`6,018,705
`[11] Patent Number:
`[45] Date of Patent:
`Jan. 25, 2000
`
`Gaudetet al.
`
`US006018705A
`
`[54]
`
`[75]
`
`MEASURING FOOT CONTACT TIME AND
`FOOT LOFT TIME OF A PERSON IN
`LOCOMOTION
`
`Inventors: Paul J. Gaudet, Tewksbury; Thomas
`P. Blackadar, Natick; Steven R.
`Oliver, Attleboro, all of Mass.
`
`Assignee: Personal Electronic Devices, Inc.,
`Wellesley Hills, Mass.
`
`5,323,650
`..
`6/1994 Fullen et al.
`« 73/172
`5,343,445
`8/1994 Cherdak ........
`.. 368/10
`5,357,696
`10/1994 Gray et al.
`wees 36/136
`5,361,778
`11/1994 Seitz .....0...
`we. 128/779
`
`5,422,628
`6/1995 Rodgers wesc ceeees 340/573
`5,437,289
`8/1995 Liveranceet al. ow. eee 128/779
`
`5,452,269
`9/1995 Cherdak 0. ce
`eeceecteeenseneees 368/10
`5,485,402
`we 364/566
`1/1996 Smith etal.
`5,526,290
`6/1996 Kanzaki
`........
`we 364/565
`5,541,860
`7/1996 Takei et al. oe eeceeeee 364/566
`5,583,776
`12/1996 Levi et al. cee eeeeceeeee 364/450
`5,623,944
`4/1997 Nashner........
`we. 128/779
`
`5,636,146
`....
`w.. 364/569
`6/1997 Flentovet al.
`Appl. No.: 08/942,802
`
`5,720,200 2/1998 Anderson et al.occeeeeeeeeeeee 73/172
`
`3/1998 Hutchings oe 364/565
`5,724,265
`Filed:
`Oct. 2, 1997
`
`
`
`
`
`Tint. C17 iceeccecccecsseeesssees GO01C 22/00; GO4F 10/00
`US. C1. cece 702/176; 702/141; 702/160;
`702/142; 368/10; 235/105
`Field of Search o....c.cccccccceeeeee 702/160, 176,
`702/144, 141, 142, 149; 368/10; 235/105
`
`Primary Examiner—MarcS. Hoff
`Assistant Examiner—Hien Vo
`
`Attorney, Agent, or Firm—Wolf, Greenfield & Sacks, P.C.
`
`[57]
`
`ABSTRACT
`
`References Cited
`
`U.S. PATENT DOCUMENTS
`
`The time period that a foot is in contact with the ground
`during a stride taken by a user, and the period that the foot
`is notin contact with the ground betweenstrides taken by the
`user are determined by processing and analyzing the output
`7/1976 Fletcheret al. wee eee eeeee 340/189
`3,972,038
`signal of an accelerometer. The accelerometer is mounted on
`
`8/1976 Sipe we
`.. 340/272
`3,974,491
`the user such that its acceleration sensing axis senses accel-
`10/1983 Wills ...
`.. 340/323
`4,408,183
`eration in a direction substantially parallel to the bottom of
`
`4,409,992 10/1983 Sidorenkoet al.
`« 128/782
`the user’s foot. The output of the accelerometer is high-pass
`4,499,394
`2/1985 Koal oe
`.. 310/330
`
`filtered, amplified, and fed to the input of a micro-controller,
`4,578,769
`3/1986 Frederick
`.. 364/565
`which monitors the signal for positive and negative signal
`
`4,649,552
`3/1987 Yukawa oo.
`eeceecseeecneeeteeees 377/24
`spikes that are indicative, respectively, of the momentthat
`4,651,446
`3/1987 Yukawaetal. ....
`w+ 36/132
`the foot of the user leaves the ground and the momentthat
`
`4,745,564
`5/1988 Tennesetal. .....
`.. 364/566
`4,763,287
`8/1988 Gerhaeuseret al.
`. 364/561
`the foot impacts with the ground. By measuring time inter-
`
`4,771,394
`9/1988 Cavanagh...........
`... 364/561
`vals between these positive and negative spikes, average
`4,774,679
`9/1988 Carlin «0.0...
`.. 364/550
`“foot contact times” and “foot loft times” of the user may be
`4,814,661
`3/1989 Ratzlaff etal.
`.. 310/328
`calculated. To derive the pace of the user, the average foot
`. 128/707
`5/1989 Thornton........
`4,830,021
`
`contact time is multiplied byafirst constant if it is less than
`
`.. 364/561
`8/1989 Bianco....
`4,855,942
`400 milli-seconds (ms) and is multiplied by a second con-
`
`.. 340/323
`9/1990 Furlong......
`.
`4,956,628
`stant if it is greater than 400 ms. This pace value may, in
`7/1991 Kato et abe oo ceeeeeee 364/561
`5,033,013
`turn, be used to calculate the distance traveled by the user.
`5,186,062
`2/1993 ROOSt oes eccserceeceeseteeeeneeeee’ 73/805.4
`
`
`12/1993 Gray wcceccccsecesecsesesceseseees 36/136
`5,269,081
`2/1994 Goldston et al. oo... eeeeeeeeee 36/137
`5,285,586
`
`
`
`38 Claims, 13 Drawing Sheets
`
`0A
`
`APPLE 1012
`
`APPLE 1012
`
`1
`
`
`
`U.S. Patent
`
`Jan. 25, 2000
`
`Sheet 1 of 13
`
`6,018,705
`
`FIG.
`
`1
`
`32
`
`10
`
`:
`
`NETWORK PROCESSING CIRCUITRY
`
`268 AUDIO OR VIBRATIONAL
`
`26A
`
`INDICATOR
`
`20A
`
`22
`
`24
`
`FOOT CONTACT TIME/FOOT LOFT TIME
`GENERATOR
`
`HEART RATE
`MONITOR
`
`RESPIRATORY
`MONITOR
`
`208
`
`FOOT CONTACT TIME/FOOT LOFT TIME
`GENERATOR
`
`2
`
`
`
`U.S. Patent
`
`Jan. 25, 2000
`
`Sheet 2 of 13
`
`6,018,705
`
`
`
`3
`
`
`
`U.S. Patent
`
`Jan. 25, 2000
`
`Sheet 3 of 13
`
`6,018,705
`
`34
`
`[MEMORY
`
`
`
`
`
`
`HIGHPASS|AMPLIFIER LOW POWER
`
`
`ACCELEROMETER FILTER|CIRCUIT MICROCONTROLLER
`
`
`
`
`4
`
`
`
`U.S. Patent
`
`Jan. 25, 2000
`
`Sheet 4 of 13
`
`6,018,705
`
`SUSWILONYO/¥/MCo
`
`118-8>
`
`H3TIOWLNOJONIIW!rmoe
`
`IMdNIQ/¥
`
`
`
`
`
`Q/¥OLLMGNI3ONIU3I3H0Y5Z
`
`BF
`
`4060¢
`
`oP
`
`
` TW4)waG/¥O1INdNISIX¥-X||
`
`
`
`
`|SSWAHOTHZH20°C
`
`0%
`
`ee
`
`02G‘9I4d
`
`ae)
`
`PM|om
`
`Aen
`
`5
`
`
`
`
`
`
`
`U.S. Patent
`
`Jan. 25, 2000
`
`Sheet 5 of 13
`
`6,018,705
`
`
` . ; . 82 . ;
`0g9ag" eH65 Boot Mea Dia! Bea 19
`22et
`2.47
`
`
`
`
`
`
`
`FIG. &
`
`
`TIME (SECONDS)
`
`8 BIT
`
`VALUE
`
`
`
`
`
`
`
`
`1D 169 22192.47195
`
`, Mon" arkhy eo 5 at " 1.17
`1.56
`11.821
`2.081
`2.34
`;
`;
`;
`A
`;
`TIME (SECONDS)
`
`6
`
`
`
`U.S. Patent
`
`Jan. 25, 2000
`
`Sheet 6 of 13
`
`6,018,705
`
` "
`
`!
`'
`0.0 0.2 0.305 0.6 08 0.9 14 12 £4 15 1.7 18 2.0 2.1 23 24
`TIME (SECONDS)
`
`8 BIT
`
`VALUE
`
`0
`
`0.00.2 0.3.0.5 06 08 03 12 22 14 £5 17 139.20 22 23 25
`
`TIME (SECONDS)
`
`7
`
`
`
`U.S. Patent
`
`Jan. 25, 2000
`
`Sheet 7 of 13
`
`6,018,705
`
`
`
`Tten! Bost wa are: 18.70=18.8618.54
`MM 18.46|)18.62118.78|18.94;
`
`
`
`TIME (SECONDS)
`
` 0
`30 0
`
`8 BIT an
`A/D
`VALUE 150
`
`100
`
`1,
`
`sat Bit are a4 Ca ns" ‘ast! Hao 18.2" Bg g7
`1.
`1.
`18.
`a]
`A
`;
`A
`;
`
`TIME (SECONDS)
`
`8
`
`
`
`U.S. Patent
`
`Jan. 25, 2000
`
`Sheet 8 of 13
`
`6,018,705
`
`FIG. 9
`20.7120.8125.9
`
`~ 3.3 a 9 6G ODO
`TIME (SECONDS)
`
`20.9.)
`
`a0 29.4 al 20.2 a4 20.4 a 20.9 a1 20.6 4 23.1 4 20.9
`;
`,
`Jd
`Bot
`TIME (SECONDS)
`
`29.9
`
`26.0
`
`9
`
`
`
`U.S. Patent
`
`Jan. 25, 2000
`
`Sheet 9 of 13
`
`6,018,705
`
`101
`
`I
`FIG. 10
`
`
`
`100~|DETECT NEGATIVE
`
`SPIKE EVENT AND
`
`
`INCREASE STEPCOUNT
`
`102~[PERFORM Ta FIFO
`SMOOTHING
`
`
`
`104~(INCREMENT TaSUN
`AND TaSTEPS
`
`106~(SLEEP FOR MININUN
`
`Tec
`
`108~(DETECT POSITIVE
`
`SPIKE EVENT
`
`
`110~[peRFORM Te FIFO
`
`
`SHOOTHING
`
`
`M12~TTNCREMENT TcSUM
`AND TeSTEPS
`
`SLEEP FOR MINIMUM
`
`114
`
`
`
`C
`
`10
`
`10
`
`
`
`U.S. Patent
`
`Jan. 25, 2000
`
`Sheet 10 of 13
`
`6,018,705
`
`FIG. 11,
`
`CALCULATE
`TcAVERAGE AND
`RESET TcSUM AND
`TeSTEPS
`
`RETURN
`
`CALCULATE
`TaAVERAGE AND
`RESET TaSUM AND
`TaSTEPS
`
`CALCULATE
`STEP FREQUENCY
`AND STEPCOUNT
`
`CALCULATE PACE
`AND DISTANCE
`
`11
`
`
`
`U.S. Patent
`
`Jan. 25, 2000
`
`Sheet 11 of 13
`
`6,018,705
`
`wot
`
`100A
`
`FIG. 12
`
`READ X-AXIS
`ACCELEROMETER
`
`1008<> TaSUM=
`
`YES
`
`TaSUM+TA
`
`STOP Ta TIMER
`
`TaSTEPS++
`
`1048
`
`START Tc TIMER
`
`STEPCOUNT +s
`
`102A
`
`ROTATE 9 Ta VALUES
`FIFO)
`
`
`SLEEP FOR TIME|106 EQUAL TO THE
`
`
`108E
`
`NO
`
`FIRST AND
`LAST Tc VALUE
`WITHIN X% OF
`EACH OTHER?
`
`]
`MIDDLE Tc VALUE
`WITHIN X% OF
`EACH OTHER?
`
`SET MIDDLE Tc
`VALUE TO MIDPOINT
`OF FIRST AND
`LAST Tc VALUES
`
`
`
`USE LEAST RECENT Tc
`FOR FUTURE
`CALCULATIONS
`
`
`
`
`
`
`
`
`
`
`112A
`
`1128
`
`114
`
`TcSUM-
`TcSUM+ Tc
`
`TCSTEPS++
`
`SLEEP FOR TIME
`EQUAL 10 THE
`MINIMUM ALLOWED
`CONTACT TIME
`
`MINIMUM ALLOWED
`CONTACT TIME
`
`READ X-AXIS
`ACCELEROMETER
`
`108A
`
`1088
`
`
`
`
`
`
`POSITIVE G
`SPIKE EVENT?
`
`
`YES
`
`STOP Tc TIMER
`
`START Ta TIMER
`
`108C
`
`1080
`
`
`
`LAST Ta VALUE
`WITHIN X% OF
`EACH OTHER?
`
`
`
`
`
`102C
`
`F
`MIDDLE Ta VALUE
`WITHIN X% OF
`EACH OTHER?
`
`
`SET MIDDLE Ta
`VALUE 10 MIDPOINT
`
`OF FIRST AND
`
`LAST Ta VALUES
`
`
`USE LEAST RECENT Ta
`FOR FUTURE
`CALCULATIONS
`
`102E
`
`12
`
`
`
`U.S. Patent
`
`Jan. 25, 2000
`
`Sheet 12 of 13
`
`6,018,705
`
`FIG. 13
`
`15
`
`116
`
`122A
`
`X SECOND RIC
`INTERRUPT?
`
`YES
`
`118A
`
`TcAVERAGE
`=ZERO?
`
`YES
`
`TcSTEPS>ZERO?
`
`3
`
`x
`
`dAVb+
`
`TcAVG)}
`
`1
`
`188
`
`STEPS-
`
`TeAVERAGE:
`TeSUM/TeSTEPS
`
`18C
`
`124A
`
`TcAVERAGE>
`400nS?
`
`NO
`
`<CremaSSTEPCOUNT«2
`
`
`
`
`
`
`
`SET TcSTEPS=0
`
`SET Tesum-0|1180 SLOPE=24.0
`
`PACE=
`TCAVERAGE«SLOPE
`
`
`DISTANCE =
`(TIME (5280/
`
`(PACE x60) ) }
`
`
`120A
`
`TaSTEPSZERO?
`
`>
`
`1208
`
`120C
`
`teod
`
`TaSUM>ZERO?
`
`
`
`
`
`TaAVERAGE=
`TaSUM/TaSTEPS
`
`SET TaSUM-0
`SET TaSTEPS=0
`
`13
`
`13
`
`
`
`U.S. Patent
`
`Jan. 25, 2000
`
`Sheet 13 of 13
`
`6,018,705
`
`0680
`
`t."hi 0
`
`mag0‘5
`0c80098°0|008
`0100|Ov
`O100089
`06300|0¢9
`06500|095
`yt914
`r0hostOn
`Wei‘ne
`cttx
`afee
`‘0oh00
`
`
`
`
`
`
`
`(SONOI3S)(91)SWILLOVINOD
`
`00
`
`00
`
`0S¢
`
`002
`
`0S}
`
`
`
`(JTIW/SSLANTH)JOVd
`
`14
`
`14
`
`
`
`6,018,705
`
`1
`MEASURING FOOT CONTACT TIME AND
`FOOT LOFT TIME OF A PERSON IN
`LOCOMOTION
`
`BACKGROUND OF THE INVENTION
`
`2
`It is therefore a general object of the present invention to
`provide a new approach to pedometry that is affordable,
`reliable, easy to use and accurate.
`SUMMARYOF THE INVENTION
`
`1. Field of the Invention
`
`The present invention relates to the monitoring of the
`orthopedic motion of a person and, more particularly, to the
`measuring of foot contact time, foot loft time, speed and/or
`pace of a person in locomotion.
`2. Discussion of the Related Art
`
`It is known that useful information may be derived from
`the measurement of the “foot contact time” of a person in
`locomotion, wherein “foot contact time” refers to the period
`of time that a foot of a person is in contact with the ground
`during a stride taken by the person. Once the foot contact
`time of a person is known,other information, such as rate of
`travel, distance traveled and ambulatory expended energy
`may be calculated based upon this measured foot contact
`time.
`
`According to the invention, a method and an apparatus are
`disclosed in which an output of an accelerometer is used to
`determine:
`(1)
`instances at which a foot of a user in
`locomotion leaves a surface, and (2) instances at which the
`foot of the user impacts the surface. By measuring the time
`difference between each instance at which the foot impacts
`the surface and the following instance at which the foot
`leaves the surface, several periods of time that the foot was
`in contact with the surface during strides taken by the user,
`1e., several foot contact times, may be measured accurately
`and reliably. By calculating an average of these several
`measured foot contact times, an average foot contact time
`may be determined, from which information such as the
`pace of the user, rate of travel, distance traveled, etc., may
`be calculated. Additionally, by measuring time differences
`between the instances at which the foot of the user leaves the
`
`10
`
`15
`
`20
`
`time has been measured by
`In the past, foot contact
`surface and the following instances at which the foot
`placing pressure-sensitive sensors or switches, such asresis-
`impacts the surface, the average period of time that the foot
`tive sensors, in both the heel and toe portions of the sole of
`wasnot in contact with the surface, 1.e., the average footloft
`25
`a shoe, and measuring a time difference betweenafirst
`time, between strides taken by the user also may be calcu-
`lated.
`signal output by the heel sensor (which indicates that the
`foot has made physical contact with the ground) and a
`second signal output by the toe sensor (which indicatesthat
`the foot has left the ground). These sensors, however, are
`subjected to a high-impact environmentinside of the shoe,
`and therefore fail frequently. In addition, inaccurate foot
`contact time measurements may result when a useris taking
`strides during which either the heel sensor or the toe sensor
`is not activated, for example, when a user is running on his
`or her toes.
`
`invention, a
`According to one aspect of the present
`method for analyzing the motion of a foot relative to a
`surface includes using an output of an accelerometer to
`determine a momentthat the foot leaves the surface.
`
`According to another aspect of the invention, the output
`signal of the accelerometer, which is indicative of the
`acceleration of the foot, is fed to a signal processing circuit
`configured to analyze the signal to determine a momentthat
`the foot leaves the surface.
`
`30
`
`35
`
`Another device well-known in the art is a pedometer. A
`pedometer typically is mounted on the waist of a user and is
`configured to count the footsteps of the user by measuring
`the number of times the user’s body moves up an down
`during footsteps taken by the user. A well-knownprior art
`pedometer design uses a weight mounted on a spring to
`count the numberof times that the user’s body moves up and
`down as the user is walking. By properly calibrating the
`pedometer according to a previously measuredstride length
`of the user,
`the distance traveled by the user may be
`measured by this device. These “weight-on-a-spring”
`pedometers, however, generally cannot measure the distance
`traveled by a runner because the weight experiences exces-
`sive bouncing during running and footsteps are often
`“double-counted” because of this bouncing, causing the
`pedometer to produce inaccurate results. These devices,
`therefore, may not be used across different training regimes
`(e.g., walking, jogging, and running).
`Anotherprior art pedometer device uses an accelerometer
`to measure the number of times that a foot impacts the
`ground whena useris in locomotion. That is, an acceler-
`ometer is mounted on a shoeso as to producea signal having
`pronounced downward going peaks that are indicative of
`moments that the foot impacts the ground. These devices
`therefore produce results similar to the prior art weight-on-
`a-spring pedometer devices in that they merely count the
`number of footsteps of a user, and must be calibrated
`accordingto the stride length of the user in order to calculate
`the distance traveled by the user. Thus, these accelerometer-
`based devices are subject to similar limitations as are the
`weight-on-a-spring devices, and are not able to measure the
`foot contact time of a user in locomotion.
`
`40
`
`45
`
`50
`
`55
`
`60
`
`65
`
`15
`
`According to another aspect, the output of the acceler-
`ometer also is used to determine a moment that the foot
`comes into contact with the surface.
`
`According to yet another aspect, a foot contact time may
`be determined based upona difference between the moment
`that the foot comes into contact with the surface and the
`moment that the foot leaves the surface, or a foot loft time
`may be determined based upona time difference between the
`momentthat the foot leaves the surface and the momentthat
`the foot comes into contact with the surface.
`
`the
`According to yet another aspect of the invention,
`measured foot contact time is used to determine the rate at
`which a user is moving relative to the surface. Further, by
`measuring the time interval that the user is in locomotion,
`the distance that the user has traveled may be determined by
`multiplying the rate at which the user is moving by the time
`interval during which the rate measurement was determined.
`According to another aspect, a method for determining a
`rate that a user is moving on foot relative to a surface
`includes the steps of: (a) determining a foot contact time of
`a user in locomotion;(b) if the foot contact time is less than
`a first amount of time, then deriving the rate at which the
`user is moving accordingto a first equation in which the foot
`contact time is a factor; and (c) if the foot contact time is
`greater than a second amountof time, which is greater than
`the first amount of time, then deriving the rate at which the
`user is moving according to a second equation in which the
`foot contact time is a factor.
`
`According to another aspect of the invention, a device for
`analyzing the motion of a foot relative to a surface includes
`an accelerometer and a signal processing circuit. The accel-
`
`15
`
`
`
`6,018,705
`
`3
`erometer is supported in relation to the foot and is configured
`and arranged to provide an output signal indicative of the
`acceleration of the foot. The signal processing circuit is
`coupled to the accelerometer to receive the output signal
`from it, and is configured to analyze the output signal to
`determine at
`least one moment
`that
`the foot
`leaves the
`surface.
`
`According to another aspect of the invention, the process-
`ing circuit also is configured to analyze the output signal to
`determine at least one momentthat the foot makes contact
`
`10
`
`with the surface. Additionally, according to yet another
`aspect, the processing circuit is configured to: (1) analyze
`the output signal to determine at least one time period that
`the foot was in contact with the surface during at least one
`stride taken by the foot; and/or (2) analyze the output signal
`to determineat least one time period that the foot was not in
`contact with the surface between strides taken by the foot.
`According to another aspect, a device for determining the
`rate at which a user in locomotion is moving includes
`processing circuitry adapted to receive information regard-
`ing a foot contact time. The processing circuitry is config-
`ured such that if the foot contact time is less than a first
`amountof time, then the processing circuitry derives the rate
`at which the user is moving according to a first equation in
`which the foot contact time is a factor, and if the foot contact
`time is greater than a second amount of time, which is
`greater than or equal to the first amount of time, then the
`processing circuitry derives the rate at which the user is
`moving according to a second equation in which the foot
`contact time is a factor.
`
`BRIEF DESCRIPTION OF THE DRAWINGS
`
`FIG. 1 is a block diagram of a network in which the
`present invention may be used;
`FIG. 2 is an illustration showing how the invention may
`be mounted with respect to a user;
`FIG. 3 is a block diagram of a system in which the
`invention may be used;
`FIG. 4 is a block diagram of one embodimentofa circuit
`according to the present invention;
`FIG. 5 is a schematic diagram of the circuit shown in FIG.
`
`4;
`
`15
`
`20
`
`30
`
`35
`
`40
`
`45
`
`50
`
`55
`
`60
`
`65
`
`4
`FIG. 13 is a more detailed flow diagram of the interrupt
`portion of the method shown in FIG. 11; and
`FIG. 14 is a graph illustrating how the pace of a user in
`locomotion may be determined based upon the average
`measured foot contact time of a foot of the user.
`
`DETAILED DESCRIPTION OF THE
`INVENTION
`
`FIG. 1 showsa block diagram of a network 70 in which
`the present invention may be used. As shown, network 70
`includes network processing circuitry 30, a memory unit 28,
`a user interface 32, a display 26A, and an audio or vibra-
`tional indicator 26B. Network processing circuitry 30 also is
`coupled to receive inputs from one or more monitoring
`devices, such as foot contact time/foot loft time generators
`20A and 20B, heart rate monitor 22, and respiratory monitor
`24. The devices shown in FIG. 1 may be linked together, for
`example, via direct wiring or capacitive coupling, by using
`radio-frequency (RF) or infa-red (IR) transmitters/receivers,
`or by any other information transmission medium known to
`those skilled in theart.
`
`Network processing circuitry 30 may include a personal
`computer, or any other device capable of processing infor-
`mation from the various inputs of network 70. Memory unit
`28 is coupled to network processing circuitry 30 and is used
`to store programming and data for network processing
`circuitry 30 and/or to log data processed by circuitry 30.
`User interface 32 also is coupled to network processing
`circuitry 30 and permits a user, e.g., a walker, jogger or
`runner, to select a particular feature implemented by opera-
`tion of a software routine,
`to input particular operating
`parameters, or to select particular outputs for display 26A
`and/or audioor vibrational indicator 26B. Heart rate monitor
`22 and respiratory monitor 24 operate according to known
`methods and supply inputs to network processing circuitry
`30.
`
`Each one of foot contact time/foot loft time generators
`20A and 20B operates accordingto the present invention and
`supplies a separate input to network processing circuitry 30.
`By receiving information from the outputs of foot contact
`time/foot
`loft
`time generators 20A and 20B, heart rate
`monitor 22, and respiratory monitor 24, as well as inputs
`from any other type of electronic health monitoring device,
`network processing circuitry 30 is able to process all such
`FIG. 6 is a pair of graphs showingsignals at two nodes of
`information and provide a user with a fitness metric, to help
`the circuit shown in FIG. 5 during a period in which a user
`the user attain a peak fitness level in the most efficient
`is walking;
`manner possible, or other health related information, useful
`FIG. 7 is a pair of graphs that compare the amplified/
`for physical therapy, recovery, etc.
`filtered output of the accelerometer according to the inven-
`FIG. 2 illustrates how a device according to the invention
`tion with data obtained using prior art resistive sensors
`may be mounted onauser. Each of devices 20A-—20C shown
`during a period that a user is walking;
`in FIG. 2 has a particular axis in which it senses acceleration,
`FIG. 8 is a pair of graphs showingsignals at two nodes of
`1e., an acceleration sensing axis. According to one embodi-
`the circuit shown in FIG. 5 during a period in which a user
`ment of the invention, each of the devices should be
`is running;
`mounted such that the acceleration sensing axis of the device
`FIG. 9 is a pair of graphs that compare the amplified/
`is oriented substantially parallel to a bottom surface of the
`filtered output of the accelerometer according to the inven-
`foot of the user. For example, device 20A is mounted on the
`tion with data obtained using prior art resistive sensors
`ankle of the user, device 20B is mounted on or within the
`during a period that a user is running;
`shoe of the user, and device 20C is mounted on the waist of
`FIG. 10 is a high-level flow diagram of a continuous-loop
`the user, with the acceleration sensing axises of the devices
`portion of a method for measuring foot contact time accord-
`being oriented as indicated by arrows 80A, 80B and 80C,
`ing to the invention;
`respectively. In each case, this positioning of the accelera-
`FIG. 11 is a high-level flow diagram of an interrupt
`tion sensing axis has been found to produce an output signal
`portion of the method for measuring foot contact
`time
`that is most strongly indicative of both: (1) the momentat
`according to the invention;
`which the foot of the user leaves the surface, and (2) the
`momentat whichthe foot of the user comesinto contact with
`FIG. 12 is a more detailed flow diagram of the
`continuous-loop portion of the method shown in FIG. 10;
`
`the surface. It is hypothesized that this is true because a large
`
`16
`
`16
`
`
`
`6,018,705
`
`5
`portion of the change in acceleration sensed by the device is
`caused by the friction between the shoe of the user and the
`surface, rather than being caused primarily by the impact of
`the shoe with the surface, as is the case with prior art
`accelerometer-based pedometers.
`FIG. 3 shows a system 72 according to the present
`invention. As shown, the system 72 includes a foot contact
`time/footloft time generator 20 (which could correspond to
`either of foot contact time/footloft time generators 20A and
`20B in FIG. 1), a memory unit 54, a user interface 58, a
`display 56A, and an audio or vibrational indicator 56B.
`According to one embodiment, foot contact time/foot loft
`time generator 20 includes a micro-controller having virtu-
`ally all circuitry, e.g., memory, timers and analog-to-digital
`(A/D) converters, on board, so that memory unit 54 need
`only be used to perform functions such as permanently
`storing data produced by foot contact time/foot loft time
`generator 20.
`User interface 58 may be activated conventionally by
`means of buttons, switches or other physically actuated
`devices, or may be voice activated using a commercially
`available voice activation device. As discussed in more
`detail below,user interface 58 may be used, for example: (1)
`to adjust any of several parameters used in a software routine
`according to the invention, (2) to select any of several
`possible outputs for the user, e.g., outputs could be displayed
`on display 56A or could provide a user with an audio or
`vibrational indication via audio or vibrational indicator 56B,
`or (3) to select features which are implemented through
`software routines called automatically responsive to user
`inputs.
`FIG. 4 shows an exemplary embodiment of the foot
`contact time/footloft time generator 20 shownin FIG. 3. As
`shown,foot contact time/footloft time generator 20 includes
`an accelerometer 34, an amplifier circuit 38 (which has a
`high-passfilter 36 included within it), and a micro-controller
`40. An output of accelerometer 34 is connected to an input
`of amplifier circuit 38, and an output of amplifier circuit 38
`is connected to an input of micro-controller 40.
`FIG. 5 showsthe foot contact time/footloft time generator
`20 shown in FIG. 4 in more detail. As shown in FIG. 5,
`output 50 of accelerometer 32 is provided to an input
`capacitor Cl included in amplifier circuit 38. Amplifier
`circuit 38 further includes operational amplifier 62 and
`resistors RI-R4. According to one embodiment, accelerom-
`eter 32 may comprise part number ADXL250, manufactured
`by Analog Devices, Inc., and operational amplifier 62 may
`comprise part number MAX418 produced by MAXIM,Inc.
`As shown in FIG. 5, resistor R1 is connected between
`input capacitor Cl and the inverting input of operational
`amplifier 62, and resistor R2 is connected in feedback
`between the inverting input
`terminal and output 52 of
`operational amplifier 62. Thus, the combination of input
`capacitor C1 and resistor R1 form a high-passfilter, and the
`position of resistors R1 and R2 place the amplifier circuit in
`an inverting configuration with a gain-factor dependent on
`the relative values of resistors R1 and R2. In the embodi-
`
`ment shown,resistor R2 has a value of one mega-ohm and
`resistor R2 has a value of 150 kili-ohms, so that the gain
`factor of the amplifier is approximately (-6.6). In addition,
`according to the embodiment shown, capacitor Cl has a
`value of 0.15 microfarads, so that high-passfilter section 36
`of amplifier circuit 38 cuts off input signal frequencies that
`are less than approximately 7.07 hertz.
`Resistor R3 is connected between VCC supply node 44
`and the non-inverting input 60 of operational amplifier 62,
`
`10
`
`15
`
`20
`
`30
`
`40
`
`45
`
`50
`
`55
`
`60
`
`65
`
`6
`and resistor R4 is hconnected between non-inverting input
`60 and ground node 42. VCCsupply node 44 is maintained
`at approximately 5 volts (e.g., regulated from a six-volt
`battery) in relation to ground node 42, andresistors R3 and
`R4 are of equal values (e.g., 50 kili-ohms each) so that the
`voltage at non-inverting input node 60 is maintained
`approximately midway between the voltage at VCC supply
`node 44 and ground (i.e., approximately 2.5 volts).
`Output 52 of amplifier circuit 38 is connected to a first
`A/D input 46 of low-power micro-controller 40, and node 60
`of amplifier circuit 38 is connected to a second A/D input 48
`of micro-controller 40. According to one cmbodiment,
`micro-controller 40 may comprise part number PIC:16C73
`manufactured by Microchip,
`Inc. This micro-controller
`includes on-board memory, A/D converters, and timers. A/D
`input 48 of micro-controller 40 serves as a zero-reference
`that is maintained at approximately 2.5 volts (as described
`above), and input 46 of micro-controller 40 serves as a
`variable input that fluctuates between 0 and 5 volts. Micro-
`controller 40 samples the voltages at inputs 46 and 48 at a
`rate of approximately 500 samples-per-second, converts
`these samples into 8-bit unsigned digital values, and calcu-
`lates the difference between the voltages at the two inputs,
`which difference is used during operation of software rou-
`tines described in more detail below.
`
`FIG. 6 shows twocurves along the same time axis. These
`curves represent the 8-bit unsigned digital values of the
`voltages at nodes 50 and 52 of the circuit shown in FIG. 5
`during a period whena user is walking. That is, curve 50W
`in FIG. 6 represents (digitally) the voltage at output 50 of
`accelerometer 32 before it
`is filtered and amplified, and
`curves 46W and 48W,respectively, represent (digitally) the
`voltages at inputs 46 and 48 of micro-controller 40 during
`the period whenthe user is walking. While each of curves
`46W, 48W and 50W shares a common time axis,
`the
`voltage-magnitude axis of curves 46W and 48W is distinct
`from the voltage-magnitude axis of curve 50W. Therefore,
`the placement of curve 50W above curves 46W and 48W is
`not intended to signify that curve 50W attains a higher
`amplitude than do curves 46W and 48W.
`As shown in FIG. 6, because amplifier circuit 38 is
`configured to have a negative gainfactor, high peak 51W of
`curve 50W corresponds with low peak 47W of curve 46W.
`High peak 49W of curve 46W, however, does not appear to
`correspond to a low peak of curve 50W. Thatis, high peak
`49W is ascertainable only after the output of accelerometer
`34 has been high-pass filtered and amplified by amplifier
`circuit 38. It is high peak 49W in curve 46W that indicates
`the momentthat the foot of the user hasleft the surface when
`the user is in locomotion.
`
`low peak 47W in curve 46W indicates the
`Similarly,
`moment that the foot of the user has impacted with the
`surface when the user is in locomotion. By measuring the
`time difference between peak 47W and peak 49W of curve
`46W,the foot contact time of the user when the user is in
`locomotion may be ascertained. As used herein, “foot con-
`tact time” refers to the period of time between when a foot
`of a user impacts a surface and whenthe foot next leaves the
`surface.
`
`time of a user in
`loft
`the foot
`In a similar manner,
`locomotion may be determined. That is, by measuring the
`time difference between high peak 49W and low peak 53W
`in curve 46W,the foot loft time of the user is ascertainable.
`As used herein, “foot loft time” refers to the period of time
`between whena foot of a user leaves a surface and when the
`foot next comes into contact with the surface.
`
`17
`
`17
`
`
`
`7
`FIG. 7 shows the correspondence, when a user is walking,
`between (1) two curves 55H and 55T producedbyresistive
`sensors mounted in the heel and toe, respectively, of a shoe
`and (2) the amplified andfiltered output of the accelerometer
`accordingto the invention. Thatis, curve 55H represents the
`output of a resistive sensor mounted in the heel of a shoe,
`curve 55T represents the output of a resistive sensor
`mounted in the toe of the shoe, and curve 46W represents the
`voltage at node 52 of circuit 20 (shown in FIG. 5). All of
`these measurements were taken while a user was walking.
`While each of curves 55H, 55T and 46W shares a common
`time axis, the voltage-magnitude axis of curves 55H and 55T
`is distinct from the voltage-magnitude axis of curve 46W.
`Therefore,
`the placement of curves 55H and 55T above
`curve 46W is not intended to signify that curves 55H and
`55T attain higher amplitudes than does curve 46W.
`As shownby the dashed lines in FIG. 7, the high to low
`transition of curve 55H (which indicates that the shoe ofthe
`user has impacted with the ground) corresponds with low
`peak 47W of curve 46W, and the low-to-high transition of
`curve 55T (which indicates that the shoe ofthe user hasleft
`the ground) corresponds with high peak 49W of curve 46W.
`Thus, the foot contact time and foot loft time measurements
`that are obtained, whena user is walking, by measuring time
`differences between high and low peaks, and vice-versa, of
`the high-pass filtered/amplified output of an accelerometer
`(mounted as described above) appear to produceresults that
`are at
`least as accurate as those produced by prior art
`resistive sensors.
`
`6,018,705
`
`8
`and (2) the amplified andfiltered output of the accelerometer
`accordingto the invention. Thatis, curve 57H represents the
`output of a resistive sensor mounted in the heel of a shoe,
`curve 57T represents the output of a resistive sensor
`mountedin the toe of the shoe, and curve 46R represents the
`voltage at node 52 of circuit 20 (shown in FIG. 5). All of
`these measurements were taken while a user was running.
`While each of curves 57H, 57T and 46R shares a common
`time axis, the voltage-magnitude axis of curves 57H and 57T
`is distinct from the voltage-magnitude axis of curve 46R.
`Therefore,
`the placement of curves 57H and 57T above
`curve 46Ris not intendedto signify that curves 57H and 57T
`attain higher amplitudes than does curve 46R.
`As shown by the dashed lines in FIG. 9, the high-to-low
`transition of curve 5714 (which indicates that the shoe of the
`user has impacted with the ground) corresponds with low
`peak 47R of curve 46R, and the low-to-high transition of
`curve 57T (whichindicates that the shoe ofthe user hasleft
`the ground) corresponds with high peak 49R of curve 46R.
`Thus, the foot contact time and foot loft time measurements
`that are obtained, whena user is running, by measuring time
`differences between high and low peaks, and vice-versa, of
`the high-pass filtered/amplified output of an accelerometer
`(mounted as described above) appear to produceresults that
`are at
`least as accurate as those produced by prior art
`resistive sensors.
`
`10
`
`15
`
`20
`
`25
`
`The output signal from accelerometer 34 (shown in FIGS.
`4 and 5)
`is analyzed by microcontroller 40 using two
`primary software routines: (1) a continuous-loop routine that
`accumulates data, e.g., foot conta