Asset health score
First Claim
Patent Images
1. A computing system comprising:
- a network interface configured to facilitate communication with a plurality of assets equipped with sensors and a plurality of computing devices;
at least one processor;
a non-transitory computer-readable medium; and
program instructions stored on the non-transitory computer-readable medium that are executable by the at least one processor to cause the computing system to;
for each respective failure type of a group of failure types;
(i) identify one or more types of abnormal-condition indicators associated with a respective failure type,(ii) identify past instances of the identified one or more types of abnormal-condition indicators at the plurality of assets equipped with sensors, wherein each past instance of a respective type of abnormal-condition indicator is indicative of a past occurrence of a respective type of abnormal condition at one asset of the plurality of assets,(iii) identify historical sensor data corresponding to the identified past instances of the identified one or more types of abnormal-condition indicators, wherein the historical sensor data for each past instance of a respective type of abnormal-condition indicator indicates operating conditions of the one asset associated with a past occurrence of a respective type of abnormal condition at the one asset, and(iv) apply a supervised machine learning technique to the identified historical sensor data to define a respective predictive model for the respective failure type that is configured to receive sensor data for the one asset as input andoutput a value indicating whether the respective failure type is likely to occur at the one asset within a given period of time in the future;
combine the respective predictive models for the group of failure types into an aggregated predictive model that is configured to receive sensor data for the one asset as input and output a health metric indicating whether at least one failure type from the group of failure types is likely to occur at the one asset within a given period of time in the future;
receive sensor data indicating operating conditions of a given asset;
apply the aggregated predictive model to the received sensor data to determine, for the given asset, a health metric indicating whether at least one failure type from the group of failure types is likely to occur at the given asset within the given period of time in the future;
detect that the health metric for the given asset satisfies threshold criteria; and
in response to the detection, cause a computing device to display a visual indicator that at least one failure type from the group of failure types is likely to occur at the given asset.
2 Assignments
0 Petitions
Accused Products
Abstract
Disclosed herein are systems, devices, and methods related to assets and asset operating conditions. In particular, examples involve defining and executing predictive models for outputting health metrics that estimate the operating health of an asset or a part thereof, analyzing health metrics to determine variables that are associated with high health metrics, and modifying the handling of abnormal-condition indicators in accordance with a prediction of a likely response to such abnormal-condition indicators, among other examples.
182 Citations
20 Claims
-
1. A computing system comprising:
-
a network interface configured to facilitate communication with a plurality of assets equipped with sensors and a plurality of computing devices; at least one processor; a non-transitory computer-readable medium; and program instructions stored on the non-transitory computer-readable medium that are executable by the at least one processor to cause the computing system to; for each respective failure type of a group of failure types; (i) identify one or more types of abnormal-condition indicators associated with a respective failure type, (ii) identify past instances of the identified one or more types of abnormal-condition indicators at the plurality of assets equipped with sensors, wherein each past instance of a respective type of abnormal-condition indicator is indicative of a past occurrence of a respective type of abnormal condition at one asset of the plurality of assets, (iii) identify historical sensor data corresponding to the identified past instances of the identified one or more types of abnormal-condition indicators, wherein the historical sensor data for each past instance of a respective type of abnormal-condition indicator indicates operating conditions of the one asset associated with a past occurrence of a respective type of abnormal condition at the one asset, and (iv) apply a supervised machine learning technique to the identified historical sensor data to define a respective predictive model for the respective failure type that is configured to receive sensor data for the one asset as input and output a value indicating whether the respective failure type is likely to occur at the one asset within a given period of time in the future; combine the respective predictive models for the group of failure types into an aggregated predictive model that is configured to receive sensor data for the one asset as input and output a health metric indicating whether at least one failure type from the group of failure types is likely to occur at the one asset within a given period of time in the future; receive sensor data indicating operating conditions of a given asset; apply the aggregated predictive model to the received sensor data to determine, for the given asset, a health metric indicating whether at least one failure type from the group of failure types is likely to occur at the given asset within the given period of time in the future; detect that the health metric for the given asset satisfies threshold criteria; and in response to the detection, cause a computing device to display a visual indicator that at least one failure type from the group of failure types is likely to occur at the given asset. - View Dependent Claims (2, 3, 4, 5, 6, 7, 8)
-
-
9. A non-transitory computer-readable medium having instructions stored thereon that are executable to cause a computing system to:
-
for each respective failure type of a group of failure types; (i) identify one or more types of abnormal-condition indicators associated with a respective failure type, (ii) identify past instances of the identified one or more types of abnormal-condition indicators at a plurality of assets equipped with sensors, wherein each past instance of a respective type of abnormal-condition indicator is indicative of a past occurrence of a respective type of abnormal condition at one asset of the plurality of assets, (iii) identify historical sensor data corresponding to the identified past instances of the identified one or more types of abnormal-condition indicators, wherein the historical sensor data for each past instance of a respective type of abnormal-condition indicator indicates operating conditions of the one asset associated with a past occurrence of a respective type of abnormal condition at the one asset, and (iv) apply a supervised machine learning technique to the identified historical sensor data to define a respective predictive model for the respective failure type that is configured to receive sensor data for the one asset as input and output a value indicating whether the respective failure type is likely to occur at the one asset within a given period of time in the future; combine the respective predictive models for the group of failure types into an aggregated predictive model that is configured to receive sensor data for the one asset as input and output a health metric indicating whether at least one failure type from the group of failure types is likely to occur at the one asset within a given period of time in the future; receive sensor data indicating operating conditions of a given asset; apply the aggregated predictive model to the received sensor data to determine, for the given asset, a health metric indicating whether at least one failure type from the group of failure types is likely to occur at the given asset within the given period of time in the future; and detect that the health metric for the given asset satisfies threshold criteria; and in response to the detection, cause a computing device to display a visual indicator that at least one failure type from the group of failure types is likely to occur at the given asset. - View Dependent Claims (10, 11, 12, 13, 14)
-
-
15. A computer-implemented method, the method comprising:
-
for each respective failure type of a group of failure types; (i) identifying one or more types of abnormal-condition indicators associated with a respective failure type, (ii) identifying past instances of the identified one or more types of abnormal-condition indicators at a given set of assets equipped with sensors, wherein each past instance of a respective type of abnormal-condition indicator is indicative of a past occurrence of a respective type of abnormal condition at one asset of the plurality of assets, (iii) identifying historical sensor data corresponding to the identified past instances of the identified one or more types of abnormal-condition indicators at the plurality of assets, wherein the historical sensor data for each past instance of a respective type of abnormal-condition indicator indicates operating conditions of the one asset associated with a past occurrence of a respective type of abnormal condition at the one asset, and (iv) applying a supervised machine learning technique to the identified historical sensor data to define a respective predictive model for the respective failure type that is configured to receive sensor data for the one asset as input and output a value indicating whether the respective failure type is likely to occur at the one asset within a given period of time in the future; combining the respective predictive models for the group of failure types into an aggregated predictive model that is configured to receive sensor data for the one asset as input and output a health metric indicating whether at least one failure type from the group of failure types is likely to occur at the one asset within a given period of time in the future; receiving sensor data indicating operating conditions of a given asset; applying the aggregated predictive model to the received sensor data to determine, for the given asset, a health metric indicating whether at least one failure type from the group of failure types is likely to occur at the given asset within the given period of time in the future; detecting that the health metric for the given asset satisfies threshold criteria; and in response to the detecting, causing a computing device to display a visual indicator that at least one failure type from the group of failure types is likely to occur at the given asset. - View Dependent Claims (16, 17, 18, 19, 20)
-
Specification