In-sensor collision detection with motion sensor

By combining motion sensor data and accelerometer data, using a finite state machine to independently detect vehicle collisions, the problem of existing systems' dependence on external signals is solved, and low power consumption and efficient collision detection is achieved.

CN118953265APending Publication Date: 2024-11-15STMICROELECTRONICS INT NV
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
CN202410588821.5
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Priority Date
2023-05-15
Filing Date
2024-05-13
Publication Date
2024-11-15

AI Technical Summary

Technical Problem

Existing automotive collision detection systems rely on signals received by external devices, which makes it impossible to determine whether a vehicle has collided in some cases and consumes a lot of power.

Method used

Activity identification based on motion sensor data and high impact detection of accelerometer data are adopted, combined with a finite state machine, independently detecting whether a vehicle has a collision, reducing dependence on external signals and reducing power consumption.

Benefits of technology

It realizes accurate detection of vehicle collision events without relying on external signals, and significantly reduces power consumption and improves system independence and efficiency.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN118953265A_ABST
    Figure CN118953265A_ABST
Patent Text Reader

Abstract

The invention relates to in-sensor collision detection with a motion sensor. A collision detection system includes first and second sensors and a processor. The processor receives first sensor data output by the first sensor, determines whether the first sensor data indicates a first class or a second class, and outputs an enable signal to the second sensor if the first sensor data indicates the second class, receiving second sensor data output by the second sensor after outputting the enable signal, determining whether the second sensor data indicates a high acceleration value, determining whether the first sensor data is indicative of the first class within a predetermined amount of time after the second sensor data is determined to be indicative of the high acceleration value, and outputting a signal indicating that a collision has occurred in response to determining that the first sensor data is indicative of the first class within the predetermined amount of time after the second sensor data is determined to be indicative of the high acceleration value.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present disclosure relates generally to collision detection systems, and more particularly to collision detection systems that can detect collisions without receiving input from an external device. Background Art

[0002] Conventionally, a car crash detection (CCD) feature is used to automatically assist a user in the event that the user is involved in a car accident. For example, if a car crash is detected, the CCD feature typically calls for medical assistance.

[0003] Existing automobile collision detection systems use so-called "high-g" accelerometers to measure peak accelerations caused by accidents, which may reach more than 100g (i.e., one hundred times the force of gravity). In addition, existing automobile collision detection systems included in smartphone devices use information from multiple external devices, including a global positioning system (GPS) receiver, multiple motion sensors, and microphones. For example, existing automobile collision detection systems use location information from a GPS receiver to determine whether a vehicle (e.g., a car) is moving, use data from a motion sensor to evaluate the peak acceleration during the impact from the accident, and use data from a microphone to capture audio signals of the accident. If the GPS receiver cannot receive GPS signals from GPS satellites, the existing automobile collision detection system may not be able to determine whether the vehicle is moving, and therefore may not be able to determine whether an accident has occurred. If the existing automobile collision detection system operates on a smartphone that is highly isolated from the surrounding environment (e.g., in a user's pocket), the existing automobile collision detection system may not be able to obtain audio data indicating that an accident has occurred. Therefore, existing automobile collision detection systems rely on signals received from external devices, which may not always be available or may not always be ideal for automobile collision detection.

[0004] In addition, the existing automobile collision detection system fuses information at the system level (eg, Android service). Therefore, the existing automobile collision detection system consumes a relatively large amount of power. Summary of the invention

[0005] The present disclosure provides collision detection systems, methods and processor readable media that do not rely on signals received from external devices and therefore do not suffer from the disadvantages of conventional CCD systems described above. In addition, the present disclosure provides collision detection systems, methods and processor readable media that consume less power than conventional CCD systems.

[0006] According to one or more embodiments of the present disclosure, a collision detection system is provided, which can be used in a smart phone to determine whether the smart phone is involved in a collision (e.g., a car collision) using a combination of activity recognition based on motion sensor data and high impact detection based on accelerometer data. In addition, according to one or more embodiments of the present disclosure, a collision detection system is provided, which can be used in a vehicle control unit (e.g., a car control unit) to determine whether the vehicle is involved in a collision using a combination of vehicle stationary detection based on motion sensor data and high impact detection based on accelerometer data.

[0007] A collision detection system according to the present disclosure is characterized by comprising: a first sensor that outputs first sensor data in operation; a second sensor that outputs second sensor data in operation; and a processor coupled to the first sensor and the second sensor. The processor in operation: receives the first sensor data, determines whether the first sensor data indicates a first class or a second class, outputs an enable signal to the second sensor in response to determining that the first sensor data indicates the second class, receives the second sensor data after outputting the enable signal to the second sensor, determines whether the second sensor data indicates a detected acceleration value greater than or equal to a threshold acceleration value, determines whether the first sensor data indicates the first class within a predetermined amount of time after the second sensor data is determined to indicate a detected acceleration value greater than or equal to the threshold acceleration value, and outputs a signal indicating that a collision has occurred in response to determining that the first sensor data indicates the first class within the predetermined amount of time after the second sensor data is determined to indicate that the detected acceleration value is greater than or equal to the threshold acceleration value.

[0008] The method for operating a collision detection system according to the present disclosure is characterized in that it includes: receiving, by a processor, first sensor data output by a first sensor; determining, by the processor, whether the first sensor data indicates a first category or a second category; in response to determining that the first sensor data indicates the second category, outputting, by the processor, an enable signal to a second sensor; receiving, by the processor, second sensor data output by the second sensor after the enable signal is output to the second sensor; determining, by the processor, whether the second sensor data indicates a detected acceleration value greater than or equal to a threshold acceleration value; determining, by the processor, whether the first sensor data indicates the first category within a predetermined time period after determining that the second sensor data indicates that the detected acceleration value is greater than or equal to the threshold acceleration value; and in response to determining that the first sensor data indicates the first category within the predetermined amount of time after determining that the second sensor data indicates that the detected acceleration value is greater than or equal to the threshold acceleration value, outputting, by the processor, a signal indicating that a collision has occurred.

[0009] A non-transitory processor-readable storage medium according to the present disclosure may be characterized as storing instructions which, when executed by a processor, cause the processor to: determine whether first sensor data output by a first sensor indicates a first category or a second category, output an enable signal to a second sensor in response to the first sensor data being determined to indicate the second category, determine whether second sensor data output by the second sensor after the enable signal is output to the second sensor indicates a detected acceleration value greater than or equal to a threshold acceleration value, determine whether the first sensor data indicates the first category within a predetermined amount of time after determining that the second sensor data indicates a detected acceleration value greater than or equal to the threshold acceleration value, and output a signal indicating that a collision has occurred in response to determining that the first sensor data indicates the first category within the predetermined amount of time after determining that the second sensor data indicates that the detected acceleration value is greater than or equal to the threshold acceleration value. BRIEF DESCRIPTION OF THE DRAWINGS

[0010] Non-limiting and non-exhaustive embodiments are described with reference to the following figures.In the drawings, like reference numerals refer to like parts throughout the various figures unless otherwise specified.

[0011] For a better understanding of the present disclosure, reference will be made to the following detailed description, which is to be read in conjunction with the accompanying drawings:

[0012] Figure 1A , 1B 1C and 1C show examples of collision detection systems according to embodiments of the present disclosure.

[0013] Figure 2A and 2BAn example of an implementation including a collision detection system according to an embodiment of the present disclosure is shown.

[0014] Figure 3A and 3B An example of a block diagram of parts of a collision detection system according to an embodiment of the present disclosure is shown.

[0015] Figure 4 An example of a finite state machine used by a collision detection system according to an embodiment of the present disclosure is shown.

[0016] Figure 5 Another example of a finite state machine used by a collision detection system according to an embodiment of the present disclosure is shown.

[0017] Figure 6 An example of a processing flow in a collision detection system according to an embodiment of the present disclosure is shown. DETAILED DESCRIPTION

[0018] According to an embodiment of the present disclosure, activity recognition based on motion sensor data can be used, based on a combination of high-shock detection and accelerometer data and a state machine. In order to detect a collision event, a high impact must be detected when the activity recognition block (or vehicle stationary detection block) is detecting the "driving" class (or "moving" class). Enable the high impact detector block when necessary to save power. After an accident, the vehicle suddenly becomes stationary. Therefore, the activity recognition block (or vehicle stationary detection block) will detect the "other" class (or "stationary" class). After the high impact is detected, a timer is enabled to give the activity recognition block (or vehicle stationary detection block) enough time to detect the "other" class (or "stationary" class). If the timer reaches a timeout value, which is higher than the waiting time required for the activity recognition block to detect the "other" class or the vehicle stationary detection block to detect the "stationary" class, the high impact event is discarded because it is considered a false positive.

[0019] Figure 1A , Figure 1B and Figure 1C An example of a collision detection system according to an embodiment of the present disclosure is shown. More specifically, Figure 1A An example of a collision detection system 100A is shown, which includes an integrated circuit device 101A having a sensor 102A, an integrated circuit device 103A having a sensor 104A, and an integrated circuit device 105A having a microprocessor 106A. The first integrated circuit device 101A, the integrated circuit device 103A, and the integrated circuit device 105A are different from each other.

[0020] The microprocessor 106A provides control signals 108A to the sensor 102A, for example, including control signals for causing the sensor 102A to enter a sleep state, control signals for causing the sensor 102A to wake up or exit a sleep state, and control signals for setting one or more detection thresholds in the sensor 102A. The sensor 102A provides data signals 110A to the microprocessor 106A, for example, including sensor data for each axis of a first accelerometer, a second accelerometer, and / or a gyroscope included in the sensor 102A. In one or more embodiments, the sensor 102A provides sensor data at a relatively low rate, for example, a rate greater than or equal to 25 Hertz (Hz) (i.e., 25 times per second). The microprocessor 106A provides control signals 112A to the sensor 104A, for example, including control signals for causing the sensor 104A to enter a sleep state, control signals for causing the sensor 104A to wake up or exit a sleep state, and control signals for setting one or more detection thresholds in the sensor 104A. The sensor 104A provides a data signal 114A to the microprocessor 106A, for example, including sensor data for each axis of the first accelerometer, the second accelerometer, and / or the gyroscope included in the sensor 104A. In one or more embodiments, the sensor 104A provides sensor data at a relatively high rate, for example, a rate greater than or equal to 400 Hz (i.e., 400 times per second). The microprocessor 106A includes at least one processor and at least one memory device storing processor-readable instructions, which, when executed by the at least one processor, perform the functions of the microprocessor 106A described herein, as described below in conjunction with Figure 2A , Figure 2B , Figure 3A , Figure 3B , Figure 4 , Figure 5 and Figure 6 Explain the functionality.

[0021] In one or more embodiments, sensor 102A is a 6-axis inertial measurement unit (IMU) model LSM6DSV16X from STMicroelectronics or an ultra-low power accelerometer model LIS2DUX12 from STMicroelectronics. In one or more embodiments, sensor 104A is a high-g, low-power 3-axis digital accelerometer model H3LIS331DL from STMicroelectronics. In one or more embodiments, microprocessor 106A is a microprocessor model STM32MP1 from STMicroelectronics. Other models of devices may be used for sensor 102A, sensor 104A, and microprocessor 106A within the scope of the present disclosure.

[0022] Figure 1B An example of a collision detection system 100B is shown, which includes an integrated circuit device 101B having a sensor 102B and a microcontroller 106B, and an integrated circuit device 103B having a sensor 104B. Figure 1B The collision detection system 100B shown is similar in many relevant respects to Figure 1A The collision detection system 100A is shown except that the sensor 102B and the microcontroller 106B are included in the same integrated circuit device 101B, while the sensor 104B is included in an integrated circuit device 103B that is different from the integrated circuit device 101B.

[0023] Sensor 102B and sensor 104B are similar to sensor 102A and sensor 104A described above in many relevant aspects. In one or more embodiments, microcontroller 106B is used to implement a machine learning core (MLC) or a finite state machine (FSM) of integrated circuit device 101B. In another embodiment, microcontroller 106B is an intelligent sensor processing unit (ISPU) of integrated circuit device 101B.

[0024] The microcontroller 106B provides a control signal 108B similar to the control signal 108A described above to the sensor 102B. The sensor 102B provides a data signal 110B similar to the control signal 110A described above to the microcontroller 106B. The microcontroller 106B provides a control signal 112B similar to the control signal 112A described above to the sensor 104B. The sensor 104B provides a data signal 114B similar to the control signal 114A described above to the microcontroller 106B. The microcontroller 106B includes at least one processor and at least one memory device storing processor-readable instructions that, when executed by the at least one processor, perform the functions of the microcontroller 106B described herein, as described below in conjunction with Figure 2A , Figure 2B , Figure 3A , Figure 3B , Figure 4 , Figure 5 and Figure 6 to explain the function.

[0025] Figure 1C An example of a collision detection system 100C is shown, which includes an integrated circuit device 101C having a sensor 102C, a sensor 104C, and a microcontroller 106C. Figure 1C The collision detection system 100C shown in FIG. 1 is similar in many relevant respects to the Figure 1A The collision detection system 100A shown in FIG. Figure 1BThe collision detection system 100B is shown in FIG. 1 , except that the sensor 102C, the sensor 104C and the microcontroller 106C are included in the same integrated circuit device 101C.

[0026] Sensor 102C and sensor 104C are similar to sensor 102A and sensor 104A described above in many relevant aspects. In one or more embodiments, microcontroller 106C is used to implement a machine learning core (MLC) or finite state machine (FSM) of integrated circuit device 101C. In another embodiment, microcontroller 106C is an intelligent sensor processing unit (ISPU) of integrated circuit device 101C.

[0027] The microcontroller 106C provides a control signal 108C similar to the control signal 108A described above to the sensor 102C. The sensor 102C provides a data signal 110C similar to the control signal 110A described above to the microcontroller 106C. The microcontroller 106C provides a control signal 112C similar to the control signal 112A described above to the sensor 104C. The sensor 104C provides a data signal 114C similar to the control signal 114A described above to the microcontroller 106C. The microcontroller 106C includes at least one processor and at least one memory device storing processor-readable instructions that, when executed by the at least one processor, perform the functions of the microcontroller 106C described herein, hereinafter in conjunction with Figure 2A , Figure 2B , Figure 3A , Figure 3B , Figure 4 , Figure 5 and Figure 6 to explain the function.

[0028] Figure 1A The collision detection system 100A shown, Figure 1B The collision detection system 100B and Figure 1C The illustrated collision detection system 100C is generally referred to herein as the collision detection system 100 unless reference is made to a specific one of these systems.

[0029] Figure 2A and Figure 2B An example of an implementation of a collision detection system including an embodiment of the present disclosure is shown. More specifically, Figure 2AA smartphone 200 including the collision detection system 100, a microprocessor 202, and a communication circuit device 204 is shown. The microprocessor 202 includes at least one processor and at least one memory device storing processor-readable instructions, which, when executed by the at least one processor, perform the functions of the microprocessor 202 described herein. The communication circuit device 204 includes a transmitter, a receiver, and an antenna. In one or more embodiments, the transmitter and the receiver included in the communication circuit device 204 are configured to communicate according to a cellular communication standard from the Third Generation Partnership Project (3GPP). For example, the transmitter and the receiver included in the communication circuit device 204 are configured to communicate according to the Long Term Evolution (LTE) communication standard, which is a fourth generation (4G) wireless communication standard, and / or to communicate according to the fifth generation (5G) New Radio (NR) wireless communication standard.

[0030] The microprocessor 202 provides control signals for configuring the collision detection system 100 to the collision detection system 100. When a collision event is detected, the collision detection system 100 provides a collision event signal 208 to the microprocessor 202. In response, the microprocessor 202 provides a signal 210 to the communication circuit device 204. The signal 210 causes the communication circuit device 204 to dial a telephone number that connects the smart phone 200 to a local medical center and / or emergency services (e.g., 911). In addition, the signal 210 may include voice data or other data (e.g., text data identifying the user of the smart phone 200). In one or more embodiments, the smart phone 200 includes a positioning device (e.g., a GPS receiver) (not shown), and the signal 210 includes data indicating the location (e.g., latitude and longitude) of the smart phone 200. The communication circuit device 204 provides a signal 212 to the microprocessor 200. The signal 212 may include audio and / or video data output to a speaker (not shown) and / or a display device (not shown) of the smart phone 200.

[0031] Figure 2B A vehicle 220 is shown, which includes a vehicle control unit 221 having a collision detection system 100, a microprocessor 222, and a communication circuit 224. The vehicle 220 may be, for example, a motorcycle, an automobile, a boat, or an airplane. The vehicle control unit 221 is connected to and controls various systems (not shown) of the vehicle 220. In one or more embodiments, the microprocessor 222 and the communication circuit 224 are similar in many relevant respects to the microprocessor 202 and the communication circuit device 204 described above.

[0032] Figure 3A and Figure 3B An example of a block diagram of various parts of a collision detection system according to an embodiment of the present disclosure is shown. More specifically, Figure 3A A method for using a smart phone (eg Figure 2A 300 is a block diagram of a processor (eg, microprocessor 106A or microcontroller 106B or 106C) used in an embodiment of a collision detection system 100 for a smartphone 200 as shown. Block diagram 300 includes an activity recognition block 302, a high impact detector block 304, and a state machine block 306.

[0033] In one example, the activity recognition block 302 is implemented in a machine learning core (MLC) of the sensor 102B (e.g., running at 30 Hz), the high impact detector block 304 is implemented in a finite state machine (FSM) of the sensor 102B (e.g., running at 960 Hz), and the state machine block 306 is implemented in a finite state machine (FSM) of the sensor 102B (e.g., running at 960 Hz). Based on the accelerometer norm and the norm over a predetermined time window (e.g., a time window of 60 samples, corresponding to 2 seconds), the accelerometer 102B may be used to generate a signal based on the accelerometer 102B norm and the accelerometer 102B norm. 2 (norm 2 ) signals to compute variance, peak-to-peak, and peak detector features. Decision tree training was performed on approximately 90 hours of collected logs, where the class "other" included scenarios where the user was stationary, walking (normal, slow, fast, stairs), running (jogging, sprinting), biking, and the class "driving" included scenarios where the user was driving a car. A very compact decision tree was used to avoid overfitting. Experiments have been conducted in which approximately 91% raw accuracy was achieved and approximately 94% meta-classification (5 consecutive windows, corresponding to 10s) accuracy was achieved. The total current consumption of sensor 102B is approximately 200 microamperes (μA), where the current consumption of the accelerometer operating in normal mode is approximately 110 μA, the current consumption of the MLC is approximately 5 μA, and the current consumption of the FSM is approximately 85 μA.

[0034] In the activity identification block 302, sensor data XL / G from an accelerometer and / or a gyroscope included in a sensor (e.g., sensor 102A, 102B, or 102C) is received, wherein the sensor data XL / G includes a value of linear acceleration for each axis of the accelerometer (e.g., specified in a number / g) and / or a value of angular velocity for each axis of the gyroscope included in the sensor (e.g., specified in a number / deg / s). In addition, in the activity identification block 302, it is determined whether the sensor data XL / G indicates a non-driving class or a driving class, and a signal indicating that the sensor data is classified as a non-driving class or a driving class is output to the state machine block 306.

[0035] In one or more embodiments, artificial intelligence (AI) techniques such as machine learning techniques are used in the activity recognition block 302. For example, the smartphone 200 including the collision detection system 100 is used in various experiments, where the sensor data XL / G is augmented to include an indication of whether the smartphone is being used while driving, and statistical methods are used to train an algorithm to classify whether the sensor data XL / G indicates a non-driving or driving class.

[0036] In the high impact detector block 304, sensor data XLsec from an accelerometer included in a sensor (e.g., sensor 104A, 104B, or 104C) is received, wherein the sensor data XLsec includes a value of linear acceleration for each axis of the accelerometer included in the sensor. In one implementation, the sensor is configured to verify whether at least one of the values ​​of the sensor data XLsec exceeds a threshold value (e.g., 30g). In another implementation, the sensor is configured to continuously output the sensor data XLsec, and the processor determines whether at least one of the values ​​of the sensor data XLsec exceeds the threshold value. The same threshold value may be used for each axis of the accelerometer, or different threshold values ​​may be used for the axes of the accelerometer. In the high impact detector block 304, if the sensor data XLsec indicates that a high impact event has been detected (e.g., the value of the linear acceleration of at least one axis of the accelerometer exceeds the threshold value), a high impact signal is provided to the state machine block 306. In one embodiment, the check may be performed on each axis of the accelerometer sensor data. For example, using a three-axis accelerometer sensor, if at least one value in the accelerometer axis exceeds a configurable threshold (e.g., 15g), a high shock signal is provided to the state machine block 306. In another embodiment, a check may be performed on the norm of the accelerometer sensor data. For example, if the norm of the accelerometer data exceeds a configurable threshold (e.g., 25g), a high shock signal is provided to the state machine block 306. When a high-g accelerometer sensor (e.g., sensor 104A) is used, the threshold may be higher.

[0037] In the state machine block 306, a finite state machine is implemented based on the signals output by the activity recognition block 302 and the high impact detector block 304. Figure 4Explained in more detail, in the state machine block 306, an enable / disable signal is output to the high impact detector block 304. If the signal output by the activity recognition block 302 indicates a non-driving class, the enable / disable signal has a state (e.g., a first voltage level or value) indicating that the high impact detector block 304 is disabled. If the signal output by the activity recognition block 302 indicates a driving class, the enable / disable signal has a state (e.g., a second voltage level or value) indicating that the high impact detector block 304 is enabled. When the enable / disable signal indicates that the high impact detector block 304 is disabled, the high impact detector block 304 does not output a high impact signal. When the enable / disable signal indicates that the high impact detector block 304 is enabled, the high impact detector block 304 outputs a high impact signal if a high impact event is detected.

[0038] In the state machine block 306, if the signal output by the activity recognition block 302 indicates the driving class and the signal output by the high impact detector block 304 indicates a high impact event, it is determined whether the signal output by the activity recognition block 302 changes to indicate the non-driving class within a predetermined period of time. If the signal output by the activity recognition block 302 changes to indicate a class other than the driving class within the predetermined period of time, a collision event signal is output. If the signal output by the activity recognition block 302 does not change to indicate a class other than the driving class within the predetermined period of time, a collision event signal is not output.

[0039] Figure 3B It is shown that in the case of a vehicle (eg Figure 2B 220) is shown in FIG. 2 . The block diagram 310 is a diagram of a processor (e.g., microprocessor 106A or microcontroller 106B or 106C) used in an embodiment of a collision detection system 100 for a vehicle 220 (e.g., FIG. 220). The block diagram 310 includes a vehicle stationary detection block 312, a high impact detector block 314, and a state machine block 316. The vehicle stationary detection block 312 is similar to the activity recognition block 302 described above in many relevant respects, except that in the vehicle stationary detection block 312, a determination is made as to whether the sensor data XL / G indicates a stationary class or a moving class, and a signal indicating whether the sensor data is classified as a stationary class or a moving class is output to the state machine block 316. The high impact detector block 314 is similar to the high impact detector block 304 described above in many relevant respects.

[0040] State machine block 316 is similar in many relevant respects to state machine block 306 described above, except that in state machine block 316, if the signal output by vehicle stationary detection block 312 indicates a motion class and the signal output by high impact detector block 314 indicates a high impact event, then a determination is made as to whether the signal output by vehicle stationary detection block 312 changes to indicate a stationary class within a predetermined period of time. If the signal output by vehicle stationary detection block 312 changes to indicate a stationary class within a predetermined period of time, then a collision event signal is output. If the signal output by vehicle stationary detection block 312 does not change to indicate a stationary class within a predetermined period of time, then no collision event signal is output. Figure 5 State machine block 316 is explained in more detail.

[0041] Figure 4 FIG. 4 shows an example of a finite state machine 400 used by a collision detection system according to an embodiment of the present disclosure. The finite state machine 400 corresponds to Figure 3A 300. Initially, the finite state machine 400 is in the activity identification state 402, where the processing of the activity identification block 302 is performed.

[0042] If the signal output by the activity recognition block 302 indicates a driving class, the finite state machine 400 transitions to an enabled high impact detector state 404, in which the signal output by the state machine block 306 indicates that the high impact detector block 304 is enabled. The finite state machine 400 then transitions to an activity recognition processing and high impact detector state 406, in which the processing of the activity recognition block 302 and the processing of the high impact detector block 304 are performed. If the signal output by the activity recognition block 302 changes to indicate a non-driving class, the finite state machine 400 transitions to a disabled high impact detector state 408, in which the signal output by the state machine block 306 indicates that the high impact detector block 304 is disabled. The finite state machine 400 then transitions back to the activity recognition state 402.

[0043] When the finite state machine 400 is in the activity recognition processing and high impact detector state 406, if the signal output by the activity recognition block 302 has not changed to indicate a class other than the driving class, and the signal output by the high impact detector block 304 indicates a high impact event, the finite state machine 400 transitions to the enable timer state 410, in which the timer is enabled. The finite state machine 400 then transitions to the check timer and activity recognition state 412, in which a process of periodically checking the timer to determine if the timer has reached a timeout value (e.g., 15 seconds) is performed, and the processing of the activity recognition block 302 is performed. The timeout value must be higher than the waiting time (e.g., 10 seconds) required for the activity recognition block 302 to detect a class other than driving.

[0044] When the finite state machine 400 is in the check timer and activity recognition state 412, if the timer reaches the timeout value, the finite state machine 400 transitions to the disable high impact detector state 408 and then transitions to the activity recognition state 402. If the timer does not reach the timeout value, and the signal output by the activity recognition block 302 changes to indicate a class other than driving, a signal indicating a collision event is output to the microprocessor 202, which causes the microprocessor 202 to place a call to emergency services and request assistance, for example, via the communication circuit device 204.

[0045] Figure 5 FIG. 5 shows an example of a finite state machine 500 used by a collision detection system according to an embodiment of the present disclosure. The finite state machine 500 corresponds to Figure 3B Block diagram 310 is shown. Initially, finite state machine 500 is in vehicle stationary detection state 502, where the processing of vehicle stationary detection block 312 is performed.

[0046] If the signal output by the vehicle stationary detection block 312 indicates a motion class, the finite state machine 500 transitions to an enabled high impact detector state 504, in which the signal output by the state machine block 316 indicates that the high impact detector block 314 is enabled. The finite state machine 500 then transitions to a vehicle stationary detection processing and high impact detector state 506, in which processing by the vehicle stationary detection block 312 and processing by the high impact detector block 314 are performed. If the signal output by the vehicle stationary detection block 312 changes to indicate a stationary class, the finite state machine 500 transitions to a disabled high impact detector state 508, in which the signal output by the state machine block 316 indicates that the high impact detector block 314 is disabled. The finite state machine 500 then transitions back to the vehicle stationary detection state 502.

[0047] When the finite state machine 500 is in the vehicle stationary detection processing and high impact detector state 506, if the signal output by the vehicle stationary detection block 312 has not changed to indicate a stationary class, and the signal output by the high impact detector block 314 indicates a high impact event, the finite state machine 500 transitions to the enable timer state 510, in which the timer is enabled. The finite state machine 500 then transitions to the check timer and vehicle stationary detection state 512, where a process of periodically checking the timer to determine if the timer has reached a timeout value (e.g., 15 seconds) is performed, and the processing of the vehicle stationary detection block 312 is performed. The timeout value must be higher than the waiting time (e.g., 10 seconds) required for the vehicle stationary detection block 312 to detect the stationary class.

[0048] When the finite state machine 500 is in the Check Timer and Vehicle Stationary Detection State 512, if the timer reaches the timeout value, the finite state machine 500 transitions to the Disable High Impact Detector State 508 and then transitions to the Vehicle Stationary Detection State 502. If the timer does not reach the timeout value and the signal output by the Vehicle Stationary Detection block 312 changes to indicate a stationary class, a signal indicating a collision event is output to the microprocessor 222, which causes the microprocessor 222 to place a call to emergency services and request assistance, for example, via the communication circuit 224.

[0049] Figure 6 An example of a processing flow 600 in a collision detection system according to an embodiment of the present disclosure is shown. The processing flow 600 includes a process 602 in which a signal output by the activity recognition block 302 is checked to determine whether it indicates a driving class. The processing flow 600 also includes a process 604 in which a signal output by the state machine block 306 indicates that the high impact detector block 304 is enabled, and a signal output by the activity recognition block 302 is checked to determine whether it indicates a class other than the driving class to disable the high impact detector block 304 and reset the processing flow 600. In addition, the processing flow 600 includes a process 606 in which a timer is enabled, and a signal output by the activity recognition block 302 is checked to determine whether it indicates a non-driving class before the timer reaches a configured timeout value. In addition, the processing flow 600 includes a process 608 of outputting a collision event signal.

[0050] The various embodiments described above can be combined to provide additional embodiments. All U.S. patents, U.S. patent application publications, U.S. patent applications, foreign patents, foreign patent applications, and non-patent publications mentioned in this specification and / or listed in the application data sheet are incorporated herein by reference in their entirety. If necessary, aspects of the embodiments can be modified to adopt the concepts of the various patents, applications, and publications to provide additional embodiments.

[0051] These and other changes can be made to the embodiments in light of the above detailed description. Generally, in the following claims, the terms used should not be interpreted as limiting the claims to the specific embodiments disclosed in the specification and claims, but should be interpreted to include all possible embodiments and the full range of equivalents to which these claims are authorized. Therefore, the claims are not limited by this disclosure.

Claims

1. A collision detection system, include: a first sensor, the first sensor being operable to output first sensor data; a second sensor, the second sensor being operable to output second sensor data; as well as a processor coupled to the first sensor and the second sensor, wherein the processor, in operation: receiving the first sensor data, determining whether the first sensor data is indicative of a first category or indicative of a second category, outputting an enable signal to the second sensor in response to determining that the first sensor data indicates the second category, receiving the second sensor data after the enable signal is output to the second sensor, determining whether the second sensor data indicates a detected acceleration value that is greater than or equal to a threshold acceleration value, determining, within a predetermined amount of time after the second sensor data is determined to indicate a detected acceleration value that is greater than or equal to the threshold acceleration value, whether the first sensor data indicates the first category; as well as In response to determining that the first sensor data indicates the first category within the predetermined amount of time after the second sensor data is determined to indicate a detected acceleration value greater than or equal to the threshold acceleration value, outputting a signal indicating that a collision has occurred.

2. The collision detection system of claim 1 , wherein in response to determining that the first sensor data indicates the second category within the predetermined amount of time after the second sensor data is determined to indicate a detected acceleration value greater than or equal to the threshold acceleration value, the processor is operative not to output a signal indicating that the collision has occurred.

3. The collision detection system according to claim 1, in: The second sensor operates while the enable signal is output by the processor, and When the enable signal is not output by the processor, the second sensor does not operate.

4. The collision detection system according to claim 1, in: the first sensor outputting the first sensor data at a first rate, The second sensor outputs the second sensor data at a second rate, and The second rate is greater than the first rate.

5. The collision detection system according to claim 1, in: The first sensor is included in a first integrated circuit device, The second sensor is included in a second integrated circuit device different from the first integrated circuit device, and The processor is located in a third integrated circuit device that is different from the first integrated circuit device and the second integrated circuit device.

6. The collision detection system according to claim 1, in: The first sensor and the processor are included in a first integrated circuit device, and The second sensor is included in a second integrated circuit device different from the first integrated circuit device.

7. The collision detection system of claim 1, wherein the first sensor, the second sensor, and the processor are included in a same integrated circuit device.

8. The collision detection system according to claim 1, in: The first sensor comprises a first accelerometer, The second sensor includes a second accelerometer, and The second accelerometer, when in operation, detects a greater amount of acceleration than the first accelerometer.

9. The collision detection system of claim 8, wherein the first sensor comprises a gyroscope, and wherein the first sensor data comprises an output from the first accelerometer and an output from the gyroscope. 10 . The collision detection system of claim 1 , wherein the first class is a class different from a driving class and the second class is a driving class, or the first class is a stationary class and the second class is a moving class.

11. A method of operating a collision detection system, the method include: receiving, by the processor, first sensor data output by the first sensor; determining, by the processor, whether the first sensor data indicates a first category or a second category; outputting, by the processor, an enable signal to a second sensor in response to determining that the first sensor data indicates the second category; receiving, by the processor, second sensor data output by the second sensor after the enable signal is output to the second sensor; determining, by the processor, whether the second sensor data indicates a detected acceleration value that is greater than or equal to a threshold acceleration value; determining, by the processor, whether the first sensor data indicates the first category within a predetermined amount of time after the second sensor data is determined to indicate a detected acceleration value that is greater than or equal to the threshold acceleration value; as well as In response to determining that the first sensor data indicates the first category within the predetermined amount of time after the second sensor data is determined to indicate a detected acceleration value greater than or equal to the threshold acceleration value, outputting, by the processor, a signal indicating that a collision has occurred.

12. The method according to claim 11, further comprising: include: determining, by the processor, that the first sensor data indicates the second category within the predetermined amount of time after the second sensor data is determined to indicate that the detected acceleration value is greater than or equal to the threshold acceleration value; wherein, in response to determining that the first sensor data indicates the second category within the expected amount of time after the second sensor data is determined to indicate that the detected acceleration value is greater than or equal to the threshold acceleration value, the processor does not output a signal indicating that the collision has occurred.

13. The method according to claim 11, further comprising: include: while the enabling signal is output by the processor, operated by the second sensor, Wherein, when the enabling signal is not output by the processor, the second sensor does not operate.

14. The method according to claim 11, further comprising: include: outputting, by the first sensor, the first sensor data at a first rate; as well as outputting the second sensor data at a second rate by the second sensor, The second rate is greater than the first rate.

15. The method according to claim 1, in: detecting a first amount of acceleration by a first accelerometer included in the first sensor; as well as detecting a second acceleration amount by a second accelerometer included in the second sensor, The second acceleration amount is greater than the first acceleration amount.

16. The collision detection system of claim 15, wherein the first sensor comprises a gyroscope, and wherein the first sensor data comprises an output from the first accelerometer and an output from the gyroscope.

17. The method according to claim 11, in, The first class is a class other than a driving class and the second class is a driving class, or the first class is a stationary class and the second class is a moving class.

18. A non-transitory processor-readable storage medium storing instructions which, when executed by a processor, cause the processor to: determining whether first sensor data output by the first sensor indicates the first category or the second category, outputting an enable signal to a second sensor in response to determining that the first sensor data indicates the second category, determining whether second sensor data output by the second sensor after the enable signal is output to the second sensor indicates a detected acceleration value that is greater than or equal to a threshold acceleration value, determining whether the first sensor data indicates the first category within a predetermined amount of time after the second sensor data is determined to indicate that the detected acceleration value is greater than or equal to the threshold acceleration value; and In response to determining that the first sensor data indicates the first category within the predetermined amount of time after determining that the second sensor data indicates that the detected acceleration value is greater than or equal to the threshold acceleration value, outputting a signal indicating that a collision has occurred.

19. The storage medium of claim 11, wherein the instructions, when executed by a processor, cause the processor to: determining, within the predetermined amount of time after the second sensor data is determined to indicate the detected acceleration value that is greater than or equal to the threshold acceleration value, that the first sensor data is indicative of the second category; in, The processor does not output the signal indicating that the collision has occurred if the first sensor data is determined to indicate the second category within the predetermined amount of time after the second sensor data is determined to indicate the detected acceleration value greater than or equal to the threshold acceleration value.

20. The storage medium according to claim 11, in, The first class is a class other than a driving class and the second class is a driving class, or the first class is a stationary class and the second class is a moving class.