ based on Wi-Fi TCP/IP protocols.|