Patent Grant
6850024
The present invention relates to a method and related systems for navigation in an enclosed area. More particularly, the invention relates to method and apparatus which can be used to cause an automated device to move and to perform predetermined tasks within an enclosed area.
The use of automated devices is widespread nowadays, and finds countless applications. For instance, robots perform very precise and delicate tasks in the construction of electronic devices, or in medicine and aviation. Robots are also used in uses which require motion, notably, for automatic warehouses, where goods are retrieved and stored by means of computer-actuated robots. Other applications include, e.g., fetching raw materials in the course of industrial manufacturing, and removing and packaging finished pieces. In everyday's life, attempts have also been made to exploit robots for lawn mowing and for vacuum cleaning.
The major drawback of mobile robots, which the art has so far been unable to overcome, is the fact that their movements are limited to well predefined paths, normally requiring that they move along rails, or that they be provided with expensive navigation signs, positioned within the area in which they move, which operate as “stations” which redefine the exact position of the robot, and from which the program may direct the robot to the next station. These intermediate signs are expensive, take up space, and are inconvenient to use, since they must be very precisely positioned and cannot be easily moved.
Another approach involves providing an area delimited by boundaries recognizable by the robot, and permitting the robot to effect a random walk therein, during which random walk it carries out its tasks. This approach entails severe drawbacks: first of all, when the robot moves within a predefined area by random walk, there is no way to ensure that the whole area will be covered by the tool which must operate thereon. As a result, even though the robot may operate for a long period of time, unworked areas may be left at the end of the operation. Secondly, if the area to be worked is irregular, or if it presents “islands ”, viz. areas which must not be worked, the random walk may lead to imperfect operation around such islands, as well as at those locations where the perimeter is of irregular shape. Thirdly, because the operation of the robot is not programmed to obtain a predetermined coverage, it is necessary to allow the random walk to go on for a long period of time, so as to increase the chances of covering a major portion of the area to be worked. This is not only energy consuming, but also leads to an increased wear of the equipment, and may also be environmentally undesirable due, e.g. to noise or other pollution caused by the operation of the robot. Even if the robot is operated by sun energy, most of the aforesaid problems are not overcome, and additional problems exists, connected with such a mode of operation. For instance, the robot may not work properly in areas of the world where sun radiation is scarce or low, and may be inoperative for substantial parts of the day, e.g., on cloudy weather.
A further approach involves preprogramming the robot with a blueprint of its designated area of operation, such as a floor map of a building in which a robot is to operate. This approach has two major drawbacks:
It is therefore clear that it would be highly desirable to be able to provide means by which automated mechanisms may move and perform their task within a predetermined area, without being hindered by the need for predefined paths and rails, or by intermediate navigation signs or preprogramming, and which may carry out their task in a predetermined manner, without relying on random occurrences and/or on unstable energy sources.
It has now been found, and this is an object of the present invention, that it is possible to free automated mechanisms operating within an enclosed zone from the need for preprogramming or predefined paths and rails, and from the need for intermediate navigation aids, and this to overcome the drawbacks of the prior art and to provide means by which a robot may perform its tasks within an enclosed area in a manner free from such limitations, with high precision and in a minimal period of time.
It is an object of the present invention to provide a navigation method which fulfills the aforementioned goals.
It is another object of the invention to provide means which can be used in systems utilizing the method of the invention.
Other objects of the invention will become apparent as the description proceeds
The method for automatically operating a robot within an enclosed area, according to the invention, comprises the steps of:
By “robot” it is meant to indicate any autonomously operating device, which may carry out pre-programmed tasks with one or more tools, while moving in the process from one location to another
According to a preferred embodiment of the invention, the location means comprise movement measuring means, such as an odometer or the like device, to measure the distance traveled by the robot, e.g., by measuring the number of revolutions of a wheel. As stated, direction finding means are also provided, so as to provide information on the direction in which the robot travels at any given time, which is needed in order to determine the coordinates of the robot on the map. The direction finding means can be of any suitable type, e.g., may comprise a compass.
While, as stated, it is an object of the invention to utilize relatively inexpensive devices for the operation of the robot, it is of course possible to employ more expensive and sophisticated equipment, without exceeding the scope of the invention. Thus, for instance, it is possible to employ range-finding means, such as a laser range-finder or RF range finders, to determine the distance of the robot from one or more given locations, at any given time, instead of, or in addition to, using an odometer or the like device to measure the distance traveled. However, any such modifications will be apparent to the skilled person, and therefore are not discussed herein in detail.
According to a preferred embodiment of the invention, the boundary which is detectable by a proximity sensor comprises a metallic wire through which electric current flows, and the proximity sensor comprises a magnetic field detector. According to another preferred embodiment of the invention, the boundary which is detectable by a proximity sensor comprises passive metallic means which is excitable by a magnetic field, and the proximity sensor comprises an electric field detector. In still another preferred embodiment of the invention the boundary which is detectable by a proximity sensor comprises passive magnetic means, and the proximity sensor comprises a magnetic field detector. Of course, the boundary may be marked by continuous or by discontinuous marking means, or by combinations thereof.
In still another alternative embodiment of the invention, the boundary which is detectable by a proximity sensor comprises a guide wire through which an acoustic signal passes, and the proximity sensor comprises an acoustic detector.
A further improvement in the precision of the determination of the actual coordinates of the robot on the map, at any given time, can be obtained by further providing on the boundaries a plurality of individually recognizable markers. Thus, when the robot reaches the boundaries, it not only identifies them by the proximity sensor, but may also receive the exact coordinates on the boundaries assigned to the specific marker it has detected. According to a preferred embodiment of the invention, when provided, the markers are substantially located at even distances from one another. Suitable markers will be easily recognized by the skilled person, and may comprise, e.g., an RF tag or magnetic tag.
As stated, according to another preferred embodiment of the invention, the distance-measuring means comprise an odometer or the like device, coupled to the wheels of the robot.
As stated, the robot, when initialized, moves along the boundaries and memorizes their shape. Such memorization may be carried out in a number of ways. For instance, the shape can be memorized by taking continuous or discontinuous readings of the compass and the odometer, and any such readings are then continuously integrated, to give the full coordinates of the boundaries.
The method of the invention can be exploited in a variety of uses, and is not limited to any particular field of application. One particularly interesting use, however, to which reference will be made also hereinafter for the purpose of exemplification, is when the robot is coupled to a lawn mower. Such robot permits to mow the lawn in the absence of the owner, and at any suitable time, or to vacuum clean any predetermined premises.
Of course, safety means should preferably be provided to ensure safe operation of the robot. for instance, automatic shut-off of the robot should be provided, coupled to logic circuitry, to ensure that the operation of the robot is discontinued if one of a number of contemplated possibilities takes place. for instance, if the measured distance traveled without encountering a boundary exceeds by a threshold value the maximal linear distance within the bounded area, as calculated from the map of the boundaries, this may mean that the robot has exited the boundaries due, for instance, to a malfunctioning of the system due to which the proximity sensor has failed to identify the boundary. Other required safety means will be easily recognized by the skilled person, according to the type of robot and the intended use thereof.
The invention is further directed to an automated robot for operation within an enclosed area, comprising:
The term “proximity sensor”, as used herein, indicates any device which is capable of detecting that the boundary of the working area is near. This may include, e.g. magnetic field detectors, acoustic signal detectors, bar code readers, resonance tag meters, transceivers, etc.
The invention also encompasses a system for automatically operating a robot within an enclosed area, comprising:
In accordance with a further embodiment of the invention, a method for automatically cutting a lawn is provided. The method includes the steps of:
The maneuver can be an S-shaped maneuver and the grass height sensor can include the following elements:
The present invention will be understood and appreciated more fully from the following detailed description taken in conjunction with the drawings in which:
The present invention will be better understood through the following illustrative and non-limitative description of preferred embodiments.
Looking now at
As stated, according to one particular embodiment of the invention, the boundaries 1, 2 and 3 may comprise a conducting wire. This type of boundary is shown in cross-section in
Taking the lawn mower as an example, but it being understood that the invention is in no way limited to its use with a lawn mower, or with any other particular device, the invention operates as follows. A coordinates system is defined, as well as a starting point.
When it is desired to teach the robot the boundaries of its task, the lawn mower is caused first to move around the boundary 1 of FIG. 1. The memory means of the robot memorize the coordinates of the boundary 1, relative to starting point S. Throughout this teaching movement, the boundary sensor positioned on the robot (not shown) senses the boundary 1. Similarly, the boundaries 2 and 3 are sensed for the first time by the robot, and memorized for future use. The robot now has an initial map of the area, similar to what is shown in
When it is desired to mow the lawn, the robot is brought to starting point S, and it is started according to a set of instructions which has been pre-programmed, and which may be different for each different task. For instance, a circular lawn may be better looking if mowed in circles, while a soccer field requires back-and-forth mowing. An automated lawn mower according to the invention may further be provided with a number of pre-set programs, from which the user can choose.
The robot, as said, is further provided with distance-measuring means, such as an odometer or the like device. However, these devices are not fully accurate, and may provide only approximate distance values for any given position. The error in the measurement of the distance may derive from a variety of reasons, e.g., the slipping of wheels on a moist lawn, uneven ground, etc., and the error may build up to quite a substantial extent, impairing the ability of the robot to complete its task with a high degree of precision. While, of course, precise measuring means exist, such as laser distance measurements, these are expensive and/or require calibration targets located in or around the working area. It is a purpose of the invention to avoid the use of such expensive and complicated distance-measuring means.
According to the invention, therefore, the robot starting a task continuously compares the distance measured by the odometer or other distance measuring device, with the distance from an earlier position to the boundaries in the angular coordinate it is following. If the boundary is detected earlier than anticipated according to this comparison (or, in other words, if the difference between the distance according to the map and the measured distance is negative), the robot continues to move until the boundary is detected. If the difference between the distance according to the map and the measured distance is positive, or in other words, if the boundaries are encountered earlier than expected, actual value of the coordinate is corrected to be that of the map.
The starting point will initially be the point “S”, and correction of distance errors will be effected relative to this point. As work proceeds, of course, the starting point may be updated to be another point within the area, e.g., a meeting point with the boundaries, for comparison purposes with the map of the area
Similarly, the robot has been pre-programmed to avoid “islands”, but will detect an island according to the actual position of the boundary detected, and will correct its present working map based on the detection of the boundary and the original map. As will be understood by the skilled person, the larger the number of bounded areas, the higher the precision of the correction of the actual working map., Therefore, the islands actually help in keeping precision and correcting the actual working map. therefore, if the working area is particularly large, it may be desirable to provide artificial islands for the purposes of map correction.
As will be appreciated by the skilled person, operating according to the preferred embodiment of the invention described above is very convenient also in respect of the boundaries, since the wire or coil may be embedded in the soil, thus avoiding any actual or even aesthetic disturbance to the working area, and the power requirements to generate a localized magnetic field are very small.
One particular process, employing Cartesian coordinates, will be described hereinafter by way of example, with reference to the flow-sheet of FIG. 5.
In
In
During the initiation process the robot performs a complete loop around the edge and memorizes the shape of the boundary as well as the position of each marker (X,Y coordinates of each individual marker). This procedure allows for the correction of both the X and the Y coordinates error, each time an edge is detected, according to the method shown in the flow-sheet of FIG. 5B.
Schematically speaking, the robot will operate according to the flow-sheet of FIG. 6.
Reference is now made to
In the present embodiment, the lawnmower, labeled 20 in
By comparing the height output of the sensors 22, the control system of the lawnmower can determine generally where the edge between cut and uncut grass is. One embodiment of a sensor 22 is illustrated in FIG. 10 and described in detail hereinbelow.
The control system maintains the desired direction until the edge of the lawn is detected, as described hereinabove. At this point, the lawnmower 20 must change direction of movement while keeping the proper percentage of uncut grass under the lawnmower 20. It is noted that the lawnmower can move both forward and backward.
In step 34, the lawnmower 20 moves forward along the edge of the cut grass until the edge of the lawn is sensed once again. This movement is indicated by the short arrows 42 of FIG. 9. Finally, the lawnmower 20 backtracks along the new swath 44. Initially and until reaching the location of the line A, the lawnmower 20 utilizes only the compass information. Once the edge of cut grass is found again (at the location of line A), the control system utilizes both the compass and the sensor output to create the new swath 44. This is indicated at step 36 of FIG. 8.
As discussed with respect to the previous embodiments, the lawnmower 20 has to return to locations of unfinished scanning, such as locations on the opposite side of a flower bed or tree. To do so, the lawnmower 20 utilizes the navigation system to head towards the desired location and, when it is close to the desired location, it additionally senses for the edge between cut and uncut grass.
Reference is now made to
The grass presses against the wing 50, which, since it is not heavy, will rotate. In turn, the wing 50 pushes against the relevant one of extensions 60. Since the other extension 60 is maintained in place by fixed pin 62, the spring 58 is tightened, thereby providing a returning force against the force of the grass.
The rotation of the wing causes the cam 64 and flexible joint 56 to rotate, which rotation is measured by the potentiometer 52. Furthermore, if the wing 50 rotates too far, protrusions 68 of cam 64 will press against a rod 70 connected to microswitch 66 which will indicate maximum travel of wing 50.
Reference is now made to
In one embodiment, shown in
The corresponding sensor 80, for both embodiments, is a gauss meter, such as the model 4048 manufactured by F.W. Bell Inc. of the USA, or any other magnetometer which senses the magnetism in the combined unit. The distances between the boundary markers 82 are defined by the strength of the magnet 92 in such a way that at any point along the marked perimeter, at least two markers are detectable by the sensor on the robot.
In a further embodiment, shown in
For all of the above embodiments, the sensor 80 determines that the lawnmower 10 has reached the boundary when the signal sensor 80 receives is at or above a threshold level which is calculated as the expected reading five to ten inches from the marker or cable.
It will be appreciated that other types of markers and their corresponding detectors are incorporated within the present invention.
All the above description and examples have been provided for the purpose of illustration, and are not intended to limit the invention in any way. Many modifications can be effected in the method and devices of the invention, without departing from its spirit.
| Number | Date | Country | Kind |
|---|---|---|---|
| 113913 | May 1995 | IL | national |
This application is a Continuation Application of U.S. Ser. No. 09/841,511, filed Apr. 23, 2001, now U.S. Patent No. 6,417,641, which is a Continuation Application of U.S. Ser. No. 08/554,691, filed Nov. 7, 1995, now U.S. Patent No. 6,255,793, both applications of which are incorporated by reference in their entirety herein. U.S. Ser. No. 08/554,691 is related to PCT/US95/07045, filed Jun. 6, 1995, and U.S. Ser. No. 08/952,639, a 371 of PCT/US95/07045, filed Jan. 9, 1998, now abandoned.
| Number | Name | Date | Kind |
|---|---|---|---|
| 2751030 | Null | Jun 1956 | A |
| 3381456 | Taylor | May 1968 | A |
| 3425197 | Kita | Feb 1969 | A |
| 3507349 | Comer et al. | Apr 1970 | A |
| 3550714 | Bellinger | Dec 1970 | A |
| 3570227 | Bellinger | Mar 1971 | A |
| 3650097 | Nokes | Mar 1972 | A |
| 3789939 | Geislinger | Feb 1974 | A |
| 3892282 | Wule | Jul 1975 | A |
| 3924389 | Kita | Dec 1975 | A |
| 3925972 | Anderson | Dec 1975 | A |
| 4119900 | Kremnitz | Oct 1978 | A |
| 4133404 | Griffin | Jan 1979 | A |
| 4180964 | Pansire | Jan 1980 | A |
| 4184559 | Rass | Jan 1980 | A |
| 4219092 | Richter | Aug 1980 | A |
| 4318266 | Taube | Mar 1982 | A |
| 4482960 | Pryor | Nov 1984 | A |
| 4500970 | Daemmer | Feb 1985 | A |
| 4545453 | Yoshimura et al. | Oct 1985 | A |
| 4566032 | Hirooka et al. | Jan 1986 | A |
| 4573547 | Yoshimura et al. | Mar 1986 | A |
| 4603753 | Yoshimura et al. | Aug 1986 | A |
| 4613804 | Swenson | Sep 1986 | A |
| 4653002 | Barry | Mar 1987 | A |
| 4658928 | Seo | Apr 1987 | A |
| 4674048 | Okumura | Jun 1987 | A |
| 4694639 | Chen et al. | Sep 1987 | A |
| 4777785 | Rafaels | Oct 1988 | A |
| 4800978 | Wasa et al. | Jan 1989 | A |
| 4809178 | Ninomiya et al. | Feb 1989 | A |
| 4809489 | Johansson | Mar 1989 | A |
| 4813218 | Claesson | Mar 1989 | A |
| 4831813 | Jonas et al. | May 1989 | A |
| 4919224 | Shyu et al. | Apr 1990 | A |
| 4933864 | Evans, Jr. et al. | Jun 1990 | A |
| 4934130 | Johansson et al. | Jun 1990 | A |
| 4962453 | Pong et al. | Oct 1990 | A |
| 4987729 | Paytas | Jan 1991 | A |
| 5001635 | Yasutomi et al. | Mar 1991 | A |
| 5007234 | Shurman et al. | Apr 1991 | A |
| 5051906 | Evans, Jr. et al. | Sep 1991 | A |
| 5068790 | Wellman | Nov 1991 | A |
| 5083256 | Trovate et al. | Jan 1992 | A |
| 5155775 | Brown | Oct 1992 | A |
| 5163273 | Wotjkowski et al. | Nov 1992 | A |
| 5170352 | McTamaney et al. | Dec 1992 | A |
| 5189612 | Lemercier et al. | Feb 1993 | A |
| 5204814 | Noonan et al. | Apr 1993 | A |
| 5213176 | Oruku et al. | May 1993 | A |
| 5216605 | Yardley et al. | Jun 1993 | A |
| 5258911 | Wellman et al. | Nov 1993 | A |
| 5321614 | Ashworth | Jun 1994 | A |
| 5351778 | Shigemi et al. | Oct 1994 | A |
| 5353224 | Lee et al. | Oct 1994 | A |
| 5369347 | Yoo | Nov 1994 | A |
| 5402051 | Fujiwara et al. | Mar 1995 | A |
| 5414625 | Hattori | May 1995 | A |
| 5426584 | Kamimura et al. | Jun 1995 | A |
| 5438817 | Nakamura | Aug 1995 | A |
| 5444965 | Colens | Aug 1995 | A |
| 5507137 | Norris | Apr 1996 | A |
| 5525882 | Asaka et al. | Jun 1996 | A |
| 5528888 | Miyamoto et al. | Jun 1996 | A |
| 5548511 | Bancroft | Aug 1996 | A |
| 5572856 | Ku | Nov 1996 | A |
| 5758298 | Guldner | May 1998 | A |
| 5956250 | Gudat et al. | Sep 1999 | A |
| 6473746 | Zakrzewski | Oct 2002 | B1 |
| Number | Date | Country |
|---|---|---|
| 37 33 939 | Apr 1988 | DE |
| 38 16 622 | Nov 1989 | DE |
| 39 16 730 | Dec 1989 | DE |
| 2 614 428 | Oct 1988 | FR |
| 2 696 569 | Apr 1994 | FR |
| 2 142 447 | Jan 1985 | GB |
| 2 148 490 | May 1985 | GB |
| 2 204 434 | Nov 1988 | GB |
| WO 9109356 | Jun 1991 | WO |
| WO 9209941 | Jun 1992 | WO |
| Number | Date | Country | |
|---|---|---|---|
| 20020140393 A1 | Oct 2002 | US |
| Number | Date | Country | |
|---|---|---|---|
| Parent | 09841511 | Apr 2001 | US |
| Child | 10147661 | US | |
| Parent | 08554691 | Nov 1995 | US |
| Child | 09841511 | US |
