Control of sensor-equipped asset based on aggregated risk values

1 . A computing platform for determining aggregated risk values, comprising:

a communication interface;

one or more processors;

a non-transitory computer-readable medium; and

program instructions stored on the non-transitory computer-readable medium that are executable by the one or more processors to cause the computing platform to:

based on operating data for a sensor-equipped asset, execute a set of predictive models, wherein each predictive model in the set of predictive models is configured to (i) evaluate a respective portion of the operating data for the sensor-equipped asset and (ii) output a respective prediction related to an operation of the sensor-equipped asset, each respective prediction having a respective confidence value and a respective severity value,

wherein the operating data comprises at least one of: (a) measure data of the sensor-equipped asset, or (b) data indicating an abnormal condition occurrence of the sensor-equipped asset;

detect a triggering event for determining an aggregated risk value for the sensor-equipped asset;

responsive to detecting the triggering event, identify (i) a set of predictions related to the operation of the sensor-equipped asset based on the operating data and (ii) a risk dataset, wherein the risk dataset defines (a) a first set of confidence levels, (b) a second set of severity levels, and (c) a risk value for each pairing of a confidence level from the first set and a severity level from the second set;

determine the aggregated risk value of the sensor-equipped asset based on the set of predictions and the risk dataset by:

for each severity level from the second set, identifying a matching associated severity value from the respective predictions;

for each severity level having one or more matching associated severity values from the respective predictions identified:

determining a severity-specific confidence level by either:

(i) if one matching associated severity value is identified, determining that the respective severity level has a severity-specific confidence level corresponding to an associated confidence value of the respective prediction, or

(ii) if two or more of the respective predictions are identified, determining that the respective severity level has a severity-specific confidence level corresponding to an aggregation of associated confidence values of the respective predictions, wherein the aggregation is determined by: for each of the two or more of the respective predictions, determining a complement value of the associated confidence values, multiplying the complement values together and thereby determining a product of the complement values, and determining a product complement value of the product of the complement values, and determining that the severity-specific confidence level corresponds to the product complement value, and

based on the severity-specific confidence level and the risk dataset, determining a severity-specific risk value for the respective severity level;

comparing the one or more severity-specific risk values to determine a highest risk value; and

determining that the highest risk value is the aggregated risk value for the sensor-equipped asset; and

cause a client device to display a visual representation of the aggregated risk value for the sensor-equipped asset;

determine that the aggregated risk value for the sensor-equipped asset meets a threshold risk criteria; and

in response to the determination, execute a remedial action for the sensor-equipped asset, wherein the remedial action comprises causing the sensor-equipped asset to modify its physical operation.

2 . The computing platform of claim 1 , wherein the aggregated risk value comprises an asset-level aggregated risk value, and wherein the set of predictive models comprises predictive models related to a plurality of subsystems of the sensor-equipped asset.

3 . The computing platform of claim 1 , wherein the aggregated risk value comprises a subsystem-level aggregated risk value, and wherein the set of predictive models comprises predictive models related to a subsystem of the sensor-equipped asset.

4 . The computing platform of claim 1 , wherein the risk dataset comprises a risk dataset that is pre-established based on user input.

5 . The computing platform of claim 1 , wherein the remedial action comprises one or more of: (a) causing a computing device to output an alert indicating that the aggregated risk value for the sensor-equipped asset meets the threshold risk criteria, (b) causing the computing device to output an indication of one or more recommended repairs to the sensor-equipped asset, or (c) transmitting, to a parts ordering system, part-order data to cause the parts-ordering system to order a component for the sensor-equipped asset.

6 . The computing platform of claim 1 , wherein the set of predictive models comprises one or more of: (i) an event-prediction model, (ii) a survival analysis model, (iii) an anomaly detection model, or (iv) a rule-based model.

7 . The computing platform of claim 1 , wherein the program instructions are further executable by the at least one processor to cause the computing platform to:

prior to receiving the operating data for the sensor-equipped asset, cause a second client device to present an interface for establishing the risk dataset.

8 . The computing platform of claim 7 , wherein:

the respective confidence value represents a likelihood of the respective prediction being correct; and

the respective severity value is from a set of predefined severity values available for assignment to predictions, wherein each severity value in the set of predefined severity values is a predefined categorization of impact severity that provides an indication of how severely the operation of the sensor-equipped asset is expected to be impacted by an occurrence of an event related to the operation of the sensor-equipped asset.

9 . The computing platform of claim 1 , wherein:

each confidence level of the first set comprises a range of confidence values;

each severity level of the second set is from a set of predefined severity levels available for assignment to predictions; and

the risk value is from a set of risk values and comprises a textual indicator that indicates a risk associated with a prediction.

10 . A computer-implemented method for determining aggregated risk values comprising:

executing, by one or more processors and based on operating data for a sensor-equipped asset, a set of predictive models, wherein each predictive model in the set of predictive models is configured to (i) evaluate a respective portion of the operating data for the sensor-equipped asset and (ii) output a respective prediction related to an operation of the sensor-equipped asset, each respective prediction having a respective confidence value and a respective severity value,

wherein the operating data comprises at least one of: (a) measure data of the sensor-equipped asset, or (b) data indicating an abnormal condition occurrence of the sensor-equipped asset;

detecting, by the one or more processors, a triggering event for determining an aggregated risk value for the sensor-equipped asset;

responsive to detecting the triggering event, identifying, by the one or more processors, (i) a set of predictions related to the operation of the sensor-equipped asset based on the operating data and (ii) a risk dataset, wherein the risk dataset defines (a) a first set of confidence levels, (b) a second set of severity levels, and (c) a risk value for each pairing of a confidence level from the first set and a severity level from the second set;

determining, by the one or more processors, the aggregated risk value of the sensor-equipped asset based on the set of predictions and the risk dataset by:

for each severity level from the second set, identifying, by the one or more processors, a matching associated severity value from the respective predictions;

for each severity level having one or more matching associated severity values from the respective predictions identified:

determining, by the one or more processors, a severity-specific confidence level by either:

(i) if one matching associated severity value is identified, determining that the respective severity level has a severity-specific confidence level corresponding to an associated confidence value of the respective prediction, or

(ii) if two or more of the respective predictions are identified, determining that the respective severity level has a severity-specific confidence level corresponding to an aggregation of associated confidence values of the respective predictions, wherein the aggregation is determined by: for each of the two or more of the respective predictions, determining a complement value of the associated confidence values, multiplying the complement values together and thereby determining a product of the complement values, and determining a product complement value of the product of the complement values, and determining that the severity-specific confidence level corresponds to the product complement value, and

based on the severity-specific confidence level and the risk dataset, determining, by the one or more processors, a severity-specific risk value for the respective severity level;

comparing, by the one or more processors, the one or more severity-specific risk values to determine a highest risk value; and

determining, by the one or more processors, that the highest risk value is the aggregated risk value for the sensor-equipped asset; and

causing, by the one or more processors, a client device to display a visual representation of the aggregated risk value for the sensor-equipped asset;

determining, by the one or more processors, that the aggregated risk value for the sensor-equipped asset meets a threshold risk criteria; and

in response to the determination, executing, by the one or more processors, a remedial action for the sensor-equipped asset, wherein the remedial action comprises causing the sensor-equipped asset to modify its physical operation.

11 . The computer-implemented method of claim 10 , wherein the remedial action comprises one or more of: (a) causing a computing device to output an alert indicating that the aggregated risk value for the sensor-equipped asset meets the threshold risk criteria, (b) causing the computing device to output an indication of one or more recommended repairs to the sensor-equipped asset, or (c) transmitting, to a parts ordering system, part-order data to cause the parts-ordering system to order a component for the sensor-equipped asset.

12 . A non-transitory computer readable storage medium comprising computer readable instructions stored thereon for determining aggregated risk values, the instructions when executed on one or more processors cause the one or more processors to:

execute, based on operating data for a sensor-equipped asset, a set of predictive models, wherein each predictive model in the set of predictive models is configured to (i) evaluate a respective portion of the operating data for the sensor-equipped asset and (ii) output a respective prediction related to an operation of the sensor-equipped asset, each respective prediction having a respective confidence value and a respective severity value,

wherein the operating data comprises at least one of: (a) measure data of the sensor-equipped asset, or (b) data indicating an abnormal condition occurrence of the sensor-equipped asset;

detect a triggering event for determining an aggregated risk value for the sensor-equipped asset;

responsive to detecting the triggering event, identify (i) a set of predictions related to the operation of the sensor-equipped asset based on the operating data and (ii) a risk dataset, wherein the risk dataset defines (a) a first set of confidence levels, (b) a second set of severity levels, and (c) a risk value for each pairing of a confidence level from the first set and a severity level from the second set;

determine the aggregated risk value of the sensor-equipped asset based on the set of predictions and the risk dataset by:

for each severity level from the second set, identifying a matching associated severity value from the respective predictions;

for each severity level having one or more matching associated severity values from the respective predictions identified:

determining a severity-specific confidence level by either:

(i) if one matching associated severity value is identified, determining that the respective severity level has a severity-specific confidence level corresponding to an associated confidence value of the respective prediction, or

(ii) if two or more of the respective predictions are identified, determining that the respective severity level has a severity-specific confidence level corresponding to an aggregation of associated confidence values of the respective predictions, wherein the aggregation is determined by: for each of the two or more of the respective predictions, determining a complement value of the associated confidence values, multiplying the complement values together and thereby determining a product of the complement values, and determining a product complement value of the product of the complement values, and determining that the severity-specific confidence level corresponds to the product complement value, and

based on the severity-specific confidence level and the risk dataset, determining a severity-specific risk value for the respective severity level;

comparing the one or more severity-specific risk values to determine a highest risk value; and

determining that the highest risk value is the aggregated risk value for the sensor-equipped asset; and

cause a client device to display a visual representation of the aggregated risk value for the sensor-equipped asset;

determine that the aggregated risk value for the sensor-equipped asset meets a threshold risk criteria; and

in response to the determination, execute a remedial action for the sensor-equipped asset, wherein the remedial action comprises causing the sensor-equipped asset to modify its physical operation.