Method and system for discerning a false positive in a fall detection signal

• US 20130197856A1
• Filed: 06/18/2012
• Published: 08/01/2013
• Est. Priority Date: 06/17/2011
• Status: Active Grant

• Current Assignee: Verizon Patent and Licensing Incorporated (Verizon Communications Inc.)
• Original Assignee: Verizon Telematics Inc. (Verizon Communications Inc.)
• Inventors: Barfield, James R., Kalinadhabhotla, Deep, Taylor, Thomas S.
• Granted Patent: US 9,638,711 B2
• Time in Patent Office: Days
• US Class Current: 702/141
• CPC Class Codes: G01P 15/00 Measuring acceleration; Measuring deceleration; Measuring shock, i.e. sudden change of acceleration
• CPC Class Codes: G01P 15/0891 with indication of predetermined acceleration values G01P15/135 takes precedence
• CPC Class Codes: G01P 15/18 in two or more dimensions

Abstract

Methods reduce the likelihood of an MPERS device falsely reporting a high acceleration event as a fall impact. The device stores acceleration data acquired before the high acceleration and afterward. If a number of samples of magnitude values from accelerometer sensors in the device acquired during as interval before the high acceleration that approach 0G exceeds a predetermined number, the high acceleration is deemed from a non-fall. Acceleration sensors can also indicate an orientation change before/after the high acceleration, and a barometric pressure sensor can do the same, to further characterize an event as a non-fall. A method compares current event data to composite data sets that have been determined from a plurality of empirically derived data sets of fall and non-fall events. High correlations can indicate falls, or non-falls, respectively. Further statistical analysis of data acquired after an event reduces the likelihood of falsely indicating a fall.

Claims

1. A method for analyzing a current event data set generated by an MPERS device to distinguish between a non-fall event and a fall event, comprising:

measuring an amount of time the event data set has an acceleration magnitude below a predetermined free fall magnitude threshold;

determining that the current event data corresponds to a fall event if the time below the predetermined free fall magnitude threshold is less than a predetermined free fall interval threshold;

determining that the current event data corresponds to a non-fall event if the time below the predetermined free fall magnitude threshold is greater than the predetermined free fall interval threshold; and

outputting an indication that a fall occurred if the current event data set corresponds to a fall event.

2. The method of claim 1 wherein the free fall interval threshold results from analyzing a plurality of empirical data sets corresponding to a plurality of fall event data sets and wherein the free fall interval threshold is the longest interval event data set acceleration magnitude from the empirical data sets indicates an acceleration magnitude below the predetermined free fall magnitude threshold.

3. The method of claim 1 further comprising detecting an acceleration magnitude that exceeds a predetermined acceleration magnitude trigger level wherein the measured amount of time the event data set has a magnitude below the predetermined free fall threshold occurs before the acceleration magnitude exceeded the predetermined acceleration magnitude trigger level.

4. The method of claim 3 further comprising performing a correlation function on the current event data and each of a plurality of composite non-fall event data sets corresponding to a plurality of groupings of non-fall event data setsif the amount of time the acceleration magnitude was below the predetermined free fall magnitude threshold was less than the predetermined free fall interval threshold;

and

outputting an indication that the current event data set indicates a fall event if the results of performing the correlation function with the current event data set and each of the plurality of composite non-fall event data sets is less than a predetermined threshold corresponding to the composite non-fall event data set.

5. The method of claim 3 further comprising performing a correlation function on the current event data and each of a plurality of composite fall event data sets corresponding to a plurality of groupings of fall event data sets if the amount of time the acceleration magnitude was below the predetermined free fall magnitude threshold was less than the predetermined free fall interval threshold;

and

outputting an indication that the current event data set indicates a fall event if the results of performing the correlation function with the current event data set and each of the plurality of composite fall event data sets is greater than a predetermined threshold corresponding to the composite fall event data set.

6. The method of claim 1 further comprising:

evaluating acceleration component values of an acceleration signal corresponding to an acceleration component, along one of a plurality of perpendicular axes, of the acceleration magnitude;

determining, for each one of the plurality of acceleration component signals, whether its values crosses its corresponding zero-acceleration axis during the free fall interval more than once; and

outputting an indication that the current event is a non-fall event.

7. The method of claim 6 further comprising:

integrating the acceleration value for each of the component acceleration signals between two zero crossings;

scaling the component'"'"'s acceleration value between the two zero crossings using the result of the integration between the two zero crossings to adjust the component'"'"'s acceleration value to a value associated with a typical fall or non-fall without spin;

computing a scaled acceleration magnitude of the scaled component acceleration signal values; and

using the scaled acceleration magnitude of the scaled component acceleration signal values as the acceleration magnitude.

8. The method of claim 1 further comprising:

storing acceleration data generated from an accelerometer device of the MPERS device generated during an event signature window, wherein the event signature window is a predetermined amount of time before an acceleration magnitude signal from the accelerometer device meets a criterion; and

storing the acceleration data as a fall event data set in a fall event data set library if the acceleration data resulted from a fall event and storing the acceleration data as a non-fall event data set in a non-fall event data set library if the acceleration data resulted from a non-fall event.

9. The method of claim 8 further comprising:

determining a maximum acceleration magnitude value from each fall event data set in the fall event data set library; and

establishing the minimum of all of the maximum acceleration magnitude values as a trigger magnitude.

10. The method of claim 9 further comprising:

determining for each fall event data set in the fall event data set library the free fall interval that precedes the maximum acceleration magnitude;

storing the minimum acceleration magnitude value during the free fall interval for each fall event data sets in the fall event data set library; and

establishing, from all of the fall event data sets in the fall event data set library the maximum of the minimum acceleration magnitude values during the respective free fall intervals as a free fall magnitude threshold.

11. An MPERS device configured to analyze a current event data set generated by an MPERS device to distinguish between a non-fall event and a fall event, the MPERS device comprising a processor configured to perform the steps of:

measuring an amount of time the event data signal has a magnitude below a predetermined free fall threshold;

determining that the event data corresponds to a fall event if the time below the predetermined free fall threshold is less than a predetermined free fall interval;

determining that the event data corresponds to a non-fall event if the time below the predetermined free fall threshold is greater than the predetermined free fall interval; and

outputting a signal that a fall occurred if the event data corresponds to a fall event.

12. The MPERS device of claim 11 wherein the free fall interval threshold results from analyzing a plurality of empirical data sets corresponding to a plurality of fall event data sets and wherein the free fall interval threshold is the longest interval event data set acceleration magnitude from the empirical data sets indicates an acceleration magnitude below the predetermined free fall magnitude threshold.

13. MPERS device of claim 11 further configured for detecting an acceleration magnitude that exceeds a predetermined acceleration magnitude trigger level wherein the measured amount of time the event data set has a magnitude below the predetermined free fall threshold occurs before the acceleration magnitude exceeded the predetermined acceleration magnitude trigger level.

14. MPERS device of claim 13 further configured for:

performing a correlation function on the current event data and each of a plurality of composite non-fall event data sets corresponding to a plurality of groupings of non-fall event data sets if the amount of time the acceleration magnitude was below the predetermined free fall magnitude threshold was less than the predetermined free fall interval threshold; and

outputting an indication that the current event data set indicates a fall event if the results of performing the correlation function with the current event data set and each of the plurality of composite non-fall event data sets is less than a predetermined threshold corresponding to the composite non-fall event data set.

15. MPERS device of claim 13 further configured for:

performing a correlation function on the current event data and each of a plurality of composite fall event data sets corresponding to a plurality of groupings of fall event data sets if the amount of time the acceleration magnitude was below the predetermined free fall magnitude threshold was less than the predetermined free fall interval threshold; and

outputting an indication that the current event data set indicates a fall event if the results of performing the correlation function with the current event data set and each of the plurality of composite fall event data sets is greater than a predetermined threshold corresponding to the composite fall event data set.

16. The device of claim 11 further configured for:

evaluating acceleration component values of an acceleration signal corresponding to an acceleration component, along one of a plurality of perpendicular axes, of the acceleration magnitude;

determining, for each one of the plurality of acceleration component signals, whether its values crosses its corresponding zero-acceleration axis during the free fall interval more than once; and

outputting an indication that the current event is a non-fall event.

17. The device of claim 16 further configured for:

integrating the acceleration value for each of the component acceleration signals between two zero crossings;

scaling the component'"'"'s acceleration value between the two zero crossings using the result of the integration between the two zero crossings to adjust the component'"'"'s acceleration value to a value associated with a typical fall or non-fall without spin;

computing a scaled acceleration magnitude of the scaled component acceleration signal values; and

using the scaled acceleration magnitude of the scaled component acceleration signal values as the acceleration magnitude.

18. A method for generating a library of event data sets, comprising:

storing acceleration data generated from an accelerometer device during an event signature window, wherein the event signature window is a predetermined amount of time before an acceleration magnitude signal from the accelerometer device meets a criterion;

wirelessly transmitting the acceleration data to a central host computer in a message;

receiving the message at a central host computer;

processing the message at the central host computer; and

storing the acceleration data as a fall event data set in a fall event data set library if the acceleration data resulted from a fall event and storing the acceleration data as a non-fall event data set in a non-fall event data set library if the acceleration data resulted from a non-fall event.

19. The method of claim 18 further comprising:

determining a maximum acceleration magnitude value from each fall event data set in the fall event data set library; and

establishing the minimum of all of maximum acceleration magnitude values is used as a trigger magnitude.

20. The method of claim 19 further comprising:

determining for each fall event data set in the fall event data set library a free fall interval that precedes the maximum acceleration magnitude;

storing the minimum acceleration magnitude value during the free fall interval for each fall event data sets in the fall event data set library; and

establishing, from all of the fall event data sets in the fall event data set library the maximum of the minimum acceleration magnitude values during the respective free fall intervals as a free fall magnitude threshold.