System and method for detection on a stationary surface

Abstract

The invention relates to a system and method for detection on a stationary surface. A method includes generating, by a first motion sensor of an electronic device, first sensor data over an acquisition time window, and determining whether the electronic device is located on a stationary surface, a semi-stationary surface, or in motion by processing the first sensor data to generate first process data. The method further includes determining whether the electronic device is in a steady state based on the first process data, the steady state indicating whether the electronic device is held on the stationary surface for a first predefined time. And, the method further includes stopping processing the first sensor data in response to determining that the electronic device has been in the steady state for a second predefined time.

Description

[0001] CROSS-REFERENCE TO RELATED APPLICATIONS

[0002] This application is a continuation-in-part of U.S. application No. 18 / 357,851, filed on July 24, 2023, which is hereby incorporated by reference into this document. Technical Field

[0003] The present disclosure relates generally to electronic devices and, in particular embodiments, to systems and methods for determining whether an electronic device is located on a stationary surface. Background Art

[0004] As electronic devices become more common and individuals become more mobile, there is a growing need to provide computing power and information anytime, anywhere. Laptops, tablet devices, and wearable electronic devices (hereinafter individually and collectively referred to as "electronic devices") can at least partially meet this need.

[0005] One problem that users often encounter when using electronic devices is high power consumption and / or poor heat dissipation, which is often manifested as heating of the electronic device. The power supply component of the electronic device may be located on the bottom surface of the electronic device (e.g., the surface below the keyboard portion of a laptop computer). During extended use or high-intensity use (e.g., during gaming), if the electronic device is in physical contact with the user (e.g., the user's knee or wrist), the base of the electronic device may overheat, burn out, or cause discomfort to the user.

[0006] In addition to the potential for harm to human skin, high temperatures in electronic devices can also adversely affect the batteries that power the electronic devices. Although batteries can operate over a wide temperature range, charging or discharging batteries when the electronic device is at high temperatures can reduce charge acceptance and shorten battery life. For example, charging or discharging lithium polymer (LiPo) batteries at high temperatures can result in the generation of gases, which can cause cylindrical batteries to vent and cause pouch cells to swell. Further, high temperatures can adversely affect the life of integrated circuits in electronic devices (e.g., provided on a printed circuit board (PCB) or implemented as a silicon-on-chip (SoC)), especially when such integrated circuits are exposed to high operating temperatures for long periods of time.

[0007] In the past, heat sinks, fans, or vents could be used to remove heat from the body of an electronic device. However, as more functionality is added to a PCB or SoC, heat becomes a more important consideration at the silicon level. Effective methods may be needed to detect whether an electronic device is on a stationary or stable surface (e.g., on a table or in a drawer) to optimize power consumption and / or heat dissipation of components within the electronic device. Summary of the invention

[0008] According to an embodiment of the present disclosure, a method includes generating first sensor data by a first motion sensor of an electronic device over an acquisition time window, and generating first process data by processing the first sensor data to determine whether the electronic device is located on a stationary surface, a semi-stationary surface, or in motion. The method also includes determining whether the electronic device is in a stable state based on the first process data, the stable state indicating whether the electronic device has remained on the stationary surface for a first predefined time. And the method also includes stopping processing of the first sensor data in response to determining that the electronic device has been in a stable state for a second predefined time.

[0009] And according to another embodiment of the present disclosure, an electronic device includes: a first motion sensor configured to generate first sensor data indicating a first type of movement of the electronic device, and a processor coupled to a memory, the memory storing instructions to be executed in the processor. When the instructions are executed, the processor generates first sensor data over an acquisition time window using the first motion sensor, and generates first process data by processing the first sensor data to determine whether the electronic device is located on a stationary surface. When the instructions are executed, the processor also determines whether the electronic device is in a stable state based on the first process data, the stable state indicating whether the electronic device has remained on the stationary surface for a first predefined time, and stops processing the first sensor data in response to determining that the electronic device has been in a stable state for a second predefined time. BRIEF DESCRIPTION OF THE DRAWINGS

[0010] For a more complete understanding of the present invention and its advantages, reference is now made to the following description in conjunction with the accompanying drawings, in which:

[0011] Figure 1 A block diagram of an electronic device including a detection system according to an embodiment is shown;

[0012] Figure 2 According to an embodiment, Figure 1 A method for extracting orientation-independent features from sensor data generated by a motion sensor of an electronic device;

[0013] Figure 3A According to an embodiment, Figure 1 sensor data generated by a motion sensor of an electronic device;

[0014] Figure 3B It shows that according to the embodiment Figure 3A a zoomed-in view of sampling times of first and second acquisition time windows of sensor data;

[0015] Figure 3C It shows that according to the embodiment Figure 3A The norm of the sensor data;

[0016] Figure 3D According to the embodiment, Figure 3A The first acquisition time window of the sensor data Figure 3A The norm of the sensor data;

[0017] Figure 4A and Figure 4B According to an embodiment, Figure 1 Mean-cross values ​​generated by different motion sensors of the electronic device for different states;

[0018] Figure 5A and Figure 5B shows the relative difference between the average crossover value and the variance for different states according to an embodiment;

[0019] FIG. 6A to FIG. 6C A block diagram is shown, which illustrates an implementation according to various embodiments Figure 1 The detection system and Figure 2 various ways of approach;

[0020] FIG. 7A to FIG. 7B Two different methods of determining stability of an on-table detection system output are illustrated according to an embodiment;

[0021] Figure 8 A block diagram is shown illustrating a method that may be used to place a table detection system into a sleep mode when an electronic device has been stable for a configurable amount of time in accordance with an embodiment;

[0022] Fig. 9 illustrates a flow chart of a method for operating a tabletop detection system according to an embodiment; and

[0023] Fig.10 A block diagram of an electronic device including a detection system according to an embodiment is shown.

[0024] Corresponding numerals and symbols in the different figures generally refer to corresponding parts unless otherwise indicated.The figures are drawn to clearly illustrate the relevant aspects of the embodiments and are not necessarily drawn to scale. DETAILED DESCRIPTION

[0025] The making and use of various embodiments are discussed in detail below. However, it should be understood that the various embodiments described herein are applicable to a variety of specific environments. The specific embodiments discussed only illustrate the specific ways to make and use the various embodiments, and should not be interpreted in a limited scope.

[0026] Various embodiments described herein are directed to effective systems and methods for determining whether an electronic device is located on a stationary or stable surface (e.g., on a stationary or stable inanimate surface, such as on a table or in a drawer). For example, such a determination can be used to optimize device performance, change the power consumption of the electronic device, and / or manage the heat dissipation of components within the electronic device. As an illustration, in various embodiments, in response to determining that the electronic device is on a stationary or stable surface (e.g., a table), the fan speed and clock frequency of electronic components (e.g., a central processing unit (CPU), a graphics processing unit (GPU), or a power supply unit) in the electronic device can be increased to achieve better performance (e.g., faster computing time); however, in response to determining that the electronic device is not on a stationary or stable surface (e.g., when the electronic device is in motion or on the user's lap), the clock frequency of the components in the electronic device can be reduced to reduce power consumption and avoid overheating of the components in the electronic device.

[0027] At the outset, it is noted that the embodiments described below are directed to systems and methods for determining whether an electronic device is located on a stationary or stable surface. The use of such determination results in electronic devices is given by way of illustration only, with examples being implementation of thermal strategies, energy conservation, and performance benchmarks. In general, the use of such determination results in controlling or altering the operation of an electronic device may be at the discretion of the manufacturer(s) of the electronic device and / or the manufacturer(s) of the electronic components of the electronic device.

[0028] As described below, the proposed method uses data from one or more motion sensors included in an electronic device. Although conventional systems and methods for determining whether an electronic device is located on a stationary or stable surface can use data from one or more motion sensors, such conventional systems and methods may have several disadvantages. For example, a motion sensor of an electronic device generates motion sensor data, and conventional systems and methods extract features from the motion sensor data that depend on the orientation of the motion sensor in the electronic device relative to multiple reference axes to determine whether the electronic device is located on a stationary or stable surface. In other words, conventional systems and methods rely on orientation-dependent features to make this determination. Exemplarily, conventional systems and methods can extract pitch, yaw, roll, and / or various acceleration components relative to a calibrated coordinate system or multiple reference axes (e.g., a three-dimensional coordinate system or a 6-axis system) from the motion sensor data, and then use such orientation-related features to determine whether the electronic device is located on a stationary or stable surface.

[0029] The use of such orientation-related features requires calibration of the motion sensors of the electronic device to reduce sensor offsets and biases (e.g., accelerometer offsets and / or gyroscope biases). Calibration is also required to generate a calibrated coordinate system or multiple reference axes, and such calibration ensures that the orientation-related features (e.g., pitch, yaw, roll, x-axis acceleration component, y-axis acceleration component, and / or z-axis acceleration component) accurately track the motion and / or orientation of the electronic device. Due to the use of orientation-related features, conventional systems and methods are not easily reconfigured or readjusted, may suffer from high latency and long convergence times (e.g., 10 seconds or longer), and have limited accuracy because such conventional systems and methods are susceptible to inter-device variations and orientation-based variations. Embodiment systems and methods are intended to avoid at least these shortcomings associated with conventional methods for determining whether an electronic device is located on a stationary or stable surface.

[0030] In general, the embodiment systems and methods described herein extract some (e.g., one or two) important features from motion sensor data, and such extracted features are orientation-independent. In other words, the features extracted from the motion sensor data do not rely on a calibrated coordinate system or multiple reference axes to obtain accuracy. In particular, the embodiment systems and methods rely on the average crossover value (explained in more detail below) and variance of the norm of the motion sensor data within each acquisition time window, and these features are orientation-independent. In addition, the embodiment systems and methods use machine learning methods to analyze the average crossover value and variance of the norm to determine whether the electronic device is located on a stationary or stable surface. In addition, the embodiment systems and methods use physical sensor data without complex processing methods (an example of this method is sensor fusion for attitude estimation, calibration, FFT, and complex filter chains). Due to the use of orientation-independent features, machine learning methods, and physical sensor data, the embodiment systems and methods have at least the following advantages: (1) easy to adjust or reconfigure; (2) have low latency and short convergence time (e.g., less than 10 seconds); (3) do not require calibration of motion sensors (thereby exhibiting immunity to device-to-device variations, accelerometer offsets, and / or gyroscope biases); and (4) have higher reliability than conventional systems and methods because orientation-independent features are used instead of orientation-dependent features.

[0031] Figure 1A block diagram of an electronic device 101 including a detection system 100 according to an embodiment is shown. The detection system 100 can be located within the electronic device 101, attached to the electronic device 101, or coupled to the electronic device 101. The detection system 100 of the electronic device 101 can be used to determine whether the electronic device 101 is on a stationary or stable surface (for example, on a table or in a drawer). As mentioned above, the electronic device 101 can be a laptop, a tablet device, or a wearable electronic device (for example, a smart watch, a mobile phone, a wireless headset, etc.). The detection system 100 includes a first motion sensor 102 and a first feature detection circuit 104 coupled to the output of the first motion sensor 102. The first feature detection circuit 104 is configured to determine one or more orientation-independent features from the output signal of the first motion sensor 102.

[0032] like Figure 1 As shown in , classification circuit 106 is coupled to the output of first feature detection circuit 104. Classification circuit 106 is configured to determine the state of electronic device 101 (e.g., assign a tag indicating whether electronic device 101 is located on a stationary or stable surface). Classification circuit 106 makes this determination based on the orientation-independent features determined by first feature detection circuit 104.

[0033] In some embodiments, detection system 100 may also include a second motion sensor 108 that measures different motion characteristics than first motion sensor 102. In such embodiments, second feature detection circuit 110 may be coupled to the output of second motion sensor 108. Similar to first feature detection circuit 104, second feature detection circuit 110 is configured to determine one or more orientation-independent features from the output signal of second motion sensor 108.

[0034] In an embodiment including a second motion sensor 108, the classification circuit 106 is configured to determine a state of the electronic device 101 (e.g., assigning a tag indicating whether the electronic device 101 is located on a stationary or stable surface), wherein such determination is made based on the orientation-independent features determined by the first feature detection circuit 104 and the orientation-independent features determined by the second feature detection circuit 110.

[0035] In some embodiments, the detection system 100 may also include a meta-classifying circuit 112 coupled to the output of the classification circuit 106. The meta-classifying circuit 112 may implement a time-based voting method that acts as a low-pass filter on the output of the classification circuit 106 to improve the overall accuracy of the detection system 100. Each component of the detection system 100 is described in further detail below.

[0036] The detection system 100 includes a first motion sensor 102, which may be an accelerometer of an electronic device 101. Figure 1 Only one first motion sensor 102 is shown in the figure, but multiple first motion sensors 102 may be included in the electronic device 101 (for example, two or more accelerometers placed at different positions of the electronic device 101). The electronic device 101 with the first motion sensor 102 may be a laptop computer, whose accelerometer is coupled or attached to the laptop computer base. As another example, the electronic device 101 with the first motion sensor 102 may be a tablet computer, whose accelerometer is included in the tablet computer. The first motion sensor 102 may be configured to sense vibration or acceleration of the electronic device 101 on each axis of motion. For example, the first motion sensor 102 may generate first sensor data 102x, 102y, 102z, which respectively indicate the vibration or acceleration of the electronic device 101 on the horizontal axis (for example, referred to as the "x-axis"), the longitudinal axis (for example, referred to as the "y-axis"), and the vertical or normal axis (for example, referred to as the "z-axis").

[0037] As will be made clear in the description below, the use of the first sensor data 102x, 102y, 102z from the first motion sensor 102 enables the embodiment system and method to determine whether the electronic device 101 is located on a stationary or stable surface. However, in other embodiments, the use of the second motion sensor 108 in combination with the first motion sensor 102 can improve detection. The second motion sensor 108 can be a gyroscope of the electronic device 101. To reiterate, the use of the second motion sensor 108 (and therefore the use of the data generated by the second motion sensor 108) is optional. For example, in a low-power or low-cost implementation of the embodiment system and method, the second motion sensor 108 (e.g., a gyroscope) and its data may not exist or be used by the classification circuit 106 to determine whether the electronic device 101 is located on a stationary or stable surface (e.g., on a table or in a drawer). The second motion sensor 108 can be configured to measure the rate at which the electronic device 101 rotates around each axis of motion. For example, the second motion sensor 108 may generate second sensor data 108x, 108y, 108z indicating rotation rates of the electronic device 101 around the x-axis, the y-axis, and the z-axis, respectively.

[0038] It is worth noting that the first sensor data 102x, 102y, 102z and the second sensor data 108x, 108y, 108z generated by the first motion sensor 102 and the second motion sensor 108, respectively, may depend at least in part on the placement or orientation of the electronic device 101. As an illustration, the electronic device 101 may be placed in an inclined plane, a horizontal plane, on a part of a human body (e.g., on a knee), or on an inanimate object (e.g., a table). The first sensor data 102x, 102y, 102z and the second sensor data 108x, 108y, 108z may indicate such placement or orientation of the electronic device 101. In addition, although the first feature detection circuit 104 and the second feature detection circuit 110 are Figure 1 Although shown as separate circuits, it should be noted that in some embodiments, a single detection circuit may implement both the first feature detection circuit 104 and the second feature detection circuit 110.

[0039] Figure 2 An embodiment method 200 is shown that may be performed by the first feature detection circuit 104 to extract or determine orientation-independent features from the first sensor data 102x, 102y, 102z. In other embodiments that optionally utilize a second motion sensor 108 (e.g., a gyroscope) in addition to the first motion sensor 102 (e.g., an accelerometer), the method 200 may also be performed by the second feature detection circuit 110 to extract or determine orientation-independent features from the second sensor data 108x, 108y, 108z. The following description is directed to an example in which the first feature detection circuit 104 performs the method 200; however, such description is equally applicable to the second feature detection circuit 110 in other embodiments that optionally utilize a second motion sensor 108 in addition to the first motion sensor 102.

[0040] In discussion Figure 2 In the details of method 200, refer to Figure 3A and Figure 3B A brief discussion of acquisition time windows is provided. Figure 3A An example of first sensor data 102 x , 102 y , 102 z generated by the first motion sensor 102 over a plurality of acquisition time windows is shown. Figure 3B Shows Figure 3A A zoomed-in view of the sampling times of the first two acquisition time windows W1 and W2 of the example. Figure 3B As illustrated in , in some embodiments, the multiple acquisition time windows are consecutive and non-overlapping time windows. However, in other embodiments, overlapping time windows are also possible. Figure 3B In the example of FIG. 1 , the first acquisition time window W1 starts at time t0 and ends at time t 49In an embodiment, such as in Figure 3A and Figure 3B In the example of , each acquisition time window has a duration of 1 second and includes 50 samples (eg, corresponding to a 50 Hz sampling frequency). Figure 3A In the example of , there are about 72 acquisition time windows and a total of about 3600 samples (i.e., 50 samples in each of the 72 acquisition time windows). Note that each sample includes a complete data set (e.g., x-axis data, y-axis data, and z-axis data). Also note that the 50 Hz sampling frequency and 1 second duration of each acquisition time window are merely examples, and other embodiments using different sampling frequencies and different time durations are contemplated. Figure 3C Shows Figure 3A The norm 302 of the first sensor data 102x, 102y, 102z in FIG. 1 , and the norm 302 at a given sampling time may indicate the size of the first sensor data 102x, 102y, 102z at the given sampling time.

[0041] For each acquisition time window W i Execute method 200. Figure 2 As shown in FIG. 2 , method 200 obtains the time window W. i The beginning of (for example, Figure 3B The first feature detection circuit 104 receives the first sensor data 102x, 102y, 102z and determines the acquisition time window W. i In some embodiments, the time window W is obtained. i The norm of each sample within is stored in a buffer included in the first detection circuit 104, although in other embodiments, the computational techniques used to determine the norm may eliminate the need for such a buffer.

[0042] In step 204, the time window W is obtained. i End, and method 200 proceeds to step 206, in which the acquisition time window W is determined. i In steps 208 and 210, the time window W is obtained. i Thus, each time an entire sample window is acquired (e.g., each time 50 samples are acquired in a 1 second time window), steps 206, 208, and 210 are triggered. The statistical data includes the acquisition time window W i The average crossover value (in step 208) within the acquisition time window W i The norm variance within (in step 210), both of which require the norm mean determined in step 206.

[0043] Referring to step 208, the average crossover value marks the acquisition time window W i The norm within the time window W i The number of crossings of the norm means within . Figure 3D A diagram is given in which the acquisition time window W is shown i norm 304 within (eg, determined in step 202) and the acquisition time window W i The norm mean 306 within (e.g., determined in step 206). Figure 3D In the example, get the time window W i The norm 304 within the acquisition time window W i The norm mean 306 within 26 times is crossed. These instances are depicted as the intersections of curve 304 and line 306. Therefore, Figure 3D The average crossover value for the example is 26.

[0044] Referring to step 210, the time window W is obtained. i The norm variance within is determined as follows:

[0045]

[0046] Where n is the acquisition time window W i The number of samples in (e.g., 50 in the case of a 50 Hz sampling frequency), x i is to obtain the time window W i The i-th norm within is 304, and x mean is to obtain the time window W i The mean of the norm 306 within.

[0047] In step 212 of method 200, the time window W is obtained. i The average cross-over value and variance of the norm within the time window W are provided to the classification circuit 106. i After the end, and in the acquisition time window W i After the mean cross-over value and variance of the norm within are determined by appropriate detection circuits, the classification circuit 106 is run. Note again that the acquisition time window W i The mean cross-sectional value and variance of the inner norm are orientation-independent features used to determine whether the electronic device 101 is located on a stationary or stable surface.

[0048] Figure 4A The figure shows the situation where the electronic device 101 is located on a stationary or stable surface (eg, a table) in 96 acquisition time windows W. iThe average crossing value 402 generated by the first feature detection circuit 104 and the average crossing value 404 generated by the second feature detection circuit 110 are shown in FIG. Figure 4A Each acquisition time window W in i has a corresponding average crossing value MC associated with the first motion sensor 102 (eg, accelerometer) A,i and a corresponding average crossing value MC associated with the second motion sensor 108 (eg, a gyroscope) G,i . Figure 4B FIG. 1 shows a case where the electronic device 101 is not located on a stationary or stable surface (eg, when on a person's lap) during the 145 acquisition time windows W. i The average crossing value 406 generated by the first feature detection circuit 104 and the average crossing value 408 generated by the second feature detection circuit 110 are shown in FIG. Figure 4B For each time window t in the example i has a corresponding mean value MC associated with the first motion sensor 102 (eg, accelerometer) A,i and a corresponding mean value MC associated with the second motion sensor 108 (eg, a gyroscope) G,i .

[0049] By comparing Figure 4A and Figure 4B The average crossover values ​​402 and 406 can be observed, through experimental observation, that the average crossover value 402 obtained by method 200 when the electronic device 101 is located on a stationary or stable surface (e.g., when on a table) is expected to be greater than the average crossover value 406 obtained by method 200 when the electronic device 101 is not located on a stationary or stable surface (e.g., when on a person's lap). The relative difference in the average crossover values ​​in the two different states is Figure 5A , and can be explained in terms of the contribution of the white noise of the first motion sensor 102 to the first sensor data 102x, 102y, 102z in the following two states: (1) when the electronic device 101 is located on a stationary or stable surface, and (2) when the electronic device 101 is not located on a stationary or stable surface.

[0050] For example, the first sensor data 102x, 102y, 102z from the first motion sensor 102 can be approximated as the white noise of the first motion sensor 102 plus the motion-related signal. When the motion-related signal is stable and changes slowly (e.g., when on a stationary or stable surface), the white noise of the first motion sensor 102 can be approximated as a signal that causes the first sensor data 102x, 102y, 102z to fluctuate frequently and randomly around its mean. In contrast, when the motion-related signal is dominant (e.g., not on a stationary or stable surface), the white noise of the first motion sensor 102 contributes less to the first sensor data 102x, 102y, 102z. Therefore, the average crossover value 402 when the electronic device 101 is on a stationary or stable surface is expected to be greater than the average crossover value 406 when the electronic device 101 is not on a stationary or stable surface.

[0051] Similarly, from Figure 4A and Figure 4B It can be observed that the average crossover value 404 obtained by the method 200 when the electronic device 101 is located on a stationary or stable surface (e.g., when on a table) is greater than the average crossover value 408 obtained by the method 200 when the electronic device 101 is not located on a stationary or stable surface (e.g., when on a person's lap). This difference in the average crossover value in the two different states can also be explained in terms of the contribution of the white noise of the second motion sensor 108 to the second sensor data 108x, 108y, 108z in the two states, as described above.

[0052] Regarding the norm variance, it has been experimentally observed that the norm variance is expected to be smaller when the electronic device 101 is located on a stationary or stable surface than when the electronic device 101 is not located on a stationary or stable surface. Figure 5B The relative difference in the norm variance between the two different states is depicted in .

[0053] Continuing to the classification circuit 106, as described above, the classification circuit 106 obtains the time window W i After it is finished and it is in the acquisition time window W i The classification circuit 106 may be configured to perform the classification based on at least each acquisition time window W. i The average crossover value and variance of the norm of the electronic device 101 are determined in the acquisition time window W i The classification circuit 106 may be a supervised machine learning classifier implemented using machine learning techniques, such as logistic regression, naive Bayes classifier, support vector machine, decision tree, boosted tree, random forest, neural network, nearest neighbor, etc. The classification circuit 106 is configured to assign a value to each acquisition time window W.i Assignment label (or decision) L i , this label L i Instruct the electronic device 101 to obtain the time window W i The use of the norm variance can improve the accuracy of the classification circuit 106, as the norm variance decreases if the electronic device 101 is on a stationary or stable surface, and increases if the electronic device 101 is not on a stationary or stable surface.

[0054] Note that supervised learning is a machine learning task that learns a function that maps input to output based on example input-output pairs. In particular, supervised learning infers functions from labeled training data that includes a set of training examples. In the supervised machine learning classifier of the classification circuit 106, by placing the electronic device 101 (including the first motion sensor 102 and the optional second motion sensor 108) on a stationary or stable surface (e.g., a table) and recording the first sensor data 102x, 102y, 102z and the second sensor data 108x, 108y, 108z for various typing intensity levels and different orientations and positions of the electronic device 101 on the stationary or stable surface, the labeled training data can be obtained. When the electronic device 101 is located on a stationary or stable surface, it is known that the first sensor data 102x, 102y, 102z and the second sensor data 108x, 108y, 108z for these various typing intensity levels and different orientations and positions are obtained. Therefore, such first sensor data 102x, 102y, 102z and second sensor data 108x, 108y, 108z are subsequently subjected to Figure 2 Method 200 to obtain the W in various acquisition time windows i The average cross-over value and the norm variance value within are calculated, and such average cross-over value and the norm variance value are then assigned a tag indicating that the electronic device 101 is located on a stationary or stable surface.

[0055] Similarly, labeled training data can also be obtained by placing the electronic device 101 on a moving or unstable surface (e.g., a person's knee) and recording first sensor data 102x, 102y, 102z and second sensor data 108x, 108y, 108z for various typing intensity levels and different orientations and positions of the electronic device 101 on a stationary or stable surface. Then, the various first sensor data 102x, 102y, 102z and the various second sensor data 108x, 108y, 108z obtained in this manner are subjected to Figure 2 Method 200 to obtain various acquisition time windows W iThe average intersection value and norm variance value within are calculated and then assigned a label to these average intersection value and norm variance values ​​indicating that the electronic device 101 is not located on a stationary or stable surface.

[0056] Figure 1 The delay of the detection system 100 shown in FIG. 1 may depend at least on the delay of the classification circuit 106, which may be equal to each acquisition time window W. i In each acquisition time window W i In the embodiment where the duration of L is 1 second, the classification circuit 106 has a delay of 1 second because one label L is output from the classification circuit 106 every second. i As described below, in embodiments that also include a meta-classification circuit 112, the latency of the detection system 100 is also affected by the latency of the meta-classifier output.

[0057] To further enhance the accuracy of determining whether the electronic device 101 is located on a stationary or stable surface, the detection system 100 may include a meta-classification circuit 112. In an embodiment, the meta-classification circuit 112 is configured to determine the output L of the classification circuit 106. i If the number of consecutive occurrences exceeds the threshold, the output of the meta-classification circuit 112 is changed (in Figure 1 Marked as L final ). Otherwise, the previous state is maintained. In this way, the meta-classification circuit 112 can be used to low-pass filter the output of the classification circuit 106 (e.g., to avoid glitches and false positives).

[0058] Using the meta-classification circuit 112 may introduce a delay to the detection system 100, and the delay of the meta-classification circuit 112 may be configured to be as small as the acquisition time window W. i In some embodiments, the duration of the electronic device 101 may be N times the duration of the electronic device 101. ... on_table , and if the number of consecutive occurrences reaches N on_table , then the output state is L final will change) or whether the output of the classification circuit 106 indicates that the electronic device 101 is not located on a stationary or stable surface (eg, where N=N not_on_table , and if the number of consecutive occurrences reaches N not_on_table , then the output state is L final In some embodiments, N not_on_table Can be different from N on_table The output of the meta-classification circuit 112 is updated according to the meta-classifier logic configuration and the configured meta-classifier output delay. In some embodiments, N on_tablecan be configured between 2 and 10, and N not_on_table Can be configured between 2 and 10.

[0059] Although the use of meta-classification circuitry 112 can improve the accuracy of determining whether electronic device 101 is located on a stationary or stable surface, this increase in accuracy comes at the expense of increased system latency. However, even though latency increases with increased accuracy, embodiment systems and methods can achieve latency of less than 10 seconds (e.g., between 4 seconds and 9 seconds), even when using meta-classification circuitry 112.

[0060] As discussed above, in low power or low cost implementations of the embodiment systems and methods, the classification circuit 106 may not use the second motion sensor 108 (e.g., gyroscope) and the data therefrom to determine whether the electronic device 101 is located on a stationary or stable surface (e.g., on a table or in a drawer). In experiments that have been conducted, it has been noted that if the classification circuit 106 only uses the average intersection value MC obtained from the first sensor data 102x, 102y, 102z i,102 Sum norm variance Var i,102 , an accuracy of about 90% can be achieved. In other words, when only the average intersection value MC obtained from the first sensor data 102x, 102y, 102z is used i,102 Sum norm variance Var i,102 When , approximately 90% of the acquisition time windows are correctly assigned the label L i Thus, high accuracy can be achieved even in low power and low cost implementations of the embodiment systems and methods, even without using the meta-classification circuit 112. It should also be noted that when using the average intersection value MC i,102 Sum norm variance Var i,102 (obtained from the first sensor data 102x, 102y, 102z) and the average crossover value MC i,108 Sum norm variance Var i,108 Even without using the meta-classification circuit 112, an accuracy of about 97% can be achieved.

[0061] In low power applications, the time window W is obtained from i The choice of which data to extract from the first feature detection circuit 104 is based on a trade-off between accuracy and power consumption. In general, the number of features determined by the first feature detection circuit 104 (and by the second feature detection circuit 110 in embodiments used in conjunction with the circuit 104) may be altered. For example, the mean of each axis may be calculated, and this may be used to determine the mean of each acquisition time window W. iAs another example, the energy of the signal received from the motion sensor can be used. However, it is worth noting that determining a greater number of features is accompanied by an increase in resource consumption (e.g., memory, execution time, and power).

[0062] The output of the meta-classification circuit 112 may be provided to a state monitor 114, which may adapt the behavior or operation of the electronic device 101. The state monitor 114 may be implemented using a controller and a memory register. The output of the classification circuit 106 and / or the output of the meta-classification circuit 112 may be stored in a memory register of the state monitor 114, and the controller of the state monitor 114 may be configured to read the contents of the memory register. In response to determining that the electronic device is on a stationary or stable surface (e.g., a table), the state monitor 114 may generate an interrupt signal 116, which may adapt the behavior or operation of the electronic device 101, for example, the fan speed and clock frequency of the electronic components (e.g., a central processing unit (CPU), a graphics processing unit (GPU), or a power supply unit) in the electronic device 101 may be increased to achieve better performance (e.g., faster computing time). Conversely, in response to determining that the electronic device is not on a stationary or stable surface (e.g., when the electronic device is in motion or on the user's knees), the interrupt signal 116 may cause the clock frequency of the components in the electronic device 101 to be reduced to reduce power consumption and avoid overheating of the components in the electronic device 101.

[0063] The embodiment systems and methods discussed above may be implemented in a variety of ways. Fig. 6A A first example is shown in which the method 200 and the classification circuit and the meta-classification circuit 112 are implemented by a controller 502 (e.g., a microcontroller) coupled to a micro-electromechanical (MEMS) system-in-package 504. The MEMS system-in-package 504 may implement the first motion sensor 102 and / or the second motion sensor 108. In addition, the controller 502 may be included in a system on a chip (SoC) 506 that is communicatively coupled to an operating system layer 508 of the electronic device 101.

[0064] Figure 6B Another example is shown in which the method 200 and the classification circuit and meta-classification circuit 112 are implemented by directly connecting the controller 502 to the operating system layer 508 (e.g., without Fig. 6A This is achieved by using the SoC 506 as an intermediate connection).

[0065] Figure 6CAnother example is shown where the method 200, as well as the classification circuit and the meta-classification circuit 112 are implemented directly in hardware connected to the operating system layer 508 (e.g., directly on the MEMS system-level package 504, assisted by software embedded in the MEMS system-level package 504). Notably, Fig. 6A The current consumption of the implementation shown in Figure 6B The current consumption of the implementation shown in Figure 6B The current consumption of the implementation shown in Figure 6C The current consumption of the implementation shown in .

[0066] The embodiment systems and methods have at least the following advantages: (1) are easy to adjust or reconfigure (e.g., due to the use of machine learning methods for the classification circuit 106); (2) have low latency and short convergence time (e.g., due to the time interval TI being split into multiple short time windows t i , each short time window is approximately 1 second and is also configurable / adjustable, so the convergence time is less than 10 seconds); (3) no calibration of the motion sensor is required (e.g., due to the use of orientation-independent features of the mean cross-value and norm variance, thereby exhibiting immunity to device-to-device variations, accelerometer offsets, and / or gyroscope bias); and (4) higher reliability than conventional systems and methods due to the use of orientation-independent features in the embodiment systems and methods. In addition, as described in reference Figure 6C As mentioned, the embodiment systems and methods may be directly implemented in hardware, thereby enabling ultra-low power consumption of the embodiment systems and methods.

[0067] As will be discussed in more detail below, a desk detection algorithm detects the orientation of an electronic device. Based on information from a motion sensor of the electronic device, the desk detection algorithm determines whether the electronic device is on a desk or not on a desk. The algorithm may also determine the stability of the electronic device. The stability of the electronic device depends on whether the device is determined to be in the same orientation within a configurable timeframe. An embodiment may also add a sleep mode to the desk detection algorithm, where the algorithm enters the sleep mode after determining that the device is in a stable orientation within a configurable timeframe. Implementing the sleep mode in the desk detection algorithm helps reduce the power consumption of the electronic device.

[0068] In further embodiments, the embodiments discussed above may be applied to implement a sleep mode for a system that determines whether an electronic device is located on a stationary or stable surface (commonly referred to as on-table detection (OTD)). These embodiments use the output from the classification circuit 106, which classifies the data obtained from the first feature detection circuit 104 and outputs an output referred to as an "unfiltered output" that is a classification of whether the electronic device is located on a stationary or stable surface (output 2, indicating on-table) or not located on a stationary or stable surface (output 1, indicating not-on-table). In explaining Figure 1 An example of such a classification circuit is discussed in the embodiment of FIG. 1 and is represented by classification circuit 106. The unfiltered output of classification circuit 106 can then be used in meta-classification circuit 112, where the counters of the two meta-classifiers are NOT-ON-TABLE and count ON-TABLE ) is used to convert the unfiltered output into what can now be called the "filtered output". The first counter count of the meta-classifier's counters NOT-ON-TABLE A counter that counts up every time the unfiltered output of the classification circuit 106 returns "not on the table"; a second counter count of the meta-classifier's counter ON-TABLE represents a counter that counts up by 1 each time the unfiltered output of the classification circuit 106 returns "on the table". The filtered output represents the output classification of the meta-classification circuit 112, which determines three possible values ​​based on the state of the meta-classifier, which are 0 (indicating undetermined), 1 (indicating not on the table), and 2 (indicating on the table). The embodiment can then use the value of the meta-classifier counter to formulate a stability condition, which depends on how the meta-classifier counter is processed by the meta-classification circuit 112. Using this stability condition, the embodiment can enable a timer-based loop mechanism to switch the OTD system from a continuous data flow mode to a sleep mode, which can change many operating parameters of the OTD system to reduce power consumption. Figure 7A-7B will with Figure 8 are used together to describe these further embodiments in more detail.

[0069] Fig. 7A A stability determination system is shown with example outputs of classification circuit 106 (unfiltered 702 and filtered 708), where meta-classifier counters 704 and 706 (derived from meta-classification circuit 112) may be used to convert the unfiltered output into a filtered output, which may be used to determine the stability of an electronic device according to an embodiment.

[0070] Fig. 7ADescribes the operation of the classification system when the counter of a meta-classifier is operated in a reset mode, wherein once the counter of the relative meta-classifier (e.g., count NOT-ON-TABLE ) increments the meta-classifier's counter (e.g., count ON-TABLE ) is reset to 0. Fig. 7A , the unfiltered output 702 is the output from the classification circuit 106, and it represents the classification (1 for not on the table, 2 for on the table) plotted in the time window. The unfiltered output is updated every time the buffer of the first motion sensor 102 is full (e.g., every 1 second). The count of the meta-classifier NOT-ON-TABLE 704 is a counter that increments by 1 each time the unfiltered output 702 is drawn in the time window (meaning the classification circuit 106 determines that the electronic device is not on the table based on the unfiltered output), up to a configurable maximum count (e.g., a maximum of 6). ON-TABLE 706 is a counter that increments by 1 each time the unfiltered output 702 is 2 drawn in the time window (meaning the classification circuit 106 determines that the electronic device is on the table based on the unfiltered output), up to a configurable maximum count (e.g., a maximum of 3). The filtered output 708 is the output from the meta-classification circuit 112 and represents the classification (0 for undetermined, 1 for not on the table, 2 for on the table) based on the counter value of the meta-classifier drawn in the time window (if the meta-classifier's counter has reached its configurable maximum value, the corresponding state determination will be returned as the filtered output state). The binary on or off stability condition 710 shows a visual depiction of how the sleep algorithm tracks the stability of the OTD output and how the unfiltered and filtered outputs and the counters of the meta-classifier correspond to the associated stability classifications. The stability condition is evaluated each time a new unfiltered output is generated (e.g., every 1 second). This stability classification can be used in the algorithm to put the OTD system into sleep mode, such as Figure 8 As shown in the figure.

[0071] For the case where the counter of the meta-classifier is reset, such as Fig. 7A As shown in FIG. 1 , the meta-classification circuit takes the unfiltered output 702 as input and uses the count NOT-ON-TABLE 704 and count ON-TABLE 706 to return the filtered output 708. This can be achieved by increasing the corresponding counter based on the returned unfiltered output 702 (a return value of 1 for the unfiltered output corresponds to increasing the count NOT-ON-TABLE 704, while the return value of 2 for the unfiltered output corresponds to increasing count ON- TABLE706). In this embodiment, if the counter of the opposite meta-classifier is increased by 1, the counter of the other meta-classifier is reset to 0.

[0072] The filtered output 708 is then determined by taking the returned values ​​of the meta-classifiers' counters 704 and 706, wherein once the corresponding meta-classifier's counter reaches the corresponding maximum value (which is a configurable limit), the corresponding value of the filtered output is changed only according to the reported value (in Fig. 7A In the embodiment shown, for count NOT-ON-TABLE 704, the value is 6, and for count ON-TABLE 706, the value is 3). The meta-classifier's counter must not exceed its configurable maximum value.

[0073] A "STABILITY" classification 710 may be determined by a counter (704 or 706) reaching its maximum value (or remaining at its maximum value), wherein the stability classification 710 will remain in a "true" state as long as the counter of one meta-classifier is equal to its maximum value and the counter of the other meta-classifier is 0. This stability classification may be used by the method to determine whether the OTD output is stable over time.

[0074] Figure 7B A stability determination system is illustrated in which example outputs of the classification circuit 106 (unfiltered 712 and filtered 718), wherein the meta-classifier's counters 714 and 716 (originating from the meta-classification circuit 112) can be used to convert the unfiltered output into a filtered output, which can be used to determine the stability of the electronic device according to an embodiment.

[0075] Figure 7B is a description of the operation of the classification system when the counter of the meta-classifier is operated in a decrementing mode, where once the counter of the relative meta-classifier (e.g., count NOT-ON-TABLE ) has been incremented by 1, and the counter of the meta-classifier (e.g., count ON-TABLE ) is reduced by 1. Figure 7B , the unfiltered output 712 is the output from the classification circuit 106 and represents the classification (1 for not on the table, 2 for on the table) plotted in the time window. The unfiltered output is updated every time the buffer of the first motion sensor 102 is full (e.g., every 1 second). The count of the meta-classifier NOT-ON-TABLE 714 is a counter that increments by 1 each time the unfiltered output 712 is drawn in the time window (indicating that the classification circuit 106 determines that the electronic device is not on the table based on the unfiltered output), until a configurable maximum count is reached (e.g., a maximum of 6). ON-TABLE716 is a counter that increments by 1 each time the unfiltered output 712 is 2 drawn in the time window (indicating that the classification circuit 106 has determined that the electronic device is on the table based on the unfiltered output), until a configurable maximum count is reached (e.g., a maximum of 3). The filtered output 718 is the output from the meta-classification circuit 112 and represents the classification (0 for undetermined, 1 for not on the table, 2 for on the table) based on the counter value of the meta-classifier drawn in the time window (if the meta-classifier's counter has reached its configurable maximum value, the corresponding state determination is returned as the filtered output state). The binary on or off stability condition 720 shows how the sleep algorithm tracks the stability of the OTD output, and how the unfiltered and filtered outputs and the counters of the meta-classifier correspond to the associated stability classifications. The stability condition is evaluated each time a new unfiltered output is generated (e.g., every 1 second). This stability classification can be used in the algorithm to put the OTD system into sleep mode, such as Figure 8 As shown in the figure.

[0076] Figure 7B Example outputs (unfiltered 712 and filtered 718) of the classification circuit 106 are shown, where the counters 714 and 716 of the meta-classifier are used to convert the unfiltered output to the filtered output. Figure 7B 720 shows how the sleep algorithm tracks the stability of the OTD output, and how the unfiltered and filtered outputs and the counters of the meta-classifiers correspond to the associated stability classification. This stability classification can be used in the algorithm to put the OTD system into sleep mode, such as Figure 8 As shown in the figure.

[0077] For the case where the counter of the meta-classifier is decremented, such as Figure 7B As shown in FIG. 1 , the meta-classification circuit takes the unfiltered output 712 as input and uses the count NOT-ON-TABLE 714 and counting ON-TABLE 716 returns the filtered output 718. This can be achieved by increasing the corresponding counter based on the returned unfiltered output 712 (a return value of 1 for the unfiltered output corresponds to an increase in count NOT-ON-TABLE 714, while the return value of 2 for the unfiltered output corresponds to the increase in count ON- TABLE 716). In this embodiment, if the counter of the opposing meta-classifier is increased by 1, the counter of the other meta-classifier will be decreased by 1.

[0078] The filter output 718 is then determined by obtaining the return values ​​of the meta-classifier counters 714 and 716, wherein once the corresponding meta-classifier counter reaches the corresponding maximum value (configurable limit), the corresponding value of the filter output is changed according to the reported value (in Figure 7B In the embodiment shown, for count NOT-ON-TABLE 714, the value is 6, and for count ON-TABLE 716, the value is 3). The meta-classifier's counter may not exceed its configurable maximum value.

[0079] A "stability" classification 720 may be determined by either counter (714 or 716) reaching its maximum value (or remaining at its maximum value), wherein the stability classification 720 will remain in a "true" state as long as the counter of one meta-classifier is equal to its maximum value and the counter of the other meta-classifier is equal to 0. This stability classification may be used by the method to determine whether the OTD output is stable over time.

[0080] In an embodiment, the stability detection system may be configured to Figure 8 The exemplary operation flow illustrated in FIG. 1 is operated to implement a sleep mode in the OTD system to prevent the motion sensor data from being continuously read and processed.

[0081] Figure 8 A block diagram is shown that illustrates an algorithm for placing the OTD system into sleep mode when the OTD output remains stable for a configurable amount of time. 802 is a gyroscope that feeds into the OTD step 806. The OTD system also takes output from the accelerometer 804. At 808, the system checks if there is a new OTD output (a "positive" result, or if there is a new OTD output, it moves to step 812) or not (the system will monitor for new OTD output and remain at box 808).

[0082] At step 812, based on the results of the analysis step of classifying the stability classification, it can be determined whether the electronic device is stable (positive result Y of block 812) or unstable (negative result N of block 812). To determine the stability classification checked in step 812, the system uses a counter 810 (counter 811) of the two meta-classifiers. NOT-ON-TABLE and count ON-TABLE ).exist Figure 7A-7B An example of such determination of stability classification is depicted in .

[0083] Due to a negative result of step 812, the operation may continue to step 816, which may include turning on the gyroscope 802, setting the accelerometer 804 streaming on (which corresponds to the accelerometer continuing to send its data to the OTD system), turning off the accelerometer wake-up condition, setting the COUNT STABILITY The counter is set to 0 (the counter counts the number of time windows that the system is in the stability classification of "true"), and the STABILITY TIMER ,STABILITY TIMER is a timer that starts only if the stability classification is true.

[0084] Since the result of step 812 is positive, the operation can continue to step 814, which can include turning off the gyroscope 802, setting the accelerometer 804 streaming off (stopping the accelerometer from sending data to the OTD system), turning on the accelerometer wake-up condition, and setting the COUNT STABILITY The counter increases by TIMEOUT. The accelerometer wake-up condition 818 is a configurable threshold (eg, 20 mg) above which, when detected by the accelerometer, a motion event is signaled to the OTD system to wake it from sleep.

[0085] After the operations performed during step 814, operations may continue to step 822, where COUNT may be determined. STABILITY The counter is equal to STABILITY TIMEOUT (Positive result of block 822, Y) still not equal to STABILITY TIMEOUT (Negative result from block 822, N). STABILITY TIMEOUT is a configurable limit (e.g., 10 seconds) that ensures that the system is in a positive stability classification within a configurable time frame.

[0086] Due to a negative result of step 822, operation may continue to step 820, where STABILITY TIMER Start with TIMEOUT. TIMEOUT is a predefined time window (e.g., equal to the OTD time window length, such as 1 second, or based on COUNT STABILITY Once set to TIMEOUT, STABILITY TIMER An elapsed condition 826 is checked to see if the timer has elapsed (positive result of block 826, Y) or has not elapsed (negative result of block 826).

[0087] Due to a negative result in step 826, operation may continue to step 826, where the system will continue to check STABILITYTIMER Has it passed?

[0088] Due to the positive result of step 826, the operation can continue to execute step 816, wherein the configuration changes made during this step have been detailed above and will not be repeated here for the sake of brevity.

[0089] Due to the positive result of step 822, operation can continue to step 824, where the sleep mode of the OTD system is enabled. Once step 824 is reached and the system has entered sleep mode, the accelerometer wake-up condition 818 can eventually be used to restart the mechanism (bring the system out of sleep mode).

[0090] Fig. 9 is a flow chart illustrating an embodiment of the present disclosure.

[0091] In an embodiment, a method includes generating, by a first motion sensor of an electronic device, first sensor data over an acquisition time window (box 910). The method includes generating first process data by processing the first sensor data to determine whether the electronic device is located on a stationary surface (box 920). The method includes determining whether the electronic device is in a stable state based on the first process data (box 930). The stable state may indicate whether the electronic device has been held on the stationary surface for a first predefined time. The method includes stopping processing of the first sensor data in response to determining that the electronic device has been in a stable state for a second predefined time (box 940).

[0092] The above-mentioned various frames can be used as shown in FIG. Figure 8 For example, in an embodiment, block 930 may be implemented as Figure 8 822 in, although typically it may include a stability classification (in box 806), a stability classification check (box 812), and a count stability check (box 822).

[0093] The method may also include generating second sensor data by a second motion sensor of the electronic device over the acquisition time window, wherein determining whether the electronic device is located on the stationary surface includes generating second process data by processing the second sensor data; and determining whether the electronic device is in a stable state includes determining based on the first process data and the second process data. The method may also include, in response to determining that the electronic device has been on the stationary surface for a second predefined time, generating a first control signal to turn off the second motion sensor. For example, in an embodiment, this may be implemented as Figure 8 Frame 812 in .

[0094] In another embodiment, the detection system of the present disclosure may include additional classification circuitry that is capable of further distinguishing the type of stationary or stable surface on which the electronic device is located through a bisection approach. For example, if it is determined that the electronic device is stable (located on a stationary or stable surface), the additional classification circuitry may further determine whether the electronic device is located on a semi-stationary surface (e.g., the user's lap) or whether the electronic device is located on a stationary surface (e.g., a table). An example embodiment of a detection system capable of further distinguishing the state (or level of motion) of an electronic device is described in Fig.10 is illustrated in and described below.

[0095] Fig.10 A block diagram of an electronic device 1001 including a detection system 1000 according to an embodiment is shown. The detection system 1000 may be located within, attached to, or coupled to the electronic device 1001. The detection system 1000 of the electronic device 1001 may be used to determine whether the electronic device 1001 is stable or in motion. In addition, Fig.10 The detection system 1000 of the embodiment of the electronic device 1001 can determine whether the electronic device 1001 is located on a stationary surface or a semi-stationary surface when the electronic device 1001 is stable. Adding a third classifier enables the detection system 1000 to distinguish between the stationary surface classification and the semi-stationary surface classification.

[0096] Including the third category (on a semi-stationary surface) can further allow the electronic device 1001 to optimize system policies (e.g., fan speeds and clock frequencies of the above-mentioned electronic components) based on whether the electronic device 1001 is on a semi-stationary surface or in motion. Specifically, a difficulty encountered in implementing the above-mentioned system policies using a system that cannot distinguish whether the electronic device 1001 is on a semi-stationary surface or in motion is that, for example, when the user is walking (or in motion), the fan speed of the electronic device 101 is increased to dissipate heat, which may be interpreted by the end user as a poor user experience. In this case, because the detection system only distinguishes whether it is on a stationary surface (on a table) or in motion (not on a table), the detection system may cause the electronic device to increase the fan speed, which will increase the noise level and pointlessly increase power consumption. By including the third category, by changing the system policy based on determining the state of the electronic device 1001, Fig.10 The electronic device 1001 can eliminate this difficulty.

[0097] Similar to Figure 1 Electronic Devices 101, Fig.10 The electronic device 1001 may also be a laptop computer, a tablet device, or a wearable electronic device (eg, a smart watch, a mobile phone, a wireless headset, etc.). Fig.10The electronic device 1001 may be the one described herein Figure 1 Specific implementation of the electronic device 101. Like-labeled elements may be as described above.

[0098] like Fig.10 As shown in , the first classification circuit 1006 is coupled to the output of the first feature detection circuit 104 and may be coupled to the output of the additional second feature detection circuit 110. The first classification circuit 1006 is configured to determine between two states of the electronic device 1001 (e.g., assign a label or category indicating whether the electronic device 1001 is stable or in motion). The first classification circuit 1006 makes this determination based on orientation-independent features determined by the first feature detection circuit 104 using the first machine learning classification. In some embodiments including the second feature detection circuit 110, the first classification circuit 1006 makes this determination based on orientation-independent features determined by both the first feature detection circuit 104 and the second feature detection circuit 110. Various embodiments of the detection system 1000 may include additional feature detection circuits for additional motion sensors.

[0099] The first classification circuit 1006 uses the first machine learning classification to determine whether the electronic device 1001 is stable or the electronic device 1001 is in motion. The first classification circuit 1006 is coupled to the second classification circuit 1007 and the meta classification circuit 1012. If the first classification circuit 1006 determines that the electronic device 1001 is in motion, the classification L a,j Directly sent to the meta-classification circuit 1012. If the first classification circuit 1006 determines that the electronic device 1001 is stable, a second classification step is performed, which determines whether the electronic device 1001 is on a stationary surface or a semi-stationary surface.

[0100] In an embodiment, the second classification circuit 1007 may be configured to be executed only when the first classification circuit 1006 determines that the electronic device 1001 is stable. Based on the orientation-independent features determined from the first feature detection circuit 104 and the second feature detection circuit 110, the second classification circuit 1007 may also be configured to determine between two states of the electronic device 1001 to classify the motion level of the electronic device 1001 using a second machine learning classification (e.g., determining whether the electronic device 1001 is on a stationary surface (desk) or on a semi-stationary surface (lap)).

[0101] The second classification circuit 1007 determines whether the electronic device 1001 is located on a stationary surface or the electronic device 1001 is located on a semi-stationary surface. Using a second machine learning classification that adaptively distinguishes between various levels of motion of the device, it can be determined whether the electronic device 1001 is in a state of being on a stationary surface or in a state of being on a semi-stationary surface. For example, the second machine learning classification can determine the classification or label of the electronic device 1001 by monitoring multiple device parameters (such as orientation-independent features provided by feature detection circuits). Then, through the coupling of the second classification circuit 1007 and the meta-classification circuit 1012, the second classification L of whether the electronic device 1001 is on a stationary or semi-stationary surface can be determined. b,j Passed to the meta-classification circuit 1012.

[0102] In an embodiment, the meta-classification circuit 1012 may be configured to receive an initial classification L j , the initial classification may be L sent from the first classification circuit 1006 a,j , or L sent from the second classification circuit 1007 b,j If the first classification circuit 1006 determines that the electronic device 1001 is in motion, then L j is L from the first classification circuit a,j And L j If the first classification circuit 1006 determines that the electronic device 1001 is stable and the second classification circuit 1007 determines that the state of the electronic device 1001 is on a stationary surface or on a semi-stationary surface, then L j is L from the second classification circuit 1007 b,j And L j It is a classification of the state of being on a stationary surface or on a semi-stationary surface.

[0103] In various embodiments, the meta-classification circuit 1012 may perform functions similar to those described above for the meta-classification circuit 112 in the figure. Again, to further enhance the accuracy of determining whether the electronic device 1001 is located on a stationary surface, a semi-stationary surface, or in motion, the detection system 1000 may include the meta-classification circuit 1012. In an embodiment, the meta-classification circuit 1012 is configured to determine the output L of the classification circuit j If the number of consecutive occurrences exceeds the threshold, the output of the meta-classification circuit 1012 is changed (in Fig.10 Marked as L final ). Otherwise, the previous state is maintained. Thus, the meta-classification circuit 1012 can be used to low-pass filter the output of the classification circuit (eg, to avoid glitches and false positives).

[0104] In an embodiment, Fig.10The classification and meta-classification circuitry within block 1020 of the electronic device 1001 may be implemented in software, where the different circuits correspond to algorithms implemented in a processor that executes instructions stored in a memory device for determining the classification of the electronic device 1001. In other embodiments, block 1020 may also include feature detection circuitry that may also be implemented in software.

[0105] In some embodiments, the initial classification L j The output L of the first classification circuit 1006 and the second classification circuit 1007 may be directly sent to the state monitor 1014 through the coupling of the first classification circuit 1006 and the second classification circuit 1007 (without including the meta classification circuit 1012). In the embodiment including the meta classification circuit 1012, the output L of the meta classification circuit 1012 final can be sent to the status monitor 1014. And, as described above for Figure 1 State monitor 1014 may function similarly to state monitor 114 described above.

[0106] The state monitor 1014 may be based on the received output (such as L j or L final) to adapt the behavior or operation of the electronic device 1001. The state monitor 1014 can be implemented using a controller and a memory register. The output of the classification circuit and / or the output of the meta-classification circuit 1012 can be stored in a storage register of the state monitor 1014, and the controller of the state monitor 1014 can be configured to read the contents of the storage register. In response to determining that the electronic device is located on a stationary surface (e.g., a table), on a semi-stationary surface (e.g., a user's knee), or in motion (e.g., the user is walking), the state monitor 1014 can generate an interrupt signal 1016, which can adapt the behavior or operation of the electronic device 1001. For example, in response to determining that the electronic device is located on a stationary surface (e.g., on a table), the interrupt signal 1016 may increase the clock frequency of electronic components (e.g., central processing unit (CPU), graphics processing unit (GPU), or power supply unit) in the electronic device 1001 to achieve better performance (e.g., faster computing time); in response to determining that the electronic device is on a semi-stationary surface (e.g., the user's knees), the interrupt signal 1016 may reduce the power (or RF power) of the RF antenna of the electronic device 1001 to meet the electromagnetic radiation requirements; in response to determining that the electronic device is in motion (e.g., when the electronic device is in motion), the interrupt signal 1016 may increase the fan speed in the electronic device 1001 to avoid overheating of components in the electronic device 1001. In various embodiments, the interrupt signal 1016 may request a component of the electronic device 1001 to generate a control signal, which may be used to change the clock frequency of the processor of the electronic device, or the fan speed, or other power settings (such as RF power).

[0107] The inclusion of a third category (on a semi-stationary surface) may ameliorate the above-described difficulty of enabling an electronic device to enable system policies that are not always optimized for every situation. Additionally, the inclusion of a third category may enable other embodiments to further control how system policies are implemented in the electronic device 1001 by enabling a third set of system policies for the device for the device. For example, a fan speed may be configured to a higher level in a state on a semi-stationary surface as compared to a state on a stationary surface. As another example, for various health reasons, when it is determined that the electronic device 1001 is located on a semi-stationary surface (e.g., on the user's lap), electromagnetic radiation emissions (such as RF power through an RF antenna (not shown) of the electronic device 1001) may be reduced to limit exposure to the user. Use Fig.10 The described embodiment 10 can also further improve the user experience of using the electronic device 1001.

[0108] In various embodiments, the first machine learning classification and the second machine learning classification may include any one or a mixture of conventional machine learning classifications, such as logistic regression classification, naive Bayes classification, support vector machine classification, decision tree classification, boosted tree classification, random forest classification, neural network classification, or nearest neighbor classification.

[0109] In additional embodiments, the detection system 1000 can be configured to distinguish the three categories without including the second classification circuit 1007. Instead, the first classification circuit 1006 can be trained to distinguish the three categories (stationary surface, semi-stationary surface, and in motion), and thus use the above Figure 1 However, it has been demonstrated that even if the single classifier circuit is allocated the same resources (classifier type, memory, computing power, etc.) as the dual classification circuit is allocated, this embodiment cannot achieve the same accuracy as that achieved by the dual classification circuit embodiment described above.

[0110] In various embodiments, the electronic device 1001 can also implement the above-mentioned sleep mode and can perform the same Figure 1 The electronic device 1001 can also implement the above-mentioned sleep mode by making the sleep mode regard any one of the classification on the stationary surface and the classification on the semi-stationary surface as the same as the classification on the desk of the sleep mode. More embodiments can also include additional motion sensors and additional feature detection circuits.

[0111] Example embodiments of the present invention are described below. Other embodiments may also be understood from the entire specification and claims submitted herein.

[0112] Example 1. A method, comprising generating first sensor data by a first motion sensor of an electronic device over an acquisition time window, and generating first process data by processing the first sensor data to determine whether the electronic device is located on a stationary surface, a semi-stationary surface, or in motion. The method also includes determining whether the electronic device is in a stable state based on the first process data, the stable state indicating whether the electronic device has remained on the stationary surface for a first predefined time. And, the method also includes stopping processing the first sensor data in response to determining that the electronic device has been in a stable state for a second predefined time.

[0113] Example 2. The method according to Example 1 further includes: in response to determining that the electronic device has been in a stable state for a second predefined time, generating additional first sensor data by the first motion sensor over a subsequent acquisition time window.

[0114] Example 3. The method according to one of Examples 1 or 2 further includes: in response to determining that the additional first sensor data exceeds a predefined threshold, determining whether the electronic device is located on a stationary surface by generating first process data by processing the additional first sensor data.

[0115] Example 4. The method according to one of Examples 1 to 3 further includes: generating second sensor data by a second motion sensor of the electronic device over an acquisition time window; wherein determining whether the electronic device is located on a stationary surface includes generating second process data by processing the second sensor data; and determining whether the electronic device is in a stable state includes making a determination based on the first process data and the second process data.

[0116] Example 5. The method of any of Examples 1 to 4, wherein the first motion sensor comprises an accelerometer and the second motion sensor comprises a gyroscope.

[0117] Example 6. The method of one of Examples 1 to 5 further includes generating a first control signal to turn off the second motion sensor in response to determining that the electronic device has been held on the stationary surface for a second predefined time.

[0118] Example 7. The method of any one of Examples 1 to 6, wherein the first predefined time is greater than the second predefined time.

[0119] Example 8. The method according to one of Examples 1 to 7 further includes: in response to determining that the additional first sensor data exceeds a predefined threshold, generating a second control signal to turn on the second motion sensor; in response to receiving the second control signal, turning on the second motion sensor; generating additional second sensor data by the second motion sensor over a subsequent acquisition time window; and generating second process data by processing the additional second sensor data.

[0120] Example 9. An electronic device, comprising: a first motion sensor configured to generate first sensor data indicating a first type of movement of the electronic device; a processor coupled to a memory, the memory storing instructions to be executed in the processor, the instructions, when executed, causing the processor to: generate first sensor data over an acquisition time window using the first motion sensor; determine whether the electronic device is located on a stationary surface by generating first process data by processing the first sensor data; determine whether the electronic device is in a stable state based on the first process data, the stable state indicating whether the electronic device has remained on the stationary surface for a first predefined time; and stop processing the first sensor data in response to determining that the electronic device has been in the stable state for a second predefined time.

[0121] Example 10. The device according to Example 9 further includes: a first feature detection circuit, which is configured to determine at least one first orientation-independent feature based on the first sensor data, wherein the at least one first orientation-independent feature includes an average crossing value associated with an acquisition time window; a first classification circuit, which is configured to determine whether the electronic device is stable or in motion based on the at least one first orientation-independent feature; and a second classification circuit, which is configured to determine whether the electronic device is on a stationary surface or on a semi-stationary surface based on the at least one first orientation-independent feature in response to the first classification circuit determining that the electronic device is stable.

[0122] Example 11. The device according to one of Examples 9 or 10 further includes: a state monitor, which is configured to: in response to determining that the electronic device is on a semi-stationary surface, generate a control signal to increase the fan speed of the electronic device and reduce the RF power of the electronic device, in response to determining that the electronic device is on a stationary surface, generate a control signal to increase the clock frequency of the electronic device and reduce the fan speed of the electronic device, and increase the RF power of the electronic device, and in response to determining that the electronic device is in motion, generate a control signal to reduce the clock frequency of the electronic device and reduce the fan speed of the electronic device.

[0123] Example 12. The device of one of Examples 9 to 11 further includes: a second motion sensor configured to generate second sensor data indicating a first type of movement of the electronic device; and a second feature detection circuit configured to determine at least one second orientation-independent feature based on the second sensor data, wherein the at least one second orientation-independent feature includes an average crossing value associated with an acquisition time window.

[0124] Example 13. A device according to one of Examples 9 to 12, wherein determining whether the electronic device is stable or in motion is based on at least one first orientation-independent feature and at least one second orientation-independent feature.

[0125] Example 14. A device according to one of Examples 9 to 13, wherein the instruction, when executed, further causes the processor to: in response to determining that the electronic device has been in a stable state for a second predefined time, generate additional first sensor data by the first motion sensor over a subsequent acquisition time window.

[0126] Example 15. A device according to one of Examples 9 to 14, wherein the instruction when executed further causes the processor to: in response to determining that the additional first sensor data exceeds a predefined threshold, determine whether the electronic device is located on a stationary surface by generating first process data by processing the additional first sensor data.

[0127] Example 16. A device according to one of Examples 9 to 15, wherein the instruction, when executed, further causes the processor to: generate second sensor data over an acquisition time window using a second motion sensor of the electronic device; wherein the instruction for determining whether the electronic device is located on a stationary surface includes instructions for generating second process data by processing the second sensor data; and determining whether the electronic device is in a stable state includes making a determination based on the first process data and the second process data.

[0128] Example 17. A device according to one of Examples 9 to 16, wherein the first motion sensor includes an accelerometer and the second motion sensor includes a gyroscope.

[0129] Example 18. A device according to one of Examples 9 to 17, wherein the instructions when executed further cause the processor to: in response to determining that the electronic device has been kept on the stationary surface for a second predefined time, generate a first control signal to turn off the second motion sensor.

[0130] Example 19. An apparatus according to one of Examples 9 to 18, wherein the first predefined time is greater than the second predefined time.

[0131] Example 20. An apparatus according to one of Examples 9 to 19, wherein the instruction, when executed, further causes the processor to: generate a second control signal to turn on the second motion sensor in response to determining that the additional first sensor data exceeds a predefined threshold; turn on the second motion sensor in response to receiving the second control signal; generate additional second sensor data by the second motion sensor over a subsequent acquisition time window; and generate second process data by processing the additional second sensor data.

[0132] Those skilled in the art will further appreciate that the various illustrative logic blocks, modules, circuits, and algorithms described in relation to the embodiments disclosed herein can be implemented as electronic hardware, instructions stored in a memory or another computer-readable medium and executed by a processor or other processing device, or a combination of the two. The equipment and processing system described herein can be adopted as examples in any circuit, hardware component, integrated circuit (IC), or IC chip. The memory disclosed herein can be a memory of any type and size, and can be configured to store any type of required information. In order to clearly illustrate this interchangeability, various illustrative components, blocks, modules, circuits, and steps have been described above from their functional aspects. How to implement this functionality depends on specific applications, design choices, and / or design constraints imposed on the entire system. A skilled person can implement the described functionality in different ways for each specific application, but this implementation decision should not be interpreted as causing deviations from the scope of the present invention.

[0133] The various illustrative logic blocks, modules, and circuits described in connection with the embodiments disclosed herein may be implemented or performed with a processor, a digital signal processor (DSP), an application specific integrated circuit (ASIC), a field programmable gate array (FPGA), or other programmable logic device, discrete gate or transistor logic, discrete hardware components, or any combination designed to perform the functions described herein. The processor may be a microprocessor, but in an alternative, the processor may be any conventional processor, controller, microcontroller, or state machine. The processor may also be implemented as a combination of computing devices, for example, a combination of a DSP and a microprocessor, a plurality of microprocessors, a combination of one or more microprocessors and a DSP core, or any other such configuration.

[0134] The embodiments disclosed herein may be embodied in hardware and in instructions stored in hardware, and may reside, for example, in a random access memory (RAM), flash memory, read-only memory (ROM), electrically programmable ROM (EPROM), electrically erasable programmable ROM (EEPROM), registers, hard disks, removable disks, CD-ROMs, or any other form of computer-readable medium known in the art. An exemplary storage medium is coupled to a processor so that the processor can read information from the storage medium and write information to the storage medium. In an alternative, the storage medium may be an integral part of the processor. The processor and the storage medium may reside in an ASIC.

[0135] Although the present invention has been described with reference to illustrative embodiments, this description is not intended to be interpreted in a limiting sense. Various modifications and combinations of the illustrative embodiments and other embodiments of the present invention will be apparent to those skilled in the art after reference to this description. Therefore, the appended claims are intended to cover any such modifications or embodiments.

Claims

1. A method comprising: Generating first sensor data by a first motion sensor of the electronic device over an acquisition time window; determining whether the electronic device is located on a stationary surface, on a semi-stationary surface, or in motion by processing the first sensor data to generate first process data; determining, based on the first process data, whether the electronic device is in a stable state, the stable state indicating whether the electronic device has been on a stationary surface for a first predefined time; as well as The processing of the first sensor data is stopped in response to determining that the electronic device has been in the stable state for a second predefined time.

2. The method according to claim 1, further comprising: In response to determining that the electronic device has been in the stable state for the second predefined time, additional first sensor data is generated by the first motion sensor over a subsequent acquisition time window.

3. The method according to claim 2, further comprising: In response to determining that the additional first sensor data exceeds a predefined threshold, determining whether the electronic device is located on the stationary surface by processing the additional first sensor data to generate first process data.

4. The method according to claim 1, further comprising: generating second sensor data by a second motion sensor of the electronic device over the acquisition time window; wherein determining whether the electronic device is located on the stationary surface comprises generating second process data by processing the second sensor data; and Determining whether the electronic device is in the stable state includes making a determination based on the first process data and the second process data. The method of claim 4 , wherein the first motion sensor comprises an accelerometer and the second motion sensor comprises a gyroscope.

6. The method according to claim 4, further comprising: In response to determining that the electronic device has remained on the stationary surface for the second predefined time, a first control signal is generated to turn off the second motion sensor. The method according to claim 6 , wherein the first predefined time is greater than the second predefined time.

8. The method according to claim 6, further comprising: In response to determining that the additional first sensor data exceeds a predefined threshold, generating a second control signal to turn on the second motion sensor; In response to receiving the second control signal, turning on the second motion sensor; generating, by the second motion sensor, additional second sensor data over a subsequent acquisition time window; as well as The second process data are generated by processing the additional second sensor data.

9. An electronic device, comprising: a first motion sensor configured to generate first sensor data indicative of a first type of movement of the electronic device; A processor coupled to a memory storing instructions to be executed in the processor, the instructions when executed causing the processor to: generating the first sensor data over an acquisition time window using the first motion sensor; determining whether the electronic device is located on a stationary surface by processing the first sensor data to generate first process data; determining, based on the first process data, whether the electronic device is in a stable state, the stable state indicating whether the electronic device has been on a stationary surface for a first predefined time; as well as The processing of the first sensor data is stopped in response to determining that the electronic device has been in the stable state for a second predefined time.

10. The device according to claim 9, further comprising: a first feature detection circuit configured to determine at least one first orientation-independent feature based on the first sensor data, wherein the at least one first orientation-independent feature comprises an average crossing value associated with an acquisition time window; a first classification circuit configured to determine whether the electronic device is stationary or in motion based on the at least one first orientation-independent feature; as well as A second classification circuit is configured to determine whether the electronic device is on a stationary surface or a semi-stationary surface based on the at least one first orientation-independent feature in response to the first classification circuit determining that the electronic device is stable.

11. The device according to claim 10, further comprising: A status monitor, the status monitor being configured to: In response to determining that the electronic device is on the semi-stationary surface, generating a control signal to increase a fan speed of the electronic device and to reduce RF power of the electronic device, In response to determining that the electronic device is on a stationary surface, generating a control signal to increase a clock frequency of the electronic device and decrease a fan speed of the electronic device, and increase the RF power of the electronic device, and In response to determining that the electronic device is in motion, a control signal is generated to reduce the clock frequency of the electronic device and reduce the fan speed of the electronic device.

12. The device according to claim 10, further comprising: a second motion sensor configured to generate second sensor data indicative of the first type of movement of the electronic device; as well as A second feature detection circuit is configured to determine at least one second orientation-independent feature based on the second sensor data, wherein the at least one second orientation-independent feature comprises an average crossing value associated with the acquisition time window.

13. The device of claim 12, wherein determining whether the electronic device is stable or in motion is based on the at least one first orientation-independent feature and the at least one second orientation-independent feature.

14. The apparatus of claim 9, wherein the instructions, when executed, further cause the processor to: In response to determining that the electronic device has been in the stable state for the second predefined time, additional first sensor data is generated by the first motion sensor over a subsequent acquisition time window.

15. The apparatus of claim 14, wherein the instructions, when executed, further cause the processor to: In response to determining that the additional first sensor data exceeds a predefined threshold, determining whether the electronic device is located on the stationary surface by processing the additional first sensor data to generate the first process data.

16. The apparatus of claim 9, wherein the instructions, when executed, further cause the processor to: generating second sensor data over the acquisition time window using a second motion sensor of the electronic device; wherein the instructions for determining whether the electronic device is located on a stationary surface include instructions for generating second process data by processing the second sensor data; and Determining whether the electronic device is in a stable state includes making a determination based on the first process data and the second process data.

17. The device of claim 16, wherein the first motion sensor comprises an accelerometer and the second motion sensor comprises a gyroscope.

18. The apparatus of claim 16, wherein the instructions, when executed, further cause the processor to: In response to determining that the electronic device has remained on the stationary surface for the second predefined time, a first control signal is generated to turn off the second motion sensor. The apparatus of claim 18 , wherein the first predefined time is greater than the second predefined time.

20. The apparatus of claim 19, wherein the instructions, when executed, further cause the processor to: In response to determining that the additional first sensor data exceeds a predefined threshold, generating a second control signal to turn on the second motion sensor; turning on the second motion sensor in response to receiving the second control signal; generating, by the second motion sensor, additional second sensor data over a subsequent acquisition time window; as well as The second process data are generated by processing the additional second sensor data.

Images