`
`111111111111111111111111111R11119111j!!1111111111111111111111111
`
`(12) United States Patent
`Pasolini et al.
`
`(10) Patent No.:
`(45) Date of Patent:
`
`US 7,698,097 B2
`Apr. 13, 2010
`
`(54) METHOD FOR CONTROLLING A
`PEDOMETER BASED ON THE USE OF
`INERTIAL SENSORS AND PEDOMETER
`IMPLEMENTING THE METHOD
`
`5/2005 Blackadar et al.
`6,898,550 B1
`1/2007 Tsuji
`7,169,084 B2 *
`7,297,088 B2 * 11/2007 Tsuji
`2001/0031031 A1* 10/2001 Ogawa et al.
`
`702/182
`482/8
`482/3
`377/24.2
`
`(75)
`
`Inventors: Fabio Pasolini, S. Martino Siccomario
`(IT); Ivo Binda, Voghera (IT)
`
`(73) Assignee: STMicroelectronics S.R.L., Agrate
`Brianza (IT)
`
`*) Notice:
`
`Subject to any disclaimer, the term of this
`patent is extended or adjusted under 35
`U.S.C. 154(b) by 707 days.
`
`(21) Appl. No.: 11/537,986
`
`(22) Filed:
`
`Oct. 2, 2006
`
`FOREIGN PATENT DOCUMENTS
`
`GB
`JP
`JP
`
`2 359 890
`63-262784
`04-192095
`
`9/2001
`10/1988
`7/1992
`
`OTHER PUBLICATIONS
`
`Tasaka, Translation of JP 63262784, published Oct. 31, 1988.*
`Tasaka, Translation of H04-192095, published Jul. 10, 1992.*
`
`(65)
`
`Prior Publication Data
`
`* cited by examiner
`
`US 2007/0143069 Al
`
`Jun. 21, 2007
`
`(30)
`
`Foreign Application Priority Data
`
`Primary Examiner Hal D Wachsman
`(74) Attorney, Agent, or Firm Lisa K. Jorgenson; Robert
`Iannucci; Seed IP Law Group PLLC
`
`Oct. 3, 2005
`
`(EP)
`
`
`
` 05425684
`
`(57)
`
`ABSTRACT
`
`(51) Int. Cl.
`GO1C 22/00
`G06F 17/40
`(52) U.S. Cl.
`
`(2006.01)
`(2006.01)
` 702/160; 702/176; 702/178;
`377/24.2
` 702/160,
`702/176, 178
`See application file for complete search history.
`
`(58) Field of Classification Search
`
`(56)
`
`References Cited
`
`U.S. PATENT DOCUMENTS
`
`A method for controlling a pedometer includes the steps of:
`generating a signal correlated to movements of a user of the
`pedometer; and detecting steps of the user on the basis of the
`signal. The method moreover envisages the steps of checking
`whether sequences of detected steps satisfy pre-determined
`conditions of regularity; updating a total number of valid
`steps if the conditions of regularity are satisfied; and prevent-
`ing the updating of the total number of valid steps if the
`conditions of regularity are not satisfied.
`
`6,175,608 B1 *
`
`1/2001 Pyles et al.
`
` 377/24.2
`
`26 Claims, 3 Drawing Sheets
`
`205
`NO
`
`NO
`
`READ Az
`
`200
`
`f 210
`
`Tc > Ts1
`
`YES
`
`FST = PD
`
`TO 120
`
`215
`YES
`
`-re' TS2
`
`Nvc =0
`Nuiv =0
`
`220
`
`NO
`
`225
`
`255
`
`YES
`
`YES
`
`STEP
`VALID.
`
`230
`
`NO
`
`235
`
`Nvc =Nvell
`
`NINV = NINV+1
`
`240
`
`YES
`
`245
`S
`
`Nyc=0
`NiNv=0
`
`NO 250
`
`Nvc = min(0, Nvc - 2)
`
`NO
`
`260
`
`110
`
`YES
`
`
`265
`
`N,T =N,TN,
`Nvc =0
`Nuw =0
`F C
`
`TO 120
`
`HTC v. Uniloc
`
`Page 1 of 10
`
`HTC Ex. 1006
`
`
`
`U.S. Patent
`
`Apr. 13, 2010
`
`Sheet 1 of 3
`
`US 7,698,097 B2
`
`10
`
`I
`
`,57
`
`z
`
`Fig.1
`
`L
`
`1
`
`INTERF
`
`ACC
`
`uc
`
`NVT
`
`Fig.2
`
`DISP.
`
`100
`
`INIT.
`NVT=0
`NVC=0
`NINV=0
`
`110
`
`Fig.3
`
`140
`
`COUNT
`
`SURV.
`
`COUNT II
`
`130
`
`ST=C
`
`120
`
`HTC v. Uniloc
`
`Page 2 of 10
`
`HTC Ex. 1006
`
`
`
`U.S. Patent
`
`Apr. 13, 2010
`
`Sheet 2 of 3
`
`US 7,698,097 B2
`
`READ Az
`
`/— 200
`
`[ 210
`
`YES
`
`FQT = PD
`
`TO 120
`
`215
`YES
`
`Tc > Ts2
`
`Nvc = 0
`N1NV = 0
` 1
`
`YES
`
`220
`
`,230
`
`205
`
`NO
`
`NO
`
`NO
`
`225
`
`255
`
`YES
`
`STEP
`VALID.
`
`NO
`
`c 235
`
`NINV = NINV+1
`
`240
`YES
`
`245
`S
`Nyc=0
`N1Nv=0
`
`NO 250
`
`Nvc = rni (0, N
`
`Nvc = Nv +1
`
`NO
`
`Ric = NT2
`
`YES
`
`265
`
`260
`
`71,
`110
`
`Fig.4
`
`Nvr = NvT + NT2
`Nvc = 0
`NiNv = 0
`FsT = C
`
`TO 120
`
`Fig. 8
`
`N =N
`
`140
`
`400
`
`N = N +1
`
`410
`
`N = N.,,_+1
`
`N =
`
`= N +1
`
`420
`
`430
`
`440
`
`YES
`
`c .--
`
`450
`
`NO
`
`TO 120
`
`HTC v. Uniloc
`
`Page 3 of 10
`
`HTC Ex. 1006
`
`
`
`U.S. Patent
`
`Apr. 13, 2010
`
`Sheet 3 of 3
`
`US 7,698,097 B2
`
`305
`NO
`
`READ Az -V -3°C)
`
`310
`
`YES
`
`Nvc=0
`Nit4v=0
`
`To 110
`
`315
`
`YES
`
`YES
`
`NO
`
`STEP
`EEC.
`
`325
`
`Nvc = Nvc41
`Nye= NV7+1
`
`NO
`
`330
`
`NVC - NT3
`
`320
`
`NO
`
`STEP
`VALID
`
`340
`
`NiNv = N
`
`YES
`
`335
`
`345
`355 YES
`
`N1Nv NT
`
`NO
`
`350
`
`Nveo
`NINV=°
`
`Nvc=0
`Nuvv=0
`
`Nvc = ming Nvc -
`Nvr= NvT+1
`
`Fig.7
`
`TO 110
`
`AZ
`
`Azp
`
`AZN
`
`I
`TR(1)
`
`Fig.6
`
`ATIci ►
`
`K-2
`
`• • •
`
`2
`I
`I
`TR(2) • • • TR(K-2)
`
`Fig.5
`
`AT
`
`ATK4
`
`K-1
`
`TR(K-1)
`
`TR(K)
`
`TA
`
`TB
`
`TV
`
`HTC v. Uniloc
`
`Page 4 of 10
`
`HTC Ex. 1006
`
`
`
`US 7,698,097 B2
`
`1
`METHOD FOR CONTROLLING A
`PEDOMETER BASED ON THE USE OF
`INERTIAL SENSORS AND PEDOMETER
`IMPLEMENTING THE METHOD
`
`BACKGROUND OF THE INVENTION
`
`1. Field of the Invention
`The present invention relates to controlling a pedometer
`based on the use of inertial sensors.
`2. Description of the Related Art
`As is known, a pedometer is a device that can be carried by
`a user and has the function of counting the number of steps
`during various walking or running activities for estimating
`accordingly the distance traveled. The indications supplied
`are useful for quantifying the motor activity performed by a
`person in the course of a given period, for instance, for clinical
`purposes, for assessing the athletic performance, or even just
`for simple personal interest.
`The reliability of a pedometer obviously depends on the
`precision in estimating the step length of the user at the
`various rates of locomotion, but also on the selectivity in
`recognizing and ignoring events not correlated to the gait,
`which, however, cause perturbations resembling those pro-
`duced by a step. For example, many pedometers are based on
`the use of inertial sensors, which detect accelerations along a
`substantially vertical axis, and recognize that a step has been
`being made by a user when the time plot of the acceleration
`signal shows given morphological characteristics. Basically,
`a step is recognized when the pedometer detects a positive
`acceleration peak (i.e., a peak directed upwards) having an
`amplitude greater than a first threshold, followed, at a dis-
`tance of some tenths of second, by a negative acceleration
`peak (directed downwards) having an amplitude greater than
`a second threshold. However, there are many random events
`that can interfere with correct recognition of the step. Impact
`or other external vibrations and given movements of the user
`can, in fact, give rise to so-called "false positives", i.e., to
`events that are recognized as steps even though in actual fact
`they are not, because the morphological characteristics pro-
`duced are compatible. Events of this type are very frequent
`also in periods of rest, when the user, albeit not walking, in
`any case performs movements that can be detected by the
`pedometer. In the majority of cases, also "isolated" steps or
`very brief sequences of steps are far from significant and
`should preferably be ignored because they are, in effect, irrel-
`evant in regard to assessment of the motor activity for which
`the pedometer is being used.
`Of course, in all these situations, the count of the steps may
`prove to be completely erroneous.
`
`BRIEF SUMMARY OF THE INVENTION
`
`One embodiment of the present invention is a method for
`controlling a pedometer and a pedometer which overcome the
`described above limitations.
`One embodiment is a method for controlling a pedometer.
`The method includes: generating a signal correlated to move-
`ments of a user of the pedometer; detecting steps of the user
`based on the signal; checking whether sequences of the
`detected steps satisfy pre-determined conditions of regular-
`ity; updating a total number of valid steps if the conditions of
`
`2
`regularity are satisfied; and preventing updating of the total
`number of valid steps if the conditions of regularity are not
`satisfied.
`
`5
`
`BRIEF DESCRIPTION OF THE SEVERAL
`VIEWS OF THE DRAWINGS
`
`For a better understanding of the invention, an embodiment
`thereof is now described, purely by way of non-limiting
`10 example and with reference to the attached plate of drawings,
`wherein:
`FIG. 1 shows a simplified and partially sectioned perspec-
`tive view of a portable electronic device incorporating a
`pedometer according to the present invention;
`FIG. 2 is a simplified block diagram of the pedometer of
`FIG. 1;
`FIG. 3 shows a flowchart corresponding to a control
`method according to the present invention executed by the
`pedometer of FIGS. 1 and 2;
`FIG. 4 is a more detailed flowchart corresponding to a first
`step of the method of FIG. 3;
`FIG. 5 is a graph that represents first quantities used in the
`method according to the present invention;
`FIG. 6 is a graph that represents second quantities used in
`the method according to the present invention;
`FIG. 7 is a more detailed flowchart corresponding to a
`second step of the method of FIG. 3; and
`FIG. 8 is a more detailed flowchart corresponding to a third
`step of the method of FIG. 3.
`
`15
`
`20
`
`25
`
`30
`
`DETAILED DESCRIPTION OF THE INVENTION
`
`With reference to FIGS. 1 and 2, a pedometer 1 is inte-
`35 grated within a portable electronic device, such as a cell
`phone 2. The pedometer 1 comprises an inertial sensor 3, a
`control unit 5, equipped with a nonvolatile-memory module
`(not illustrated herein), a display 6, and a communication
`interface 8, all housed on a card 9, which is, in turn, fixed
`40 within a casing 10 of the cell phone 2. In the embodiment
`described herein, the control unit 5 performs control func-
`tions of the pedometer 1 and, moreover, presides over bi-
`directional communication and over handling of the func-
`tions envisaged for the cell phone 2. Likewise, the display 6,
`45 which is obviously arranged so as to be visible from the
`outside of the casing 10, can be used for displaying both
`information regarding the pedometer 1 and, more in general,
`information regarding the operation of the cell phone 2.
`The inertial sensor 3 is a linear accelerometer of a MEMS
`50 (micro-electromechanical systems) type and is mounted on
`the card 9 so as to have a detection axis Z substantially parallel
`to a longitudinal axis L of the casing 10 of the cell phone 2. In
`practice, the detection axis Z and the longitudinal axis L are
`substantially horizontal, when the cell phone 2 is resting on a
`55 surface, and substantially vertical or slightly inclined with
`respect to the vertical when the cell phone 2 is handled. The
`inertial sensor 3 supplies at output an acceleration signal Az,
`which is correlated to the accelerations undergone by the
`inertial sensor 3 itself along the detection axis Z.
`The control unit 5 receives and processes the acceleration
`signal A, as explained in detail hereinafter for identifying and
`counting a total number of valid steps N„ made by a user
`wearing or carrying the pedometer 1, for example, on his belt
`or on his shoulder. In addition, the control unit 5 is preferably
`65 configured for generating an estimate of the distance traveled
`by the user and other data, such as, for example, estimates of
`the average speed during movement and energy consumption.
`
`60
`
`HTC v. Uniloc
`
`Page 5 of 10
`
`HTC Ex. 1006
`
`
`
`US 7,698,097 B2
`
`TV= [TR(K-1)+AT K_,- TA, TR(K-1)-FATK_ I-FTB]
`
`where TA and TB are complementary portions of the valida-
`tion interval TV. In the embodiment of the invention
`described herein, the complementary portions TA, TB are
`45 defined as follows, for the generic current step K:
`
`4
`after the test on the state flag FST of block 120 of FIG. 3, the
`surveying procedure is executed, block 140). Otherwise (out-
`put NO from block 205), the duration of the time interval T,
`is compared with a second time threshold Ts2, shorter than the
`first time threshold Ts, and equal, for example, to 3 s (block
`215). If the second time threshold Ts2 has been exceeded
`(output YES from block 215), the number of valid control
`steps N„ and the number of invalid steps N iNv are set to zero
`(block 220); then a step-recognition test is carried out (block
`0 225). Otherwise (output NO from block 215), the control unit
`5 directly executes the step-recognition test.
`In the step-recognition test of block 225, the control unit 5
`verifies whether the time plot of the acceleration signal Az
`(i.e., the sequence of the samples acquired) has pre-deter-
`of the total number of valid steps N„; a second counter, 1 5 mined characteristics. In particular (FIG. 5), a step is recog-
`nized if the acceleration signal Az shows a positive peak,
`higher than a positive acceleration threshold Azp, followed by
`a negative peak, smaller than a negative acceleration thresh-
`old AzN, and if the negative peak falls within a time window
`(block 110), based upon the sampling of the acceleration 2 o TW of pre-determined amplitude and, moreover, located at a
`pre-determined distance after the positive peak.
`If the control unit 5 does not recognize an event corre-
`sponding to a step (output NO from block 225), a new sample
`of the acceleration signal Az is read (block 200). If, instead,
`a sequence of steps that are close to one another, which satisfy 2 5 the step-recognition test is passed (output YES from block
`225), the control unit 5 executes a first validation test, corre-
`sponding to the regularity of the individual step (block 230).
`With reference also to FIG. 6, the validation occurs when the
`duration ATK of a current step K is substantially homoge-
`dure terminates when a time interval T, that has elapsed from 3 o neous with respect to the duration ATK_, of an immediately
`preceding step K-1 (the duration of a generic step is deter-
`mined by the time that has elapsed between an instant of
`recognition of the step of which the duration is evaluated and
`an instant of recognition of the step that immediately pre-
`larity has been recognized, and to a second value PD, if the 3 5 cedes it). More precisely, the last step recognized is validated
`if the instant of recognition of the current step TR(K) falls
`within a validation interval TV, defined with respect to the
`instant of recognition of the immediately preceding step
`TR(K-1), in the following way:
`
`3
`The total number of valid steps N„ and the other data possi-
`bly produced are sent to the display 6.
`The communication interface 8 in this case is based on the
`transceiver system (known and not shown) of the cell phone 2
`and, preferably, also comprises a port (also known and not 5
`shown) for communication with a computer. The communi-
`cation interface 8 can thus be used both for downloading the
`data produced by the pedometer 1 (amongst which at least the
`total number of valid steps N„) and for uploading operating
`parameters for the pedometer 1 into the control unit 5.
`The control unit 5 is configured for executing a control
`procedure, as illustrated with reference to FIGS. 3-8.
`Upon switching-on of the pedometer 1, an initialization
`step is executed (block 100, FIG. 3), in which a first counter
`
`hereinafter referred to as number of valid control steps N„;
`and a third counter, hereinafter referred to as number of
`invalid steps NiNv, are set to zero.
`The control unit 5 then executes a first counting procedure
`
`signal Az at a pre-determined frequency, for example 25 Hz.
`In this step, the user is considered at rest and the control unit
`5 is considered as waiting to recognize, on the basis of the
`acceleration signal Az, sequences of events corresponding to
`
`pre-determined conditions of regularity described in detail
`hereinafter. When a sequence of steps corresponding to a
`regular gait of the user is recognized, the first counting pro-
`cedure is interrupted. Alternatively, the first counting proce-
`
`the last step recognized is longer than a first time threshold
`Ts „, for example 10 s. On exit from the first calculation
`procedure, the control unit 5 sets a state flag F„to a first value
`C, if a sequence of steps that satisfies the conditions of regu-
`
`first time threshold Ts , has been exceeded.
`At the end of the first counting procedure, the control unit
`5 checks whether the state flag FST has been set at the first
`value C (block 120), i.e., whether a sequence of steps has been
`recognized. If so (output YES from block 120), a second 40
`counting procedure is executed (block 130). The user is con-
`sidered to be moving, and a first counter, hereinafter referred
`to as total number of valid steps N„,, is incremented when-
`ever an event corresponding to a step is recognized. Further-
`more, the control unit 5 checks the regularity of the sequences
`of steps, as explained hereinafter, and, when an interruption in
`the locomotion is detected, the second counting procedure is
`terminated, and execution of the first counting procedure
`resumes (block 110).
`If, instead, the state flag FST has the second value PD, the
`pedometer 1 is set in a low-consumption wait state ("power
`down" state), and the control unit 5 executes a surveying
`procedure (block 140). The surveying procedure terminates
`when a variation of the d.c. component of the acceleration
`signal Az is detected, i.e., when the cell phone 2 that includes
`the pedometer 1 is moved. The control unit 5 then returns to
`execution of the first calculation procedure (block 110).
`The first counting procedure is illustrated in greater detail
`in FIG. 4.
`Initially, the control unit 5 reads a sample of the accelera-
`tion signal Az (block 200) and then evaluates whether the time
`interval T, that has elapsed from the last step recognized is
`higher than the first time threshold Ts „, i.e., whether the step
`recognition fails for a period longer than the first time thresh-
`old Ts , (block 205). If so (output YES from block 205), the
`state flag F ST is set at the second value PD (block 210) and the
`first counting procedure is terminated (in this eventuality,
`
`TA=ATK,12
`
`TB=ATK_,
`
`so
`
`Consequently, the validation interval is asymmetrical with
`respect to the instant TR(K-1)+ATK_1 and has an amplitude
`equal to 3ATK_1/2. The validation interval TV could, how-
`ever, be symmetrical and have a different amplitude. In prac-
`tice, it is verified that the last step recognized is compatible
`55 with the frequency of the last steps made previously.
`If the verification yields a negative result (output NO from
`block 230), the number of invalid steps N iNv is incremented
`by one (block 235) before being compared with a first pro-
`grammable threshold number NT1, for example 3 (block 240).
`60 If the number of invalid steps N iNv has reached the first
`threshold number NT1 (output YES from block 240), both the
`number of invalid steps N iNv , and the number of valid control
`steps N„ are set to zero (block 245), and the first counting
`procedure is resumed, with reading of a new sample of the
`65 acceleration signal Az (block 200). If, instead, the number of
`invalid steps NiNz is smaller than the first threshold number
`N il (output NO from block 240), the number of valid control
`
`HTC v. Uniloc
`
`Page 6 of 10
`
`HTC Ex. 1006
`
`
`
`US 7,698,097 B2
`
`5
`steps N„ is decremented (block 250). In the embodiment
`described herein, the decrement is equal to two. If the result of
`the decrement operation is negative, the number of valid
`control steps N„ is set to zero (in practice, the updated value
`of the number of valid control steps N„ is equal to the
`smaller between zero and the previous value of the number of
`valid control steps N„, decreased by two). Then, the control
`unit 5 reads a new sample of the acceleration signal A, (block
`200).
`If the first validation test of block 230 is passed, the number
`of valid control steps N„ is incremented by one (block 255),
`and then the control unit 5 executes a first test on regularity of
`the sequence of steps recognized (block 260). The first regu-
`larity test is based upon a first condition of regularity and
`envisages comparing the number of valid control steps N„
`with a second programmable threshold number N7.2 greater
`than the first threshold number NT, (for example, 8). In prac-
`tice, the first condition of regularity is satisfied when there is
`a significant prevalence of steps spaced in a substantially
`uniform way, at the most interrupted sporadically by a num-
`ber of irregular steps smaller than the first threshold number
`NT,. If the number of valid control steps N„ is smaller than
`the second threshold number NT, (output NO from block
`260), the first condition of regularity is not satisfied, and the
`first regularity test indicates that there has not yet been iden-
`tified a sequence of steps corresponding to a sufficiently
`regular gait, and hence the control unit 5 acquires once again
`a new sample of the acceleration signal Az (block 200), with-
`out the total number of valid steps N„ being incremented.
`Otherwise (output YES from block 260), a sequence of steps
`is recognized that satisfies the first condition of regularity, and
`the first regularity test is passed. The number of invalid steps
`N iNv and the number of valid control steps N„ are set to zero,
`whereas the total number of valid steps N vT is updated and
`incremented by a value equal to the second threshold number
`1\17-2 (block 265). Furthermore, the state flag F„ is set at the
`count value, and the first counting procedure is terminated. In
`this case, after the test on the state flag of block 120 of FIG. 3,
`the second counting procedure is executed (block 130).
`In practice, the first counting procedure enables the pedom-
`eter 1 to remain waiting for a sequence of events correspond-
`ing to a sequence of steps that satisfies the first condition of
`regularity. The regularity of the gait is considered sufficient
`when the number of valid control steps N„ reaches the sec-
`ond threshold number N72. The events considered irregular or
`a waiting time that is too long between two successive steps
`cause the decrement (block 250) or the zeroing (blocks 220
`and 245) of the number of valid control steps N„, so that the
`first counting procedure resumes from the start. As long as the
`pedometer 1 is in the waiting condition, the total number of
`valid steps N„ is not incremented because the user is still
`considered as at rest. However, when the first regularity test
`(block 260) is passed, the total number of valid steps N vT is
`immediately updated so as to take into account the valid steps
`(equal to N7.2) that make up the sequence considered as being
`regular. Isolated events and sequence of steps that are in any
`case too short are thus advantageously ignored, whereas
`counting of the steps promptly resumes also in the case of
`isolated irregularities (for example, due to a non-homoge-
`neous acceleration or to a loss of balance at the start of
`locomotion).
`The possibility of programming the value of the first
`threshold number NT, and of the second threshold number
`NT2 enables modification of the sensitivity of the pedometer
`in recognizing an initial sequence of steps. For example, the
`user can program lower values of the first threshold number
`NT, and of the second threshold number N7.2 (for example 2
`
`6
`and 4, respectively) when he remains for a long time in a
`closed environment, for example an office or a room, where it
`would not in any case be possible to maintain a regular gait for
`a long time. In this way, shorter sequences of steps are vali-
`5 dated and counted. Instead, during a more constant and
`intense activity, such as running, the gait remains constant for
`a long time, and hence the first threshold number NT, and the
`second threshold number N7.2 can be programmed with
`higher values (for example, 4 and 12, respectively). Step
`10 sequences that are shorter and not very significant in relation
`to the activity performed can be ignored.
`FIG. 7 illustrates in detail the second counting procedure
`(executed in block 130 of FIG. 3).
`The control unit 5 initially reads a sample of the accelera-
`15 tion signal Az (block 300), and then evaluates whether the
`time interval T, that has elapsed from the last step recognized
`is higher than the first second time threshold Ts2 (block 305).
`If so (outputYES from block 205), the number of invalid steps
`N iNv and the number of valid control steps N„ are zeroized
`20 (block 310), and the second counting procedure is terminated.
`Otherwise (output NO from block 305), a step-recognition
`test is carried out (block 315), identical to the step-recogni-
`tion test of block 225 of FIG. 3. Also in this case, then, step
`recognition is based upon the detection of a positive peak of
`25 the acceleration signal Az followed by a negative peak that
`falls in the time window TW (see FIG. 5).
`If the control unit 5 does not recognize an event corre-
`sponding to a step (output NO from block 315), a new sample
`of the acceleration signal Az is read (block 300). If, instead,
`30 the step-recognition test is passed (output YES from block
`315), a second validation test is made, corresponding to the
`regularity of the individual step (block 320). The second
`validation test is altogether similar to the first validation test
`carried out in block 230 of FIG. 3. Also in this case, then, the
`35 last step recognized is validated if the instant of recognition of
`the current step Tc(K) falls within the validation interval TV
`defined above. In practice, it is verified that the last step
`recognized is compatible with the frequency of the last steps
`made previously.
`If the check yields a positive result (outputYES from block
`320), the control unit 5 updates the total number of valid steps
`N„ and the number of valid control steps N„, incrementing
`them by one (block 325). The number of valid control steps
`N„ is then compared with a third programmable threshold
`45 number NT3 (block 330), which, in the embodiment described
`herein, is equal to the second threshold number N72. If the
`number of valid control steps N„ is smaller than the second
`threshold number NT2 (output NO from block 330), the con-
`trol unit 5 once again directly acquires a new sample of the
`so acceleration signal Az (block 300), whereas otherwise (out-
`put YES from block 330), the number of invalid steps N iNv
`and the number of valid control steps N„ are set to zero
`(block 335) prior to acquisition of a new sample A.
`If, instead, the second validation test of block 320 is nega-
`55 tive, the number of invalid steps N iNv is incremented by one
`(block 340) before being compared with a fourth program-
`mable threshold number NT4 (block 345), which, in the
`present embodiment, is equal to the first threshold number
`N il . If the number of invalid steps NiN, is smaller than the
`60 fourth threshold number NT4 (output NO from block 345), the
`number of valid control steps N„ is decremented (block
`350), here by two. Also in this case, if the result of the
`decrement operation is negative, the number of valid control
`steps N„ is set to zero (the updated value of the number of
`65 valid control steps N„ is equal to the smaller between zero
`and the previous value of the number of valid control steps
`N„, decreased by two). Then, the control unit 5 reads a new
`
`40
`
`HTC v. Uniloc
`
`Page 7 of 10
`
`HTC Ex. 1006
`
`
`
`US 7,698,097 B2
`
`7
`sample of the acceleration signal Az (block 300). If the num-
`ber of invalid steps
`has reached the fourth threshold
`number NT4 (output YES from block 345), the number of
`invalid steps NiNv and the number of valid control steps N„
`are set to zero (block 355), and the second counting procedure
`is terminated.
`In practice, the second counting procedure is based on a
`second condition of regularity, which is satisfied as long as
`sporadic irregular steps occur within sequences of steps
`spaced in a substantially homogeneous way. More precisely,
`the second condition of regularity is satisfied as long as the
`number of invalid steps NiNv is smaller than the fourth thresh-
`old number NT4. Consequently, the second counting proce-
`dure continues to update and increment the total number of
`valid steps N„ as long as the gait of the user is kept regular.
`Possible isolated irregularities are ignored and do not inter-
`rupt or suspend updating of the count, which is, instead,
`interrupted when prolonged pauses occur or in the presence
`of significant discontinuities in locomotion. However, if the
`gait becomes regular again, even with a different rhythm, also
`the count promptly resumes, because the first counting pro-
`cedure is once again executed. This prevents a significant
`number of steps from being neglected.
`The surveying procedure executed in block 140 of FIG. 3
`will now be described in greater detail, with reference to FIG.
`8.
`
`When the surveying procedure is started, a current mean
`value AzM of the acceleration signal Az is stored in the non-
`volatile-memory module (not illustrated) of the control unit 5
`(block 400). The current mean value Azm represents an esti-
`mate of the DC component of the acceleration signal Az,
`which, when the cell phone 2 containing the pedometer 1 is
`stationary, is determined substantially by the contribution of
`the acceleration of gravity along the detection axis Z. In
`practice, then, the current mean value AzM provides an esti-
`mate of the position of the cell phone 2 and of the pedometer
`1.
`After storage of the current mean value A, the pedometer
`1 is set in a low-consumption operating condition (power-
`down condition), in which at least the inertial sensor 3 is
`inactive (block 410).
`A waiting cycle is then carried out (block 420), for example
`of the duration of 10 s, after which all the functions of the
`pedometer 1 are re-activated ("power on", block 430).
`The control unit 5 acquires from the inertial sensor 3 a
`number of samples of the acceleration signal Az sufficient for
`estimating an updated mean value A' (block 440), which is
`then compared with the current mean value Azm previously
`stored (block 450).
`If the updated mean value A? departs from the current
`mean value Azm (output NO from block 450), the surveying
`procedure is interrupted, and the first counting procedure
`indicated in block 110 of FIG. 3 is executed. If, instead, the
`updated mean value AY is substantially unvaried with
`respect to the current mean value A
`(output YES from
`block 450), the surveying procedure proceeds and the pedom-
`eter 1 is set again in the low-consumption operating condition
`(block 410).
`Clearly, the use of the surveying procedure enables a dras-
`tic reduction in the power consumption when the pedometer
`1 is not used and, hence increases the autonomy thereof. If, as
`in the embodiment described, the pedometer 1 is integrated in
`a portable device with which it shares the use of resources, for
`example the control unit 5, the surveying procedure entails
`further advantages. In fact, the de-activation of the functions
`linked to the pedometer 1 frees the shared resources for use by
`
`8
`the active functions, which can thus access the resources
`themselves in a more efficient way.
`Finally, it is evident that modifications and variations can
`be made to the device described herein, without thereby
`5 departing from the scope of the present invention, as defined
`in the annexed claims.
`In particular, the control procedure described can be used
`to advantage in a stand-alone pedometer or in any case one
`integrated in a further portable device, but with stand-alone
`io and non-shared resources.
`Furthermore, the conditions of regularity used to enable or
`prevent counting of the steps recognized can be different from
`the ones described. For example, a sequence of steps can be
`considered regular when possible steps recognized and not
`15 validated are separated by at least one pre-determined number
`of consecutive validated steps. Again, a sequence of a pre-
`determined number of validated or non-validated steps (se-
`quence of fixed length) can be considered regular when the
`validated steps are at least a given percentage of the steps of
`20 the sequence.
`Finally, the inertial sensor can be of the type with two or
`three axes of detection. In this case, step recognition can
`advantageously be performed by selecting the acceleration
`signal corresponding to the detection axis nearest to the ver-
`25 tical. The nearer the detection axis used is to the vertical, in
`fact, the greater the amplitude of the signal useful for step
`recognition. The detection axis is selected on the basis of the
`value of the DC component of the respective acceleration
`signal, which is correlated to the contribution of the accelera-
`30 tion of gravity. The detection axis nearest to the vertical is the
`axis along which the contribution of the acceleration of grav-
`ity is greater. The pedometer can then be used independently
`of how it is oriented.
`The invention claimed is:
`1. A method for controlling a pedometer, the method com-
`prising:
`generating a signal correlated to movements of a user of the
`pedometer;
`detecting steps of the user based on said signal;
`checking whether sequences of the detected steps indicate
`w