Digital twins for energy efficient asset maintenance
A system for using digital twins for scalable, model-based machine predictive maintenance comprises a plurality of digital twins and a simulation platform. The plurality of digital twins correspond to plurality of remotely located physical machines. Each respective digital twin comprises: product nameplate data corresponding to a unique physical machine, one or more simulation models, and a database comprising run time log data collected from sensors associated with the unique physical machine. The simulation platform is configured to process simulation models corresponding to the plurality of digital twins using a plurality of multiprocessor computer systems.
1. An automation system for using digital twins for scalable, model-based machine predictive maintenance, the automation system comprising:
a plurality of digital twins corresponding to plurality of remotely located physical machines; and
a simulation platform configured to process a plurality of simulation models corresponding to the plurality of digital twins using a plurality of multiprocessor computer systems, wherein each respective digital twin comprises:
product nameplate data corresponding to a unique physical machine of the plurality of remotely located physical machines,
one or more simulation models of the plurality of simulation models, the one or more simulation models corresponding to the unique physical machine, and
a database comprising run time log data collected from sensors associated with the unique physical machine,
wherein each respective digital twin is configured to:
calculate a domain-specific Condition Index (CI) value indicative of energy efficiency of the unique physical machine corresponding to the respective digital twin, wherein an observer block operation is configured within the one or more simulation models corresponding to the unique physical machine to calculate the domain-specific CI value, wherein the domain-specific CI value is calculated using a probabilistic framework based on automatic inference to account at least for sensor noise, insufficient hardware sensors, and unknown characteristic curves representative of performance of the unique physical machine over a period of time,
wherein the observer block operation configured within the one or more simulation models calculates the domain-specific CI value using a plurality of individual observer blocks connected serially, in parallel, or using a combination of serial and parallel connections,
wherein the observer block operation is configured to generate estimates based on for the insufficient hardware sensors, and the unknown characteristic curves,
identify one or more required maintenance tasks for the unique physical machine based on the calculated domain-specific CI value, and
send a notification of the one or more required maintenance tasks to an operator device.
2. The automation system of claim 1 , further comprising:
a data platform configured to process a plurality of data query tasks using the plurality of multiprocessor computer systems.
3. The automation system of claim 2 , wherein the data platform utilizes a map-reduce programming model to process each of the plurality of data query tasks.
4. The automation system of claim 1 , wherein each simulation model included in the plurality of digital twins is implemented using a Bayesian filtering framework.
5. The automation system of claim 1 , wherein each respective digital twin comprises a web service interface configured to facilitate communication between the respective digital twin and one or more remote devices.
6. The automation system of claim 5 , wherein the system further comprises a mobile device interface configured to facilitate monitoring of the plurality of remotely located physical machines via the plurality of digital twins.
7. The automation system of claim 5 , wherein the system further comprises a sensor interface configured to facilitate transfer of the run time log data from plurality of physical machines to the plurality of digital twins.
8. The automation system of claim 7 , wherein the run time log data is collected from plurality of physical machines to the plurality of digital twins in real-time.
9. The automation system of claim 1 , wherein each respective digital twin further comprises:
a multimedia database configured to store maintenance data associated with the unique physical machine and inspection data associated with the unique physical machine.
10. The automation system of claim 1 , wherein the simulation platform is configured to execute each respective simulation model of the plurality of simulation models using a plurality of simulation engines executing in parallel across a plurality of processors on the plurality of multiprocessor computer systems.
11. A computer-implemented method for using a digital twin for scalable machine maintenance of a remotely located digital twin, the method comprising:
generating, by a computer system, a digital twin corresponding to a physical machine in a digital twin repository stored on the computer system by aggregating observer blocks of serial or parallel digital twin components;
receiving, by the computer system, sensor data from the physical machine;
storing, by the computer system, the sensor data in association with the digital twin;
identifying, by the computer system, one or more required maintenance tasks for the physical machine; and
sending, by the computer system, a notification of the one or more required maintenance tasks to an operator device,
wherein the digital twin comprises a simulation model to simulate operation of the physical machine based on the sensor data, wherein a simulated operation of the physical machine estimates energy efficiency of the physical machine,
calculating a domain-specific Condition Index (CI) value indicative of energy efficiency of the physical machine corresponding to the respective digital twin, wherein an observer block operation is configured within the simulation model corresponding to the physical machine to calculate the domain-specific CI value, wherein the domain-specific CI value is calculated using a probabilistic framework based on automatic inference to account at least for sensor noise, insufficient hardware sensors, and unknown characteristic curves representative of performance of the unique physical machine over a period of time,
wherein the observer block operation configured within the simulation model calculates the domain-specific CI value using a plurality of individual observer blocks connected serially, in parallel, or using a combination of serial and parallel connections,
wherein the observer block operation is configured to generate estimates to account for the insufficient hardware sensors, and the unknown characteristic curves,
wherein the one or more required maintenance tasks are identified for the physical machine based on the calculated domain-specific CI value of the physical machine.
12. The method of claim 11 , wherein the digital twin is generated based on one or more of a manual associated with the physical machine or a datasheet associated with the physical machine received from the physical machine's original equipment manufacturer.
13. The method of claim 11 , further comprising:
receiving, by the computer system, maintenance results generated in response to performance of the one or more required maintenance tasks by a maintenance engineer; and
storing, by the computer system, the maintenance results in association with the digital twin.
14. The method of claim 11 , further comprising:
receiving, by the computer system, inspection results generated in response to performance of an inspection of the physical machine; and
storing, by the computer system, the inspection results in association with the digital twin.
15. The method of claim 11 , further comprising:
executing the simulation model on the computer system using a plurality of simulation engines operating in parallel to identify the one or more required maintenance tasks.
16. An automation system for using a digital twin for scalable machine maintenance, the system comprising:
an embedded computing sensor co-located with a physical machine and configured to:
collect monitoring data from the physical machine; and
transfer the monitoring data to a data center; and
a digital twin of the physical machine located at the data center and configured to:
identify one or more required maintenance tasks for the physical machine based on the monitoring data, and
send a notification of the one or more required maintenance tasks to an operator device,
wherein the digital twin comprises a simulation model to simulate operation of the physical machine based on the monitoring data, wherein a simulated operation of the physical machine estimates energy efficiency of the physical machine,
calculating a domain-specific Condition Index (CI) value indicative of energy efficiency of the physical machine corresponding to the respective digital twin, wherein an observer block operation is configured within the simulation model corresponding to the physical machine to calculate the domain-specific CI value,
wherein the domain-specific CI value is calculated using a probabilistic framework based on automatic inference to account at least for sensor noise, insufficient hardware sensors, and unknown characteristic curves representative of performance of the unique physical machine over a period of time,
wherein the calculated CI value is calculated using a probabilistic framework configured to account at least for sensor noise, insufficient hardware sensors, and unknown characteristic curves representative of performance of the physical machine over a period of time,
wherein the observer block operation configured within the simulation model calculates the domain-specific CI value using a plurality of individual observer blocks connected serially, in parallel, or using a combination of serial and parallel connections,
wherein the observer block operation is configured to generate estimates to account for the insufficient hardware sensors, and the unknown characteristic curves, wherein the one or more required maintenance tasks are identified for the physical machine based on the calculated domain-specific CI value of the physical machine.