WO2016085405A1 - Fall detection system, device and method - Google Patents
Fall detection system, device and method Download PDFInfo
- Publication number
- WO2016085405A1 WO2016085405A1 PCT/SG2014/000565 SG2014000565W WO2016085405A1 WO 2016085405 A1 WO2016085405 A1 WO 2016085405A1 SG 2014000565 W SG2014000565 W SG 2014000565W WO 2016085405 A1 WO2016085405 A1 WO 2016085405A1
- Authority
- WO
- WIPO (PCT)
- Prior art keywords
- processor
- accelerometer
- mobile device
- altitude
- threshold values
- Prior art date
Links
Classifications
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
- H04L67/00—Network arrangements or protocols for supporting network services or applications
- H04L67/01—Protocols
- H04L67/10—Protocols in which an application is distributed across nodes in the network
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04W—WIRELESS COMMUNICATION NETWORKS
- H04W4/00—Services specially adapted for wireless communication networks; Facilities therefor
- H04W4/60—Subscription-based services using application servers or record carriers, e.g. SIM application toolkits
-
- G—PHYSICS
- G08—SIGNALLING
- G08B—SIGNALLING OR CALLING SYSTEMS; ORDER TELEGRAPHS; ALARM SYSTEMS
- G08B21/00—Alarms responsive to a single specified undesired or abnormal condition and not otherwise provided for
- G08B21/02—Alarms for ensuring the safety of persons
- G08B21/04—Alarms for ensuring the safety of persons responsive to non-activity, e.g. of elderly persons
- G08B21/0407—Alarms for ensuring the safety of persons responsive to non-activity, e.g. of elderly persons based on behaviour analysis
- G08B21/043—Alarms for ensuring the safety of persons responsive to non-activity, e.g. of elderly persons based on behaviour analysis detecting an emergency event, e.g. a fall
-
- G—PHYSICS
- G08—SIGNALLING
- G08B—SIGNALLING OR CALLING SYSTEMS; ORDER TELEGRAPHS; ALARM SYSTEMS
- G08B21/00—Alarms responsive to a single specified undesired or abnormal condition and not otherwise provided for
- G08B21/02—Alarms for ensuring the safety of persons
- G08B21/04—Alarms for ensuring the safety of persons responsive to non-activity, e.g. of elderly persons
- G08B21/0438—Sensor means for detecting
- G08B21/0446—Sensor means for detecting worn on the body to detect changes of posture, e.g. a fall, inclination, acceleration, gait
Definitions
- the present invention relates to a system, device and method for detecting a fall.
- a fall detection device is an assistive device whose main objective is to raise an alert when someone has fallen.
- a problem with the current fall detection devices is that the frequency of false detections is high.
- a false detection is when the device incorrectly determines that the user has fallen where in actual fact, he has not.
- the object of the invention is to provide a solution that overcomes the above disadvantage or at least provide a novel system, device and method for reducing the frequency of the false detections.
- the object of the invention is also to solve the problem when a user falls but he is out of range of his mobile device so he is unable to make a call to his family and friends to ask for help.
- a fall detection system comprising a mobile device, the mobile device comprising a mobile device processor and a mobile device wireless module; and a fall detection device.
- the fall detection device comprises a wireless module for wireless communication with the mobile device wireless module; a memory module comprising accelerometer threshold values, an altitude threshold value and a time interval; and an accelerometer and a barometer.
- the fall detection device further comprises a processor, the processor configured to read and write to the memory module and receive accelerometer readings from the accelerometer and altitude readings from the barometer, and configured to send the accelerometer readings and the altitude readings to the mobile device processor when the processor has determined that the accelerometer readings exceed the accelerometer threshold values and the altitude readings exceed the altitude threshold value.
- the fall detection device further comprises an alarm adapted to be activated by the processor for the time interval, and adapted to be deactivated prior to the expiry of the time interval.
- the mobile device processor is configured to use a machine learning algorithm and the accelerometer readings and the altitude readings to determine when a fall event has occurred, and when the fall event has occurred, to instruct the processor to activate the alarm for the time interval.
- the processor when the alarm is deactivated prior to the expiry of the time interval, the processor will send instructions to the mobile device processor to calculate new accelerometer threshold values and new altitude threshold values; and wherein the processor is further configured to receive the new accelerometer threshold values and the new altitude threshold values from the mobile device processor, and further configured to write the new accelerometer threshold values and the new altitude threshold values into the memory module.
- the mobile device processor is further configured to use gait parameters of a user to determine when the fall event has occurred.
- the fall detection device further comprises an air inlet and a connecting pipe terminating in the interior of the fall detection device and proximate the barometer, and wherein the connecting pipe comprises a plurality of bends.
- the fall detection device further comprises a push button and the alarm is deactivated prior to the expiry of the time interval by the push button being pressed.
- the fall detection system further comprises a cloud server, wherein the cloud server is configured to send an alert to a designated responder.
- the cloud server is further configured to update the machine learning algorithm.
- a fall detection method comprising the steps of receiving with a processor, accelerometer readings from an accelerometer and altitude readings from a barometer; determining with the processor, that the accelerometer readings exceed accelerometer threshold values and the altitude readings exceed altitude threshold values; sending from the processor to a mobile device processor, the accelerometer readings and the altitude readings; and determining when a fall event has occurred by using the mobile device processor, a machine learning algorithm and the accelerometer readings and the altitude readings, and when the fall event has occurred, instructing the processor to activate an alarm for a time interval.
- the fall detection method further comprises the steps of sending from the processor to the mobile device processor, instructions that the alarm has been deactivated prior to the expiry of the time interval; calculating with the mobile device processor, new accelerometer threshold values and new altitude threshold values; receiving with the processor, the new accelerometer threshold values and the new altitude threshold values from the mobile device processor; and writing with the processor, the new accelerometer threshold values and the new altitude threshold values into a memory module.
- the step of determining when a fall event has occurred further comprises the use of gait parameters.
- a fall detection system comprising a mobile device; a cloud server wirelessly connected to the mobile device and a fall detection device.
- the fall detection device comprises a wireless module for wireless communication with the mobile device; a memory module comprising accelerometer threshold values, an altitude threshold value and a time interval; and an accelerometer and a barometer.
- the fall detection device further comprises a processor, the processor configured to read and write to the memory module and receive accelerometer readings from the accelerometer and altitude readings from the barometer, and configured to send the accelerometer readings and the altitude readings to the cloud server via the mobile device when the processor has determined that the accelerometer readings exceed the accelerometer threshold values and the altitude readings exceed the altitude threshold value; and an alarm adapted to be activated by the processor for the time interval, and adapted to be deactivated prior to the expiry of the time interval.
- the cloud server is configured to use a machine learning algorithm and the accelerometer readings and the altitude readings to determine when a fall event has occurred, and when the fall event has occurred, to instruct the processor via the mobile device to activate the alarm for the time interval.
- the processor when the alarm is deactivated prior to the expiry of the time interval, the processor will send instructions to the cloud server via the mobile device to calculate new accelerometer threshold values and new altitude threshold values; and wherein the processor is further configured to receive the new accelerometer threshold values and the new altitude threshold values from the cloud server via the mobile device, and further configured to write the new accelerometer threshold values and the new altitude threshold values into the memory module.
- a fall detection device for use with a mobile device, the mobile device comprising a mobile device processor and a mobile device wireless module.
- the fall detection device comprises a wireless module for wireless communication with the mobile device wireless module; a memory module comprising accelerometer threshold values, an altitude threshold value and a time interval; and an accelerometer and a barometer.
- the fall detection device further comprises a processor, the processor configured to read and write to the memory module and receive accelerometer readings from the accelerometer and altitude readings from the barometer, and configured to send the accelerometer readings and the altitude readings to the mobile device processor when the processor has determined that the accelerometer readings exceed the accelerometer threshold values and the altitude readings exceed the altitude threshold values; and an alarm adapted to be activated by the processor for the time interval, and adapted to be deactivated prior to the expiry of the time interval.
- the processor is further configured to activate the alarm for the time interval upon receiving instructions of a fall event from the mobile device processor; send instructions to the mobile device processor when the alarm is deactivated prior to the expiry of the time interval; receive new accelerometer threshold values and new altitude threshold values from the mobile device processor; and write the new accelerometer threshold values and the new altitude threshold values into the memory module.
- Figure 1 is a diagram illustrating a fall detection system
- Figure 2 is a block diagram illustrating the main components in a fall detection device in accordance with a preferred embodiment of the invention
- Figure 3 is a block diagram illustrating the main components in a mobile device in accordance with a preferred embodiment of the invention.
- Figure 4 illustrates the process flow when a false detection has been detected in accordance with the preferred embodiment
- Figure 5 illustrates the process flow when a positive detection has been detected
- Figure 6 illustrates the process flow when the fall detection device is out of range with the user's mobile device and when a false detection has been detected in accordance with the preferred embodiment
- Figure 7 illustrates the process flow when the fall detection device is out of range with the user's mobile device and when a positive detection has been detected
- Figure 8 illustrates the interior of the fall detection device
- Figure 1 shows an exemplary fall detection system by using a fall detection device 101, mobile device 103, cloud server 104 and designated responder 105.
- Fall detection device 101 can be attached to accessory 102 to facilitate it being worn by a user.
- Accessory 102 can be a chain or necldace or the like.
- Fall detection device 101 is communicatively connected to mobile device 103.
- Mobile device 103 can be a smartphone, a tablet, a mobile phone or the like.
- Mobile device 103 is communicatively connected to cloud server 104.
- fall detection device 101 can be communicatively connected to cloud server 104.
- FIG. 2 shows the main components of fall detection device 101.
- Fall detection device 101 comprises processor 201, accelerometer 202, barometer 203 and wireless module 204.
- accelerometer 202 can provide acceleration readings.
- barometer 203 can provide altitude readings.
- processor 201 is a microcontroller.
- Fall detection device 101 also comprises memory module 206, battery 207 and input and output ports 209.
- fall detection device 101 also comprises push button 205 and alarm 208.
- alarm 208 can be a siren.
- alarm 208 can be a blinking light emitting diode (LED) display.
- alarm 208 can be a vibrator. When pressed, pushbutton 205 is configured to deactivate alarm 208.
- Processor 201 is configured to receive acceleration data from accelerometer
- Processor 201 is also configured to receive altitude data from barometer 203. Processor 201 is also configured to read and write data to memory module 206. Memory module 206 contains accelerometer threshold values, altitude threshold values and a predetermined time interval.
- Wireless module 204 can wirelessly send and receive data via any suitable wireless technology Icnown in the art.
- the wireless technology is Bluetooth 4.0 or Bluetooth Low Energy (BLE).
- BLE Bluetooth Low Energy
- the wireless technology is ZigBee, Wifi, etc.
- wireless module 204 can perform a range/iBeacon function with mobile device 103, such that when fall detection device 101 and mobile device 103 are separated beyond a certain range, an alarm will alert the user of this. This functionality helps to ensure that the user brings fall detection device 101 and mobile device 103 along with him wherever he goes.
- FIG. 3 shows the main components of mobile device 103.
- Mobile device 103 comprises mobile device processor 301 and mobile device wireless module 302.
- mobile device processor 301 has more processing power than processor 201.
- mobile device processor 301 implements a machine learning algorithm.
- FIG. 4 illustrates a method for when a false detection has been detected in accordance with the preferred embodiment.
- processor 201 of fall detection device 101 obtains acceleration readings from accelerometer 202. This may be triggered by the acceleration readings exceeding acceleration thresholds, and these acceleration readings are then sent from accelerometer 202 to processor 201 via an interrupt. These acceleration thresholds are stored in accelerometer 202 and can be individual thresholds in the X, Y and Z axis directions.
- processor 201 compares the acceleration readings with the accelerometer threshold values stored in memory module 206. This comparison can done using Signal Magnitude Vector (SV) calculations as shown below:
- SV Signal Magnitude Vector
- a x is the acceleration in the X axis
- a y is the acceleration in the Y axis
- a z is the acceleration in the Z axis
- SV is a magnitude of the acceleration data.
- the SV may be calculated in 2 dimensions instead of 3 dimensions. It is usually the Z-axis and Y-axis that has larger variations during a fall as compared to the X-axis:
- a moving average of the SV is computed. Crossings are detected based on peak swing of the SV about the moving average. For example, if the accelerometer threshold value is 120, and SV 2D exceeds 120, this will be deemed as a crossing.
- processor 201 is also capable of detecting a pattern, for example when the user is running. This is done by obtaining acceleration readings over a particular frequency. In such an event, processor 201 will raise the accelerometer threshold values (such that a fall event will not be wrongly triggered. This will be more apparent later).
- the frequency of the acceleration data is used in an adaptive filtering technique. Based on the dominant frequency of the acceleration, a frequency band (speed of running or activity) is selected to filter the acceleration data.
- the frequency band may be dynamically changed to one of several frequency bands when a significant frequency change is detected in the dominant frequency of the acceleration data(slowing down or speeding up of activity).
- activities are detected based on the acceleration threshold-filtered and the frequency-filtered modulus.
- the acceleration data is stored in a 320-slot of lobit data for XYZ axis First In First Out (FIFO) buffer with fixed intervals/50 Hz frequency and the data can be filtered accordingly.
- FIFO First In First Out
- step 403 if the acceleration readings exceed the accelerometer threshold values, processor 201 will obtain altitude readings from barometer 203.
- step 404 processor 201 compares the altitude readings with the altitude threshold value stored in memory module 206.
- the altitude threshold value is pre-calculated based upon the user's gait parameters.
- the gait parameter taken into consideration is the user's height. For example, if the user is 1.8 meters tall, the altitude threshold value may be set at approximately 1.55 meters. And if for example, the user is 1.5 meters tall, the altitude threshold value may be set at approximately 1.2 meters.
- step 405 if the altitude readings exceed the altitude threshold value, processor 201 will transmit the acceleration readings and the altitude readings via wireless module 204 to mobile device 103.
- step 406 based on the acceleration readings and the altitude readings, mobile device processor 301 will determine if it is a fall event. This determination can be made using a machine learning algorithm.
- the machine learning algorithm can be Naive Bayes classification algorithm or support vector machines or the like.
- the inputs to the machine learning algorithm can be the acceleration readings and the altitude readings.
- the gait parameters of the user can also be input to the machine learning algorithm. Gait parameters refer to the specific traits and patterns of movement of the user, such as height, gender, weight, age or activity levels of the user.
- a decision tree induction algorithm may be used to distinguish between subjects with high and low risk using the determined gait parameters.
- step 407 if a fall event has been determined by mobile device processor 301, mobile device processor 301 will send instructions to processor 201 to activate alarm 208 for a predetermined time interval.
- the predetermined time interval can be 20 seconds, 40 seconds or 60 seconds.
- the alarm can be de-activated prior to the expiry of the predetermined time interval.
- the user of fall detection device 101 can deactivate the alarm prior to the expiry of the predetermined time interval by pressing push button 205.
- step 408 if the alarm is de-activated prior to the expiry of the predetermined time interval, processor 201 will send instructions to mobile device processor 301 informing that it is a false detection, and to update the machining learning algorithm as well as cloud server 104.
- the new altitude threshold value and the new accelerometer threshold values will be calculated by mobile device processor 301.
- the new altitude threshold value may be an adjustment of the previous altitude threshold value i.e. a decrease of the previous altitude threshold value by a small margin.
- the new accelerometer threshold values can be calculated by using a K factor algorithm and factoring in the gait parameters of the user.
- Factoring in the gait parameters is advantageous as they help to address the possible difference in orientation of accelerometer 202. This is because a user would have certain preferences or habits on how he wears or carries fall detection device 101. Depending on the user's wearing habits, the X, Y and Z accelerometer threshold values will be adjusted accordingly.
- step 410 the new threshold values will then be sent to fall detection device
- the advantage is that simple threshold calculation is done on processor 201 of fall detection device 101 while extensive computation of the machine learning algorithm is done on mobile device processor 301 on mobile device 103, maximising battery consumption efficiency, as mobile device 103 usually has better computational power and longer battery life. Furthermore, as false detections result in the threshold values being adjusted, recalculated and "fine-tuned", this minimises the probability of future false detections.
- Figure 5 illustrates a method for when a positive detection has been detected.
- step 501 processor 201 of fall detection device 101 obtains acceleration readings from accelerometer 202. This step is similar to step 401 as previously described.
- step 502 processor 201 compares the acceleration readings with the accelerometer threshold values. This step is similar to step 402 as previously described.
- step 503 if the acceleration readings exceed the accelerometer threshold values, processor 201 will obtain altitude readings from barometer 203. This step is similar to step 403 as previously described.
- step 504 processor 201 compares the altitude readings with the altitude threshold value. This step is similar to step 404 as previously described.
- step 505 if the altitude readings exceed the altitude threshold value, processor 201 will transmit the acceleration readings and the altitude readings via wireless module 204 to mobile device 103. This step is similar to step 405 as previously described.
- step 506 based on the acceleration readings and the altitude readings, mobile device processor 301 will determine if it is a fall event. This step is similar to step 406 as previously described.
- step 507 if a fall event has been determined by mobile device processor 301, mobile device processor 301 will send instructions to processor 201 to activate the alarm 208 for the predetermined time interval. This step is similar to step 407 as previously described.
- step 508 if the alarm is not deactivated prior to the expiry of the predetermined time interval, processor 201 will send instructions to mobile device processor 301 informing that it is a positive detection, and to update the machining learning algorithm being implemented by mobile device processor 301 .
- the advantage is that the machining learning algorithm is being trained with positive detections and false detections (see figure 4).
- step 509 mobile device 103 updates cloud server 104, and mobile device
- Designated responder 105 can be the next of kin of the user.
- mobile device 103 can also send its Global Positioning System (GPS) data to designated responder 105.
- GPS Global Positioning System
- Cloud server 104 acts as a central repository for the user data.
- cloud server 104 also has a processor to implement a machine learning algorithm.
- Cloud server 104 functions in a similar capacity to mobile device 103, though it stores user data and machine learning models for all users, as opposed to mobile device 103 which stores user data and the machine learning model for just that one user. Cloud server 104 could upon gathering more data, update the machine learning algorithm in mobile device processor 301 to attain higher accuracy.
- steps 601, 602, 603 and 604 and 605 are similar to steps
- step 606 if there is a delivery failure to mobile device 103, processor 201 of fall detection device 101 will increase its signal strength of wireless module 204 until it connects to another mobile device.
- processor 201 of fall detection device 101 will transmit the acceleration readings and the altitude readings via wireless module to the another mobile device.
- step 608 the another mobile device will then send the acceleration readings and altitude readings to cloud server 104.
- step 609 if a fall event has been determined by cloud server 104, cloud server 104 will instruct the another mobile device to send instructions to processor 201 of fall detection device 101 to activate alarm 208 for the predetermined time interval.
- step 610 if the alarm is de-activated prior to the expiry of the predetermined time interval, processor 201 will send instructions to cloud server 104 via the another mobile device informing that it is a false detection, and to update the machining learning algorithm.
- the new altitude threshold value and the new accelerometer threshold values will be calculated by cloud server 104.
- the new altitude threshold value may be an adjustment of the previous altitude threshold value i.e. a decrease of the previous altitude threshold value by a small margin.
- the new accelerometer threshold values can be calculated by using a K factor algorithm and factoring in the gait parameters of the user.
- steps 701, 702, 703 and 704 and 705 are similar to steps
- step 706 if there is a delivery failure to mobile device 103, processor 201 of fall detection device 101 will increase its signal strength of wireless module 204 until it connects to another mobile device.
- processor 201 of fall detection device 101 will transmit the acceleration readings and the altitude readings via wireless module 204 to the another mobile device.
- step 708 the mobile device processor of the another mobile device will then send the acceleration readings and altitude readings to cloud server 104.
- step 709 if a fall event has been determined by cloud server 104, cloud server 104 will instruct the mobile device processor of the another mobile device to send instructions to processor 201 of fall detection device 101 to activate alarm 208 for the predetermined time interval.
- step 710 if alarm 208 is not deactivated prior to the expiry of the predetermined time interval, processor 201 of fall detection device 101 will instruct the mobile device processor of the another mobile device to send instructions to cloud server 104, informing that it is a positive detection, and to update the machining learning algorithm. Cloud server 104 will then send an alert to designated responder 105.
- fall detection device 101 is waterproof.
- the difficulty in making fall detection device 101 waterproof is that barometer 203 requires exposure to the surrounding air in order to function properly. This is for the pressure in fall detection device 101 to be equalized with the environment so that barometer 203 can provide accurate pressure and altitude readings.
- fall detection device 101 has air inlet 801 with connecting pipe 802 terminating in the interior of fall detection device 101 as shown in figure 8.
- air inlet 801 has a small diameter.
- air inlet 801 has a diameter of 0.5mm.
- connecting pipe 802 has a plurality of bends such that the bends act as water flow abutting areas to negotiate any water flow away from the interior of the fall detection device 101.
- connecting pipe 802 has at least one "S"-shaped bend.
- connecting pipe 802 terminates near barometer 203.
Abstract
A fall detection system comprising a mobile device and a fall detection device. The mobile device comprising a mobile device processor and a mobile device wireless module; the fall detection device comprising a wireless module and a memory module comprising accelerometer threshold values, an altitude threshold value and a time interval. The fall detection device further comprises an accelerometer and a barometer; a processor configured to write to the memory module, and to send the accelerometer readings and the altitude readings to the mobile device processor, when the processor has determined that the accelerometer readings exceed the accelerometer threshold values and the altitude readings exceed the altitude threshold value; and an alarm adapted to be activated by the processor for a time interval, and adapted to be deactivated prior to the expiry of the time interval. Wherein the mobile device processor is configured to use machine learning algorithm.
Description
FALL DETECTION SYSTEM, DEVICE AND METHOD
FIELD OF THE INVENTION
[0001] The present invention relates to a system, device and method for detecting a fall.
BACKGROUND
[0002] A fall detection device is an assistive device whose main objective is to raise an alert when someone has fallen.
[0003] A problem with the current fall detection devices is that the frequency of false detections is high. A false detection is when the device incorrectly determines that the user has fallen where in actual fact, he has not.
[0004] Therefore, the object of the invention is to provide a solution that overcomes the above disadvantage or at least provide a novel system, device and method for reducing the frequency of the false detections.
[0005] The object of the invention is also to solve the problem when a user falls but he is out of range of his mobile device so he is unable to make a call to his family and friends to ask for help.
SUMMARY OF INVENTION
[0006] The invention will now be described in detail with reference to the accompanying drawings.
[0007] According to a first aspect of the invention, a fall detection system is described, the fall detection system comprising a mobile device, the mobile device comprising a mobile device processor and a mobile device wireless module; and a fall detection device. The fall detection device comprises a wireless module for wireless communication with the mobile device wireless module; a memory module comprising accelerometer threshold values, an altitude threshold value and a time interval; and an accelerometer and a barometer. The fall detection device further comprises a processor, the processor configured to read and write to
the memory module and receive accelerometer readings from the accelerometer and altitude readings from the barometer, and configured to send the accelerometer readings and the altitude readings to the mobile device processor when the processor has determined that the accelerometer readings exceed the accelerometer threshold values and the altitude readings exceed the altitude threshold value. The fall detection device further comprises an alarm adapted to be activated by the processor for the time interval, and adapted to be deactivated prior to the expiry of the time interval. Wherein the mobile device processor is configured to use a machine learning algorithm and the accelerometer readings and the altitude readings to determine when a fall event has occurred, and when the fall event has occurred, to instruct the processor to activate the alarm for the time interval. Wherein when the alarm is deactivated prior to the expiry of the time interval, the processor will send instructions to the mobile device processor to calculate new accelerometer threshold values and new altitude threshold values; and wherein the processor is further configured to receive the new accelerometer threshold values and the new altitude threshold values from the mobile device processor, and further configured to write the new accelerometer threshold values and the new altitude threshold values into the memory module.
[0008] Preferably, the mobile device processor is further configured to use gait parameters of a user to determine when the fall event has occurred.
[0009] Preferably, the fall detection device further comprises an air inlet and a connecting pipe terminating in the interior of the fall detection device and proximate the barometer, and wherein the connecting pipe comprises a plurality of bends.
[0010] Preferably, the fall detection device further comprises a push button and the alarm is deactivated prior to the expiry of the time interval by the push button being pressed.
[0011] Preferably, the fall detection system further comprises a cloud server, wherein the cloud server is configured to send an alert to a designated responder.
[0012] Preferably, the cloud server is further configured to update the machine learning algorithm.
[0013] According to a second aspect of the invention, a fall detection method is described, the fall detection method comprising the steps of receiving with a processor, accelerometer readings from an accelerometer and altitude readings from a barometer;
determining with the processor, that the accelerometer readings exceed accelerometer threshold values and the altitude readings exceed altitude threshold values; sending from the processor to a mobile device processor, the accelerometer readings and the altitude readings; and determining when a fall event has occurred by using the mobile device processor, a machine learning algorithm and the accelerometer readings and the altitude readings, and when the fall event has occurred, instructing the processor to activate an alarm for a time interval. The fall detection method further comprises the steps of sending from the processor to the mobile device processor, instructions that the alarm has been deactivated prior to the expiry of the time interval; calculating with the mobile device processor, new accelerometer threshold values and new altitude threshold values; receiving with the processor, the new accelerometer threshold values and the new altitude threshold values from the mobile device processor; and writing with the processor, the new accelerometer threshold values and the new altitude threshold values into a memory module.
[0014] Preferably, the step of determining when a fall event has occurred further comprises the use of gait parameters.
[0015] According to a third aspect of the invention, a fall detection system is described, the fall detection system comprising a mobile device; a cloud server wirelessly connected to the mobile device and a fall detection device. The fall detection device comprises a wireless module for wireless communication with the mobile device; a memory module comprising accelerometer threshold values, an altitude threshold value and a time interval; and an accelerometer and a barometer. The fall detection device further comprises a processor, the processor configured to read and write to the memory module and receive accelerometer readings from the accelerometer and altitude readings from the barometer, and configured to send the accelerometer readings and the altitude readings to the cloud server via the mobile device when the processor has determined that the accelerometer readings exceed the accelerometer threshold values and the altitude readings exceed the altitude threshold value; and an alarm adapted to be activated by the processor for the time interval, and adapted to be deactivated prior to the expiry of the time interval. Wherein the cloud server is configured to use a machine learning algorithm and the accelerometer readings and the altitude readings to determine when a fall event has occurred, and when the fall event has occurred, to instruct the processor via the mobile device to activate the alarm for the time interval. Wherein when the alarm is deactivated prior to the expiry of the time interval, the processor will send instructions to the cloud server via the mobile device to calculate new accelerometer threshold values and new altitude threshold values; and wherein the processor is further configured to receive the new accelerometer threshold
values and the new altitude threshold values from the cloud server via the mobile device, and further configured to write the new accelerometer threshold values and the new altitude threshold values into the memory module.
[0016] According to a fourth aspect of the invention, a fall detection device for use with a mobile device is described, the mobile device comprising a mobile device processor and a mobile device wireless module. The fall detection device comprises a wireless module for wireless communication with the mobile device wireless module; a memory module comprising accelerometer threshold values, an altitude threshold value and a time interval; and an accelerometer and a barometer. The fall detection device further comprises a processor, the processor configured to read and write to the memory module and receive accelerometer readings from the accelerometer and altitude readings from the barometer, and configured to send the accelerometer readings and the altitude readings to the mobile device processor when the processor has determined that the accelerometer readings exceed the accelerometer threshold values and the altitude readings exceed the altitude threshold values; and an alarm adapted to be activated by the processor for the time interval, and adapted to be deactivated prior to the expiry of the time interval. Wherein the processor is further configured to activate the alarm for the time interval upon receiving instructions of a fall event from the mobile device processor; send instructions to the mobile device processor when the alarm is deactivated prior to the expiry of the time interval; receive new accelerometer threshold values and new altitude threshold values from the mobile device processor; and write the new accelerometer threshold values and the new altitude threshold values into the memory module.
[0017] The invention will now be described in detail with reference to the accompanying drawings.
BRIEF DESCRIPTION OF THE DRAWINGS
[0018] In order that embodiments of the invention may be fully and more clearly understood by way of non-limitative examples, the following description is taken in conjunction with the accompanying drawings in which like reference numerals designate similar or corresponding elements, regions and portions, and in which:
[0019] Figure 1 is a diagram illustrating a fall detection system;
[0020] Figure 2 is a block diagram illustrating the main components in a fall detection device in accordance with a preferred embodiment of the invention;
[0021] Figure 3 is a block diagram illustrating the main components in a mobile device in accordance with a preferred embodiment of the invention;
[0022] Figure 4 illustrates the process flow when a false detection has been detected in accordance with the preferred embodiment;
[0023] Figure 5 illustrates the process flow when a positive detection has been detected;
[0024] Figure 6 illustrates the process flow when the fall detection device is out of range with the user's mobile device and when a false detection has been detected in accordance with the preferred embodiment;
[0025] Figure 7 illustrates the process flow when the fall detection device is out of range with the user's mobile device and when a positive detection has been detected;
[0026] Figure 8 illustrates the interior of the fall detection device;
[0027] Exemplary, non-limiting embodiments of the present application will now be described with references to the above-mentioned figures.
DETAILED DESCRIPTION
[0028] Referring to the drawings, Figure 1 shows an exemplary fall detection system by using a fall detection device 101, mobile device 103, cloud server 104 and designated responder 105.
[0029] Fall detection device 101 can be attached to accessory 102 to facilitate it being worn by a user. Accessory 102 can be a chain or necldace or the like. Fall detection device 101 is communicatively connected to mobile device 103. Mobile device 103 can be a smartphone, a tablet, a mobile phone or the like. Mobile device 103 is communicatively connected to
cloud server 104. Alternatively, fall detection device 101 can be communicatively connected to cloud server 104.
[0030] Figure 2 shows the main components of fall detection device 101. Fall detection device 101 comprises processor 201, accelerometer 202, barometer 203 and wireless module 204. Preferably, accelerometer 202 can provide acceleration readings. Preferably, barometer 203 can provide altitude readings. Preferably, processor 201 is a microcontroller. Fall detection device 101 also comprises memory module 206, battery 207 and input and output ports 209. Preferably, fall detection device 101 also comprises push button 205 and alarm 208. Preferably, alarm 208 can be a siren. Preferably, alarm 208 can be a blinking light emitting diode (LED) display. Preferably, alarm 208 can be a vibrator. When pressed, pushbutton 205 is configured to deactivate alarm 208.
[0031] Processor 201 is configured to receive acceleration data from accelerometer
202. Processor 201 is also configured to receive altitude data from barometer 203. Processor 201 is also configured to read and write data to memory module 206. Memory module 206 contains accelerometer threshold values, altitude threshold values and a predetermined time interval.
[0032] Wireless module 204 can wirelessly send and receive data via any suitable wireless technology Icnown in the art. Preferably the wireless technology is Bluetooth 4.0 or Bluetooth Low Energy (BLE). Preferably the wireless technology is ZigBee, Wifi, etc. Preferably, wireless module 204 can perform a range/iBeacon function with mobile device 103, such that when fall detection device 101 and mobile device 103 are separated beyond a certain range, an alarm will alert the user of this. This functionality helps to ensure that the user brings fall detection device 101 and mobile device 103 along with him wherever he goes.
[0033] Figure 3 shows the main components of mobile device 103. Mobile device 103 comprises mobile device processor 301 and mobile device wireless module 302. Preferably, mobile device processor 301 has more processing power than processor 201. Preferably, mobile device processor 301 implements a machine learning algorithm.
[0034] Figure 4 illustrates a method for when a false detection has been detected in accordance with the preferred embodiment.
[0035] In step 401, processor 201 of fall detection device 101 obtains acceleration readings from accelerometer 202. This may be triggered by the acceleration readings exceeding acceleration thresholds, and these acceleration readings are then sent from accelerometer 202 to processor 201 via an interrupt. These acceleration thresholds are stored in accelerometer 202 and can be individual thresholds in the X, Y and Z axis directions.
[0036] In step 402, processor 201 compares the acceleration readings with the accelerometer threshold values stored in memory module 206. This comparison can done using Signal Magnitude Vector (SV) calculations as shown below:
SV = J¾ 2 + ay 2 + az 2,
wherein ax is the acceleration in the X axis, ay is the acceleration in the Y axis and azis the acceleration in the Z axis, and SV is a magnitude of the acceleration data. In cases where the variations are more pronounced in a particular axis, the SV may be calculated in 2 dimensions instead of 3 dimensions. It is usually the Z-axis and Y-axis that has larger variations during a fall as compared to the X-axis:
[0037] Preferably, a moving average of the SV is computed. Crossings are detected based on peak swing of the SV about the moving average. For example, if the accelerometer threshold value is 120, and SV2D exceeds 120, this will be deemed as a crossing.
[0038] Preferably, processor 201 is also capable of detecting a pattern, for example when the user is running. This is done by obtaining acceleration readings over a particular frequency. In such an event, processor 201 will raise the accelerometer threshold values (such that a fall event will not be wrongly triggered. This will be more apparent later).
[0039] Preferably, the frequency of the acceleration data is used in an adaptive filtering technique. Based on the dominant frequency of the acceleration, a frequency band (speed of running or activity) is selected to filter the acceleration data. The frequency band may be dynamically changed to one of several frequency bands when a significant frequency change is detected in the dominant frequency of the acceleration data(slowing down or speeding up of activity). Preferably, activities are detected based on the acceleration threshold-filtered and the frequency-filtered modulus. The acceleration data is stored in a
320-slot of lobit data for XYZ axis First In First Out (FIFO) buffer with fixed intervals/50 Hz frequency and the data can be filtered accordingly.
[0040] In step 403, if the acceleration readings exceed the accelerometer threshold values, processor 201 will obtain altitude readings from barometer 203.
[0041] In step 404, processor 201 compares the altitude readings with the altitude threshold value stored in memory module 206.
[0042] The altitude threshold value is pre-calculated based upon the user's gait parameters. In this case, the gait parameter taken into consideration is the user's height. For example, if the user is 1.8 meters tall, the altitude threshold value may be set at approximately 1.55 meters. And if for example, the user is 1.5 meters tall, the altitude threshold value may be set at approximately 1.2 meters.
[0043] In step 405, if the altitude readings exceed the altitude threshold value, processor 201 will transmit the acceleration readings and the altitude readings via wireless module 204 to mobile device 103.
[0044] In step 406, based on the acceleration readings and the altitude readings, mobile device processor 301 will determine if it is a fall event. This determination can be made using a machine learning algorithm. The machine learning algorithm can be Naive Bayes classification algorithm or support vector machines or the like. The inputs to the machine learning algorithm can be the acceleration readings and the altitude readings. Preferably, the gait parameters of the user can also be input to the machine learning algorithm. Gait parameters refer to the specific traits and patterns of movement of the user, such as height, gender, weight, age or activity levels of the user. A decision tree induction algorithm may be used to distinguish between subjects with high and low risk using the determined gait parameters.
[0045] In step 407, if a fall event has been determined by mobile device processor 301, mobile device processor 301 will send instructions to processor 201 to activate alarm 208 for a predetermined time interval. The predetermined time interval can be 20 seconds, 40 seconds or 60 seconds. The alarm can be de-activated prior to the expiry of the predetermined time interval. The user of fall detection device 101 can deactivate the alarm prior to the expiry of the predetermined time interval by pressing push button 205.
[0046] In step 408, if the alarm is de-activated prior to the expiry of the predetermined time interval, processor 201 will send instructions to mobile device processor 301 informing that it is a false detection, and to update the machining learning algorithm as well as cloud server 104.
[0047] In step 409, the new altitude threshold value and the new accelerometer threshold values will be calculated by mobile device processor 301. The new altitude threshold value may be an adjustment of the previous altitude threshold value i.e. a decrease of the previous altitude threshold value by a small margin.
[0048] The new accelerometer threshold values can be calculated by using a K factor algorithm and factoring in the gait parameters of the user. An illustration of the K factor algorithm is as follows. Assume that the accelerometer threshold values are X = 100 (in the X axis direction), Y = 10 (in the Y axis direction), and Z - 10 (in the Z axis direction) for a user who is 60 years old, is of male gender, weighs 80kg, and runs quite often. While the user is "running", the following peak acceleration values are detected i.e. X peak - 120 and Y peak = 12 and Z peak = 12. To obtain the new accelerometer threshold values, the peak acceleration values are summed with the initial accelerometer values and are multiplied by a K factor, the K factor having a value of 0.5 =>
Xnew = (Xpeak + Xinitial threshold) * K = (120 + 100) * 0.5 = 110
Ynew = (Ypeak + Yinitial threshold) * K = (12 + 10) * 0.5 = 11
Znew = (Zpeak + Zinitial threshold) * K = (12 + 10) * 0.5 = 11
[0049] Factoring in the gait parameters is advantageous as they help to address the possible difference in orientation of accelerometer 202. This is because a user would have certain preferences or habits on how he wears or carries fall detection device 101. Depending on the user's wearing habits, the X, Y and Z accelerometer threshold values will be adjusted accordingly.
[0050] In step 410, the new threshold values will then be sent to fall detection device
101 and stored on memory module 206, replacing the previous threshold values.
[0051] The advantage is that simple threshold calculation is done on processor 201 of fall detection device 101 while extensive computation of the machine learning algorithm is done on mobile device processor 301 on mobile device 103, maximising battery consumption
efficiency, as mobile device 103 usually has better computational power and longer battery life. Furthermore, as false detections result in the threshold values being adjusted, recalculated and "fine-tuned", this minimises the probability of future false detections.
[0052] Figure 5 illustrates a method for when a positive detection has been detected.
[0053] In step 501, processor 201 of fall detection device 101 obtains acceleration readings from accelerometer 202. This step is similar to step 401 as previously described.
[0054] In step 502, processor 201 compares the acceleration readings with the accelerometer threshold values. This step is similar to step 402 as previously described.
[0055] In step 503, if the acceleration readings exceed the accelerometer threshold values, processor 201 will obtain altitude readings from barometer 203. This step is similar to step 403 as previously described.
[0056] In step 504, processor 201 compares the altitude readings with the altitude threshold value. This step is similar to step 404 as previously described.
[0057] In step 505, if the altitude readings exceed the altitude threshold value, processor 201 will transmit the acceleration readings and the altitude readings via wireless module 204 to mobile device 103. This step is similar to step 405 as previously described.
[0058] In step 506, based on the acceleration readings and the altitude readings, mobile device processor 301 will determine if it is a fall event. This step is similar to step 406 as previously described.
[0059] In step 507, if a fall event has been determined by mobile device processor 301, mobile device processor 301 will send instructions to processor 201 to activate the alarm 208 for the predetermined time interval. This step is similar to step 407 as previously described.
[0060] In step 508, if the alarm is not deactivated prior to the expiry of the predetermined time interval, processor 201 will send instructions to mobile device processor 301 informing that it is a positive detection, and to update the machining learning algorithm being implemented by mobile device processor 301 . The advantage is that the machining learning algorithm is being trained with positive detections and false detections (see figure 4).
[0061] In step 509, mobile device 103 updates cloud server 104, and mobile device
103 sends an alert or makes a call to designated responder 105. Designated responder 105 can be the next of kin of the user. Preferably, mobile device 103 can also send its Global Positioning System (GPS) data to designated responder 105.
[0062] Cloud server 104 acts as a central repository for the user data. Preferably, cloud server 104 also has a processor to implement a machine learning algorithm. Cloud server 104 functions in a similar capacity to mobile device 103, though it stores user data and machine learning models for all users, as opposed to mobile device 103 which stores user data and the machine learning model for just that one user. Cloud server 104 could upon gathering more data, update the machine learning algorithm in mobile device processor 301 to attain higher accuracy.
[0063] It is also envisaged that in the event that mobile device 103 is unavailable, a crowd sourcing function in processor 201 of fall detection device 101 will trigger other mobile devices in the vicinity. Other mobile devices would also have mobile device processor 301 and mobile device wireless module 302. These other mobile devices act as a medium for the acceleration readings and altitude readings to be sent to cloud server 104. Figure 6 illustrates this process flow with a false detection while Figure 7 illustrates this process flow with a positive detection.
[0064] Referring to figure 6, steps 601, 602, 603 and 604 and 605 are similar to steps
401, 402, 403, 404 and 405 as previously described.
[0065] In step 606, if there is a delivery failure to mobile device 103, processor 201 of fall detection device 101 will increase its signal strength of wireless module 204 until it connects to another mobile device.
[0066] Once connection has been established between fall detection device 101 and the another mobile device, in step 607, processor 201 of fall detection device 101 will transmit the acceleration readings and the altitude readings via wireless module to the another mobile device.
[0067] In step 608, the another mobile device will then send the acceleration readings and altitude readings to cloud server 104.
[0068] In step 609, if a fall event has been determined by cloud server 104, cloud server 104 will instruct the another mobile device to send instructions to processor 201 of fall detection device 101 to activate alarm 208 for the predetermined time interval.
[0069] In step 610, if the alarm is de-activated prior to the expiry of the predetermined time interval, processor 201 will send instructions to cloud server 104 via the another mobile device informing that it is a false detection, and to update the machining learning algorithm.
[0070] In step 611, the new altitude threshold value and the new accelerometer threshold values will be calculated by cloud server 104. The new altitude threshold value may be an adjustment of the previous altitude threshold value i.e. a decrease of the previous altitude threshold value by a small margin.
[0071] The new accelerometer threshold values can be calculated by using a K factor algorithm and factoring in the gait parameters of the user. An illustration of the K factor algorithm is as follows. Assume that the accelerometer threshold values are X = 100 (in the X axis direction), Y = 10 (in the Y axis direction), and Z = 10 (in the Z axis direction) for a user who is 60 years old, is of male gender, weighs 80kg, and runs quite often. While the user is "running", the following peak acceleration values are detected i.e. X peak = 120 and Y peak = 12 and Z peak = 12. To obtain the new accelerometer threshold values, the peak acceleration values are summed with the initial accelerometer values and are multiplied by a K factor, the K factor having a value of 0.5 =>
Xnew = (Xpeak + Xinitial threshold) * K = (120 + 100) * 0.5 = 110
Ynew = (Ypeak + Yinitial threshold) * K = (12 + 10) * 0.5 - 11
Znew = (Zpeak + Zinitial threshold) * K - (12 + 10) * 0.5 = 11
[0072] Factoring in the gait parameters is advantageous as they help to address the possible difference in orientation of accelerometer 202. This is because a user would have certain preferences or habits on how he wears or carries fall detection device 101. Depending on the user's wearing habits, the X, Y and Z accelerometer threshold values will be adjusted accordingly.
[0073] In step 612, the new threshold values will then be sent to fall detection device
101 via the another mobile device and stored on memory module 206, replacing the previous threshold values.
[0074] Referring to figure 7, steps 701, 702, 703 and 704 and 705 are similar to steps
501, 502, 503, 504 and 505 as previously described.
[0075] In step 706, if there is a delivery failure to mobile device 103, processor 201 of fall detection device 101 will increase its signal strength of wireless module 204 until it connects to another mobile device.
[0076] Once connection has been established between fall detection device 101 and the another mobile device, in step 707, processor 201 of fall detection device 101 will transmit the acceleration readings and the altitude readings via wireless module 204 to the another mobile device.
[0077] In step 708, the mobile device processor of the another mobile device will then send the acceleration readings and altitude readings to cloud server 104.
[0078] In step 709, if a fall event has been determined by cloud server 104, cloud server 104 will instruct the mobile device processor of the another mobile device to send instructions to processor 201 of fall detection device 101 to activate alarm 208 for the predetermined time interval.
[0079] In step 710, if alarm 208 is not deactivated prior to the expiry of the predetermined time interval, processor 201 of fall detection device 101 will instruct the mobile device processor of the another mobile device to send instructions to cloud server 104, informing that it is a positive detection, and to update the machining learning algorithm. Cloud server 104 will then send an alert to designated responder 105.
[0080] It is possible to further classify the positive detection into the following: "fall from chair", "collapsing into chair", "fall from bed", "collapsing into the bed", "resting against a wall then sliding down".
[0081] Preferably, fall detection device 101 is waterproof. However, the difficulty in making fall detection device 101 waterproof is that barometer 203 requires exposure to the surrounding air in order to function properly. This is for the pressure in fall detection device 101 to be equalized with the environment so that barometer 203 can provide accurate pressure and altitude readings. For this reason, fall detection device 101 has air inlet 801 with connecting pipe 802 terminating in the interior of fall detection device 101 as shown in figure 8. Preferably, air inlet 801 has a small diameter. Preferably, air inlet 801 has a diameter of 0.5mm. Preferably, connecting pipe 802 has a plurality of bends such that the bends act as water flow abutting areas to negotiate any water flow away from the interior of the fall detection device 101. Preferably, connecting pipe 802 has at least one "S"-shaped bend. Preferably, connecting pipe 802 terminates near barometer 203.
[0082] While exemplary embodiments pertaining to the invention have been described and illustrated, it will be understood by those skilled in the technology concerned that many variations or modifications involving particular design, implementation or construction are possible and may be made without deviating from the inventive concepts described herein.
Claims
CLAIMS l. A fall detection system comprising :
a mobile device, the mobile device comprising a mobile device processor and a mobile device wireless module; and a fall detection device, the fall detection device comprising:
a wireless module for wireless communication with the mobile device wireless module; a memory module comprising accelerometer threshold values, an altitude threshold value and a time interval; an accelerometer and a barometer; a processor, the processor configured to read and write to the memory module and receive accelerometer readings from the accelerometer and altitude readings from the barometer, and configured to send the accelerometer readings and the altitude readings to the mobile device processor when the processor has determined that the accelerometer readings exceed the accelerometer threshold values and the altitude readings exceed the altitude threshold value; and an alarm adapted to be activated by the processor for the time interval, and adapted to be deactivated prior to the expiry of the time interval; wherein the mobile device processor is configured to use a machine learning algorithm and the accelerometer readings and the altitude readings to determine when a fall event has occurred, and when the fall event has occurred, to instruct the processor to activate the alarm for the time interval; wherein when the alarm is deactivated prior to the expiry of the time interval, the processor will send instructions to the mobile device processor to calculate new accelerometer threshold values and new altitude threshold values; and wherein the processor is further configured to receive the new accelerometer threshold values and the new altitude threshold values from the mobile device processor, and further configured to write the new accelerometer threshold values and the new altitude threshold values into the memory module.
2. The fall detection system of claim l wherein the mobile device processor is further configured to use gait parameters of a user to determine when the fall event has occurred.
3. The fall detection system of claims 1 or 2 wherein the fall detection device further comprises an air inlet and a connecting pipe terminating in the interior of the fall detection device and proximate the barometer, and wherein the connecting pipe comprises a plurality of bends.
4. The fall detection system of any one of the preceding claims wherein the fall detection device further comprises a push button and the alarm is deactivated prior to the expiry of the time interval by the push button being pressed.
5. The fall detection system of any one of the preceding claims further comprising a cloud server, wherein the cloud server is configured to send an alert to a designated responder.
6. The fall detection system of claim 5 wherein the cloud server is further configured to update the machine learning algorithm.
7. A fall detection method comprising the steps of :- receiving with a processor, accelerometer readings from an accelerometer and altitude readings from a barometer; determining with the processor, that the accelerometer readings exceed accelerometer threshold values and the altitude readings exceed altitude threshold values; sending from the processor to a mobile device processor, the accelerometer readings and the altitude readings; determining when a fall event has occurred by using the mobile device processor, a machine learning algorithm and the accelerometer readings and the altitude readings, and when the fall event has occurred, instructing the processor to activate an alarm for a time interval;
sending from the processor to the mobile device processor, instructions that the alarm has been deactivated prior to the expiry of the time interval; calculating with the mobile device processor, new accelerometer threshold values and new altitude threshold values; receiving with the processor, the new accelerometer threshold values and the new altitude threshold values from the mobile device processor; and writing with the processor, the new accelerometer threshold values and the new altitude threshold values into a memory module.
8. The fall detection method of claim 7 wherein the step of determining when a fall event has occurred further comprises the use of gait parameters.
9. A fall detection system comprising : a mobile device; a cloud server wirelessly connected to the mobile device; and a fall detection device, the fall detection device comprising: a wireless module for wireless communication with the mobile device; a memory module comprising accelerometer threshold values, an altitude threshold value and a time interval; an accelerometer and a barometer; a processor, the processor configured to read and write to the memory module and receive accelerometer readings from the accelerometer and altitude readings from the barometer, and configured to send the accelerometer readings and the altitude readings to the cloud server via the mobile device when the processor has determined that the accelerometer readings exceed the accelerometer threshold values and the altitude readings exceed the altitude threshold value; and an alarm adapted to be activated by the processor for the time interval, and adapted to be deactivated prior to the expiry of the time interval;
wherein the cloud server is configured to use a machine learning algorithm and the accelerometer readings and the altitude readings to determine when a fall event has occurred, and when the fall event has occurred, to instruct the processor via the mobile device to activate the alarm for the time interval; wherein when the alarm is deactivated prior to the expiry of the time interval, the processor will send instructions to the cloud server via the mobile device to calculate new accelerometer threshold values and new altitude threshold values; and wherein the processor is further configured to receive the new accelerometer threshold values and the new altitude threshold values from the cloud server via the mobile device, and further configured to write the new accelerometer threshold values and the new altitude threshold values into the memory module.
10. A fall detection device for use with a mobile device, the mobile device comprising a mobile device processor and a mobile device wireless module, the fall detection device comprising : a wireless module for wireless communication with the mobile device wireless module; a memory module comprising accelerometer threshold values, an altitude threshold value and a time interval; an accelerometer and a barometer; a processor, the processor configured to read and write to the memory module and receive accelerometer readings from the accelerometer and altitude readings from the barometer, and configured to send the accelerometer readings and the altitude readings to the mobile device processor when the processor has determined that the accelerometer readings exceed the accelerometer threshold values and the altitude readings exceed the altitude threshold values; and an alarm adapted to be activated by the processor for the time interval, and adapted to be deactivated prior to the expiry of the time interval; wherein the processor is further configured to : activate the alarm for the time interval upon receiving instructions of a fall event from the mobile device processor;
send instructions to the mobile device processor when the alarm is deactivated prior to the expiry of the time interval; receive new accelerometer threshold values and new altitude threshold values from the mobile device processor; and write the new accelerometer threshold values and the new altitude threshold values into the memory module.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
PCT/SG2014/000565 WO2016085405A1 (en) | 2014-11-28 | 2014-11-28 | Fall detection system, device and method |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
PCT/SG2014/000565 WO2016085405A1 (en) | 2014-11-28 | 2014-11-28 | Fall detection system, device and method |
Publications (1)
Publication Number | Publication Date |
---|---|
WO2016085405A1 true WO2016085405A1 (en) | 2016-06-02 |
Family
ID=56074784
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
PCT/SG2014/000565 WO2016085405A1 (en) | 2014-11-28 | 2014-11-28 | Fall detection system, device and method |
Country Status (1)
Country | Link |
---|---|
WO (1) | WO2016085405A1 (en) |
Cited By (5)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN106228752A (en) * | 2016-09-14 | 2016-12-14 | 努比亚技术有限公司 | A kind of intelligence nurse method and device |
TWI635465B (en) * | 2017-12-11 | 2018-09-11 | 宏達國際電子股份有限公司 | Electronic device and drop warning method |
CN113124924A (en) * | 2020-01-10 | 2021-07-16 | 手持产品公司 | Device drop detection using machine learning |
US20230127067A1 (en) * | 2021-10-27 | 2023-04-27 | International Business Machines Corporation | Multi-dimensional mapping and user cognitive profile based device control and channel assignment |
US20240077374A1 (en) * | 2022-09-02 | 2024-03-07 | Honeywld Technology Corp. | Single-barometer device, method for fall detection, and system thereof |
Citations (1)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US20130054180A1 (en) * | 2011-08-29 | 2013-02-28 | James R. Barfield | Method and system for detecting a fall based on comparing data to criteria derived from multiple fall data sets |
-
2014
- 2014-11-28 WO PCT/SG2014/000565 patent/WO2016085405A1/en active Application Filing
Patent Citations (1)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US20130054180A1 (en) * | 2011-08-29 | 2013-02-28 | James R. Barfield | Method and system for detecting a fall based on comparing data to criteria derived from multiple fall data sets |
Non-Patent Citations (1)
Title |
---|
BIANCHI ET AL.: "Barometric Pressure and Triaxial Accelerometry-Based Falls Event Detection", IEEE TRANSACTIONS ON NEURAL SYSTEMS AND REHABILITATION ENGINEERING, vol. 18, no. 6, December 2010 (2010-12-01), pages 619 - 627, XP011373977, DOI: doi:10.1109/TNSRE.2010.2070807 * |
Cited By (6)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN106228752A (en) * | 2016-09-14 | 2016-12-14 | 努比亚技术有限公司 | A kind of intelligence nurse method and device |
TWI635465B (en) * | 2017-12-11 | 2018-09-11 | 宏達國際電子股份有限公司 | Electronic device and drop warning method |
CN113124924A (en) * | 2020-01-10 | 2021-07-16 | 手持产品公司 | Device drop detection using machine learning |
US20230127067A1 (en) * | 2021-10-27 | 2023-04-27 | International Business Machines Corporation | Multi-dimensional mapping and user cognitive profile based device control and channel assignment |
US11856592B2 (en) * | 2021-10-27 | 2023-12-26 | International Business Machines Corporation | Multi-dimensional mapping and user cognitive profile based device control and channel assignment |
US20240077374A1 (en) * | 2022-09-02 | 2024-03-07 | Honeywld Technology Corp. | Single-barometer device, method for fall detection, and system thereof |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
WO2016085405A1 (en) | Fall detection system, device and method | |
US9576460B2 (en) | Wearable smart device for hazard detection and warning based on image and audio data | |
US8933801B2 (en) | Fall detection system and method | |
KR100988459B1 (en) | Apparatus and method for fall-down detection | |
CN106473749A (en) | For detecting the device that falls, system and method | |
US20160278647A1 (en) | Misalignment detection of a wearable device | |
KR102141367B1 (en) | Device and method for classifying the activity and/or counting steps of a user | |
CN103810817B (en) | A kind of detection alarm method of the wearable human paralysis device of falling detection alarm | |
US20170035367A1 (en) | Personal safety monitoring using a multi-sensor apparatus | |
JP2010508945A (en) | System for preventing fall and method for preventing fall using this system | |
KR20170004269A (en) | Apparatus and method for fall-down detection | |
US20170154521A1 (en) | Methods and systems for providing personal emergency alerts and context aware activity notifications | |
KR20170112778A (en) | System and method for fall detection using smart band | |
US20190066478A1 (en) | Personal safety monitoring using a multi-sensor apparatus | |
WO2019097248A1 (en) | Improvements in or relating to fall detectors and fall detection | |
CN205103993U (en) | Intelligence human body guardianship device of tumbleing | |
JP2016177459A (en) | Fall detection terminal and program | |
KR101990032B1 (en) | A wearable smart device for distinguishing dangerous situation and a method for distinguishing dangerous situation using the same | |
JP2016177437A (en) | Fall detection terminal and program | |
US10136393B2 (en) | Control method for real-time scene detection by a wireless communication apparatus | |
CN111724569B (en) | Alarm method, alarm device and storage medium | |
US20210306774A1 (en) | Selectively Collecting and Storing Sensor Data of a Hearing System | |
KR100961476B1 (en) | Behavior pattern recognition device and its method | |
CN112438726A (en) | Fall alarm method based on wearable device, wearable device and storage medium | |
EP3796282A2 (en) | Device, system and method for fall detection |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
121 | Ep: the epo has been informed by wipo that ep was designated in this application |
Ref document number: 14906778 Country of ref document: EP Kind code of ref document: A1 |
|
NENP | Non-entry into the national phase |
Ref country code: DE |
|
32PN | Ep: public notification in the ep bulletin as address of the adressee cannot be established |
Free format text: NOTING OF LOSS OF RIGHTS PURSUANT TO RULE 112(1) EPC (EPO FORM 1205A DATED 11.10.2017) |
|
122 | Ep: pct application non-entry in european phase |
Ref document number: 14906778 Country of ref document: EP Kind code of ref document: A1 |