United States Patent
`
`[19]
`
`[11] Patent Number:
`5,204,814
`[45] Date of Patent: Apr. 20, 1993
`Noonan et a1.
`
`
`
`|||ll|llllllllIlllllll|||||lllllllllllllllllllllllllllllllllllll|||||||||||
`US005204814A
`
`[54] AUTONOMOUS LAWN MOWER
`[75]
`Inventors: Thomas H. Noonan, Havertown;
`John Fisher, Avondale, both of Pa.;
`Barry Bryant, New Baltimore, Mich.
`
`_
`.
`[73] ASSlgnee' Mobot, Inc., Havertown, Pa‘
`[2]] Appl. No.: 611,679
`
`OTHER PUBLICATIONS
`
`“
`,
`gisggwslkjggér ge’sgzuesg‘tge Weekend, Electronic
`’
`p.
`'
`’ p.
`'
`Primary Examiner—Thomas G. Black
`Attorney, Agent, or Firm—Sixbey, Friedman, Leedom &
`Ferguson
`
`[22] Filed:
`
`Nov. 13, 1990
`
`[57]
`
`ABSTRACT
`
`Int. 0.5 ......................... GOéF 15/50; B62D 1/28
`[51]
`
`364/424.02; 180/168
`52
`1
`]
`[58] Field of Search ...................... 364/424.01, 424.02;
`180/167-170
`
`[56]
`
`.
`References Cited
`U.S. PATENT DOCUMENTS
`
`3,095,939 4/1960 Hine .................................... 180/167
`3,482,644 12/1969 Krieger et al.
`...... 180/79.l
`
`3,550,714 10/1964 Bellinger ...........
`.. ISO/79.1
`
`3,563,327 11/1968 Mier ...........
`18/79.1
`
`----- 180/93
`3,734,229
`5/1973 Comer
`
`idler """""361721290111.
`3,302,315): 1:;13;:
`2119900 10/1978 Kiem'nitaw
`"
`318/587
`
`..
`4,137,984 2/1979 Jennings et al.
`..... 180/98
`
`4,306,329 12/1981 Yokoi .......................... 180/167
`7/1985 Mackinnon et a1. ........ 364/424.02
`4,530,056
`4,626,993 12/1986 Okuyama et a1.
`364/424.02
`
`7/1987 PCTdUC .................... 364/424.02
`4,679,152
`4,694,639 9/1987 Chen et a1. ................ 56/102
`
`4
`4,730,690
`3/1988 McNutt et al.
`33%;;
`4,742,283
`5/1988 Bolger et a1.
`180/1 68
`4,800,978
`1/1989 Wasa et a1.
`
`4 847 774 7/1989 Tomikawa et a1.
`i... 364/449
`
`4,990,841
`5/1991 Elder ........................ 318/587
`2/1991 Ohkura ........................... 364/426.01
`4,993,507
`
`Disclosed is an automatic, self propelled lawn mower
`that references its position within a cutting area usin
`g
`electronically stored path and terrain information as a
`primary navigation system, senses a non magnetic, non-
`current carrying metallic
`uide path as a secondary
`g
`navigation system, senses underground metallic refer-
`ences for use as position reference points as a further
`navigation system, incorporates an ultrasonic obstacle
`detection system to stop the vehicle if unexpected ob-
`stacles are encountered. The vehicle further includes a
`cow-catcher front bumper arrangement to divert small
`objects away from the cutting blades, contains bumper
`switches to shut off the vehicle upon contact with large
`objects, contains an inclinometer sensor to halt the vehi-
`cle during unexpected tilting, incorporates a radio com-
`munications link to notify an off-board monitoring sys-
`tem of problems, contains an on-board power genera—
`tion system using an internal combustion engine, gener-
`atof, and a battery arrangement. The lawn mower also
`provides a navigation controller and servo motor posi-
`tioning system that plans path motion, using sensory
`data and stored map data and to turn the drive wheels to
`PTOPe11 the “hide in the deSired manner-
`
`28 Claims, 7 Drawing Sheets
`
`
`METAL
`MAP DATABASE
`
`
`
`DETECTOR
`target detectionts 1 52.53.54)
`
`
`
` Path VECIOFS
`
`ELEMENTS
`
` azmuth correction (i,j)
`
`
`Navigation Controller
`
`
`
`
`COORDINATE
`
`PATH SEGMENT PLANNER
`TRANSFORMATION
`
`
`
`PROCESSOR
`
`
`target postx,y,heading)
`
`
`
`wneel position and velocity
`ULTRASONIC
`
`
`
`
`
`(ei,w1,92,w2)
`
`
`COMPUTER
`
`MOTOR CONTROLLER
`
`20
`
`Ex
`feature data
`
`TILT SENSOR
`
`21
`
`39
`.
`I-—/
`
`;
`.
`2
`
`2
`
`Silver Star Exhibit 1012
`
`Silver Star Exhibit 1012
`
`

`

`' US. Patent
`
`Apr. 20, 1993
`
`Sheet 1 of 7
`
`5,204,814
`
`FIG. 1A
`
`FIG. 1B
`
`_
`
`FIG. 1c
`
`.
`
`
`
`305—43 ‘
`
`Silver Star Exhibit 1012 - 2
`
`Silver Star Exhibit 1012 - 2
`
`

`

`US. Patent
`
`Apr. 20, 1993
`
`Sheet 2 of 7
`
`5,204,814
`
`FIG. 2A
`
`7
`
`FIG. 28
`
`SENSOR NUMBER
`
`DIRECTlON
`OF TRAVEL
`
`HEN—I.
`
`DISTANCE
`
`O. A
`
`
`
`
`
`
`
`
`
`m mz (nO :0 <>r— c: ma)
`
`
`
`Silver Star Exhibit 1012 - 3
`
`Silver Star Exhibit 1012 - 3
`
`

`

`US. Patent
`
`Apr. 20, 1993
`
`‘Sheet 3 of 7
`
`5,204,814
`
`FIG. 4
`
`
`
`
`CONTROLLER
`
`12
`
`III
`
`
`
`0 0 “ 0
`
`'
`
`DUAL SERVO
`
`A:
`
`m 3
`
`.2
`
`X
`
`I
`.
`l
`:
`: A 24
`- I DEE.
`l
`[J
`g
`|—-
`
`CONTROLLER 13
`
`Egg
`D
`
`16
`
`————— LOGIC POWER
`
`—- SIGNAL
`
`—— HIGH POWER
`
`Silver Star Exhibit 1012 - 4
`
`30
`
`ULTRASONIC
`
`(C
`
`’é
`
`INDICATOR
`
`'E'
`
`'I a o
`
`Silver Star Exhibit 1012 - 4
`
`

`

`Hg. 5
`
`MAP DATABASE
`
`Path vectors
`&
`
`feature data
`
`_
`
`azmuth correction (le
`
`METAL
`
`ELEMENTS
`
`DETECTOR
`
`target detection(s I 52.53.54)
`ATION
`
`Navigation Controller
`
`3.9
`
`PATH SEGMENT PLANNER
`
`COORDINATE
`
`TRANSFORM
`
`PROCESSOR
`
`target pos(x,y.heading)
`
`TILT SENSOR
`
`ULTRASONIC
`commeR
`
`wneel position and velocity
`(3"W"°2'W2’
`
`‘
`
`
`
`mama'S'fl
`
`£661‘oz'Jdv
`
`L30V199‘18
`
`1718‘1702‘5
`
`Silver Star Exhibit 1012 - 5
`
`

`

`US. Patent
`
`Apr. 20, 1993
`
`Sheet 5 of 7
`
`5,204,814
`
`
`
`fig. 6
`
`Silver Star Exhibit 1012 - 6
`
`Silver Star Exhibit 1012 - 6
`
`

`

`US. Patent
`
`5,204,814
`
`
`
`7n}. 7
`
`Silver Star Exhibit 1012 - 7
`
`

`

`US. Patent
`
`Apr. 20, 1993
`
`Sheet 7 of 7
`
`5,204,814
`
`
`
`71g. 3
`
`Silver Star Exhibit 1012 - 8
`
`Silver Star Exhibit 1012 - 8
`
`

`

`1
`
`AUTONOMOUS LAWN MOWER
`
`FIELD OF THE INVENTION
`
`5,204,814
`
`to self-
`invention relates in general
`The present
`propelled, fully automatic vehicles, and more specifi-
`cally to a computer controlled, self guiding, autono-
`mous lawn mower with an onboard power generation
`system, sensory systems, computer controlled naviga-
`tion system and safety systems.
`BACKGROUND OF THE INVENTION
`
`Self guided lawn mowers, as well as navigation sys-
`tems in which a vehicle follow a guide path are known
`in the art. U.S. Pat. No. 4,679,152, issued to Perdue
`discloses a navigation system and method for guiding a
`mobile robot employing a plurality of sensors that re-
`lates to guiding a vehicle to a charging station. Such a
`guiding system is used to allow a mobile robot to detect
`and mate with a floating charging unit and does not
`relate to the guidance of a unit about a predetermined
`path in order to cover an entire area.
`U.S. Pat. No. 4,137,984 issued to Jennings et al. and
`U.S. Pat. No. 4,730,690 issued to McNutt et al., each
`disclose bumper mechanisms to prevent forceful colli-
`sions with obstacles but which are not intended or opti-
`mized for lawn mowers. These devices sense the pres-
`ence of objects and guide the vehicle around such ob
`jects. Again, however, such devices do not relate to the
`guidance of a unit about a predetermined path in order
`to cover an entire area.
`‘
`Numerous patents exists describing navigation sys-
`tems where vehicles and transporters follow a wire or
`guide path. For instance, U.S. Pat. No. 4,800,978 issued
`to Wasa discloses a strip of magnetic material which is
`detected by detection circuits. Similarly, U.S. Pat. No.
`4,742,283 issued to Bolger, U.S. Pat. No. 3,482,644 is-
`sued to Krieger et al., U.S. Pat. No. 3,563,327 issued to
`Mier, and U.S. Pat. No. 3,734,229 issued to Comer de-
`scribe additional guidance systems. Each of these sys-
`tems, however, require either a permanent magnet or an
`electricity conducting electromagnet as a guide path,
`and they do not discuss a system for following a passive
`guide path.
`U.S. Pat. Nos. 3,800,902 to Keller, 3,550,714 to Be]-
`linger, and 3,095,939 to Hine each disclose automated
`lawn mowers which use a variety of navigation systems.
`However,
`these systems do not
`include following a
`buried underground metallic path or a navigation sys-
`tem which includes a marker identification system. U.S.
`Pat. No. 4,694,639 issued to Chen et al. discloses a ro-
`botic lawn mower wherein the path or route is stored in
`a memory device but does not consider or disclose a
`system capable of accurately recalibrating the vehicles
`position to actual landmarks along its path of travel.
`Clearly, there is a need for a navigation system for
`guiding an automatic lawn mower which is capable of
`recalibrating the units position along a guide path and
`correct the misalignment of the unit.
`SUMMARY OF THE INVENTION
`
`An autonomous lawn mower (MOBOT TM) in ac-
`cordance with the present invention is a self contained,
`self powered, and self navigating vehicle designed pri-
`marily for lawn mowing although its design is applica-
`ble to other repetitive lawn and garden functions such
`as dispensing fertilizers and the like. The autonomous
`lawn mower, also referred to in this disclosure as lawn
`
`10
`
`15
`
`20
`
`25
`
`30
`
`35
`
`45
`
`50
`
`55
`
`60
`
`65
`
`2
`mower and vehicle, consists of a structural chassis,
`cutting apparatus, gas engine, generator, batteries, elec-
`tric drive motors, navigation sensors, control computer,
`and safety systems. The autonomous lawn mower
`drives along a predetermined cutting path by steering
`itself using a stored map depicting the lawn surface. The
`cutting route is stored in the autonomous lawn mower’s
`computer memory as a collection of path vectors and
`arcs. In addition, terrain and navigation information is
`also stored in the map database. During cutting opera-
`tions, the autonomous lawn mower attempts to traverse
`a given stored route by controlling its drive motors to
`move the vehicle in either straight lines or arcs that
`describe path segments comprising the cutting route.
`Navigating purely from the stored map information,
`however, is not sufficient to precisely control the vehi-
`cle. Factors such as wheel slippage, mechanical inaccu-
`racies, and control errors, which produce cumulative
`position errors, may eventually cause the vehicle to
`deviate from its programmed route. Consequently, in
`order to maintain correct cutting over the entire travel
`route, the autonomous lawn mower includes methods to
`recalibrate its position with actual reference points in
`the lawn.
`
`Position referencing is accomplished by comparing
`the vehicles perceived position and operating state,
`derived from sensory data, with information it expects
`from its stored map. The autonomous lawn mower
`determines its actual position by identifying under-
`ground metallic references placed at specific locations
`in the lawn. Once a reference is located, the autono-
`mous lawn mower finds the position coordinates of the
`reference from its stored map and recalibrates its own
`position for further navigation. Alternately, the autono-
`mous lawn mower has the ability to follow an under-
`ground metallic guide path. This is used when a very
`precise route must be maintained. Metallic references
`and guide paths are passive meaning that they do not
`conduct electric current or aCt as magnets. Since guide
`paths and references do not require electric power, they
`can be easily installed or removed. The guide paths and
`references are detected with an array of inductive metal
`detector sensors mounted near the front of the vehicle.
`Each element signals the presence of a metallic target
`(i.e. a guide path or reference) directly beneath it. Since
`multiple sensors are spaced across the width of the
`vehicle, the locations of the metallic targets with re-
`spect to the vehicle can be determined.
`The map information is stored in a database within
`the vehicles’ control system. It provides the autono-
`mous lawn mower with knowledge of the approaching
`terrain so it can slow down before negotiating difficult
`cutting areas (i.e. hills, sharp turns, etc.). The autono-
`mous lawn mower monitors differences between the
`information contained in the map database and the
`sensed position of guidepaths and references and auto-
`matically shuts itself off if a correlation can not be
`made, signifying that the autonomous lawn mower is
`lost. This feature prohibits the autonomous lawn mower
`from deviating from its stored path more then a short
`distance.
`The autonomous lawn mower includes two front
`drive wheels and two non powered trailing wheels
`which swivel. Both front wheels are powered by their
`own electric servo gear motor. A computer-synchro-
`nizes the motors to move the vehicle in the desired
`direction. Actuating the drive wheels in the same direc-
`
`Silver Star Exhibit 1012 - 9
`
`Silver Star Exhibit 1012 - 9
`
`

`

`5,204,814
`
`3
`tion but at different speeds causes the vehicle to turn at
`a given radius. Driving the wheels in opposite direction
`pivots the vehicle to quickly change its direction. The
`chassis of the vehicle supports a number of subsystems
`including a gasoline powered motor, generator, batter-
`ies, navigation computer, drive means, and grass cut—
`ters. The gasoline engine turns the cutting blades
`through power transmission means and also turns a
`generator to charge the batteries. Sufficient energy is
`supplied from the generator to keep the batteries fully
`charged during a cutting operation. The cutting mecha-
`nism consists of a reel type cutter mounted in a carriage
`that pivots from the chassis to follow the contour of the
`ground. The reel cutting blades enable a wide cutting
`path in a compact arrangement that requires minimal
`added length to the vehicle. The reel cutters used in the
`autonomous lawn mower offer safety advantages over
`conventional rotary blades since the cutting radius is
`less. This decreases the blade tip velocity and signifi-
`cantly lowers the kinetic energy generated. The reel
`cutter also has multiple cutting blades spaced closely
`together making the arrangement less likely than a ro-
`tary blade arrangement to chop up objects that get
`lodged in the cutting area. The autonomous lawn
`mower is equipped with a smooth housing that facili-
`tates easy cleaning as well as providing protection to
`the internal electronics of the vehicle. A rear panel,
`attached with quick release clips, can be removed to
`provide access to the cutting. arrangement.
`Since the vehicle is designed to operate unattended,
`several means are included to stop the vehicle if it en-
`counters obstructions or people. The autonomous lawn
`mower’s navigation system is designed to sense obsta-
`cles via both non-contact and contact means and to shut
`down the vehicle if objects are sensed. Ultrasonic sen-
`sors are used to identify objects in the vehicle’s immedi-
`ate path and signal the vehicle to stop before contact is
`made. If for some reason the vehicle strikes an object, a
`“cowcatcher” type bumper will try to push it out of the
`way. If the object is heavy, however, the bumper will
`deflect and trip a emergency stop switch to halt the
`vehicle and stop the cutting blades. Contact switches
`are also placed on the underside of the bumper to sense
`objects that escape other detection means before they
`enter the cutting blades. A tilt sensor is used to indicate
`if the vehicle is on too steep of a grade or if it is tipping.
`Other safety features include a flashing light and exter-
`‘ nal emergency stop switches. The Emergency stop
`switches are placed on both sides the vehicle in easily
`accessible locations to allow individuals to stop the
`vehicle. A light and audible warning device is mounted
`on the top of the vehicle to indicate its approach. The
`light
`is particularly useful when the vehicle is used
`during night operation.
`If the vehicle stops while cutting, a built in radio
`system signals the potential problem to a remote radio
`receiver so a person can be dispatched to remedy the
`problem.
`The autonomous lawn mower uses three principle
`navigation systems: 1) navigating from a preestablished
`stored map, 2) following an underground guide path,
`and 3) navigating by sensing actual underground path
`references. The stored map is maintained in the autono-
`mous lawn mower’s computer memory as a collection
`of path segments represented as vectors and arcs that
`describe the cutting route. Also stored therein is terrain
`and navigation information related to each path seg-
`ment. This information includes:
`the grade of lawn
`
`10
`
`15
`
`20
`
`25
`
`3O
`
`35
`
`45
`
`50
`
`55
`
`65
`
`4
`surface, the presence of any underground references,
`the type of reference (if present), and information con-
`trolling the operation of the ultrasonic system. During
`cutting operations the autonomous lawn mower at—
`tempts to follow a cutting route one path segment at a
`time by controlling its drive motors to move the vehicle
`in either a straight line or are that prescribe to a desired
`path segment. By referencing the map database,
`the
`autonomous lawn mower obtains knowledge of the
`approaching terrain so that it can slow down before
`negotiating difficult cutting areas (i.e. hills, sharp turns,
`etc).
`The autonomous lawn mower of the present inven-
`tion also knows when to expect the presence of guide
`paths and underground references. In these cases, the
`autonomous lawn mower senses the guide path or refer-
`ence to accurately determine its coordinate position
`within a cutting area. After the autonomous lawn
`mower recalibrates its position, it can travel a predeter-
`mined safe distance using only path segment informa-
`tion from the map data base before another recalibration
`operation is required.
`The autonomous lawn mower follows a guide path as
`a principle navigation system for precise operation. The
`guide path is buried at a shallow depth beneath the
`surface of the lawn along the desired route the vehicle
`will follow. The path is installed by cutting a narrow
`groove along the desired route that the vehicle will
`travel. The groove is not deep or wide enough to seri-
`ously impact the appearance of the lawn. The guide
`path, which will typically be a wire or metal tape is laid
`into the groove. The navigation control system used by
`the autonomous lawn mower guides the vehicle along
`the desired travel route defined by the stored map while
`checking data from the metal sensors to see ensure that
`the vehicle is centered on the guide path. If the guide
`path is sensed to the left or right of the vehicle’s center,
`the appropriate position corrections are made to recen-
`ter the vehicle. A single continuous guide path can be
`used on the entire lawn surface or separate sections of
`guide path may be installed. The stored map provides
`information on when guide path following should be
`invoked.
`
`The autonomous lawn mower senses underground
`metallic references to determine its position at various
`locations along a cutting route. Unlike guide path fol-
`lowing, underground references do not provide con~
`stant position information but are used instead to pro-
`vide intermittant position corrections. Two types of
`underground references are used: uniquely shaped
`markers, and spaced objects. A marker is sufficiently
`large that it can be sensed by several metal detectors at
`once. As the autonomous lawn mower drives over a
`marker, the metal detector elements supply data to the
`autonomous lawn mower’s control computer. Data is
`stored as a two dimensional array. The number of data
`values in each array row is equal to the number of metal
`detector elements and the number of array columns is
`equal to the number of times the sensor elements are
`sensed. The two dimensional data array is filtered to
`eliminate bad sensor values by averaging sensor values
`in adjacent columns. After filtering, the data represents
`a low resolution image of the buried target. The shape
`of the marker is compared with stored shapes in the
`autonomous lawn mower’s map data base to locate a
`match. Once a specific marker is detected, the autono-
`mous lawn mower resets its estimated X and Y position
`to the Cartesian coordinates assigned to the marker in
`
`Silver Star Exhibit 1012 - 10
`
`Silver Star Exhibit 1012 - 10
`
`

`

`5,204,814
`
`5
`the map database. The presence of a marker can also be
`used to signal the vehicle to perform a given function
`such as disabling the cutting blades.
`Metallic objects spaced at preset incrementslare also
`used as position references. Objects may be metal disks,
`plates, or other consistant shaped reference. Each time
`the vehicle encounters an object it stores its current
`coordinates as well as the distance the object is sensed
`from the centerline of the vehicle. Distances between
`objects reflect unique values that can be correlated to
`the distances describing the layout and position of ob-
`jects on the stored map. This procedure allows the
`vehicles internal position to be recalibrated on the fly.
`Like guide path following, object identification can also
`supply data to the vehicle in order to center it on the
`cutting path. If, for instance, a metal detector element
`located near the outside of the vehicle senses an object,
`the direction and distance the vehicle has deviated from
`the center of the cutting path can be determined.
`Guide paths, markers, and spaced objects are sensed
`using plural metal detector elements. Each sensor ele-
`ment operates in a similar manner to conventional metal
`detectors in that they contain an excitation and a detec-
`tion coil. Both coils are excited with turned oscillators
`and coupled to produce a hetrodyne frequency caused
`by the two coils resonating at slightly different frequen-
`cies. The presence of the metallic guide path or a metal—
`lic target under a sensor alters the hetrod yne frequency.
`This frequency change is monitored and interpreted as
`the presence of the metallic guide path or target under-
`neath a sensor element. The sensing elements each mon—
`itor a small area. The spacing of the sensors is optimized
`for sensing metallic guide paths and references. In guide
`path following, the metallic guide path is always sensed
`by at least one sensor element so that as a signal disap-
`pears from one sensor, it will be picked up by an adja-
`cent sensor as long as the guide path is within the sens-
`ing area of the metal detector array. These, as well as
`_ additional advantages of the present
`invention, will
`become apparent from the following detailed descrip—
`tion taken along with the several figures.
`
`BRIEF DESCRIPTION OF THE DRAWINGS
`
`10
`
`IS
`
`20
`
`25
`
`30
`
`35
`
`4s
`
`lA—lC represent configurations of metallic
`FIGS.
`guide paths describing entire cutting routes or path
`segments contained within a cutting route.
`FIGS. 2A and 2B represent, by way of an illustrated
`example and table, the autonomous lawn mower identi-
`fying a uniquely shaped metallic marker as a means of 50
`underground metallic referencing.
`FIG. 3 represents, by way of an illustrated example,
`the autonomous lawn mower sensing objects at preset
`distances as another form of underground metallic ref-
`erencing disk detection.
`FIG. 4 represents a functional block diagram of the
`electronic subsystems used by the autonomous lawn
`mower as well as the direction of data flow.
`FIG. 5 represents a simplified block diagram of the
`Navigation controller showing the major functional
`modules.
`
`55
`
`60
`
`FIG. 6 represents a rear perspective view of the au-
`tonomous lawn mower depicting the basic carriage
`design and position of the cutting blades.
`FIG. 7 represents a front perspective view of the
`autonomous lawn mower with the front panel removed
`to reveal the basic positioning of the major drive com-
`ponents and the metal detector sensors.
`
`65
`
`6
`FIG. 8 represents a front perspective View of the
`autonomous lawn mower showing the housings, front
`bumper and location of ultrasonic sensors.
`DETAILED DESCRIPTION OF THE
`PREFERRED EMBODIMENT
`
`In FIGS. lA-lC, three different layouts of guide
`paths are illustrated. Paths l and 2 provide a means for
`the autonomous lawn mower to cover a completely
`bounded area. Path 1 shows parallel rows followed by
`180 degree turns at the end of each row. Path 2 is a
`concentric guide path layout where the vehicle spirals
`towards a center point or in an alternate configuration
`away from a center point. Path 3 represents a short path
`section installed to enable the vehicle to navigate in a
`narrow area between obstacles 4. Path sections, as de-
`picted in path 3, may be placed intermittently within a
`cutting area.
`FIGS. 2A and 2B depicts the basic concept of navi-
`gating by identifying uniquely shaped markers as posi-
`tion references. A marker consists of metallic plates cut
`into unique shapes. The preferred method of identifying
`a marker is to first allow the vehicle 6 to center itself
`along a section of guide path 7 prior to sensing the
`marker 5, then, with the vehicle centered, drive over
`the marker 5 collecting data from the metal detector
`elements.
`
`The complexity of the target that can be identified
`depends on the amount of data collected in the 2-D
`array which is dictated by the following relationship:
`SV = N L/d where
`
`SV is the number of sensor values contained in a
`two-dimensional array describing the shape of the tar—
`get;
`N is the number of inductive sensors in an array;
`(:1 is the distance the vehicle travels between sensor
`array updates;
`L ‘is the path length the vehicle travels while the
`sensor elements within the metal detector array are
`being updated.
`'
`FIG. 2B contains a table showing the related sensor
`data from metal detector elements 8. Typically, data
`will be sampled several times before inserting a value
`into an array value in an effort to filter out erroneous
`values. The autonomous lawn mower’s navigation con-
`troller uses the marker data to correlate the shape of
`target to a mathematical description of a similar marker
`stored in the map database. If a target is recognized
`from the map database the vehicles position can be
`recalibrated. Markers can also be used for other func-
`tions in addition to position referencing. A given shape
`marker can be a signal for the vehicle to perform a
`given function.
`In FIG. 3, the autonomous lawn mower 9 detects
`underground objects 11 spaced at preset distances d1
`and d2. The distance values can be correlated to stored
`map information depicting arrangements of stored ob-
`jects to determine if a correlation exists. If so, the vehi-
`. cle position can be recalibrated to the stored map. Ob-
`jects spaced sequentially along a path in an established
`pattern can be decoded to signal a given operation as
`well as a position reference. The metal detector array 10
`can sense the objects if the vehicle has deviated from
`the center of the path where the objects are installed. As
`objects are sensed by the outside metal detector ele-
`ments 10, the autonomous lawn mower determines the
`
`Silver Star Exhibit: 1012 - 11
`
`Silver Star Exhibit 1012 - 11
`
`

`

`7
`appropriate correction to recenter itself on the cutting
`path.
`FIG. 4 is a block diagram of the the autonomous lawn
`mower‘s major electronic subsystems. The autonomous
`lawn mower’s brain consists of a microcontroller 12
`containing non-volatile memory, random access mem-
`ory, a microprocessor, and input/output data communi-
`cations ports. The microcontroller controls the dual
`axis servo motor controller 13, which in turn controls
`the two drive amplifiers 14 that power the two drive
`motors 15. Wheel rotational sensor 16, provides posi-
`tion and angular velocity information to the dual axis
`servo controller 13 and the microcontroller 12. The
`
`dual axis servo motor controller uses a current limiting
`technique to maintain the vehicle’s speed within a safe
`operating range. A engine RPM indicator 17 is moni-
`tored by the microcontroller to sense if the engine has
`bogged down or stalled. If the engine begins to slow
`down as a result of thick grass,
`the microprocessor
`automatically slows down the speed of the lawn mower
`until the engine begins to respond normally again. Ob-
`stacles in the path of the vehicle are sensed with ultra-
`sonic sensors 19 and an ultrasonic computer 20. The
`ultrasonic computer consists of a dedicated micro‘
`processor to collect and analyze data from the ultra»
`sonic sensors 19. The ultrasonic computer informs the
`microcontroller with values representing the distance
`to the nearest object from the center of the vehicle. A
`tilt sensor 2] updates the microcontroller with the angle
`of the vehicle and senses if the vehicle is in danger of
`tipping over. Several contact switches 22 are mounted
`to the surface and underside of the front bumper 54 to
`sense impact with obstacles. The bumper contact
`switches 22, mounted underneath the bumper, are used
`to sense if the vehicle is trying to climb over an obstacle.
`The metal detector elements 23 sense the position of the
`underground metallic references and the metallic guide
`path. A spark enable relay 18 is opened by the mi-
`crocontroller 12 to halt the engine and attached cutting
`blades if the vehicle encounters a blocked path,
`the
`vehicle is lost, if a bumper switch 22 is opened, an emer-
`gency stop switch 30 is depressed, the vehicle tilt sensor
`21 indicates an unsafe operating state, or some other
`system failure. The emergency stop switches are di-
`rectly wired to the spark enable relay 18. A starter 26,
`powered from the batteries 29, is used to start the engine
`25. The engine, once running, drives via a power trans-
`mitting means, the cutting blades 44 as well as a genera-
`tor 27 that recharges the batteries 29. Drive power for
`the servo amplifiers 13 is taken directly from the batter-
`ies 29. Logic power for all sensors and controllers is
`generated from a DC/DC :converter 28 connected to
`the batteries 29.
`To guide the autonomous lawn mower to its starting
`position on a lawn, and to drive the vehicle through a
`cutting route during a learning operation, an operator
`pendant 24 is used. The operator pendant includes a
`joystick or similar means to control the speed and direc-
`tion of the vehicle and function buttons to enter data
`
`into the microprocessor 12. The manual control pen—
`dant can be attached directly to the servo amplifier to
`power the vehicle even if the microcontroller is inoper-
`ative. A radio transmitter 31, is used to communicate
`with a self contained, off-board receiver 32. The re-
`ceiver contains a panel to display information corre—
`sponding to the operating status of the vehicle and the
`nature of a problem if it occurs. A safety light and audi-
`ble warning indicator 36 are activated by the micro-
`
`IO
`
`15
`
`20
`
`25
`
`30
`
`35
`
`45
`
`50
`
`55
`
`65
`
`5,204,814
`
`8
`processor to warn individuals of the vehicles approach.
`A diagnostic communications port 37 enables a data
`input terminal or other computer to communicate data
`with the microprocessor 12.
`Referring to FIG. 5 the autonomous lawn mower
`guides itself along the cutting route using a micro-
`processor-based navigation controller 39 consisting of a
`map database 33, a path segment planner 34, and a coor-
`dinate transformation processor 35. The navigation
`controller is implemented with the microcontroller 12.
`The map data base is stored on-board the vehicle in
`non-volatile electronic memory such as EPROMs, EE—
`PROMs, Static Ram, or similar means. The functional
`modules of the navigation controller (i.e. the map data-
`base 33, path segment planner 34, and coordinate trans-
`formation processor 39) may be implemented as soft-
`ware routines. Altemately,
`the functional modules
`could be implemented using specialized electronic hard-
`ware. Sensory data from the metal detector elements 23,
`tilt sensor 21, and the ultrasonic computer 20 serves as
`inputs to the navigation controller 39. Drive wheel
`rotational position information from the wheel position
`encoders 16, monitored by the dual servo motor con-
`troller 12, also serve as inputs to the microcontroller.
`Outputs from the navigation controller 39 drive the
`dual motor controller 13 to command the autonomous
`lawn mower’s drive motors 15 to a new target position.
`The map database 33 contains a list of path vectors,
`arcs, and terrain feature data which describe the cutting
`path for a given lawn. The information from the map
`database 33 is loaded into the autonomous lawn mow-
`er’s memory prior to the beginning of a cutting opera-
`tion. Alternate map databases can be accessed by the
`autonomous lawn mower so that it can out several dif—
`ferent lawns.
`
`The map database 33 is created by the autonomous
`lawn mower after a cutting route is planned and under-
`ground metallic references are installed. The autono-
`mous lawn mower is placed into a learning operation by
`instructing the microcontroller 12 from the operator
`pendant 24. During the learning mode, the autonomous
`lawn mower is manually driven through the desired
`cutting path. If underground guide paths are used, the
`autonomous lawn mower automatically follows the
`guide path without external manual control. If under
`ground metallic references are used (i.e. uniquely
`shaped markers 5 or objects 11), the autonomous lawn
`mower recognizes them and automatically stores their
`position. The autonomous lawn mower remembers all
`turns that it makes as well as the distances between the
`turns and creates a map database. The navigation con»
`troller 39 accomplishes this by transforming the drive
`wheel rotational position data from wheel position en-
`coders 16 into the vehicles’ current Cartesian coordi-
`nate position using the coordinate transformation pro-
`cessor 35. The data generated from the coordinate
`transformation processor 35 consists of X and Y coordi-
`nates as well as a heading. The coordinate transforma-
`tion processor 35 is capable of reducing the amount of
`Cartesian data by converti