Automated identification of device status and resulting dynamic modification of device operations

• US 10,353,764 B1
• Filed: 11/08/2018
• Issued: 07/16/2019
• Est. Priority Date: 11/08/2018
• Status: Active Grant

• Current Assignee: Curinos Incorporated
• Original Assignee: Amplero Inc.
• Inventors: Cazzanti, Luca, Downs, Oliver B., Danielson, Matthew G.
• Primary Examiner(s): Lottich, Joshua P
• Application Number:
  US16/184,900
• Time in Patent Office: 250 Days
• CPC Class Codes: G06F 11/008 Reliability or availability analysis
• CPC Class Codes: G06F 11/0721 within a central processing unit [CPU]
• CPC Class Codes: G06F 11/0742 in a data processing system embedded in a mobile device, e.g. mobile phones, handheld devices
• CPC Class Codes: G06F 11/0748 in a remote unit communicating with a single-box computer node experiencing an error/fault remote testing G06F11/2294
• CPC Class Codes: G06F 11/079 Root cause analysis, i.e. error or fault diagnosis in a hardware test environment G06F11/22; in a software test environment G06F11/36
• CPC Class Codes: G06F 11/0793 Remedial or corrective actions recovery from an exception in an instruction pipeline G06F9/3861; by retry G06F11/1402; for recovering from a failure of a protocol instance or entity H04L69/40
• CPC Class Codes: G06N 20/00 Machine learning
• CPC Class Codes: G06N 7/01 Probabilistic graphical models, e.g. probabilistic networks
• CPC Class Codes: H04L 41/0661 by reconfiguring faulty entities
• CPC Class Codes: H04L 41/147 for predicting network behaviour
• CPC Class Codes: H04L 41/149 for prediction of maintenance
• CPC Class Codes: H04L 43/0817 by checking functioning

Abstract

Techniques are described for automatically and dynamically modifying ongoing operations of computing devices in device-specific manners, such as based on an automated identification of a computing device'"'"'s status (e.g., identifying a likely ongoing or imminent failure of a smart phone or other computing device based on a series of observed hardware states of the computing device, and taking automated corrective actions to prevent or otherwise mitigate such device failure, such as by modifying configuration settings on the computing device or on associated systems). The techniques may include, for each of multiple device status outcomes of interest (e.g., device failure versus device non-failure), generating a state-space outcome model representing devices that reach that status outcome within a time period of interest, and using such outcome models to identify a likely ongoing or imminent outcome of a current device, with corresponding automated corrective actions then taken.

Claims

1. A computer-implemented method comprising:

tracking, by a configured computing system, device operations of multiple mobile devices, including storing multiple observed hardware attribute states for each mobile device at each of a plurality of times, determining whether each of the mobile devices has a subsequent device failure after the plurality of times, and separating the multiple mobile devices into a first plurality of failed mobile devices that have the subsequent device failure and a second plurality of non-failed mobile devices that do not have the subsequent device failure;

generating, by the configured computing system and based on an automated analysis of first data generated about the failed mobile devices from the tracking, one or more failure model data structures with information correlating first observed hardware attribute states of the failed mobile devices with subsequent device failure, including modeling additional first non-observable hardware attribute states of the failed mobile devices that cause the first observed hardware attribute states, and determining probabilities of occurrences of the first observed hardware attribute states and of the first non-observable hardware attribute states in light of device operations of the failed mobile devices;

generating, by the configured computing system and based on an automated analysis of second data generated about the non-failed mobile devices from the tracking, one or more non-failure model data structures with information correlating second observed hardware attribute states of the non-failed mobile devices with a lack of subsequent device failure, including modeling additional second non-observable hardware attribute states of the non-failed mobile devices that cause the second observed hardware attribute states, and determining probabilities of occurrences of the second observed hardware attribute states and of the second non-observable hardware attribute states in light of device operations of the non-failed mobile devices;

identifying, by the configured computing system, and based on the one or more failure model data structures and the one or more non-failure model data structures, an upcoming device failure of an additional mobile device, including determining that a third sequence of observed hardware attribute states of the additional mobile device over multiple times corresponds to the first non-observable hardware attribute states of the failed mobile devices; and

controlling, by the configured computing system, ongoing operations for the additional mobile device to prevent the upcoming device failure of the additional mobile device, including modifying configuration settings of the additional mobile device to alter future hardware attribute states of the additional mobile device to correspond to the second non-observable hardware attribute states of the non-failed mobile devices.

2. The computer-implemented method of claim 1 wherein the identifying of the upcoming device failure of the additional mobile device includes determining that the third sequence of observed hardware attribute states of the additional mobile device matches the first non-observable hardware attribute states of the failed mobile devices, and includes predicting that the additional mobile device will have a future device failure based at least in part on the third sequence of observed hardware attribute states of the additional mobile device and on the determined probabilities of the occurrences of the first observed hardware attribute states and of the first non-observable hardware attribute states.

3. The computer-implemented method of claim 2 wherein the controlling of the ongoing operations for the additional mobile device to prevent the upcoming device failure includes modifying use of memory and communication capabilities on the additional mobile device.

4. A computer-implemented method comprising:

obtaining, by a configured computing system, a failure model representing a first plurality of mobile devices that experience a failure status after a period of time, and a non-failure model representing a second plurality of mobile devices that do not experience a failure status after the period of time, wherein the failure model is generated using first sequences of observed device attribute states of the first plurality of mobile devices over the period of time and includes information relating the observed device attribute states of the first sequences to first non-observable device attributes associated with the failure status, and wherein the non-failure model is generated using second sequences of observed device attribute states of the second plurality of mobile devices over the period of time and includes information relating the observed device attribute states of the second sequences to second non-observable device attributes associated with the non-failure status;

using, by the configured computing system, the failure model and the non-failure model to predict a future failure status of an additional mobile device, including determining that a third sequence of observed device attribute states of the additional mobile device over multiple times matches the first non-observable device attributes of the failure model, to correlate the third sequence of observed device attribute states with the future failure status; and

controlling, by the configured computing system, ongoing operations for the additional mobile device to mitigate the future failure status of the additional mobile device, including modifying configuration of the additional mobile device to alter future device attribute states of the additional mobile device.

5. The computer-implemented method of claim 4 further comprising using, by the configured computing system, the failure model and the non-failure model to predict a future non-failure status of a further device, including determining that a fourth sequence of observed device attribute states of the further device matches the second non-observable device attributes of the non-failure model and does not match the first non-observable device attributes of the failure model, to correlate the fourth sequence of observed device attribute states with the future non-failure status,

and wherein ongoing operations of the further device are not modified based of the predicted future non-failure status.

6. The computer-implemented method of claim 4 wherein the controlling of the ongoing operations for the additional mobile device to mitigate the future failure status of the additional mobile device includes altering the future device attribute states of the additional mobile device to match the second non-observable device attributes of the non-failure model.

7. The computer-implemented method of claim 4 wherein the controlling of the ongoing operations for the additional mobile device to mitigate the future failure status of the additional mobile device includes changing one or more configuration settings of the additional mobile device that cause altering of the future device attribute states of the additional mobile device.

8. The computer-implemented method of claim 4 wherein the controlling of the ongoing operations for the additional mobile device to mitigate the future failure status of the additional mobile device further includes changing one or more types of interactions performed by one or more other computing devices with the additional mobile device.

9. The computer-implemented method of claim 4 wherein the controlling of the ongoing operations for the additional mobile device to mitigate the future failure status of the additional mobile device includes at least one of modifying future use of memory on the additional mobile device or modifying future use of communication capabilities of the additional mobile device.

10. The computer-implemented method of claim 4 wherein the obtaining of the failure model and the non-failure model includes tracking the first sequences of observed device attribute states and the second sequences of observed device attribute states, and includes generating the failure model and the non-failure model based on the tracking.

11. The computer-implemented method of claim 10 wherein the first and second sequences of observed device attribute states includes, at each of multiple times within the period of time and for each of at least some of the first and second plurality of mobile devices, attribute information about multiple of attribute states that include packet loss status, device restart status, application load time status, application crash status, connection speed status, dropped call status, screen capacitance status and accelerometer data status, and wherein the generating of the failure model and the non-failure model based on the tracking includes using the attribute information.

12. The computer-implemented method of claim 10 wherein the first and second sequences of observed device attribute states includes, at each of multiple times within the period of time and for each of at least some of the first and second plurality of mobile devices, attribute information for packet loss status, device restart status, application load time status, application crash status, connection speed status, dropped call status, screen capacitance status and accelerometer data status, and wherein the generating of the failure model and the non-failure model based on the tracking includes using the attribute information.

13. The computer-implemented method of claim 10 wherein the first and second sequences of observed device attribute states includes, at each of multiple times within the period of time and for each of at least some of the first and second plurality of mobile devices, attribute information for multiple of attribute states that include use of different types of communication capabilities, versions of executing software, an operating system in use, and characteristics of user interactions with one or more input devices, and wherein the generating of the failure model and the non-failure model based on the tracking includes using the attribute information.

14. The computer-implemented method of claim 10 wherein the first and second non-observable device attributes include multiple of a cracked screen, a data port failure, a charging port failure and an antenna failure, and wherein the generating of the failure model and the non-failure model based on the tracking includes modeling effects of the multiple of the first and second non-observable device attributes.

15. The computer-implemented method of claim 4 wherein the obtained failure model is segmented into multiple failure sub-models that are associated with distinct groups of device characteristics corresponding to subsets of the first plurality of mobile devices, and wherein the using of the failure model and the non-failure model to predict the future failure status of the additional mobile device includes selecting, by the configured computing system, one of the multiple failure sub-models to use based at least in part on the associated device characteristics of the selected one failure sub-model matching device characteristics of the additional mobile device, and matching the observed device attribute states of the third sequence to a subset of the observed device attribute states of the first sequences that are part of the selected one failure sub-model.

16. The computer-implemented method of claim 4 wherein the predicted future failure status of the additional mobile device includes a security failure for the additional mobile device, and wherein the controlling of the ongoing operations includes initiating use of additional security measures on the additional mobile device.

17. The computer-implemented method of claim 4 wherein the controlling of the ongoing operations includes initiating use of a further mobile device to replace the additional mobile device.

18. The computer-implemented method of claim 4 wherein the configured computing system is a server system separated from the additional mobile device by one or more computer networks, and wherein the using of the failure model and the non-failure model to predict a future failure status of an additional mobile device is performed for each of a plurality of additional mobile devices.

19. The computer-implemented method of claim 4 wherein the configured computing system is the additional mobile device, and wherein the using of the failure model and the non-failure model to predict the future failure status includes executing an automated operations manager system on the additional mobile device to perform the controlling of the ongoing operations for the additional mobile device.

20. The computer-implemented method of claim 4 further comprising notifying, by at least one of the configured computing system or the additional mobile device, a user of one or more configuration modifications for the additional mobile device, and wherein the modifying of the configuration of the additional mobile device is based at least in part on a response of the user to the notifying.

21. The computer-implemented method of claim 4 wherein the modifying of the configuration of the additional mobile device includes selecting one or more modification actions to perform based on at least one of information specific to the additional mobile device to customize the selecting to the additional mobile device or information specific to a user of the additional mobile device to customize the selecting to the user.

22. The computer-implemented method of claim 4 wherein the using of the failure model and the non-failure model to predict the future failure status of the additional mobile device further includes determining that the third sequence of observed device attribute states has a fit to the first non-observable device attributes of the failure model that exceeds a defined threshold and does not have a fit to the second non-observable device attributes of the non-failure model that exceeds the defined threshold.

23. The computer-implemented method of claim 4 wherein the additional mobile device is at least one of a smartphone, a portable gaming device, a tablet computer or a laptop computer, wherein the failure model further includes information about determined probabilities of occurrences of the observed device attribute states of the first sequences and of the first non-observable device attributes, and wherein the using of the failure model and the non-failure model to predict the future failure status of the additional mobile device includes using the determined probabilities for the third sequence of observed device attribute states to determine that a likelihood of the additional mobile device failing by a specified time exceeds a defined threshold.

24. A non-transitory computer-readable medium having stored contents that cause one or more computing systems to perform automated operations including at least:

obtaining, by the one or more computing systems, a failure model generated using first sequences over time of observed device attribute states of a first plurality of computing devices each having a failure status after the first sequences, and a non-failure model generated using second sequences over time of observed device attribute states of a second plurality of computing devices each having a non-failure status after the second sequences, wherein the failure model includes information relating the observed device attribute states of the first sequences to first non-observable device attributes associated with the failure status, and wherein the non-failure model includes information relating the observed device attribute states of the second sequences to second non-observable device attributes associated with the non-failure status;

using, by the one or more computing systems, the failure model and the non-failure model to predict a future failure status of an additional computing device, including determining that a third sequence of observed device attribute states of the additional computing device over multiple times matches the first non-observable device attributes of the failure model, to correlate the third sequence of observed device attribute states with the future failure status; and

controlling, by the one or more computing systems, ongoing operations to mitigate the future failure status of the additional computing device.

25. The non-transitory computer-readable medium of claim 24 wherein the additional computing device is a mobile device, wherein the controlling of the ongoing operations includes initiating use of a further computing device to replace the additional computing device, wherein the using of the failure model and the non-failure model to predict the future failure status of the additional computing device further includes determining that the third sequence of observed device attribute states has a fit to the first non-observable device attributes of the failure model that exceeds a defined threshold and does not have a fit to the second non-observable device attributes of the non-failure model that exceeds the defined threshold, and wherein the stored contents include software instructions that, when executed, program the one or more computing systems to perform the automated operations.

26. The non-transitory computer-readable medium of claim 24 wherein the stored contents further cause the one or more computing systems to perform additional automated operations including at least:

using, by the one or more computing systems, a third model and a fourth model to predict a future health impairment diagnosis of an indicated patient, wherein the third model includes information relating, for first health attribute states observed over time for a first group of target patients that each has a subsequent health impairment diagnosis, the first health attribute states to first non-observable health attributes associated with the health impairment diagnosis, wherein the fourth model includes information relating, for second health attribute states observed over time for a second group of target patients that each has a subsequent health diagnosis of non-impairment, the second health attribute states to second non-observable health attributes associated with the health diagnosis of non-impairment, and wherein the using of the third and fourth models includes determining that additional health attribute states observed for the indicated patient over a plurality of times matches the first non-observable health attributes of the third model, to correlate the additional health attribute states with the future health impairment diagnosis; and

providing, by the one or more computing systems, information about the predicted future health impairment diagnosis, to cause further actions to be taken for the indicated patient to address the future health impairment diagnosis.

27. The non-transitory computer-readable medium of claim 24 wherein the stored contents further cause the one or more computing systems to perform additional automated operations including at least:

using, by the one or more computing systems, a third model and a fourth model to predict that an indicated person in a physical store will participate in a future activity of an indicated type, wherein the third model includes information relating, for first activity attribute states observed over time for a first group of target people that each participates in a subsequent activity of the indicated type, the first activity attribute states to first non-observable activity attributes associated with participating in the activity of the indicated type, wherein the fourth model includes information relating, for second activity attribute states observed over time for a second group of control people that each does not participate in a subsequent activity of the indicated type, the second activity attribute states to second non-observable activity attributes associated with not participating in the activity of the indicated type, and wherein the using of the third and fourth models includes determining that additional activity attribute states observed for the indicated person over a plurality of times matches the first non-observable activity attributes of the third model, to correlate the additional activity attribute states with participating in the future activity of the indicated type; and

providing, by the one or more computing systems, information about the predicted participation in the future activity of the indicated type, to cause further actions to be taken for the indicated person that correspond to the future activity of the indicated type.

28. A system comprising:

one or more hardware processors of one or more computing systems; and

one or more memories with stored instructions that, when executed, cause the system to perform automated operations including at least;

obtaining a failure model generated using first sequences over time of observed device attribute states of a first plurality of computing devices each having a failure status after the first sequences, and a non-failure model generated using second sequences over time of observed device attribute states of a second plurality of computing devices each having a non-failure status after the second sequences, wherein the failure model includes information relating the observed device attribute states of the first sequences to first non-observable device attributes associated with the failure status, and wherein the non-failure model includes information relating the observed device attribute states of the second sequences to second non-observable device attributes associated with the non-failure status;

using the failure model and the non-failure model to predict a future failure status of an additional computing device, including determining that a third sequence of observed device attribute states of the additional computing device over multiple times matches the first non-observable device attributes of the failure model, to correlate the third sequence of observed device attribute states with the future failure status; and

controlling ongoing operations of the additional computing device to mitigate the future failure status of the additional computing device.

29. The system of claim 28 wherein the additional computing device is a mobile device, wherein the predicted future failure status of the additional computing device includes a security failure for the additional computing device, wherein the controlling of the ongoing operations includes initiating use of additional security measures on the additional computing device, and wherein the stored instructions include software instructions that, when executed, program the one or more computing systems to perform the automated operations.

30. The system of claim 28 wherein the using of the failure model and the non-failure model to predict the future failure status of the additional computing device further includes determining that the third sequence of observed device attribute states has a fit to the first non-observable device attributes of the failure model that exceeds a defined threshold and does not have a fit to the second non-observable device attributes of the non-failure model that exceeds the defined threshold.