Haircare Appliance

The present invention relates to a haircare appliance.

A haircare appliance may comprise a sensor for sensing the proximity of an object to the haircare appliance. However, the sensor is typically incapable of discriminating between hair and other objects.

According to a first aspect of the present invention, there is provided a haircare appliance comprising: a body for engaging hair in use; a sensor arrangement configured to output a plurality of signals, each signal being indicative of a presence of an object at a respective region of the body; and a control unit configured to determine whether the object is hair based on temporal differences between the signals.

This may allow for a reliable determination of whether an object engaging a body of a haircare appliance is hair. For example, hair detection based only on a signal associated with a single region of the body may be prone to returning false positives, i.e. prone to returning a determination that hair is present when in fact some other object is present, such as a table or a user's hand. Sensing the presence of an object at multiple regions of the body independently may not significantly improve this situation. However, by determining whether an object is hair based on temporal differences between a plurality of signals each of which being indicative of a presence of an object at a respective region of the body, such false positives can be reduced or eliminated, i.e. it can be more reliably determined whether the object is hair. In particular, rather than a static evaluation of whether an object is hair, use of the temporal differences between the signals allows for the dynamics of the engagement of the object with the body to be incorporated into the determination. Since the dynamics of the engagement of hair with the body may be relatively specific, for example follow certain paths and/or move relative to the body at certain rates or with certain other characteristics, the specificity and reliability with which hair presence at the body can be determined may accordingly be improved. Improving the reliability of a determination of whether an object engaging a body of a haircare appliance is hair is in itself an advantage, but this may also, in turn, allow for numerous other benefits to be provided for, such as improved styling, improved energy efficiency, and improved product safety, as described in more detail below.

As mentioned, each signal is indicative of a presence of an object at a respective region of the body. In examples, each signal may be binary. For example, each signal may be logically low when an object is not present at the respective region and may be logically high when an object is present at the respective region. In other examples, each signal may be non-binary, for example have a range or continuum of values. For example, the magnitude of amplitude of each signal may indicate the proximity of the object to the respective region or otherwise a degree to which the object is present at the respective region. In some examples, an object being present at a region may correspond to the object being in contact with the region. In other examples, the object being present at a region may correspond to the object being near or proximate to the region.

In some examples, a first signal of the plurality of signals changes in response to an object present at a first region of the body; a second signal of the plurality of signals changes in response to an object present at a second region of the body; and the control unit is configured to determine whether the object is hair based on a temporal difference between the changes in the first signal and the second signal.

It will be appreciated that, in some examples, more than two signals may be used. For example, each signal of the plurality of signals may change in response to an object present at a respective region of the body; and the control unit may be configured to determine whether the object is hair based on the temporal differences between the changes in the plurality of signals. For example, when an object becomes present at a region, the respective signal may increase in response. The dynamics with which hair becomes present at different regions of the body when the haircare appliance is in use may be reliably differentiated from that of other objects. Accordingly, basing the determination on temporal differences between changes that occur in response to an object being present at the respective regions may allow for reliable and/or robust determination of whether the object is hair.

In some examples, the control unit is configured to determine, based on changes in the signals, a sequence in which an object becomes present at different respective regions of the body, and the control unit is configured to determine whether the object is hair based on the determined sequence. For example, the sequence in which the object becomes present at different respective regions may be inferred from the sequence in which the changes in the signals for the respective regions occur. In some examples, the sequence may comprise an order in which an object becomes present at different regions and/or a time between an object becoming present at different regions. Determining whether the object is hair based on the determined sequence may allow for the way in which hair is engaging with the body in use, and specifically the movement of hair relative to the body, to be incorporated into the determination. This may accordingly improve the reliability of the determination.

In some examples, the control unit is configured to determine that the object is hair responsive to the determined sequence corresponding to a predefined sequence. For example, the predefined sequence may be a sequence that has been measured or otherwise observed to occur when hair engages the body of the haircare appliance in use. In some examples, there may be a plurality of predefined sequences. Determining that the object is hair responsive to the determined sequence corresponding to a predefined sequence may allow, for example, for one or more ways in which hair is expected to engage or interact with body when the haircare appliance is in use to be incorporated into the determination. For example, if an object is engaging the body in a way that has been observed or is otherwise expected when the body engages hair, then the chances of the object being hair are relatively high, whereas if an object is engaging with the body in a way that is different to what has been observed or is otherwise expected when the body engages hair, then the chances of the object being hair are relatively low. This may reduce the possibility that the object is determined to be hair when the haircare appliance is in fact not engaging hair, and hence improve the reliability of the hair presence determination.

In some examples, the body comprises a curved portion and the regions are distributed so as to follow a curve of the curved portion. Certain haircare appliances such a styling wands, wraps and brushes may have a curved portion over or around which hair may be engaged in use. Hair may be formed of numerous relatively flexible strands which may accordingly readily follow a curve of the curved portion in use. Accordingly, having regions for which signals are output distributed so as to follow a curve of the curved portion may allow for the wrapping motion of hair along the curve to be encoded into the determination of whether the object is hair. Similarly to as above, this may improve the reliability of the determination.

In some examples, the curved portion has the shape of a cylinder or cone and the regions are distributed around a circumference of the cylinder or cone. This may allow for the degree to which an object is wrapped, or whether an object is fully wrapped, around the body to be encoded into the determination of whether the object is hair. Similarly to as above, this may improve the reliability of the determination.

In some examples, the sensor arrangement comprises a plurality of sensors, each sensor being located at a respective region of the body and being configured to output one of the plurality of signals. This may provide for a robust way to provide the plurality of signals. For example, each sensor may output a signal independently of the other sensors. This may for example allow a clear and reliable signal as to the presence of an object at each region. Moreover, this may allow for the determination as to whether the object is hair to be carried to be robust with respect to the failure of one of the sensors. In some examples, one or more of the sensors are non-contact sensors. For example, a non-contact sensor may be configured to output a signal indicative of an object being present at the sensor without the object necessarily coming into physical contact with the sensor itself. Examples of non-contact sensors include a capacitance sensor (e.g. either a self-capacitance sensor or a mutual capacitance sensor), a light sensor, and an ultrasonic sensor. In some examples, one or more of the sensors may be a contact sensor, i.e. one which is configured to output a signal indicative of an object being present at the sensor only when an object comes into physical contact with the sensor itself. Examples of contact sensors include touch sensors, force sensors, and pressure sensors. Another example of a contact sensor is an element coupled with a temperature sensor configured to measure the temperature of the element and whereby contact of an object with the element causes a change in the temperature of the element and thereby a change in a signal output by the temperature sensor. Other types of sensor may be used.

Non-contact sensors allow for determination of whether an object is hair even when the object is not physically touching the sensors. However, a haircare appliance using non-contact sensors to detect the presence of hair, and in particular one in which those sensors are positioned so as to sense into free space, may be particularly prone to false positives as hair could be erroneously detected even when the body is not touching any object. Accordingly, the present invention may be particularly effective in improving the reliability of hair presence detection for a haircare appliance which uses non-contact sensors.

In some examples, each signal comprises a temporal series of values or the control unit samples each signal as a temporal series of values, and each of the values is indicative of the degree to which an object is present at the respective region at a given time. This may allow for the determination to be readily carried out by a microcontroller or other computer processor, which may provide for a cheap, simple, flexible and/or low weight means by which to determine whether an object is hair. This is for example as compared to using hardwired circuit logic, which may nonetheless in some other examples be used.

In some examples, the control unit is configured to determine whether the object is hair based on temporal differences between the plurality of signals that occur within a given time window. This may allow the determination of whether the object is hair to be made for a specific time period. This may allow, for example, for the determination to be made periodically and/or repeatedly at successive times, which may in turn allow for the reliable updating of the determination as time progresses. Where each signal comprises or is sampled as a temporal series of values, the control unit may be configured to determine whether the object is hair based on values of the plurality of signals that occur within the given time window.

In some examples, the control unit is configured to: between successive determinations of whether the object is hair, move the given time window so as to include the most recent values and remove the oldest values of the plurality of signals. This may allow for the determination of whether the object is hair to be made in an efficient manner. For example, by including only the most recent values of the plurality of signals, the control unit can provide up-to-date hair presence determinations in relatively quick succession, while the processing and memory resource consumption by the control unit can be kept at or below an appropriate level.

In some examples, the control unit is configured to determine whether the object is hair using a trained machine learning model. For example, the trained machine learning model may be or comprise a regression model. In some examples, the machine learning model may be a trained neural network, although other trained machine learning models may be used. In some examples, the control unit may be configured to: input the plurality of signals into the trained machine learning model to obtain an output, and determine whether the object is hair based on the output. For example, the trained machine learning model may have been trained to, based on an input of a plurality of such signals, output a determination of whether an object is hair. For example, the machine learning model may have been trained based on a training data set comprising a plurality of training data samples, each training data sample comprising a plurality of such signals and a label indicating whether an object associated with the plurality of signals is hair. In some examples the training data set may comprise positive training data samples comprising a plurality of such signals and a label indicating that the object associated with the plurality of signals is hair, and negative training data samples comprising a plurality of such signals and a label indicating that the object associated with the plurality of signals is not hair.

Using a trained machine learning model may allow for a reliable and/or flexible determination of whether the object is hair. For example, this may be as compared to applying a hard-coded algorithm to the plurality of input signals to determine whether the object is hair. For example, such a hard-coded algorithm or set of rules would be inflexible with respect to timings of changes in signals, which timings were not contemplated when the rules were written. However, the trained machine learning model may generalise from training samples on which it has been trained, and hence be more flexible with respect to such uncontemplated timings. As another example, using a hard-coded algorithm or set of rules would require the set of rules to be written, which is not only labour intensive but may necessarily involve assumptions on the way in which hair interacts with the body of the haircare appliance and may not account for the precise way in which hair interacts with the body. On the other hand, using a machine learning model, which for example may have been trained using training data including actual signal values for when it is known hair is engaging with the body (and e.g. for when hair is not engaging with the body, and/or for when the body is engaging something other than hair, such as a hand), may automatically encode the precise way in which hair actually interacts with the body in use, and hence may allow for a more reliable determination of whether the object is hair.

In some examples, the control unit is configured to: concatenate the values of each of the plurality of signals included in the given time window with one another, thereby to obtain an input for the machine learning model. This may allow for the time series of data to be converted into a column vector, which in turn may allow for the ready input of the plurality of signals in a given time window into a machine learning model, for example a neural network having an input layer in this format.

In some examples, the control unit is configured to control an operating mode of the haircare appliance in response to the determination. This may allow the haircare appliance to operate more precisely with respect to whether hair is engaging the body of the haircare appliance. For example, this may allow for improved styling or other functionality, improved energy saving, and/or improved product safety. For example, controlling the haircare appliance to operate in a styling mode (which may e.g. involve heating to a certain temperature and/or an airflow being expelled at a certain flow rate) precisely when hair is engaged with the body may allow for the more precise styling of hair. As another example, operating the appliance in an idle mode when it is not determined that hair is present and operating the appliance in a styling mode when it is determined that the object is hair, may allow for power consumption to be reduced as compared to the appliance being in the styling mode even when not engaging hair. As another example, operating the appliance at a lower temperature when it is determined the object is not hair and operating the appliance at a higher temperature when it is determined the object is hair, may allow for the reduction of risk of a user accidently burning themselves or other objects e.g. when the appliance is not actively being used to style hair. As another example, operating the appliance to expel or generate an airflow having a lower flow rate when it is determined the object is not hair and operating the appliance to expel or generate an airflow having a higher flow rate when it is determined the object is hair, may allow for the reduction of risk of a user inadvertently directing airflow at an unintended object.

In some examples, the control unit may comprise a controller that is configured to both determine whether the object is hair and, in response, control the operating mode of the haircare appliance. In other examples, the control unit may comprise a first controller configured to determine whether the object is hair, a second controller configured to control the operating mode of the haircare appliance. In this latter case, the first controller may be configured to output, for example to the second controller, a signal indicative of the result of the determination of whether the object is hair.

In some examples, the haircare appliance expels an airflow, and the control unit is operable to control one or more of a flow rate and a temperature of the airflow in response to the determination. Similarly to as mentioned above, this may allow for improved styling or other functionality, improved energy efficiency, and/or improved product safety.

In some examples, the control unit is configured to: operate the haircare appliance in a first mode in response to determining that the object is not hair; and operate the haircare appliance in a second mode in response to determining that the object is hair, wherein operation in the first mode consumes a lower electrical power than operation in the second mode. Similarly to as mentioned above, this may allow for improved energy efficiency. Improving energy efficiency (e.g. reducing overall energy consumption) may be particularly important in battery operated appliances, which have a limited energy storage capacity. As such these features may allow for an improved run-time of the device, and/or for a smaller battery to be used, which may reduce the weight of the appliance.

In some examples, the haircare appliance comprises a heater, and the control unit is configured to operate the heater at a first temperature in the first mode and to operate the heater at a second temperature in the second mode, the second temperature being greater than the first temperature. Similarly to as mentioned above, this may allow for improved styling or other functionality, improved energy efficiency, and/or improved product safety.

In some examples, the haircare appliance comprises an air inlet, an air outlet, and an air flow generator for generating an airflow from the air inlet to the air outlet, and the control unit is configured to operate the airflow generator at a first flow rate in the first mode and to operate the airflow generator at a second flow rate in the second mode, the second flow rate being greater than the first flow rate. Similarly to as mentioned above, this may allow for improved styling or other functionality, improved energy efficiency, and/or improved product safety.

According to a second aspect of the present invention, there is provided a control unit for a haircare appliance, the control unit configured to: receive a plurality of signals, each signal being indicative of a presence of an object at a respective region of a body of the haircare appliance; and determine whether the object is hair based on temporal differences between the signals. This may allow for similar advantages to as described above for the first aspect. In some examples the control unit may be the same as described above with reference to the first aspect. In some examples, the control unit may be located in a main part, for example a handle, of a haircare appliance. In some examples, the signals may be received from an attachment, for example an attachment that includes the body, which attachment is attachable to and/or detachable from the main part. This may allow for improved flexibility in the functioning of the haircare appliance. The signals may be received wirelessly or over wires or conductive tracks, for example.

According to a third aspect of the present invention, there is provided a method of determining whether hair is present at a body of a haircare appliance, the method comprising: receiving a plurality of signals, each signal being indicative of a presence of an object at a respective region of the body; and determining whether the object is hair based on temporal differences between the signals. This may allow for similar advantages to as described above for the first aspect. In some examples, the method may be performed by a control unit of a haircare appliance. In other examples, the method may be performed by a separate entity for example remotely from the haircare appliance. For example, the signals may be transmitted to and received by a remote processor, and the determination may be made by the remote processor. In some examples, the result of the determination may be transmitted to and received by the haircare appliance, which may in turn, for example, control an operating mode on the basis of the determination.

As used herein, like reference numerals denote like features.

Referring to, there is illustrated a haircare applianceaccording to an example.

The haircare appliancecomprises a bodyfor engaging hair in use (hair is not shown inbut see e.g. the hair tressindescribed below). The bodycomprises a plurality of regions A, B, C. The haircare appliancecomprises a sensor arrangement. The sensor arrangementis configured to output a plurality of signals, each signal being indicative of a presence of an object at a respective region A, B, C of the body(an object is not shown in, but see e.g. the hair tressindescribed below). The haircare appliancecomprises a control unit. As described in more detail below, the control unitis configured to determine whether the objectis hair based on temporal differences between the signals.

In this example, the haircare appliancecomprises a main partand the bodyis provided as an attachmentthat is attachable and/or detachable from the main part. The main partcomprises an airflow generatorand a heater. The airflow generatoris configured to generate an airflow from an air inletto an air outlet. In this example, the air inletis in the main part, and the air outletis in the body. The heateris located downstream of the airflow generator. The heatermay be configured to heat the airflow so as to provide a heated airflow at the air outlet. The bodyof this example is shown in more detail in. In this example, the bodyis generally in the form of a cylinder or a cone. In this example, each region A, B, C corresponds to an elongate member. In this example these regions/members A, B, C are distributed around a circumference (seein) of the cylinder or cone forming the body. In this example, an air outletis provided by a gap in between each member A, B, C which directs airflow across the surface of an adjacent member A, B, C. As is known per se, in use, the airflow expelled from these air outletscauses, via the Coanda effect, hair, or more specifically a hair tress, in the vicinity of the bodyto wrap around the body, where it can be styled. In some examples, the control unitmay be configured to control an operating mode of the haircare appliance, for example control one or more of a flow rate and a temperature of the airflow, for example control a flow rate of the airflow generated by the airflow generatorand/or control a temperature at which the heateris operated. In some examples, the control unitmay be configured to control the operating mode in response to the determination that the objectis hair. As explained in more detail below, this may allow for improved styling, improved energy efficiency, and/or improved product safety of the haircare appliance.

As mentioned, the haircare appliancecomprises a sensor arrangementconfigured to output a plurality of signals, each signal being indicative of a presence of an object at a respective region A, B, C of the body. In this example, the sensor arrangementcomprises a plurality of sensors,,, each sensor,,being located at a respective region A, B, C of the bodyand being configured to output one of the plurality of signals. In this example, each sensor,,is positioned so as to sense radially outwardly from the body(i.e. radially outwardly of the region A, B, C to which the sensor corresponds). In some examples, one or more of the sensors,,may be non-contact sensors. For example, a non-contact sensor may be configured to output a signal indicative of an objectbeing present at the sensor,without the objectnecessarily coming into physical contact with the sensor,,itself. Examples of non-contact sensors include a capacitance sensor (e.g. either a self-capacitance sensor or a mutual capacitance sensor), a light sensor, and an ultrasonic sensor. Non-contact sensors may allow for determination of whether an object is hair even when the object is not physically touching the sensors. In other examples, one or more of the sensors,,may be a contact sensor, i.e. one which is configured to output a signal indicative of an objectbeing present at the sensor,,only when an object comes into physical contact with the sensor itself. Examples of contact sensors include touch sensors (e.g. capacitance sensors or resistance sensors), force sensors, and pressure sensors. Another example of a contact sensor is an element coupled with a temperature sensor configured to measure the temperature of the element and whereby contact of an object with the element causes a change in the temperature of the element and thereby a change in a signal output by the temperature sensor. Other types of sensor may be used. In some examples, one or more of the sensors,,may be any sensor whose output is affected by the presence of hair, which allows for a high flexibility for product integration. For example, where the regions are heated, a sensor resistant to high temperature may be chosen (such as the heated element coupled with a temperature sensor described above).

In some examples, each signal may be non-binary, for example have a range or continuum of values. For example, the magnitude or amplitude of each signal may indicate the proximity of the objectto the respective region A, B, C or otherwise a degree to which the objectis present at the respective region A, B, C. In other examples, each signal may be binary. For example, each signal may be logically low when an objectis not present at the respective region A, B, C and may be logically high when an objectis present at the respective region A, B, C. In some examples, the objectbeing present at a region A, B, C may correspond to the objectbeing near or proximate to the region A, B, C. In other examples, an objectbeing present at a region may correspond to the objectbeing in contact with the region A, B, C.

The sensor arrangementmay be connected to the control unitby wired or wireless means, in order to provide the plurality of signals to the control unit. In examples where the bodyis provided by an attachment and the control unitis in the main part, the bodyand the main partmay form a connection, which may be wired or wireless, over which the signals may be transmitted from the sensor arrangementto the control unit. In some examples, the sensor arrangementmay comprise the sensors,,and the signals output by each sensor may be provided directly to the control unitas the plurality of signals. In other examples, the sensor arrangementmay comprise the sensors,,and a multichannel reader (not shown), and the plurality of signals may be provided to the control unitby the multichannel reader (not shown). For example, the multichannel reader may sample each of the signals output by the sensors as a temporal series of values and provide these temporal series of values to the control unitas the plurality of signals. In any case, the sensor arrangementis configured to output a plurality of signals, each signal being indicative of a presence of an object at a respective region A, B, C of the body.

Referring to,illustrates the way in which hair, or more specifically a hair tress, may wrap around the cylindrical bodyof the haircare applianceaccording to the example ofat different successive times T, T, T, T; andillustrates a plot of the plurality of signals sA, sB, sC, sD, sE, sF output by the sensor arrangementfor a respective one of the plurality of regions A, B, C, D, E, F of the body, as a function of time, according to an example. It is noted that regions D, E, F (and their associated sensors) are present on the bodyillustrated in, but are not visible in. As can be seen from, the regions A to F are distributed around the circumference of the cylindrical bodyin the order A, B, C, D, E, F. In this example, the larger the degree to which an objectis present at a given one of the regions A-F, the larger the signal sA-sF that is output corresponding to that given region A-F.

At time T, the hair tressis present at regions A and B, but not at regions C to F. Accordingly, at time T, the values of the signals sA and sB have increased, but the those of the other signals sC to sF have not. At a later time T, the hair tresshas wrapped further around the body, and is now present at regions A to C, but not at regions D to F. Accordingly, at time T, the values of the signals sA to sC have increased, but those of the other signals sD to sF have not. At a still later time T, the hair tresshas wrapped still further around the body, and is now present at regions A to D, but not at regions E and F. Accordingly, at time T, the values of the signals sA to sD have increased, but those of the other signals sE to sF have not. At a yet still later time T, the hair tresshas wrapped yet still further around the body, and is now present at regions A to F. Accordingly, at time T, the values of the signals sA to sF have all increased.

As mentioned, the control unitis configured to determine whether the objectis hair based on temporal differences between the signals sA-sF output by the sensor arrangement. Temporal differences may be considered as differences between the signals that occur over a period of time and may for example be contrasted with purely static or instantaneous differences. For example, as illustrated above in, each signal of the plurality of signals sA-sF may change (e.g. a value of the signal may increase) in response to an objectbeing present at a respective region A, B, C of the body. The control unitmay be configured to determine whether the objectis hair based on the temporal differences between these changes in the plurality of signals sA-sF. For example, the temporal difference between two signals may comprise or be based on a time at which a change in one signal (e.g. an increase in the value of one signal) occurs relative to a time at which a change in the other signal (e.g. an increase in the value of the other signal) occurs. For example, the temporal difference between signals may comprise differences in the time at which each change occurred relative to the other changes (which may for example be expressed as a sequence or time-order in which the changes in signals occur), a time in between each change and/or other time-related differences. For example, based on the sequence in which the changes in the signals occur, the time in between each change and/or other temporal differences between the signals, the control unitmay determine whether the objectis hair. For example, when the temporal differences between the signals output by the sensor arrangementcorrespond to temporal differences known or expected to occur when hair is engaging the body(e.g. as illustrated in), the control unitmay determine that the objectis hair. For example, when changes in the signals correspond to those which would occur when an object wraps around the circumferenceof the body, the control unitmay determine that the object is hair. It is noted that in the schematic diagram ofthe signals sA to sF are shown as increasing to a similar value but it will be appreciated that this need not necessarily be the case and that the signals may in principle increase by different amounts but that the control unitmay nonetheless determine whether the object is hair based no temporal differences between the signals.

Use of the temporal differences between the signals allows for the dynamics of the engagement of the objectwith the bodyto be incorporated into the determination by the control unitof whether the objectis hair. Since the dynamics of the engagement of hairwith the bodymay be relatively specific, for example follow certain paths and/or move relative to the bodyat certain rates or with certain other characteristics, the specificity and reliability with which hair presence at the bodycan be determined may accordingly be improved. Improving the reliability of a determination of whether an objectengaging a body of a haircare appliance is hair is in itself an advantage, but this may also, in turn, allow for numerous other benefits to be provided for, such as improved styling, improved energy efficiency, and improved product safety, as described in more detail below.

As mentioned, in some examples, each signal of the plurality of signals sA-sF may change in response to an objectpresent at a respective region A, B, C of the body; and the control unitmay be configured to determine whether the objectis hair based on the temporal differences between the changes in the plurality of signals. The dynamics with which hairbecomes present at different regions A, B, C of the body when the haircare applianceis in use may be reliably differentiated from that of other objects. Accordingly, basing the determination on temporal differences between changes in the signals that occur in response to an object being present at the respective regions A, B, C may allow for reliable and/or robust determination of whether the object is hair.

In some examples, the control unitis configured to determine, based on changes in the signals, a sequence in which an objectbecomes present at different respective regions A-F of the body, and the control unitis configured to determine whether the objectis hair based on the determined sequence. For example, the sequence in which the objectbecomes present at different respective regions may be inferred from the sequence in which the changes (e.g. increases) in the signals for the respective regions A-F occur. For example, a change (e.g. an increase) in the signal for respective regions A, B, C, D may occur at times tA, tB, tC, tD respectively, and for example in the case that tD>tC>tB>tA, it can be inferred that an objectbecomes present at the respective regions in the order A, B, C, D. Moreover from the difference in time between tA, tB, tC, tD the difference in time between the object becoming present at the regions A, B, C, D respectively can be inferred.

In some examples, an object may be determined as becoming present at a given region A, B, C, D when the corresponding signal sA, sB, sC, sD increases to above a threshold value.

In some examples, the sequence in which an objectbecomes present at different respective regions A-F of the bodymay comprise an order in which the objectbecomes present at different regions A, B, C, D and/or a time between an objectbecoming present at different regions A, B, C, D. Determining whether the object is hair based on the determined sequence may allow for the way in which hair is engaging with the bodyin use, and specifically the movement of hairrelative to the body, to be incorporated into the determination of whether the object is hair. This may accordingly improve the reliability of the determination.

In some examples, the control unitis configured to determine that the objectis hair responsive to the determined sequence corresponding to a predefined sequence. For example, the predefined sequence may be a sequence that has been measured or otherwise observed to occur when hairengages the bodyof the haircare appliancein use. In some examples, there may be a plurality of predefined sequences. Determining that the object is hair responsive to the determined sequence corresponding to a predefined sequence may allow, for example, for one or more ways in which hair is expected to engage or interact with bodywhen the haircare applianceis in use to be incorporated into the determination of whether the object is hair. For example, if an objectis engaging the bodyin a way that has been observed or is otherwise expected when the bodyengages hair, then the chances of the objectbeing hair are relatively high, whereas if an object is engaging with the bodyin a way that is different to what has been observed or is otherwise expected when the bodyengages hair, then the chances of the object being hair are relatively low. This may reduce the possibility that the object is determined to be hair when the haircare applianceis in fact not engaging hair, and hence improve the reliability of the hair presence determination. For example, the predefined sequence may comprise an order in which hair becomes (and e.g. remains) present at different regions A, B, C and/or a time in between an object becoming present at different respective regions. For example, a predefined order for the bodyof the example ofmay be A, B, C, D, E, F (or indeed B, C, D, E, F, A; C, D, E, F, A, B; D, E, F, A, B, C; E, F, A, B, C, D; F, A, B, C, D, E; or part of any one of those sequences) which follows in sequential order the path that hair is expected to follow when it wraps around the body. As another example, the predefined time in between the object becoming present at different regions may be the condition that tB occurs within a predetermined time range relative to tA, and tC occurs within a predetermined time range relative to tB etc.; or for example that tA, tB, and tC all occur within a certain time of one another.

In some examples, each signal comprises a temporal series of values or the control unitmay sample each signal as a temporal series of values, and each of the values is indicative of the degree to which an objectis present at the respective region A-F at a given time. This may allow for the determination by the control unitto be readily carried out by a microcontroller or other computer processor, which may provide for a cheap, simple, flexible and/or low weight means by which to determine whether an object is hair. This is for example as compared to using hardwired circuit logic, which as mentioned below may nonetheless in some other examples be used.

In some examples, the control unitmay be configured to determine whether the objectis hair by applying an algorithm to the plurality of signals. In some examples, the algorithm may be implemented by a computer program executing on a processor such as a microcontroller. In other examples, the algorithm may be implemented by control logic circuitry. In either case algorithmic logic may be applied to the plurality of signals to determine whether or not the object is hair based on temporal differences between the signals. For example, the signals associated with respective regions A, B, C may be monitored and a change (e.g. an increase) in the signal value may be observed at respective times tA, tB, tC. A logic may be applied that if these changes occurred in a certain temporal order or sequence (e.g. tA>tB>tC, or tC>tB>tC) and/or the difference in time between the times tA and tB, and/or tB and tC is within a certain range, then the object is determined as hair, otherwise the object is determined as not hair. Other algorithmic logic may be applied.

In some examples, the control unitmay use other means to determine whether the objectis hair based on temporal differences between the signals. For example, the determination of whether the objectis hair may be made using a trained machine learning model, e.g. by inputting the plurality of signals into a trained machine learning model which outputs the determination based on the input signals. For example, values representing each of the plurality of signals as a function of time may be input together into the machine learning model, which may, using an inferred function derived from its training, map these input values onto a determination of whether or not the object is hair.

In some examples, the control unitis configured to determine whether the object is hair based on temporal differences between the plurality of signals that occur within a given time window (see e.g. the time windowof). This may allow the determination of whether the object is hair to be made for a specific time period. This may allow, for example, for the determination to be made periodically and/or repeatedly at successive times, which may in turn allow for the reliable updating of the determination as time progresses. Where each signal comprises or is sampled as a temporal series of values, the control unitmay be configured to determine whether the object is hair based on values of the plurality of signals that occur within the given time window. As an example, the time windowmay be on the order of 1 or 2 seconds, for example 1.5 seconds. Each signal may comprise or be sampled as a temporal series of values where the time between each value in the series is 125 ms. In the example where the time windowis 1.5 second, the time windowwill accordingly include the signal values of each of the plurality of signals at 13 different moments in time.

In some examples, the control unitis configured to: between successive determinations of whether the objectis hair, move the given time windowso as to include the most recent values and remove the oldest values of the plurality of signals. This may allow for the determination of whether the objectis hair to be made in an efficient manner. For example, by including only the most recent values of the plurality of signals, the control unit can provide up-to-date hair presence determinations in relatively quick succession, while the processing and memory resource consumption by the control unitcan be kept at or below an appropriate level. As an example, every 125 ms the oldest values of the signals are removed and the newest values of the signals are added. Between each successive determination, the time windowmay be moved on by 125 ms. Other example time windows and sampling rates may be used.

In examples where each signal comprises or is sampled as a temporal series of values, and where the control unitis configured to determine whether the objectis hair using a trained machine learning model, the control unitmay be configured to concatenate the values (of the temporal series of values) of each of the plurality of signals sA-sF included in the given time windowwith one another, thereby to obtain an input for the machine learning model. This may allow for the time series of data to be converted into a column vector, which in turn may allow for the ready input of the plurality of signals in a given time window into a machine learning model, for example a neural network having an input layer in this format.

For example, a column vector may be formed where the first six entries or elements are the signal values for each of the six signals sA-sF at the earliest sample time included in the given time window, the following six entries are the signal values for each of the six signals sA-sF at the second earliest sample time included in the given time window, and so on until the last six entries are the signal values for each of the six signals aS-sF at the latest sample time included in the given time window. In the case of the given time windowincluding 13 sample times, this would result in a column vector with 78 elements. Referring to, there is a shown a plotillustrating the signal value of each element of the column vector as a function of element number or position in the column vector, according to an example. As can be seen, the values, and their position within the column vector, encode the changes in the signals sA-sF occurring as a function of time within the given time window, and as such encode the temporal differences between those signals. The conversion of the plurality of signals into this format may provide for the ready input of the plurality of signals into a machine learning model, for example a neural network having an input layer in this format (e.g. a neural network having 78 nodes in its input layer).