Patent
1996-12-17
1998-08-11
MacDonald, Allen R.
G05B 1904
Patent
active
057939349
DESCRIPTION:
BRIEF SUMMARY
BACKGROUND OF THE INVENTION
The most varied fields of use can be conceived for autonomously operating units. They are particularly suitable for use in danger areas and for remote sensing, but they are also capable of highly varied activities in buildings. There, they can carry out activities such as, for example, industrial vacuum cleaners, transport vehicles in the production industry or, not least, as mobile multipurpose robots. In executing these different activities, the autonomous mobile unit is confronted, however, with the problem of having to draw up a map of surroundings which are at first unknown, and of being able to use this map to locate itself at any given instant in its working environment. To solve this problem, such autonomous robots mostly have a control computer and sensors by means of which they interact flexibly with their environment. Examples of such sensors are laser distance scanners, video cameras and ultrasonic sensors.
The robot's operating procedure of orientating itself while traveling and simultaneously building up a map of the unknown surroundings poses the problem that there is a mutual functional relationship between drawing up the map of the surroundings and locating the robot. An important role is played here by, in particular, the type and accuracy of the sensors which the robot uses to survey the path it has covered and to locate obstacles in the surroundings. For example, the path covered from a starting point is determined with the aid of a wheel sensor. On the other hand, the distance from obstacles which occur is measured with the aid of distance sensors, and said obstacles are entered as landmarks in the map of the surroundings. Because of the mutual functional relationship between the measuring procedure for determining the distance of obstacles and the procedure for measuring the path distance covered in conjunction with drawing up the map and with the errors which the measuring sensors have, these errors accumulate as a function of the path distance covered by the robot.
The autonomous mobile unit can therefore no longer be manipulated sensibly from a specific limit.
A method which addresses this problem and indicates a solution for it was advanced by W. D. Rencken in the article "Concurrent Localisation and Map Building for Mobile Robots Using Ultrasonic Sensors", Proc. of the 1993 IEEE/RSJ. International Conference on Intelligent Robots and Systems, Yokohama, Jap. Jul. 26 to Jul. 30, 1993, pages 2192 to 2197. The known measuring errors of the sensors used are used there for the purpose of correcting a predicted landmark position, found with the aid of the internal map, as a function of a path distance covered. The absolute measuring error which occurs during the movement of the autonomous mobile unit is thereby reduced.
A further method for orientating self-propelled mobile units in unknown surroundings consists in that the unit builds up a two-dimensional grid of its surroundings and provides individual cells of this grid with occupancy values. The occupancy values assigned per grid cell represent the occurrence of obstacles in the surroundings.
Such a method is specified by the published document "Histogrammic in-motion mapping for mobile robot obstacle avoidance", IEEE Transactions on Robotics Automation, Vol. 7, No. 4, August 1991, by J. Borenstein and Yoram Koren. It is described there how ultrasonic sensors can be used to draw up a map of the surroundings of a self-propelled mobile unit.
The process of drawing up a map while the robot is possibly continuing to travel and being repositioned is time consuming and requires the control computer to perform computations. This hampers the robot in carrying out an activity it has been assigned.
It is therefore extremely desirable for an autonomous mobile unit not to use too much time for orientation tasks when performing a task defined by the user. However, it is also important in this case that within the contest of the task which has been set it can always maintain a defined measure of accuracy of orientatio
REFERENCES:
patent: 5001635 (1991-03-01), Yasutomi et al.
patent: 5006988 (1991-04-01), Borenstein
patent: 5229941 (1993-07-01), Hattori
patent: 5275568 (1994-01-01), Pfuetze
patent: 5377106 (1994-12-01), Drunk
patent: 5525883 (1996-06-01), Avitzour
D'Orazio, "mobile robot position determination using visual landmarks," IEEE Transactions on industrial electronics, v41 n6 Dec. 1994.
Borenstein, "real time obstacle avoidance for fast mobile robots," IEEE transactions on systems, man and cybernetics v19 n5 Sep./Oct. 1989.
IROS '93, Proceedings of the 1993 IEEE/RSJ International Conference on Intelligent Robots and Systems, Yokohama, Japan, Jul.26-30, 1993, New York, W. D. Rencken, Concurrent Localization and Map Building for Mobile Robots Using Ultrasonic Sensors, pp. 2192-2197.
IEEE Transactions on Robotics and Automation, vol. 7, No. 4, Aug. 1991, Short Papers, J. Borenstein et al, Histogramic-Motion Mapping for Mobile Robot Obstacle Avoidance, pp. 535-539.
Soldat Und Technik (1993), H.W. Scriba, Robotik Fur Das Gefechtsfeld, pp. 368-373.
MacDonald Allen R.
Siemens Aktiengesellschaft
Smith Jeffrey S.
LandOfFree
Method for the orientation, route planning and control of an aut does not yet have a rating. At this time, there are no reviews or comments for this patent.
If you have personal experience with Method for the orientation, route planning and control of an aut, we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Method for the orientation, route planning and control of an aut will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-398706