Continous monitoring, advance alerts, and control of key performance indicator variables
View Patent ↗Embodiments of the invention are directed to collecting, by a computer system, sensor data of a manufacturing system, the sensor data being measured at intervals smaller than a time interval of a target measurement of the manufacturing system. The sensor data is determined to have a relationship to the target measurement. A synthetic target measurement is generated at an interval smaller than the time interval based on the relationship. An advance warning is automatically generated for the target measurement based on the synthetic target measurement within the interval smaller than the time interval.
1. A computer-implemented method comprising:
collecting, by a computer system, sensor data of a manufacturing system, the sensor data being measured at a plurality of intervals smaller than a time interval “T” of a target measurement of the manufacturing system, wherein the sensor data is determined to have a relationship to the target measurement;
generating, by the computer system, a synthetic target measurement at a first interval of the plurality of intervals based on the relationship, the first interval being smaller than the time interval “T”, wherein the computer system employs a machine learning model to generate the synthetic target measurement, the machine learning model comprising a neural network trained on training data comprising historical sensor data and historical target measurement data, wherein training the machine learning model comprises fitting a regression model that relates the sensor data being measured at the plurality of intervals to the historical target measurement data at the time interval “T” through unknown intermediate quality results in which the historical target measurement data are based on equal volumes sampled uniformly over time;
automatically generating, by the computer system, an advance warning for the target measurement based on the synthetic target measurement within the interval smaller than the time interval; and
causing a control system of the manufacturing system to change operation of an actuator in response to the advance warning.
2. The computer-implemented method of claim 1 , wherein one or more set points associated with the manufacturing system are automatically revised in response to the advance warning for the target measurement based on the synthetic target measurement.
3. The computer-implemented method of claim 1 , wherein the advance warning is generated in response to the synthetic target measurement being outside a predetermined range.
4. The computer-implemented method of claim 1 , wherein modifications are made to the manufacturing system to bring the synthetic target measurement within a predetermined range.
5. The computer-implemented method of claim 1 , wherein the target measurement is a quality-related variable corresponding to a physical material process outflow stream of the manufacturing system, the target measurement comprising a non-instantaneous, time-aggregated nature due to mixing equal sampled volumes in a vessel where the equal sampled volumes are collected at multiple instants in time throughout the time interval and measured at an end of the time interval, thereby obtaining the target measurement.
6. The computer-implemented method of claim 5 , wherein:
the target measurement is a composite measurement of an aggregated total volume of a sample; and
the synthetic target measurement is a generated value for a point in time based on the sensor data and is not a measurement of the aggregated total volume of the sample, the synthetic target measurement representing a state of an individual sampled volume at the point in time within the time interval.
7. The computer-implemented method of claim 5 , wherein:
the target measurement is a composite measurement of an aggregated total volume of a sample;
an aggregation function that produces the composite measurement from the equal sampled volumes that make up the aggregated total volume is unknown and is learned automatically as a function of the target measurement across the equal sampled volumes; and
the synthetic target measurement is a generated value for a point in time based on the sensor data and is not a measurement of the aggregated total volume of the sample, the synthetic target measurement representing a state of an individual sampled volume of the equal sampled volumes at the point in time within the time interval.
8. The computer-implemented method of claim 5 , wherein:
the target measurement is a composite measurement of an aggregated total volume of a sample;
an aggregation function that produces the composite measurement from the equal sampled volumes that make up the aggregated total volume is known to be an average of the target measurement across the equal sampled volumes; and
the synthetic target measurement is a generated value for a point in time based on the sensor data and is not a measurement of the aggregated total volume of the sample, the synthetic target measurement representing a state of an individual sampled volume of the equal sampled volumes at the point in time within the time interval.
9. The computer-implemented method of claim 5 , wherein:
the target measurement is a composite measurement of an aggregated total volume of a sample;
an aggregation function that produces the composite measurement from the equal sampled volumes that make up the aggregated total volume is known from a user-specified function of the target measurement across the equal sampled volumes; and
the synthetic target measurement is a generated value for a point in time based on the sensor data and is not a measurement of the aggregated total volume of the sample, the synthetic target measurement representing a state of an individual sampled volume of the equal sampled volumes at the point in time within the time interval.
10. A system comprising:
a non-transitory memory having computer readable instructions; and
one or more processors for executing the computer readable instructions, the computer readable instructions controlling the one or more processors to perform operations comprising:
collecting sensor data of a manufacturing system, the sensor data being measured at a plurality of intervals smaller than a time interval “T” of a target measurement of the manufacturing system, wherein the sensor data is determined to have a relationship to the target measurement;
generating a synthetic target measurement at a first interval of the plurality of intervals based on the relationship, the first interval being smaller than the time interval “T”, wherein the computer system employs a machine learning model to generate the synthetic target measurement, the machine learning model comprising a neural network trained on training data comprising historical sensor data and historical target measurement data, wherein training the machine learning model comprises fitting a regression model that relates the sensor data being measured at the plurality of intervals to the historical target measurement data at the time interval “T” through unknown intermediate quality results in which the historical target measurement data are based on equal volumes sampled uniformly over time;
automatically generating an advance warning for the target measurement based on the synthetic target measurement within the interval smaller than the time interval; and
causing a control system of the manufacturing system to change operation of an actuator in response to the advance warning.
11. The system of claim 10 , wherein one or more set points associated with the manufacturing system are automatically revised in response to the advance warning for the target measurement based on the synthetic target measurement.
12. The system of claim 10 , wherein the advance warning is generated in response to the synthetic target measurement being outside a predetermined range.
13. The system of claim 10 , wherein modifications are made to the manufacturing system to bring the synthetic target measurement within a predetermined range.
14. The system of claim 10 , wherein the target measurement is a quality-related variable corresponding to a physical material process outflow stream of the manufacturing system, the target measurement comprising a non-instantaneous, time-aggregated nature due to mixing equal sampled volumes in a vessel where the equal sampled volumes are collected at multiple instants in time throughout the time interval and measured at an end of the time interval, thereby obtaining the target measurement.
15. The system of claim 14 , wherein:
the target measurement is a composite measurement of an aggregated total volume of a sample; and
the synthetic target measurement is a generated value for a point in time based on the sensor data and is not a measurement of the aggregated total volume of the sample, the synthetic target measurement representing a state of an individual sampled volume at the point in time within the time interval.
16. A computer program product comprising a non-transitory computer readable storage medium having program instructions embodied therewith, the program instructions executable by a processor to cause the processor to perform operations comprising:
collecting sensor data of a manufacturing system, the sensor data being measured at a plurality of intervals smaller than a time interval “T” of a target measurement of the manufacturing system, wherein the sensor data is determined to have a relationship to the target measurement;
generating a synthetic target measurement at a first interval of the plurality of intervals based on the relationship, the first interval being smaller than the time interval “T”, wherein the computer system employs a machine learning model to generate the synthetic target measurement, the machine learning model comprising a neural network trained on training data comprising historical sensor data and historical target measurement data, wherein training the machine learning model comprises fitting a regression model that relates the sensor data being measured at the plurality of intervals to the historical target measurement data at the time interval “T” through unknown intermediate quality results in which the historical target measurement data are based on equal volumes sampled uniformly over time;
automatically generating an advance warning for the target measurement based on the synthetic target measurement within the interval smaller than the time interval; and
causing a control system of the manufacturing system to change operation of an actuator in response to the advance warning.
17. The computer program product of claim 16 , wherein one or more set points associated with the manufacturing system are automatically revised in response to the advance warning for the target measurement based on the synthetic target measurement.