System and method for estimating remaining useful lifetime of electrical machines

Abstract

A system for determining the remaining useful lifetime of an electric machine is provided. The system includes an electrical machine, sensors that capture sensor data associated with the components of the electrical machine, and processors that perform operations of the system. The system receives, such as during an evaluation interval, signals from the sensors including sensor data associated with the components of the electrical machine. Based on the sensor data, the system determines a remaining useful lifetime for each component of the electric machine, such as by utilizing a physics-based model, and / or generates a prediction for the remaining useful lifetime for the components of the electric machine by utilizing an artificial intelligence model. Once the determination and / or prediction is made, the system outputs the remaining useful lifetime determination and / or prediction via an interface. Based on the remaining useful lifetime, actions can be initiated to maintain the electric machine.

Description

SYSTEM AND METHOD FOR ESTIMATING REMAINING USEFUL LIFETIME OFELECTRICAL MACHINESFIELD OF THE TECHNOLOGY

[0001] At least some embodiments disclosed herein relate to electrical machines, electric machine health analysis technologies, artificial intelligence technologies, component lifetime estimation technologies, and more particularly, but not limited to, a system and method for estimating remaining useful lifetime of electrical machines.BACKGROUND

[0002] Electrical machines, such as motors or generators, are utilized to convert between electricity and mechanical energy as torque associated with a rotating motor shaft. Generators transform the input mechanical power into output electricity, and motors transform the input electrical energy to mechanical energy. Motors can be used for a variety of different applications, such as, but not limited to, industrial fans, blowers and pumps, machine tools, factory equipment, conveyors, vehicles, industrial equipment, appliances, and the like. A variety of different types of motors exist, which utilize different types of electrical power. Electrical machines can be utilized in a variety of applications to create products, facilitate business operations, and perform useful services. As a result, electrical machines can often be critical to the performance of a business. Electrical machines can have a variety of different failure modes. For example, possible electrical machine failure modes can include bearing degradation due to inherent bearing currents, corrosion, metal fatigue, lubrication contamination, rotor bar cracking due to material fatigue at cycling loads, or age-related degradation of dielectrics and shorting of conductors. To attempt to address the variety of possible failure modes for electrical machines, businesses often deploy preventative inspection and maintenance programs before unplanned outages or catastrophic failures occur. While such programs can be helpful in reducing electrical machine downtime, such programs canbe expensive because preventative maintenance activities need to occur rather frequently as very little may be known about the electrical machine operating conditions.

[0003] Technologies and techniques for alleviating costs and downtime associated with electrical machine failures exist, however, such technologies often only detect imminent electrical machine failures. These technologies and techniques detect the onset of failure, and an immediate machine shutdown may be required. Additionally, such technologies fail to provide a comprehensive view of the electrical machine’s components, and electrical machines often fail shortly right after detection of an imminent failure. When electrical machines fail shortly after detection of an imminent failure, there is typically insufficient time to mitigate business production losses from a downed electrical machine and process incorporating the electrical machine. Existing techniques often involve using periodic inspection performed by field engineers, which requires manual labor and time. Such periodic inspections are not only costly, but can be hazardous depending on the location and placement of the electrical machines. While existing electrical machine failure detection techniques and technologies provide various benefits, such techniques and technologies may be enhanced to provide enhanced electrical machine monitoring and condition-detection capabilities, reduced electrical machine downtime, reduced costs, and enhanced electrical machine failure and lifetime predictive capabilities.SUMMARY

[0004] A system and accompanying methods for estimating the remaining useful lifetime of an electrical machine, such as a motor, is provided. In certain embodiments, the system and methods provide functionality to minimize deployment of preventative maintenance activities for electrical machines and their corresponding components. The system and methods provide longterm electrical machine remaining useful lifetime estimates that are updated continuously based on operating and environmental conditions, such as by utilizing any number of sensors embedded in the electrical machines. More specifically, the system and methods are capable of predicting the remaining useful lifetime of industrial electrical machines based on physics-based models, measurements from sensors embedded in the electrical machines, and historic databases of failures from identical or similar electrical machines. Notably, in certain embodiments, the system and methods do not require or rely on the detection of impending failures of components of electricalmachines from abnormalities identified in sensor signals obtained from such embedded sensors. The system and methods can utilize motor-design information, such as design information in the form of physics-based models, which can constitute equations related to the materials used for the electrical machines, such as, but not limited to, stator and rotor construction, and empirical relations from manufacturers of rolling element bearings, shaft seals, dielectric coatings, insulators, lubricants, electrical steels, and carbon grounding or commutator brushes. In certain embodiments, for example, the physics-based models can include common constitute relations and equations for reaction moments, stresses, and forces exerted to individual electrical machine components.

[0005] In certain embodiments, various types of embedded sensors can be utilized to facilitate the estimations of remaining useful lifetime of an electrical machine according to the system and method. For example, such sensors can include, but are not limited to, air gap flux sensing coils, vibration sensors, temperature sensors, and / or other sensors that can be utilized to collect the operational motor parameters to be utilized as inputs to the physics-based models. Exemplary operating parameters can include, but are not limited to, motor currents, voltages, vibration information, shaft speed, winding temperature, output torque, among other parameters. In certain embodiments, the system and methods can determine a remaining useful lifetime estimate based on the sensor data, which can include instantaneous, historic, and / or extrapolated electrical machine (e.g., motor) loads. In certain embodiments, the system and methods can also be utilized to generate a dynamic database that stores results from warranty claims, repair logs, teardowns, and Failure Modes and Effect Analyses (FMEA) of failed or warrantied electrical machines. In certain embodiments, the information stored in the database can be utilized to train artificial intelligence models (e.g., machine learning models) to enhance the estimation capability of the system and methods. In certain embodiments, for example, real-time sensor data associated with electrical machine components can be uploaded into the database use cloud-computing capabilities, and such data can improve the predictive capability of the artificial intelligence models on a more rapid basis.

[0006] In certain embodiments, a system for estimating the remaining useful lifetime of electrical machines is provided. In certain embodiments, the system can include an electrical machine, one or more sensors configured to capture sensor data associated with one or more components of the electrical machine, and one or more processors configured to perform variousoperations. Tn certain embodiments, the one or more processors can be configured to receive, such as during an evaluation interval, one or more signals from the one or more sensors. In certain embodiments, the one or more signals can include the sensor data. In certain embodiments, the one or more processors can be configured to determine, based on the sensor data and by utilizing a physics-based model, one or more input parameters for use in calculating a remaining useful lifetime for the one or more components of the electrical machine. In certain embodiments, the one or more processors can also be configured to calculate, by utilizing the physics-based model, the remaining useful lifetime for the one or more components of the electrical machine. In certain embodiments, the remaining useful lifetime for the at least one component can be calculated based on the one or more input parameters and one or more electrical machine specifications specified for the electrical machine. In certain embodiments, the one or more processors can be configured to output the remaining useful lifetime via an interface. In certain embodiments, in addition to providing the capability of determining remaining useful lifetime using physics-based models, the system and methods can also predict the remaining useful lifetime for the one or more components by utilizing an artificial intelligence model trained based on physics-based models, measurements from sensors embedded in the electrical machines, and historic databases of failures from identical or similar electrical machines.

[0007] In certain embodiments, another system for estimating the remaining useful lifetime of electrical machines is provided. In certain embodiments, the system can include an electrical machine, one or more sensors configured to capture sensor data associated with one or more components of the electrical machine, and one or more processors that are configured to perform a variety of operations. In certain embodiments, for example, the one or more processors can be configured to receive, during an evaluation interval, one or more signals from the one or more sensors. In certain embodiments, the one or more signals can include the sensor data. In certain embodiments, the one or more processors can be configured to determine, by utilizing an artificial intelligence model and based on one or more manufacturing parameters associated with the electrical machine, a comparable electrical machine that is comparable to the electrical machine. In certain embodiments, historical data associated with the comparable electrical machine can be utilized to train the artificial intelligence model to predict a remaining useful lifetime for the one or more components of the electrical machine. In certain embodiments, the one or more processors can be configured to predict, by utilizing the artificial intelligence model, the remaining usefullifetime for the one or more components of the electrical machine. In certain embodiments, the remaining useful lifetime for the one or more components can be calculated based on the sensor data being compared with the historical data associated with the comparable electrical machine. In certain embodiments, the one or more processors can be configured to output the remaining useful lifetime predicted by the artificial intelligence model via an interface.

[0008] In certain embodiments, a method for estimating the remaining useful lifetime of electrical machines is provided. In certain embodiments, the method can be performed by utilizing a memory that stores instructions and a processor that executes the instructions to perform the various operations of the method. In certain embodiments, various components, devices, and / or parts of an electrical machine can perform the method. In certain embodiments, the method can include receiving, during an evaluation interval, one or more signals from one or more sensors. In certain embodiments, the one or more signal can include sensor data associated with an electrical machine. In certain embodiments, the method can include predicting, by utilizing an artificial intelligence model, a remaining useful lifetime for the one or more components of the electrical machine. In certain embodiments, the remaining useful lifetime for the one or more components can be calculated based on the sensor data. In certain embodiments, the method can include outputting the remaining useful lifetime predicted by the artificial intelligence model via an interface. In certain embodiments, the method can include providing a reward to the artificial intelligence model in accordance with an accuracy of the remaining useful lifetime prediction to modify a prediction capability of the artificial intelligence model. In certain embodiments, the method can include predicting, for a next evaluation interval, a next remaining useful lifetime for the at least one component of the electrical machine by utilizing the prediction capability.BRIEF DESCRIPTION OF THE DRAWINGS

[0009] The embodiments of the present disclosure are illustrated by way of example and not limited by the figures of the accompanying drawings in which like references indicate similar elements. Embodiments of the present disclosure will be described in even greater detail below based on the exemplary figures. The present disclosure is not limited to the exemplary embodiments. All features described and / or illustrated herein can be used alone or combined in different combinations in embodiments of the present disclosure. The features of variousembodiments of the present disclosure will become apparent by reading the following detailed description with reference to the attached drawings which illustrate the following:

[0010] FIG. 1 is an exemplary perspective view of an electrical machine, such as a motor, for which a remaining useful lifetime can be determined according to embodiments of the present disclosure.

[0011] FIG. 2 is a schematic diagram illustrating various components of an electrical machine for which a remaining useful lifetime can be determined according to embodiments of the present disclosure.

[0012] FIG. 3 is an exemplary process flow for determining remaining useful lifetime for rotor bar components of an electrical machine by utilizing a physics-based model according to embodiments of the present disclosure.

[0013] FIG. 4 is an exemplary process flow for determining remaining useful lifetime for bearing components of an electrical machine by utilizing a physics-based model according to embodiments of the present disclosure.

[0014] FIG. 5 is an exemplary process flow for determining remaining useful lifetime for insulation components of an electrical machine by utilizing a physics-based model according to embodiments of the present disclosure.

[0015] FIG. 6 illustrates exemplary histograms from which parameters for determining the remaining useful lifetime for an electrical machine can be extrapolated according to embodiments of the present disclosure.

[0016] FIG. 7 is an exemplary schematic illustrating remaining useful lifetime estimation by utilizing a machine learning model and industry standard information according to embodiments of the present disclosure.

[0017] FIG. 8 is an exemplary process flow for determining remaining useful lifetime for components of an electrical machine by a utilizing machine learning model and / or physics-based models according to embodiments of the present disclosure.

[0018] FIG. 9 is an exemplary process flow for determining remaining useful lifetime for components of an electrical machine by utilizing a machine learning model and / or physics-based models according to embodiments of the present disclosure.

[0019] FIG. 10 is an exemplary system illustrating system the can be utilized to facilitate estimation of remaining useful lifetime of an electrical machine according to embodiments of the present disclosure.

[0020] FIG. 11 illustrates an exemplary method for determining remaining useful lifetime for electrical machines according to embodiments of the present disclosure.

[0021] FIG. 12 illustrates a schematic diagram of a machine in the form of a computer system within which a set of instructions, when executed, can cause the machine to facilitate determination of remaining useful lifetime according to embodiments of the present disclosure.DETAILED DESCRIPTION

[0022] The present disclosure describes various embodiments of systems and methods for determining and estimating the remaining useful lifetime of electrical machines, such as by utilizing physics-based models, machine learning models, or a combination thereof. The system and accompanying methods, for example, can be utilized in estimating the remaining useful lifetime of various types of electrical machine, such as, but not limited to, motors, generators, factory machines, computing machines, consumer machines, other types of machines, or a combination thereof. In certain embodiments, the system and methods provide functionality to minimize the need for preventative maintenance activities for electrical machines and their corresponding components. In certain embodiments, the system and methods provide long-term electrical machine remaining useful lifetime estimates that are updated continuously, such as by utilizing any number of sensors embedded in the electrical machines. The system and methods are capable of predicting the remaining useful lifetime of industrial electrical machines based on physics-based models, measurements from sensors embedded in the electrical machines, and historic databases of failures from identical or similar electrical machines. In certain embodiments, the system and methods may not require or rely on the detection of impending failures of components of electrical machines from abnormalities identified in sensor signals obtained from such embedded sensors. The system and methods can utilize motor-design information, such as design information in the form of physics-based models, which can constitute equations related to the materials used for the electrical machines, such as, but not limited to, stator and rotor construction, and empirical relations from manufacturers of rolling element bearings, shaft seals,dielectric coatings, insulators, lubricants, electrical steels, and carbon grounding brushes. In certain embodiments, for example, the physics-based models can include common constitute relations and equations for reaction moments, stresses, and forces exerted to individual electrical machine components.

[0023] In certain embodiments, various types of embedded sensors can be utilized to facilitate the estimations of remaining useful lifetime of an electrical machine according to the system and method. For example, such sensors can include, but are not limited to, air gap flux sensing coils, partial discharge sensors, strain gauges, vibration sensors, temperature sensors, and / or other sensors that can be utilized to collect the operational motor parameters and environmental conditions to be utilized as inputs to the physics-based models. Exemplary operating parameters can include, but are not limited to, motor currents, voltages, vibrations, magnetic flux, shaft speed, winding temperature, output torque, axial and radial shaft loads, strains and stresses in the shaft and housing among other parameters. Exemplary environmental conditions can include, but are not limited to, ambient temperature, relative humidity, airborne particulate count, surrounding electromagnetic fields, elevation above sea level, seismic vibrations, air pressure, among other parameters. In certain embodiments, the system and methods can determine a remaining useful lifetime estimate based on the sensor data, which can include instantaneous, historic, and / or extrapolated electrical machine (e.g., motor) operating loads and environmental conditions. In certain embodiments, the system and methods can also be utilized to generate a dynamic database that stores results from warranty claims, repair logs, customer surveys, customer photos, teardowns, and Failure Modes and Effect Analyses (FMEA) of failed or warranted electrical machines. In certain embodiments, the information stored in the database can be utilized to train artificial intelligence models (e.g., machine learning models) to enhance the estimation capability of the system and methods. In certain embodiments, for example, real-time sensor data associated with electrical machine components can be uploaded into the database use cloud-computing capabilities, and such data can improve the predictive capability of the artificial intelligence models on a more rapid basis.

[0024] In certain embodiments, a system for estimating the remaining useful lifetime of electrical machines is provided. In certain embodiments, the system can include an electrical machine, one or more sensors configured to capture sensor data associated with one or more components and / or the operating environment of the electrical machine, and one or moreprocessors configured to perform various operations. In certain embodiments, the one or more processors can be configured to receive, such as during an evaluation interval, one or more signals from the one or more sensors. In certain embodiments, the one or more signals can include the sensor data. In certain embodiments, the one or more processors can be configured to determine, based on the sensor data and by utilizing a physics-based model, one or more input parameters for use in calculating a remaining useful lifetime for the one or more components of the electrical machine. In certain embodiments, the one or more processors can also be configured to calculate, by utilizing the physics-based model, the remaining useful lifetime for the one or more components of the electrical machine. In certain embodiments, the remaining useful lifetime for the at least one component can be calculated based on the one or more input parameters, one or more assumptions for one or more future operating loads and / or environmental conditions, and one or more electrical machine specifications specified for the electrical machine. In certain embodiments, for example, the one or more assumptions can be related to usage of the electrical machine, loads applied to the electrical machine, and the environmental conditions present and / or affecting the electrical machine. In certain embodiments, the one or more processors can be configured to output the remaining useful lifetime via an interface. In certain embodiments, in addition to providing the capability of determining remaining useful lifetime using physics-based models, the system and methods can also predict the remaining useful lifetime for the one or more components by utilizing an artificial intelligence model trained based on physics-based models, measurements from sensors embedded in the electrical machines, and historic databases of failures from identical or similar electrical machines.

[0025] In certain embodiments, the one or more assumptions for the one or more future operating loads and / or environmental conditions can be an extrapolation of measured load and environmental condition histograms. In certain embodiments, the one or more assumptions for the one or more future operating loads and / or environmental conditions can include an extrapolation of the measured load and environmental condition histograms. In certain embodiments, the one or more processors can be further configured to initiate an evaluation interval for evaluating the remaining useful lifetime of the electrical machine, the at least one component of the electrical machine, or a combination thereof. In certain embodiments, the one or more processors can be further configured to assume electrical machine load histograms and environmental condition histograms for one or more periods of time during which the evaluationinterval was not running. In certain embodiments, the one or more processors can be further configured to calculate an updated remaining useful lifetime for the one or more components based on calculating incremental damage to the one or more components based on the one or more input parameters and one or more prior damage values calculated for the one or more components during a prior evaluation interval. In certain embodiments, the one or more processors can be further configured to determine a shortest remaining useful lifetime for a component of the one or more components by comparing each remaining useful lifetime for each component of the one or more components. In one or more embodiments, the one or more processors can be further configured to output the shortest remaining useful lifetime as the remaining useful lifetime of the electrical machine via the interface, and provide a signal to indicate whether a component of the one or more components having the shortest remaining lifetime is to be repaired or replaced.

[0026] In certain embodiments, the sensor data collected by the one or more sensors can include operating parameters and / or environmental conditions that include a current draw associated with the one or more components, a phase voltage and imbalance associated with the one or more components, a temperature associated with the one or more components, an acoustic pattern associated with the one or more components, a temperature pattern associated with the one or more components, an air gap flux associated with the electrical machine, stray flux associated with the electrical machine, a housing vibration associated with the electrical machine, strain and stress on a body of the electrical machine, ambient air temperature, relative humidity, airborne particulate count and size distribution, electromagnetic fields, elevation, air pressure, seismic vibrations, or a combination thereof. In certain embodiments, greasing information can also be measured and added to the machine learning algorithm as additional data. For example, if the electrical machine is greased, the model can adjust the lifetime estimate accordingly. In certain embodiments, the one or more processors can be further configured to initiate a next evaluation interval, receive one or more additional signals from the one or more sensors that can include additional sensor data, determine, based on the additional sensor data and by utilizing the physicsbased model, one or more new input parameters for use in calculating a current remaining useful lifetime for the one or more components of the electrical machine, and calculate the current remaining useful lifetime for the one or more components of the electrical machine based on the one or more new input parameters, the remaining useful lifetime from the evaluation interval, and the one or more electrical machine specifications.

[0027] In certain embodiments, the one or more processors can be configured to train an artificial intelligence model to generate a remaining useful lifetime prediction for the electrical machine, the one or more components, or a combination thereof. In certain embodiments, the artificial intelligence model can be trained based on the sensor data, the one or more input parameters, the one or more electrical machine specification, one or more electrical machine specifications for another electrical machine having a correlation to the electrical machine, or a combination thereof. In certain embodiments, the one or more processors can be further configured to adjust the remaining useful lifetime prediction generated by the artificial intelligence model based on reward function inputs and the remaining useful lifetime determined using the physics-based model. For example, in certain embodiments, the one or more processors can be configured to adjust the remaining useful lifetime prediction generated by the artificial intelligence model based on unsupervised learning (or other type learning) reward function inputs and the remaining useful lifetime determined using the physics-based model.

[0028] In certain embodiments, another system for estimating the remaining useful lifetime of electrical machines is provided. In certain embodiments, the system can include an electrical machine, one or more sensors configured to capture sensor data associated with one or more components of the electrical machine and its environmental conditions, and one or more processors that are configured to perform a variety of operations. In certain embodiments, for example, the one or more processors can be configured to receive, during an evaluation interval, one or more signals from the one or more sensors. In certain embodiments, the one or more signals can include the sensor data. In certain embodiments, the one or more processors can be configured to determine, by utilizing an artificial intelligence model and based on one or more manufacturing parameters associated with the electrical machine, a comparable electrical machine that is comparable to the electrical machine. In certain embodiments, historical data associated with the comparable electrical machine can be utilized to train the artificial intelligence model to predict a remaining useful lifetime for the one or more components of the electrical machine. In certain embodiments, the one or more processors can be configured to predict, by utilizing the artificial intelligence model, the remaining useful lifetime for the one or more components of the electrical machine. In certain embodiments, the remaining useful lifetime for the one or more components can be calculated based on the sensor data being compared with the historical data associated with the comparable electrical machine. In certain embodiments, the one or more processors can beconfigured to output the remaining useful lifetime predicted by the artificial intelligence model via an interface.

[0029] In certain embodiments, the one or more processors can be further configured to cumulate prior damage to the one or more components of the electrical machine determined from a prior evaluation interval with current damage determined during the evaluation interval when determining the remaining useful lifetime for the one or more components of the electrical machine. In certain embodiments, the one or more processors can be further configured to select a strategy from a plurality of strategies for determining the remaining useful lifetime of the one or more components of the electrical machine. In certain embodiments, the one or more processors can be further configured to weight a first portion of the sensor data over a second portion of the sensor data when determining the remaining useful lifetime of the one or more components. In certain embodiments, the one or more processors can be further configured to determine an accuracy of the remaining useful lifetime predicted by the artificial intelligence model based on a comparison with a result of an inspection of the one or more components of the electrical machine. In certain embodiments, the one or more processors can be further configured to provide a reward to the artificial intelligence model in accordance with the accuracy.

[0030] In certain embodiments, the one or more processors can be further configured to utilizing one or more industry standards associated with the electrical machine, the one or more components, or a combination thereof, to determine the remaining useful lifetime. In certain embodiments, the one or more processors can be further configured to optimize an electrical machine maintenance schedule for the electrical machine, the one or more components, or a combination thereof, based on the remaining useful lifetime. In certain embodiments, the one or more processors can be further configured to provide an electrical machine health indication associated with the electrical machine, the one or more components, or a combination thereof, via the interface and based on the remaining useful lifetime.

[0031] In certain embodiments, a method for estimating the remaining useful lifetime of electrical machines is provided. In certain embodiments, the method can be performed by utilizing a memory that stores instructions and a processor that executes the instructions to perform the various operations of the method. In certain embodiments, various components, devices, and / or parts of an electrical machine can perform the methods and operative functionality described in the present disclosure. In certain embodiments, the methods and operative functionality describedin the present disclosure can be utilized to determine the remaining useful lifetime of such various components, devices, and / or parts of an electrical machine. In certain embodiments, the method can include receiving, during an evaluation interval, one or more signals from one or more sensors. In certain embodiments, the one or more signal can include sensor data associated with an electrical machine. In certain embodiments, the method can include predicting, by utilizing an artificial intelligence model, a remaining useful lifetime for the one or more components of the electrical machine. In certain embodiments, the remaining useful lifetime for the one or more components can be calculated based on the sensor data. In certain embodiments, the method can include outputting the remaining useful lifetime predicted by the artificial intelligence model via an interface. In certain embodiments, the method can include providing a reward to the artificial intelligence model in accordance with an accuracy of the remaining useful lifetime prediction to modify a prediction capability of the artificial intelligence model. In certain embodiments, the method can include predicting, for a next evaluation interval, a next remaining useful lifetime for the at least one component of the electrical machine by utilizing the prediction capability. In certain embodiments, the method can further include scheduling a repair or replacement of the one or more components based on the remaining useful lifetime indicating onset of a failure of the one or more components.

[0032] In certain embodiments, a method for estimating the remaining useful lifetime of electrical machines is provided. In certain embodiments, the method can be performed by estimating the future usage pattern (i.e., future operating parameters) and future environmental conditions based on prior usage history of the electrical machine. In certain embodiments, operating parameters and environmental conditions that were present during a period of time may be represented using histograms. For calculating the remaining useful life of the electrical machine, the future operating parameters and environmental conditions may be extrapolated from the histograms and / or present sensor data.

[0033] Based on at least the present disclosure, the system and accompanying methods can provide functionality that reduces or eliminates the labor involved in electrical machine evaluation by measuring electrical machine performance in the field on a regular basis. Additionally, the system and accompanying methods provide the ability to estimate the remaining hours of electrical machine operating life, such as based on current and past electrical machine usage profiles. The system and accompanying methods can also provide the ability to ascertain which component ofan electrical machine will be the first to fail, such as depending on motor usage and failures that may have been encountered in this past. The machine learning models of the system and methods also provide a new techniques in which to calculate electrical machine design lifetime and can be utilized to optimize future motor designs. The system and methods can utilize machine learning training databanks to provide real-life examples of electrical-machine usage and weaknesses, which can facilitate the development of new and improved types of electrical machines. The machine learning databanks provided by the present disclosure can also indicate how an electrical machine is utilized at a customer site and can provide custom-tailored suggestions for improved electrical machine maintenance to prolong to the useful lifetime of the electrical machine and custom-tailored future product suggestions to satisfy a current usage profile for the electrical machine. In certain scenarios, the machine learning databank can include information generated by the system and methods that can enable the suggestion of possible changes to industrial standards to manufacture improved versions of electrical machines. Still further, the system and methods can factor in usage parameters currently assumed to be not feasible in accurately linking to electrical machine lifetime, such as, but not limited to, grease quality (e.g., acidity, moisture content, trapped particulate count, material, and size distribution, etc.), temperature, environmental conditions, and manufacturing and assembly precision.

[0034] Referring now to Figure 1, a perspective view of an exemplary electric machine 100 (e g., an electric motor) for which a remaining useful lifetime can be determined and / or predicted according to embodiments of the present disclosure. For the purposes of the present disclosure, the terms electric machine and electrical machine can be utilized interchangeably. In certain embodiments, the electric machine 100 can be a machine utilized in a manufacturing process, a packaging process, a factory-based process, a distribution process, any other type of process, or a combination thereof. Additionally, in certain embodiments, the electric machine 100 can be part of a larger or more complex machine and can be configured to facilitate the operative functionality of the larger or more complex machine, such as by generating rotational motion to drive machinery and components. In certain embodiments, the perspective view of the electric machine 100 illustrates a rotational axis 110 and an electric machine configuration supporting alternating current synchronous or asynchronous (in case of induction motors) operation. In certain embodiments, the electric machine 100 can include a plurality of components including, but not limited to, a rotating electric machine shaft 102, an electric machine enclosure 104, mounting feet106, eyehooks 108, a rotational axis 110, a terminal box 112, power leads 1 14, a neutral or ground lead 115, a mitigation component 130, other components, or a combination thereof. In Figure 1, the exemplary electric machine 100 can be a rotating electrical machine that is configured to convert electrical energy to mechanical energy. Torque may be transmitted via a rotating electric machine shaft 102 to connected loads. In certain embodiments, the electric machine shaft 102 can protrude from the forward end of the electric machine enclosure 104 that encloses and houses the internal operating components of the electric machine 100. In certain embodiments, the electric machine enclosure 104 may be made from any suitable structural material such as, but not limited to, cast iron, steel, aluminum, plastics or other suitable materials, and the electric machine enclosure 104 may be configured according to various frame sizes that determine the location and arrangement of mounting features, such as mounting feet 106, eyehooks 108, or other mounting features. In certain embodiments, the electric machine enclosure 104 may be designated in accordance with any of several enclosure types, such as open drip proof (ODP) or totally enclosed fan cooled (TEFC) that determine how the electric machine 100 is constructed to interact with the operating environment to provide for cooling and protect the internal components against contaminants, such as moisture and dust. In certain embodiments, the electric machine shaft 102 can be supported to rotate with respect to and defines a rotational axis 110 of the electric machine 100.

[0035] In certain embodiments, the electric machine 100 can receive power from an external power source, such as external power source 103. In order to receive electric current from an external power source, the electric machine 100 may include a conduit box or terminal box 112 located at an appropriate location on the electric machine enclosure 104 from which a plurality of power leads 114, such as insulated conductive wires, can extend. The power leads 114 may be electrically connected to and complete a circuit with the external power source that provides electricity having appropriate electrical characteristics and properties for operation of the electric machine 100. In certain embodiments, the electric machine 100 can be configured to operate on poly-phase, alternating current power source. In a poly-phase power system, the plurality of power leads 114 may each supply alternating current and voltage of the same frequency to the electric machine 100, however, the alternating current conducted in each power lead may be out of phase with that in the other power leads. Accordingly, the cyclic oscillations between 0°-360° of alternating current in each power lead 114 may be delayed or advanced with respect to that in theother power leads. As an example, a three-phase electric machine (e g., motor) 100 may include three power leads 114 that conduct alternating currents that are 120° out of phase with each other and a fourth neutral or ground lead 115 that may be connected to an electrical ground, for example, the electric machine frame, and that serves as a reference. This type of connection is referred to as star (Y) connection. An alternative connection configuration that may also be used is the delta connection which does not have the ground connection. In certain embodiments, a three-phase electric machine 100 may include additional power leads, such as power leads for connecting to and / or powering one or more external accessories (e.g., user-accessible power ports). For example, the electric machine 100 may include primary and auxiliary coils (e.g., windings). The primary coils (e.g., stator windings) may be powered via the power leads 114. The primary coils may be coupled to the auxiliary coils (e.g., windings) such that when powered by the power leads 114, the primary coils induce voltages and currents in the auxiliary coils. Based on the foregoing, the auxiliary coils may be electrically connected to accessory devices such as the user-accessible power ports (e.g., additional power leads within the terminal box 112 that are configured to power user devices) and / or sensor devices.

[0036] In order to actuate rotation of the electric machine shaft 102, the electric machine 100 may include a rotor and a stator. In certain embodiments, the rotor can be generally cylindrical in shape and can be assembled about the extension of the shaft 102 that is located within the enclosure 104. The rotor can be configured to electromagnetically interact with an annular stator in which the rotor is disposed. The cylindrical rotor and the annular stator can be concentrically aligned with the rotational axis 110 of the electric machine 100 defined by the electric machine shaft 102. In certain embodiments, the annular stator may be fixedly disposed concentrically around the rotor and can be spaced apart and separated therefrom by an annular air gap. In certain embodiments, the stator can include a stator core that may be made from a magnetically permeable material, such as iron or steel. The stator core may be made from a plurality of annularly shaped core laminations that are axially arranged as a stack and extend coaxially along the rotational axis 110. The stator core may be fixed to and enclosed in the electric machine enclosure 104, which may include fins, water cooling jackets, and other components to facilitate cooling.

[0037] In order to accommodate the conductive coils (e.g., the primary coils or stator windings) that conduct current to generate the magnetic field, the stator core may include a plurality of stator teeth that are radially arranged in the circumferential direction around therotational axis 110 and circumferentially separated from each other by stator slots radially disposed into the inner cylindrical surface of the stator core. Hence, between each two adjacent stator teeth, there can be disposed a stator slot so that the teeth and slots circumferentially alternate about the inner cylindrical surface of the stator core. The alternating stator teeth and stator slots may axially extend along the axial length of the stator core with respect to the rotational axis 110. The conductive coils (e.g., primary windings or stator coils or windings) may be elongated wires of copper or other conductive material that are wound or looped about the stator teeth and accommodated in the stator slots. The conductive windings may be wound around a stator tooth or a plurality of stator teeth a number of successive times, each time being referred to as a “turn.” The total number of turns of the conducting winding about the same stator tooth or stator teeth forms a “coil.” For example, in certain embodiments, a coil may be formed from multiple turns of the conductive coils. In certain embodiments, any type of coil formed in any type of manner can also be utilized as well. The conductive wires of the conductive coils may then be directed around additional stator teeth that are spaced from the initial coil in a continuous manner until the conductive coils circumscribe the inner circumference of the stator core. The path and geometry of the conductive coils around the stator core may be referred to as the “winding (or coil) pattern,” and the winding pattern can take various arrangements and may determine the electrical characteristics and operating principles of the electric machine 100.

[0038] For example, the winding pattern may assign or allocate the coils by phases and by pole-phase groups. The phases may include the coils that are electrically connected in series to the same electrical phase of the poly-phase power source. For example, in a three-phase power system, for the electric machine 100 to receive three-phase power, a first phase conductor may be associated with “A” phase current, a second phase conductor may be associated with “B” phase current, and a third phase conductor may be associated with “C” phase current. The phase conductors may be electrically connected with the power leads 114. The series of coils that are electrically connected to a respective one of the first, second, and third phase conductors may be referred to as a phase. The number of coils included with each phase can be dependent upon the number of stator teeth and stator slots.

[0039] Operatively, when the first, second, and third phase conductors are energized from a three-phase power system with alternating electric current that is 120° degrees out of phase by the respective conductor, the current flowing in the plurality of phases generates a magnetic field thatcircumferentially rotates around the rotational axis 110. As the polarity of one phase connected to the first conductor begins to change, e.g., from north to south, due to the periodic reversal of the direction of the alternating current associated with phase “A”, the polarity of the adjacent phase may become stronger because it is connected to the second or third phase conductor carrying current 120° degrees out of phase with the first conductor. The combined changing polarity from all phases can produce a circumferentially rotating magnetic field around the rotor. For an induction machine, this rotating magnetic field crosses through the air gap and induces voltage and consequently current in the rotor conductors. The rotor field due to rotor conductor current lags behind the stator field, and hence the rotor undergoes a torque that causes it to rotate in the direction of the rotating magnetic field. In the case of permanent magnet rotors, the rotor fields due to magnet poles experiences torque due to the rotating stator field and may rotate in synchronous speed with the stator field. In the case of a synchronous reluctance motor, the rotor may be constructed with variable reluctances having same number of reluctance variations as of stator number of poles. The rotating stator field from the stator and variable reluctance from the rotor creates rotational torque for the synchronous reluctance rotor to rotate at synchronous speed. The synchronous speed in turn depends on the fundamental frequency of the supplied voltage to the motor phases. The rotor is thus caused to rotate with respect to the rotational axis 110. However, while aspects of the disclosure may be described with respect to poly-phase alternating current power systems, aspects of the disclosure will also be applicable to other types of power systems and electric machine configurations.

[0040] In embodiments where the electric machine is an electric generator, an externally excited electromagnet residing on a rotor is rotated by mechanical power source or prime mover to produce a rotating electric field. In a three-phase generator example, three sets of stationary coils are placed 120 degrees apart spatially. The rotating field crossing the air gap induces voltage in the stator coils and generates three phase electrical power. The rotating field also cross the harvesting coil, hence generates voltage and electric power for the sensing system. The same principle also works for more than three phase generators system as well.

[0041] Referring now also to Figure 2, a schematic diagram illustrating various components of an electric machine 100 including a sensor device 220 for measuring sensor data associated with the electric machine 220 according to embodiments of the present disclosure is provided. In certain embodiments, the electric machine 100 and the sensor device 220 can be located in anenvironment, such as a factory. In certain embodiments, the sensor device 220 can be incorporated into, onto, or otherwise in proximity to the electric machine 100. In certain embodiments, the sensor device 220 can be housed in the terminal box 112 of the electric machine 100. In certain embodiments, the electric machine 100 can include any number of components 202, which may include, but are not limited to, bearings, stators, rotors, sensors, stator teeth, commutators, field windings, brushes, armatures, shafts, cooling channels, fans, armature cores, coils, windings, insulation, housings, any other components, or a combination thereof. In certain embodiments, the electric machine 100 can include a plurality of primary coils 204 (e.g., primary windings), a plurality of auxiliary coils 206 (e.g., auxiliary windings), and internal electronic circuitry. The electric machine 100 may include an internal compartment that houses the internal electronic circuitry, such as sensors, a rectifier, super capacitors, and / or other components. The internal electronic circuitry can include sensors and a rectifier.

[0042] As described herein, the electric machine 100 may include a rotor and a stator with any number of primary coils 204. In certain embodiments, each of the primary coils 204 may be connected to one of the power leads 114, which may be configured to provide alternating-current (AC) power to the electric machine 100. In certain embodiments, each of the primary coils 204 may be associated with a phase of the three-phases for the electric machine 100. The primary coils 204 may be configured to generate a magnetic field based on power received from an external power source via the power leads 114. The rotor of the electric machine 100 may include magnets that respond to the generated magnetic field of the primary coils 204, thereby causing the electric machine shaft 102, which is attached to the rotor, to rotate. In certain embodiments, the electric machine shaft 102 may be operatively coupled to a load (e.g., fan, pump, etc.), and electric machine 100 may provide power to the load (e.g., fan, pump, etc.) based on the rotation of the electric machine shaft 102. Furthermore, the electric machine 100 may include a plurality of auxiliary coils 206. The current being provided to the primary coils 204 may be transferred as electrical energy to the auxiliary coils 206. The primary and auxiliary coils 204 and 206 may be linked by electromagnetic flux and configured to transfer energy between the primary coils 204 to the auxiliary coils 206. The primary coils 204 may produce a magnetic flux, which may cause an induced voltage and / or current to be generated in the auxiliary coils 206. The amount of induced voltage and / or current in the auxiliary coils 206 may be based on the number of turns in the primarycoils 204 when compared to the number of turns in the secondary or auxiliary coils 206 and the projected areas of the coils that the magnetic field passes through.

[0043] The sensors (e.g., components 202) utilized with the electric machine 100 may be any type of sensors that are configured to measure sensor information and / or provide the sensor information to the sensor device 220, the system 1100, and / or other devices and systems. The sensors may be electrically connected to the auxiliary coils 206 such that the auxiliary coils 206 provide energy (e.g., the induced current) to the sensors. The sensors may be and / or may include any type of sensors that are configured to measure and / or acquire information associated with the electric machine 100 and / or an environment in which the electric machine 100 resides. The sensors may be configured to measure the operational variables of the electric machine 100, such as the air gap flux, frequency, speed, input current, stator temperature, faulty conditions, rotor eccentricity, output torque, stresses and strains in the shaft and housing, and other variables and / or conditions. In certain embodiments, the sensors may be configured to measure bearing performance (e.g., vibrations, radial / axial forces, speeds ,etc.), trends in electric machine characteristics, and / or rotor eccentricity, and provide these measurements to the sensor device 220, the system 1100, or a combination thereof.

[0044] The sensor device 220 or system may be utilized to monitor the health condition of the electric machine 100 and determine the remaining useful lifetime of components of the electric machine 100. For example, the health condition may relate to the components 202 of the electric machine 100, such as, but not limited to, whether a component is damaged, at risk for damage, failing, overheating, overcooling, operating outside of required or desired specifications, experiencing excess pressure, experiencing excess humidity or condensation, infiltrated with dirt or dust, has any other condition, or a combination thereof. In certain embodiments, the sensor device 220 can include any number of sensors 222, microprocessors 224, memory devices 226, interfaces 228, indicators 230, input devices 232, actuators 234, communication devices 236, any other components or a combination thereof. In certain embodiments, the sensors 222 can include, but are not limited to, cameras, accelerometers (e.g., 1-axis or 3-axis MEMS or piezo accelerator for sensing the vibrations of the electric machine bearings and / or other components), motion sensors, acoustic / audio sensors (e.g., directional ultrasonic microphones for sensing the airborne noise that is caused by structure borne vibrations at the bearings), air pressure sensors, temperature sensors (e.g., thermocouples, RTDs a.k.a. a resistance temperature detector), light sensors,gyroscopes, chemical sensors, any type of sensors, or a combination thereof. In certain embodiments, the sensor device 220 can be affixed directly to the electric machine frame within or outside of the terminal box 112. The foregoing can improve the signal-to-noise ratio of the bearing vibration measurements by optimizing the coupling between the bearing and the sensor. In certain embodiments, the sensor device 220 may be located some distance away from the electrical machine frame to measure environmental conditions such as ambient air temperature, relative humidity, seismic vibrations, ambient air pressures, other environmental conditions, or a combination thereof. In certain embodiments, the sensor device 220 can include suitable visual indicators and signal paths may be provided for electric machine health visualization and notification, and optional connectivity to a cloud service or smartphone application, respectively.

[0045] The sensors 222 of the sensor device 220 can measure sensor data, such as, but not limited to, by measuring currents, voltages, magnetic fields, temperatures and / or signals occurring in the electric machine 100. The sensor data can be compared to baseline data for a particular component and may be utilized to trigger output of alerts and / or initiating of actions to counter the conditions detected. In certain embodiments, the microprocessor 224 can be a combination of software, hardware, or a combination thereof, and may be utilized to execute instructions from the memory device 226. The microprocessor 224 can also be configured to analyze sensor data to determine whether a condition exists and can store sensor data and / or determinations in the memory device 226. In certain embodiments, the microprocessor 224 can provide sensor data and determinations and / or analyses to any component of the electric machine 100, any component of the sensor device 220, and / or the system 1100.

[0046] In certain embodiments, the memory device 226 can be configured to store sensor data, remaining useful lifetime predictions and / or determinations, determinations relating to whether a condition exists, pervious load history, alerts, sensor calibration data (e.g., sensor sensitivities and gains), baseline sensor data (e.g., a baseline for illustrating the safe operating range or acceptable operating range for a component), any other information, or a combination thereof. In certain embodiments, the interface 228 can be a user interface, a touchscreen, a screen with input devices (e.g., buttons), a display, or other interface to enable a user to perceive alerts, sensor data, actions to counteract conditions detected, conditions detected, or a combination thereof. In certain embodiments, the indicator 230 (or output devices) can be any type of indicator, such as, but not limited to, lighting devices (e.g., LEDs), audio devices (e.g., speaker or microphone), hapticdevices, vibration devices, information screens, any other types of indicator devices, or a combination thereof. In certain embodiments, the input device 232 can be a button or other input device that may be configured to activate or deactivate the sensor device 220, reset baseline sensor data, reset the sensor device 200, disconnect the electric machine, and / or perform any actions with respect to the sensor device 220 and / or electric machine 100. In certain embodiments, the actuator 234 can be any type of actuator that may be configured to provide or produce motion for moving or controlling a device or part of a device to cause an action. For example, the actuator 234 can be a mechanism by which a component of the electric machine 100 is moved or otherwise put into motion.

[0047] In certain embodiments, the sensor device 220 can also include any number of communication devices 236. Such communications devices 236 can include, but are not limited to, antennas, communication chips, cellular devices, wireless devices (including wireless interfaces), gateways, communication interfaces, any type of communication device or a combination thereof. In certain embodiments, the communication devices can be loT, Bluetooth, Zigbee, Z-wave, Wireless HART, cellular, LoRA, NB-IoT, etc., or any combination thereof. In certain embodiments, the communication devices 236 can receive or transmit sensor data, health condition determinations, firmware updates, any type of data, or a combination thereof, to the electric machine 100, the sensor device 220, the system 1100, other devices and / or systems, or a combination thereof.

[0048] Referring now also to Figures 3, 4, and 5, and more specifically to Figure 3, an exemplary process flow 300 for determining remaining useful lifetime for components, such as rotor bar components, of an electrical machine 100 by utilizing a physics-based model, specifically the model for the rotor bars of the electrical machine 100, according to embodiments of the present disclosure is shown. In certain embodiments, for example, the system 1100 can include an electric machine 100 for which a remaining useful lifetime is to be determined, one or more sensors 220 (e.g., embedded in the electric machine 100 and / or in range or proximity of the electric machine 100), computing resources (e.g., microprocessors 224, memory devices 226, system 1100 (e.g., cloud-server, edge computer, embedded processor, first user device 1102, etc.), a human machine interface (e.g., touchscreen display, handheld device (e.g., first user device 1102), etc.), a physicsbased model(s), any other components and / or systems, or a combination thereof. The process flow 300 can, in certain embodiments, operate and / or execute continuously on a computing resourceand can include exchanging data with the human machine interface 324. In certain embodiments, the process flow 300 can begin at block 302, which can include starting a new evaluation interval to evaluate the remaining useful lifetime of the electric machine 100 as whole and / or the remaining useful lifetime of specific components of the electric machine 100.

[0049] At block 304, the process flow 300 can include obtaining signals from the various sensors being utilized to monitor the electric machine and the environmental conditions of the electrical machine 100, which, for example, can be a squirrel cage induction motor In certain embodiments, the sensors (e.g., sensors 220) can be embedded in the electric machine 100, in proximity to the electric machine 100, or a combination thereof. The sensors can measure sensor data, which can include parameters of interest that can be utilized to determine the remaining useful lifetime of components and / or the electric machine 100 itself. In certain embodiments, for example, the parameters can be operational parameters (e.g., relating to the functional capability and / or outputs of a component), parameters associated with the characteristics of components (e.g., materials, shape, size, etc.), environmental parameters associated with the environmental surroundings of the machine, any other types of parameters, or a combination thereof. In certain embodiments, exemplary parameters can include, but are not limited to, current draw (e.g., by a component and / or the electric machine 100), phase voltages and imbalance, stator end winding temperatures, bearing temperatures, air gap flux, housing vibrations, housing strains and stresses, rotor shaft strains and stresses, air pressure measurements, humidity readings, motion readings, any other parameters, or a combination thereof. In certain embodiments, sensor data measurements can be converted to derived values. For example, generated motor torque can be extracted from current or voltage measurements and / or from an embedded air gap flux sensing coil. Stresses can be calculated from the strains, and strains can be calculated from a relative resistance change of a strain gauges. At block 306, various specifications of the electric machine 100 can be fed into a physic-based model of the electric machine 100. Such specifications can include any information associated with the electric machine 100, such as, but not limited to, information on the components of the electric machine 100, information identifying materials of the components, information identifying functionality and stress values for the components, information identifying any characteristics of the electric machine 100, or a combination thereof.

[0050] At block 308, the outputs from block 304 can be supplied to the portion of the physicsbased model of the electric machine that pertains to the rotor bars. In certain embodiments, aphysics-based model of an electrical machine 100 can constitute equations related to the materials and geometry used for the various components of the electric machine 100 (e.g., stator and rotor materials and dimensions), empirical relations from manufacturers of components of the electric machines 100 (e.g., dynamic load ratings of the bearings), component specifications related to expected performance and / or characteristics of properly-functioning components, or a combination thereof. The physics-based models of the electrical machine may use moment and force balancing equations and a simplified geometry. In certain embodiments, the physics-based model may be based on functions (e.g., polynomials, exponential, logarithmic functions) that were fitted to a structural Finite Element Analysis (FEA) for the structural machine components that was performed during the design phase of the electrical machine. In certain embodiments, at block 308, the physics-based model of the electric machine 100 can be utilized to calculate the input parameters needed for the remaining useful lifetime calculations of the various components of the electric machine. In certain scenarios, not all parameters utilized for the remaining useful lifetime calculations may be available from the sensor measurements. For example, remaining parameters, such as maximum overhung and axial shaft loads and motor speed-torque curves, can also be provided by the user through the use of QR codes, RFID tags, and the human machine interface 324 during electric machine commissioning.

[0051] In certain embodiments, blocks 310, 312, 314, 316 can be utilized to calculate, for rotor bar components of the electric machine 100, the mechanical stresses and incremental damage Rt = that is based on the fatigue limit of the rotor bar material and the number of cycles the rotor has been exposed to during the evaluation interval. In further detail, at block 310, the process flow 300 can include calculating the characteristic stresses of components (e.g., rotor bar components) using the equations pertaining to the rotor bars of the physics model. At block 312, the process flow 300 can include determining the fatigue limit Ntfor the characteristic stress from the S -N curve for materials utilized for the components (e.g., aluminum or copper). At block 314, the process flow 300 can include calculating the number of load cycles nLin the evaluation interval. In certain embodiments, if the electrical machine was operated without the evaluation interval looping continuously, the load cycles and the characteristic stresses that occurred during that time can be considered as well in the calculation. At block 316, the process flow 300 calculating the incremental damage Rt= —. Block 318 can include summing up the incremental damagesevidenced by the sensor data, including pre-existing damage values from prior evaluation cycles and stored in memory (e.g., memory 226), such as according to a Palmgren-Miner damage accumulation model. The updated summed up total damage can be expressed as X Based on the foregoing calculations, the total damage of each component being analyzed can be computed. At block 320, the process flow 300 can include computing the remaining useful lifetime for each component under evaluation. The remaining useful life of the components (e.g., rotor bar components) can be calculated by assuming future operating conditions, for example future speeds and torques. In certain embodiments, the remaining useful lifetime can be based on an extrapolation of a histogram of electrical machine 100 loads (e.g., motor loads) containing present and prior load information. The assumed future operating conditions can also be the average of the prior operating conditions. In certain embodiments, the calculation can be expressed in any time-based format, such as hours, minutes, days, years, number of cycles, and / or any other type of format. Exemplary calculations are provided in the present disclosure. In certain embodiments, the calculation of the remaining useful lifetime for the components and / or the electrical machine 100 as a whole can factor in tracked electrical machine idle time and operational time. At block 322, the process flow 300 can include comparing the degradation of each individual component of the electric machine 100 being evaluated and can output the most critical value as the remaining useful lifetime of the electric machine 100 overall. For example, the remaining useful lifetime for the component with the shortest remaining useful lifetime can be utilized as the remaining useful lifetime of the entire electric machine 100. At block 324, the remaining useful lifetime for the electric machine 100 and / or each of the individual components can be provided to a human machine interface, such as a touchscreen display for visualization or perception. Based on the remaining useful lifetime, one or more actions can be initiated, such as an action to replace one or more components, an action to repair one or more components, and / or any other actions. If there is a threshold amount of remaining useful lifetime remaining, no action may be performed until the remaining useful lifetime is calculated to satisfy the threshold.

[0052] Referring now more specifically to Figure 4, an exemplary process flow 400 for determining remaining useful lifetime for components of an electrical machine 100 by utilizing a physics-based model (which can be part of Method A) according to embodiments of the present disclosure is shown. Figure 4, for example, can be utilized to determine the remaining useful lifetime of bearing components of an electric machine 100. In certain embodiments, for example,the system 1100 can include an electric machine 100 for which a remaining useful lifetime is to be determined, one or more sensors 220 (e.g., embedded in the electric machine 100 and / or in range or proximity of the electric machine 100), computing resources (e.g., microprocessors 224, memory devices 226, system 1100 (e.g., cloud-server, edge computer, embedded processor, first user device 1102, etc.), a human machine interface (e.g., touchscreen display, handheld device (e.g., first user device 1102), etc.), a physics-based model(s), any other components and / or systems, or a combination thereof. As with process flow 300, the process flow 400 can, in certain embodiments, operate and / or execute continuously on a computing resource and can include exchanging data with the human machine interface 424. In certain embodiments, the process flow 400 can begin at block 402, which can include starting a new evaluation interval to evaluate the remaining useful lifetime of the electric machine 100 as whole and / or the remaining useful lifetime of specific components of the electric machine 100, such as bearing components.

[0053] At block 404, the process flow 400 can include obtaining signals from the various sensors being utilized to monitor the electric machine 100, which, for example, can be a squirrel cage induction motor and / or stator winding isolation, as in process 300. In certain embodiments, the sensors (e.g., sensors 220) can be embedded in the electric machine 100, in proximity to the electric machine 100, or a combination thereof. At block 406, various specifications of the electric machine 100 can be fed into a physics-based model of the electric machine 100, including those relating to bearings and / or other components of interest. Such specifications can include any information associated with the electric machine 100, such as, but not limited to, information on the components of the electric machine 100, information identifying materials of the components, information pertaining to the dynamic load rating C, construction of the bearing and associated bearing form factor v (ball bearing vs. roller bearing, see example below), information identifying functionality and stress values for the components, information identifying any characteristics of the electric machine 100, or a combination thereof. At block 408, the outputs from block 404 can supplied to the physics-based model of the electric machine 100, which can be an electrical machine model pertaining to the bearings. In certain embodiments, the physics-based model of an electrical machine 100 can constitute equations related to the materials used for the various components of the electric machine 100 (e.g., stator and rotor materials), empirical relations from manufacturers of components of the electric machines 100, component specifications related to expected performance and / or characteristics of properly-functioning components, or acombination thereof. In certain embodiments, at block 408, the physics-based model of the electric machine 100 can be utilized to calculate the input parameters needed for calculating the remaining useful lifetime values of the various components of the electric machine. As indicated in the present disclosure, not all parameters utilized for the remaining useful lifetime calculations may be available from the sensor measurements. For example, remaining parameters, such as overhung and axial shaft loads and motor speed-torque curves, can also be provided by the user through the use of QR codes, RFID tags, and the human machine interface 424 during electric machine commissioning.

[0054] Blocks 410, 412, 414, 416 can be utilized to calculate the bearing forces and incremental bearing damagethat is based on the L10calculation methods provided by the bearing manufacturers and standards, such as ISO 281, which relates to bearings. Blocks 418 and 420 are utilized to sum up the incremental damages, including pre-existing damage values, and calculating the remaining useful lifetimes of the bearing components. In further detail, at block 410, the process flow 400 can include calculating axial and radial bearing forces at each bearing, such as by utilizing the sensor data obtained from the sensors of the electric machine and the information about additional overhung and axial shaft loads provided by the user. At block 412, the process flow can include calculating an equivalent dynamic bearing force, retrieval of the dynamic bearing load rating, and calculation of the bearing life expectancy Lw iin millions of revolutions according to ISO 281. At block 414, the process flow 400 can include calculating the number of revolutions n during the interval based on the measured motor speed co and length of the evaluation time T (see example). At block 416, the process flow 400 can include calculating the incremental bearing damage Si =10~6m / Lio.i. At block 418, the process flow 400 can include determining an updated total damage Si. At block 420, the process flow can include computing the remaining useful lives for each of the bearing components. At block 422, the process flow 400 can include determining which of the bearing components has the shortest remaining life, which can serve as the remaining useful life of the electric machine 100 as a whole. At block 424, the process flow 400 can include providing the remaining useful life determinations for perception via a user interface.

[0055] An exemplary calculation of remaining useful lifetime for bearing components can be as follows:Bearing Remaining Useful Life (RUL) estimationConstants:Bearing form factor v 3Dynamic load rating [kN] C 10Evaluation time [s] T 600Interval 1 2 3 4 5Another bearing calculation can be as follows: Lio = (C / P)v; P = XFr+YFa; where v=3 (form factor for ball bearings), C=10 kN (dynamic load rating, T= 600 s (interval time), X,Y (from bearing catalog / specifications). For a first interval 1, Pi=7 kN => Lio,i = ( 10 / 7)3= 2.91 [in 106revs]; wi = (1800 / 60)s-1= 30s’1=> Ni=WiT= 18* 103revs. Damage: E =E^ 'L' io.i = Ni / Lio,i = (18*103) / (2.91*106); Remaining useful lifetime after interval 1 : Ni / Lio.i + Nrui,i / Lio.i = 1, Nmi,i = (l- S))Lio,i. For a second interval 2, P2 = lOkN => Lio, 2= 1.0; W2 = (1200 / 60)s’1= 20s’1=> N2 = 12*103revs; Damage:Remaining useful lifetime after interval 2: E + Nrui,2 / Lio,2=l; Nrui,2 = (1-E ) Lio, 2.

[0056] Referring now more specifically to Figure 5, an exemplary process flow 500 for determining remaining useful lifetime for components of an electrical machine 100 by utilizing a physics-based model (which can be part of Method A) according to embodiments of the present disclosure is shown. Figure 5, for example, can be utilized to determine the remaining useful lifetime of insulation components of an electric machine 100. In certain embodiments, for example, the system 1100 can include an electric machine 100 for which a remaining useful lifetime is to be determined, one or more sensors 220 (e.g., embedded in the electric machine 100 and / or in range or proximity of the electric machine 100), computing resources (e.g., microprocessors 224, memory devices 226, system 1100 (e.g., cloud-server, edge computer, embedded processor, first user device 1102, etc.), a human machine interface (e.g., touchscreendisplay, handheld device (e.g., first user device 1 102), etc ), a physics-based model(s), any other components and / or systems, or a combination thereof. As with process flows 300, 400, the process flow 500 can, in certain embodiments, operate and / or execute continuously on a computing resource and can include exchanging data with the human machine interface 524.

[0057] With regard to insulation components, in certain embodiments, the life of insulation can be halved for each 10 degrees Celsius increase in the operating temperature above the nominal temperature. Thermal aging in insulation can be expressed as the rate at which temperature- induced changes (i.e., deterioration) occur. In certain scenarios, the foregoing phenomenon can follow the Arrhenius chemical rate model, which describes how chemical reactions are influenced by temperature. In the case of insulation, higher temperatures accelerate the aging process, leading to degradation over time. In light of foregoing, the lifespan of insulation subjected to higher temperatures can be expressed as, L = B e<^' / / c7’ where, L = time to reach a specified endpoint or lifetime; B = a constant specific to an insulation material and assembly which can be determined experimentally; (f> = the activation energy; k = Boltzmann’s constant; and T = absolute temperature in K. Repetitive start and stop of the electric machine 100 can also deteriorate the insulation life of the winding. In certain scenarios, when the electric machine 100 starts under load, the electric machine 100 may draw five to eight times the nominal current. This may result in high short-term copper losses and heat build-up. In certain scenarios, even greater draw may occur. For example, for a 0.25 HP motor start current of 9.1 A, the no-load current can be 0.7 A, which is thirteen times the nominal current. If the electric machine 100 is then stopped and subsequently restarted before the electric machine 100 has had a chance to cool down, a cumulative heat energy can build up. Repetitive starts and stops in a short interval of time can have an adverse effect on motor winding life as well. The frequency of starts and stops and the inertia of the load being accelerated, and the thermal path to dissipate heat to ambient. Additionally, when electric machines 100 operate at higher altitudes, electric machines 100 can experience increased temperature rises compared to those operating at sea level. This occurs because the ambient air is less dense, resulting in reduced heat dissipation.

[0058] In certain embodiments, the process flow 500 can begin at block 502, which can include starting a new evaluation interval to evaluate the remaining useful lifetime of the electric machine 100 as whole and / or the remaining useful lifetime of specific components of the electric machine 100, such as insulation components. In certain embodiments, the evaluation loop can start uponactivation of the electric machine 100 so that the evaluation is conducted from new and to avoid losing historic load data for the electric machine 100.

[0059] At block 504, the process flow 500 can include obtaining signals from the various sensors being utilized to monitor the electric machine 100. In certain embodiments, the sensors (e g., sensors 220) can be embedded in the electric machine 100, in proximity to the electric machine 100, or a combination thereof. At block 506, various specifications of the electric machine 100 can be fed into a physics-based model of the electric machine 100, including those relating to insulation components and / or other components of interest. Such specifications can include any information associated with the electric machine 100, such as, but not limited to, information on the components of the electric machine 100, information identifying materials of the components, information identifying functionality and stress values for the components, information identifying any characteristics of the electric machine 100, or a combination thereof. At block 508, the outputs from block 504 can supplied to the physics-based model of the electric machine 100. In certain embodiments, the physics-based model of an electrical machine 100 can constitute equations related to the materials used for the various components of the electric machine 100 (e.g., stator and rotor materials), empirical relations from manufacturers of components of the electric machines 100, component specifications related to expected performance and / or characteristics of properly-functioning components, or a combination thereof. In certain embodiments, the equations of the physics model can describe the heat flux and temperature distribution in the insulation material as a function of measured temperatures at certain locations. The equations may be based on fits to numerical results from a FEA or thermal measurements. In certain embodiments, at block 508, the physics-based model of the electric machine 100 can be utilized to calculate the input parameters needed for calculating the remaining useful lifetime values of the various components of the electric machine.

[0060] Blocks 510, 512, 514, 516 of process flow 500 can be utilized to calculate the available and estimated information relevant to insulation remaining useful life. Block 518 can involve updating a lookup table for remaining useful lifetime based on the insulation material (or other characteristics) for the electric machine. Block 520 includes calculating the remaining useful lifetime of the insulation components applying all applicable and calculated derating factors. In further detail, block 510 can include estimating the electric machine temperature , Ti using the equations from the physics model. Block 512 can include record a time duration tn, for operationat temperature Ti and idle time duration tn, at idle temperature Tit. At block 514, the process flow 500 can include calculating the average idle temperature TT and average operation temperature TR from previous operation history. At block 516, the process flow 500 can include accumulating all previous time durations Xfor respective temperatures for previous intervals and idle times. At block 518, the process flow 500, for temperature T the total lifetime can be calculated: Ln = B exp((p / BTi ). At block 520, the process flow 500 can include computing the remaining insulation life operating at future operation temperature TR and idle temperature TTin the future Lrem = average( LTR , LT ) - ^ -At block 522, the process flow 522 can include determining insulation component with the shortest remaining life. At block 524, the process flow 524 can include outputting the determinations for remaining useful life, such as via a human machine interface.

[0061] As can be seen from the example table for the bearings above, the calculations of the remaining useful bearing life can assume that the speeds and loads at the current evaluation interval will continue at the same level in the future (i.e., will remain constant at the level measured). As a result, the remaining useful life estimates can increase or decrease from interval to interval. The remaining useful life for the bearings after the 1stinterval is 2.90 (millions of revolutions) and decreases to 0.98 after the 2lldinterval. The remaining useful life increases again to 1.91 after the 3rdinterval. If the sensor measurements (i.e., load, speed, temperatures, etc.) in a subsequent evaluation interval decrease or increase, the calculation of the remaining useful life from that evaluation interval can also decrease or increase because it can be assumed that the present conditions will remain constant into the future. This calculation method can be desirable for immediately reflecting the effects of a load change on the remaining useful life. Similar statements can be made about the rotor bars (Fig. 3) and the winding insulation (Fig. 5) calculations.

[0062] In certain embodiments, the estimate of future operating conditions can be based on the assumption that future loads and environmental conditions will follow a distribution corresponding to the previously measured loads and environmental conditions. The previous measurements can be binned and plotted on histogram. Example histograms for torque and speed are shown in Figure 6. The histograms 602, 604 show how the speed and torque are distributed over time respectively. Other measured or derived parameters may also be plotted on a histogram and can have a distribution that can be used for assuming future conditions.

[0063] In certain embodiments, the evaluation loop automatically loops repeatedly while theelectrical machine is running. In this case, the operating conditions consisting of torque, temperature, vibrations, speed, and sound are measured by the sensors, read out by the processor, and logged into memory. The estimate of the remaining useful life of the electrical machine can be calculated each time the evaluation loop executes. The future loads can be assumed to have a distribution following the prior load distribution or be the same as the measured load in the current evaluation interval.

[0064] In certain embodiments, the human machine interface 324 / 424 / 524 can be utilized to initiate an evaluation loop for evaluating the electric machine 100, and can enable a user to start or stop the evaluation loop. For example, in block 302 / 402 / 502, the evaluation can be initiated by a user input or automatically based on completion of a previous loop or other condition occurring. In certain embodiments, the evaluation loop can start as soon as the electric machine 100 is started or activated so that the evaluation is conducted from new and to avoid losing historic load data for the calculation of remaining useful life. The previous measurements of previously applied loading and previously present environmental conditions can be stored in memory and can be retrieved next time the electric machine is in use so that the remaining useful lifetime is estimated again.

[0065] The time periods during which the evaluation loop is NOT running but the electrical machine is operating can be tracked / logged by saving the times at which the machine is turned off or on. The next time the lifetime calculations are performed (because a user initiates the evaluation loop), the loads from when the evaluation loop was NOT running can be included in the calculation of the incremental damages and the total accumulated damages. The loads and environmental conditions can be assumed to have the same distribution as the loads and environmental conditions that were measured when the evaluation cycle was running. This method assumes there is some prior data from prior evaluation intervals from which some histograms of loads and environmental conditions (i.e., a load history) can be established. While the estimation loop was NOT running can be done based on a load profde that was established from prior usage history. In certain embodiments, the remaining useful lifetime can be the expected amount of time remaining before a particular component (or the electric machine 100 itself) fails, needs to be replaced, needs to be repaired, and / or has certain operative capabilities.

[0066] Referring now also to Figure 7, an exemplary schematic illustrating remaining useful lifetime estimation by utilizing a system 700 including an artificial intelligence and / or machine learning model (Method B) according to embodiments of the present disclosure is shown. Thesystem 700 can be included within system 1000 and / or any other system of the present disclosure. In certain embodiments, the system 700 can provide predictions for remaining useful lifetime for components of an electric machine and for an entire electric machine 100 as whole. In certain embodiments, the system 700 can include one or more machine learning models 702, which can be a supervised machine learning model that is fed labeled training data for training, an unsupervised machine learning model that does not utilize labeled training data, semi-supervised machine learning models, a model that utilizes reinforcement learning, any type of machine learning model, a neural network, or a combination thereof. In certain embodiments, the machine learning models can incorporate and / or utilize any type of neural network. The machine learning model 702 can be trained with training data 706 (e.g., experimental data and / or historical data associated with the comparable electric machines 100 that can have a threshold similarity and / or functionality in common with the electric machine 100 under evaluation and data that has labels corresponding to the failure type and time to failure). The training data can include electric machine specifications, information regarding materials of components (these are the labels of the data) indicating whether a component has deteriorated or not, any other training data (e.g., from prior predictions that have been determined to be accurate), or a combination thereof. In certain embodiments, the machine learning model 702 can be trained with training data 706 to perform predictions on remaining useful lifetime of components of electrical machines, such as electrical machine 100. In certain embodiments, data can refer to known input-output pairs of remaining useful lifetime values for different operating and environmental conditions for different electrical machines. The training data may be collected from lab tests or measurements in the field. In certain embodiments, the machine learning model 702 can be fed with calculated data 703, which can include industry standard and / or theoretical training data. The calculated data may be based on simulations (i.e., FEA), empirical relations, or analytical equations. Additionally, the machine learning model 702 can be fed with real-time data 704 (e.g., sensor data and / or other types of data) measured for the electric machine 100 and / or the electric machine’s components, such as by the sensors 220, such as during an evaluation interval. The machine learning model 702 can analyze the sensor data for the electric machine 100 and can predict and / or estimate the remaining motor lifetime for the components and / or the electric machine 100 as a whole at 708. In certain embodiments, the predictions can be generated by factoring in the training data, the industry standards, and / or the real-time data. Additionally, test data might be used to test the performanceof the machine learning model before it gets deployed. The test data is similar to the training data, except for that it is not being used for training the model so that the performance of the model can be tested with new data that the model has not seen before.

[0067] In certain embodiments, the supervised machine learning model can be trained using statistical methods (e.g., stochastic gradient decent) using a large database of information containing information on electric machine failures (e.g., data labeled as a failure, impending failure, no failure, etc. under different operating conditions and for different electrical machines). In certain embodiments, the database can be continuously or periodically updated over time using the sensors and can include additional information, such as, but not limited to, information relating to warranty claims, customer surveys (e.g., photos of failed motors and application specifics supplied by customers), maintenance calls, etc. In certain embodiments, the machine learning model can utilize manufacturing parameters (i.e. electric machine component dimensional analysis, electric machine assembly parameters, part quality control data, and usage parameters (i.e. torque loads, temperature, and / or vibration data, air quality index, altitude) to match the customer’s electric machine, to an electric machine that had failed previously, and had close (e.g., threshold correlation) or identical manufacturing and usage parameters. The result of this approach can be utilized to predict lifetime remaining (i.e., until motor failure occurs). In certain embodiments, the electric machine sensor system can include and collect any number and / or types of motor usage parameters, such as current, voltage, vibration, speed, temperature, and output torque. In certain embodiments, a supervised, deep-learning Machine Learning (ML) algorithm can be utilized by the model to estimate remaining useful life in the electric machine based on present and historic data and without human intervention. In certain embodiments, the machine learning model can predict the lifetime of a given electric machine based on sensor feedback and correlation of the sensor signal with a database of historical data from previously failed comparable electric machines.

[0068] Referring now also to Figure 8, an exemplary process flow 800 for determining remaining useful lifetime for components of an electrical machine by utilizing machine learning model and / or physics-based models according to embodiments of the present disclosure is shown. In the process flow 800, one or more parameters can be fed into the machine learning model. For example, measurement data including, but not limited to, motor vibrations, speeds, temperatures (including temperature difference from ambient), torques, shaft axial and radial (i.e., overhung)loads, manufacturing parameters, altitude above sea level, air quality, air density, relative humidity, and / or any other parameters can be fed as inputs to the machine learning model. Measured data (e.g., sensor data for components of an electric machine being evaluated) can be compared side-by-side with data from training examples to a comparable electric machine from a training dataset. Damages that were introduced to the motor components since a previous monitoring event can be generated and added to the existing cumulative damage of the electric machine components. To compute an expected remaining life of the motor, or more preferentially specific component (drive / non-drive bearing, rotor bar, etc.) remaining life, time until each component failure is calculated: Remaining hours = (1 - cumulative damage) / damage rate (damage / h), where cumulative damage is the fraction of damage already affecting the component: 0 corresponds to a damage-free component and 1 corresponds to a damaged component at the end of its useful life.

[0069] In certain embodiments, the estimated lifetime reduction during the measured interval can be calculated, such as based on the training dataset electric machine life. The method benefits from being independent of motor lifetime calculations of a physics-based model, which, in certain scenarios, can be too restrictive or lax. For example, in certain scenarios, if bearings are pressed at a higher load during their manufacture, their life can be reduced. In certain scenarios, imperceptible supplier specific defects on components for a specific batch of supplies within a timeframe can impact the lifetime of a motor. Such defects might not be captured through inspection, and can lead to some deviation in lifetime calculation if only model based on industry standards (i.e., manufacturer recommended calculation methods) is used for prediction. Field-data driven machine learning based methods can capture these failure mechanisms and improve the prediction system over the time because they capture effects that may not be included or unknown when performing the manufacture recommended or industry standard methods. Such information can be unique to each electric machine and can be fed into the machine learning model. In certain embodiments, if any component of an electrical machine 100 is detected to have a defect, such as a manufacturing defect, the defective component can be replaced prior to providing the electrical machine 100 to a customer, or the component can be rejected from inclusion in the electrical machine 100 prior to being provided to a customer or end user. In certain embodiments, there can be some level of variance in the remaining useful lifetime estimate based on manufacturing variability that can occur when components of electrical machines 100 are produced. In certainembodiments, the machine learning model can utilize different rules. In certain embodiments, the rule can be part of the model, where specific lifetime estimation calculation approach can be selected. For example, the rule can be matching similarly loaded test dataset electric machine to the electric machine being measured. In certain embodiments, the rule may be grouping the electric machines based on the environment conditions and loads. The rules can include multiple grouping ways of the electric machine that is being measured and failed electric machines in the testing database. The rule may also use different coefficients to the different historic electric machine failures for the different manufacture process and measurement data to emphasize most important factors affecting motor component life and deemphasize non important factors. A goal of the rule selection can be to match the electric machine that has already failed, and its lifetime is known to the electric machine that is being measured. Each electric machine component lifetime can be measured using this embodiment, a simplified method that focuses on just the key components (DE, NDE bearings, windings, rotor bar), or even a more simplified method focusing on the electric machine and not its components can be used. Simplified methods may be used if thorough dataset is missing failure root cause analysis information.

[0070] At block 802, the process flow 800 can include starting a new evaluation interval for predicting the remaining useful lifetime for components of an electric machine 100 and / or the electric machine 100 itself. At block 804, the process flow 800 can include obtaining and / or reading real-time sensor data from sensors of the electric machine 100. In certain embodiments, the sensor data can be obtained from sensors in proximity to and / or embedded in the electric machine 100. At block 806, the process flow 800 can include performing various actions on the sensor data. For example, the actions can include conducting clean up of the sensor data (i.e., omitting / deleting erroneous data), filtering of the sensor data (i.e., to reduce the effects of measurement or shot noise), binning, mapping, framing, time stamping, providing data attributes, conducting domain conversions (i.e., transform time domain data to frequency domain), and / or other actions. At block 808, the process flow 800 can include selecting a component of the electric machine 100. At block 810, the process flow 800 can include performing a lookup of remaining useful life from a previous cycle or evaluation interval. At block 812, the process flow 800 can include selecting a strategy (i.e., a calculation method) for predicting the remaining useful lifetime. To facilitate selection of the strategy, a component i or failure mode j rule-based system can be utilized. The rule-based system can include optional new rules that can be uploaded into the rule-based system. In certain embodiments, additionally, to facilitate the selection of the strategy, knowledge base information and / or rule based information can be accessed at 816. For example, in terms of knowledge base information, new knowledge and factors for component i or failure mode j, manufacturing processing parameters and factors, electric machine design parameters, supplier m historical lifetime factors, failure modes of historically-failed components, installation features, mounting orientation, ambient temperature, pressure, relative humidity, altitude above sea level and / or other factors, and component i or failure mode j lifetime correlation with other components and failure modes can be utilized in selecting a strategy. In certain embodiments, at 814, new rules can be provided to the rule base, which can be as source of information for selection of the strategy. In certain embodiments, the rule base can be a physics-based set of equations corresponding to a particular electrical machine. Additionally, at 814, the physics-based model lifetime information and determinations can also be utilized for the rule base, which can also be utilized for the selection strategy.

[0071] At block 818, once the strategy (i.e., the calculation method) is selected, the machine learning model can estimate the lifetime Li for a component i and Lj for a failure mode j . Once the estimate for the first component is completed, the process flow 800 can revert back to 808 if iteration through all components have not been completed at 820, and can select a next component and predict the lifetime for the next component. The process can be repeated until the iteration through all components is completed. At block 822, the process flow 800 can predict the remaining useful lives for all the components. Based on the predicted remaining useful lives, the process flow 800 can proceed to determine whether there is a failure onset for the component, such as by determining whether the remaining useful lifetime is near zero or at a low value. If there is no failure onset for any components, the process 800 can initiate a next evaluation interval at 802 and proceed through the process flow 800. If, however, at 824, there is a failure onset, the process flow 800 can proceed to block 826, which can include conducting an optional manual inspection to confirm failure of the component. If the inspection confirms the failure, the machine learning model can be rewarded for the accurate prediction at block 828. If, however, the prediction is inaccurate based on the inspection, the process flow 800 can penalize the machine learning model at 828. In certain embodiments, the reward to the model can be that the parameters for the machine learning model are not modified, and a penalty can be provided to the machine learning model when modifications (e g., corrections) are made to the machine learning model. Based on theaccuracy, the rule base can be updated and can be utilized to facilitate estimation strategy selection on a next interval.

[0072] Referring now also to Figure 9, an exemplary process flow 900 (Method C) for determining remaining useful lifetime for components of an electrical machine by utilizing machine learning models and / or physics-based models according to embodiments of the present disclosure is shown. For example, at 902 environmental condition information can be utilized for learning electric machine usage patterns at 908. Additionally, operating load information 904 can be utilized for learning of electric machine usage patterns at 908. Furthermore, electric machine conditions 906, such as electric machine temperatures and / or other sensor measurements can also be utilized for learning of electric machine usage patterns 908. Electric machine conditions 906 can include, but are not limited to, temperatures, vibrations, sound / acoustic information, pressure measurements, strain measurements, any other conditions, or a combination thereof. The information from 908 can be provided to a machine learning model capable of performing lifetime estimations for components at 910. At 910, either method A or method B can be utilized. For example, in certain embodiments, when sufficient history of failures is not available, it can be recommended to start with method A. Component level models and understanding of the physics (e.g., materials: steels, dielectrics, lubricants, etc. components: bearings, shafts, rotor bars, seals, carbon electrodes) can be utilized. In method A, a physics model may be relied on because the library / database of known failures needed for the other methods may be too small. For example, rotor bar failures, winding defects, corona discharge (sensed with partial discharge equipment), imbalance or windings shorting (picked up by an imbalance of the signals from the sensor), and / or other issues can be predicted in the context of physics-based models. In certain embodiments, method B can be used to calculate expected motor life based on a combination of both physics model and machine learning from training dataset, as shown in Figure 8. The calculation of component RUL can be initially based on established empirical or analytical equations following industry standards, material properties, sensed operating and environmental conditions, and dimensions. With more training examples of a particular electric machine, its materials and manufacturing procedures, the calculation can become more heavily based on the machine learning evaluations. This may be particularly true if the industry standards do not provide an accurate expected motor lifetime prediction.

[0073] While Method B can be utilized to predict electric machine lifetime, Method B canrequire a large training database, which can take a long time to develop, to provide sufficient confidence that the predicted lifetime is accurate for an electric machine operating in previously untested environment or condition, or when an electric machine with different design parameters is used. To mitigate the lack of data in the initial period of Method A application, Method B can be required. Method B can be a modified machine learning algorithm that uses both lifetime estimated by a physics model and lifetime estimates from the training database. In certain embodiments, the motor operating parameters, (e.g., speed and torque) and environmental conditions (e.g., rel. humidity and ambient temperature) can be provided by sensors. The parameters can be fed into a central computing unit, which may be an edge unit or a cloud computing system. Similarly, if the machine learning method is used, then the current load is weighted into the model and the remaining lifetime is computed with equations developed by the model, using the training example.

[0074] Depending on database and available information, it can be recommended to use either Method A and B but add an element of unsupervised learning. In certain embodiments, a process can start using an unsupervised ML method without the database of failures from method B. The reward function can take care of adjusting the agent which is giving the predictions, as shown in Figure 8. Method C can be concerned with the prediction of future usage of the electric machine. The foregoing enables accurate lifetime predictions with either a physics model or a machine learning / large data approach (i.e., methods A or B). With method C, the electric machine loads can be recorded constantly, and this historic data can be used to predict future motor loads. Method C extrapolates the measured motor usage data into the future. This extrapolation can be done in several ways. One way to do this is to “learn” the usage with machine learning algorithms. The algorithm may utilize histograms on which prior and current torques, speeds are logged. The more data is fed into the histograms, the more confident predictions about usage of the motor for the rest of its design life can be made.

[0075] For example, in an exemplary use-case scenario, an electric machine “Ml” is used in an aggregate production facility. The loads that are being recorded show very high torque ripples from rocks being crushed. The loads are present only during regular business hours, and the motor is not used at night. A second electric machine “M2” is used in a wastewater treatment plant to drive a clarifier. This motor sees almost constant torques with minimum ripple and no downtime, i.e., it runs day and night. Although the hours of operation of Ml can only be about 1 / 3 of that ofM2, the severe service that the electric machine is subjected to may cause a physics model or a machine learning model to predict much shorter remaining lifetimes for Ml than for M2. If, on the other hand, operating conditions where to change for either of the two motors, this change would be reflected in the load histograms and the lifetime estimations. In certain embodiments, a drawback of the histogram described above is that it may not retain information about the order in which loads are applied which might be important for some damage accumulation models. Therefore, the machine learning model of electric machine usage might also employ strip charts or spectrograms (in frequency domain) that keep track of the time. In certain scenarios, window functions can be used to periodize the data (divide it up into equal periods of nearly identical data).

[0076] At 912, maintenance downtime can be initiated if the output of block 810 shows zero RUL left (Output 1 )In certain embodiments, at 912, the remaining useful lifetime can indicate that the electrical machine has failed or has reached and / or is imminently about to reach its end of life. At 914, an inspection can be conducted to confirm accuracy of lifetime estimations or discover inaccuracies of the lifetime estimations performed by the model. At 916, the process flow 900 can communicate the predicted component lives from block 910 (Output 2) that are remaining until required maintenance downtime (when RUL is zero) to the user via the HMI. At 918, the inspection information from 914 and predictions of component lives from 916 can be utilized for evaluating the accuracy of the model predictions. The evaluation results at 918 can be used to create a reward function at 922. At 924, an agent can be provided that makes changes to block 910 based on a reward function from block 922 so that the RUL estimates from 910 are more accurate.

[0077] In certain embodiments, additional information can be utilized in the physics-based models and / or artificial intelligence models. For example, information about a supplier m for a specific electrical machine component can be recorded and tracked in a data library. The state of this component at the end of electrical machine life can be updated and if the inspection finds electrical machine failure due to the specific component, such information can be recorded in the database. Such component i is expected to have the shortest lifetime in an electrical machine which is exposed to the identical operation conditions. If supplier m is changed to supplier n for the component z, such information can be recorded and initially, where there is no information about the expected component lifetime, the lifetime can be assumed to be the same as in the old supplier m. Over time, the information from the new supplier n (because of electrical machine failures) is updated, indicating whether the new supplier component i is the first one to fail or not.Information from the two suppliers m and n over time allow to better predict expected life of a specific component (or the whole electrical machine if the component is the first to fail from the electrical machine assembly).

[0078] In certain embodiments, still additional features and functionality can be provided. For example, electric machine lifetime predictions can be made using machine learning algorithm and training modules with historic motor lifetime and usage database, where the lifetime calculation can be based on various types of measurable factors / parameters, such as, but not limited to realtime torque (e.g., real-time data can mean measurements with acceptable delay while the motor is operating), vibration pattern, electric machine housing or internal component strains, acoustic pattern for components, temperature pattern for components, and / or electric machine parameters (e.g., current, voltage, speed, torque, etc.). In certain embodiments, motor and individual component lifetime calculation can start with historical lifetime estimates from comparable electric machine designs. In certain embodiments, industrial electric machine and individual component lifetime prediction can be performed, where industrial standards are used together with machine learning algorithm with a training databank. In certain embodiments, electric machine lifetime predictions can be made, where no sensors may be on the electric machine directly, but sensors can be in a motor, couplers, or bearings. In certain embodiments, electric machine and individual component lifetime calculations can be conducted, where individual motor manufacturing / supplier data (e.g., assembly steps, materials characteristics curve, and / or environmental condition) can be fed into the model. In certain embodiments, electric machine and individual component lifetime predictions, where one, two, or a plurality of sensors are feeding data into a local or remote computation unit for calculation and data storage can be utilized. In certain embodiments, motor maintenance schedule optimization, based on real time usage of the electric machine, where load data and / or grease quality and / or supplied voltage are continuously monitored can be utilized. In certain embodiments, electric machine lifetime monitoring, where lifetime can be shown in cumulative time used for the electric machine, and remaining electric machine use time, for the current or average load scenario can be utilized. In certain embodiments, electric machine health indication on the electric machine or in remote human machine interface based on remaining lifetime using a color scheme can be utilized. In certain embodiments, motor lifetime prediction algorithms, which can be optimized and tuned based on field results can also be utilized.

[0079] Referring now also to Figure 10, an exemplary system 1000 for estimating remaininguseful lifetime according to embodiments of the present disclosure is provided. In certain embodiments, the exemplary system 1000 can be utilized to supplement and / or augment the functionality that can be integrated into the electric machine 100 and / or components of the electric machine 100. Notably, the system 1000 can be configured to support, but is not limited to supporting, monitoring systems and services, cloud computing systems and services, electric machine-monitoring services and applications, electric machine health condition applications and services, sensor applications and services, mitigation capability applications and services, firewall systems and services, data analytics systems and services, data collation and processing systems and services, artificial intelligence services and systems, machine learning services and systems, neural network services, mobile applications and services, content delivery services, satellite services, telephone services, voice-over-internet protocol services, software as a service (SaaS) applications, platform as a service (PaaS) applications, operations management applications and services, productivity applications and services, and / or any other computing applications and services. Notably, the system 1000 can include a first user 1001, who can utilize a first user device 1002 to access data, content, and services, or to perform a variety of other tasks and functions. As an example, the first user 1001 can utilize first user device 1002 to transmit signals to access various services and content, such as those available on an internet, on other devices, and / or on various computing systems. In certain embodiments, the first user 1001 can utilize the first user device 1002 to interact and / or control an electric machine 100, a sensor device 220 or system, or a combination thereof. As another example, the first user device 1002 can be utilized to access an application, devices, and / or components of the system 1000 that provide any or all of the operative functions of the system 1000, the electric machine 100, the sensor device 220, or a combination thereof. In certain embodiments ,the first user 1001 can be a user that seeks to determine the remaining useful life of the electric machine 100 and / or components of the electric machine 100 so that potential repairs, component replacements, and / or other actions can be performed prior to the electric machine 100 and / or corresponding components failing.

[0080] In certain embodiments, the first user 1001 can be a person, a robot, a humanoid, a program, a computer, any type of user, or a combination thereof, that can be located in a particular location or environment, such as an environment including one or more electric machines 100, sensor devices 220, or a combination thereof. In certain embodiments, the first user 1001 can be a person that can want to utilize the first user device 1002 to conduct various types of activitiesand / or control devices and systems of the system 1000. For example, an activity can include, but is not limited to, accessing digital resources, such as, but not limited to, application content, video content, audio content, haptic content, audiovisual content, virtual reality content, augmented reality content, any type of content, or a combination thereof. In certain embodiments, other activities can include, but are not limited to, accessing various types of applications, such as to perform work, control the operative functionality of the electric machines 100, control the sensor devices 220, activate components and / or devices to respond to a condition detected by the sensor devices 220, or a combination thereof. In certain embodiments, the activities can include training and / or activating artificial intelligence models to predict remaining useful lifetimes for electric machines 100 and / or components of the electric machines 100.

[0081] In certain embodiments, the first user device 1002 can include a memory 1003 that includes instructions, and a processor 1004 that executes the instructions from the memory 1003 to perform the various operations that are performed by the first user device 1002. In certain embodiments, the processor 1004 can be hardware, software, or a combination thereof. The first user device 1002 can also include an interface 1005 (e.g., screen, monitor, graphical user interface, etc.) that can enable the first user 1001 to interact with various applications executing on the first user device 1002 and to interact with the system 1000. In certain embodiments, the first user device 1002 can be and / or can include a computer, any type of sensor, a laptop, a set-top-box, a power outlet such as a USB-C outlet, a tablet device, a phablet, a server, a mobile device, a smartphone, a smart watch, a voice-controlled-personal assistant, a physical monitoring device (e g., camera, sensors, etc.), an internet of things device (loT), appliances, an autonomous vehicle, and / or any other type of computing device. Illustratively, the first user device 1002 is shown as a computer in Figure 10. In certain embodiments, the first user device 1002 can be utilized by the first user 1001 to control, access, and / or provide some or all of the operative functionality of the system 1000.

[0082] In addition to using first user device 1002, the first user 1001 can also utilize and / or have access to any number of additional user devices. As with first user device 1002, the first user 1001 can utilize the additional user devices to transmit or receive signals to access various services and content and / or access functionality provided by the system 1000. The additional user devices can include memories that include instructions, and processors that executes the instructions from the memories to perform the various operations that are performed by the additional user devices.In certain embodiments, the processors of the additional user devices can be hardware, software, or a combination thereof. The additional user devices can also include interfaces that can enable the first user 1001 to interact with various applications executing on the additional user devices and to interact with the system 1000. In certain embodiments, the first user device 1002 and / or the additional user devices can be and / or can include a computer, any type of sensor, a laptop, a set-top-box, a power outlet such as a USB-C, a tablet device, a phablet, a server, a mobile device, a smartphone, a smart watch, an autonomous vehicle, and / or any other type of computing device, and / or any combination thereof. Sensors can include, but are not limited to, cameras, motion sensors, acoustic / audio sensors, pressure sensors, temperature sensors, light sensors, accelerometers, gyroscopes, any type of sensors, or a combination thereof.

[0083] The first user device 1002, the electric machine 100, sensor devices 220, and / or additional user devices can belong to and / or form a communications network 1033. In certain embodiments, the communications network 1033 can be a local, mesh, and / or other network that enables and / or facilitates various aspects of the functionality of the system 1000. In certain embodiments, the communications network can be formed between the first user device 1002 and additional user devices through the use of any type of wireless or other protocol and / or technology. For example, user devices can communicate with one another in the communications network by utilizing any protocol and / or wireless technology, satellite, fiber, communication technologies (e g., loT, Bluetooth, Zigbee, Z-wave, Wireless HART, cellular, LoRA, NB-IoT, etc.), or any combination thereof. Notably, the communications network 1033 can be configured to communicatively link with and / or communicate with any other network of the system 1000 (e.g., communications network 1035) and / or outside the system 1000.

[0084] In certain embodiments, the first user device 1002, the electric machines 100, the sensor devices 220, and additional user devices belonging to the communications network 1033 can share and exchange data with each other via the communications network 1033. For example, the user devices can share information relating to the various components of the devices, information associated with content accessed and / or attempting to be accessed by the first user 1001 of the user devices, information identifying the locations of the devices, information indicating the health condition and / or remaining useful lifetimes of the components 202 of the electric machines 100, information relating to replacement components for replacing existing components of the electric machine 100, information indicating the types of sensors that are contained in and / or on thedevices, information identifying the applications being utilized on the devices, information identifying how the devices are being utilized, information identifying user profiles for users of the devices, information identifying device profiles for the devices, information identifying the number of devices in the communications network 1033, information identifying devices being added to or removed from the communications network 1033, any other information, or any combination thereof.

[0085] In addition to the first user 1001, the system 1000 can also include a second user 1021. In certain embodiments, the second user 1021 can be similar to the first user 1001 and can seek to access content, applications, systems, and / or devices. Additionally, the second user 1021 can control the operative functionality of the electric machines 100, control the sensor devices 220, activate components and / or devices, such as to respond to a condition detected by the sensor devices 220, or a combination thereof, such as by utilizing the second user device 1022. In certain embodiments, the second user device 1022 can be utilized by the second user 1021 to transmit signals to request various types of resources, content, services, and data provided by and / or accessible by communications network 1035 or any other network in the system 1000. In further embodiments, the second user 1021 can be a robot, a computer, a vehicle, a humanoid, an animal, any type of user, or any combination thereof. The second user device 1022 can include a memory 1023 that includes instructions, and a processor 1024 that executes the instructions from the memory 1023 to perform the various operations that are performed by the second user device 1022. In certain embodiments, the processor 1024 can be hardware, software, or a combination thereof. The second user device 1022 can also include an interface 1025 (e.g., screen, monitor, graphical user interface, etc.) that can enable the first user 1001 to interact with various applications executing on the second user device 1022 and, in certain embodiments, to interact with the system 1000, electric machines, 100, and / or sensor devices 220. In certain embodiments, the second user device 1022 can be a computer, a laptop, a set-top-box, a tablet device, a phablet, a server, a mobile device, a smartphone, a smart watch, an autonomous vehicle, and / or any other type of computing device. Illustratively, the second user device 1022 is shown as a mobile device in Figure 10. In certain embodiments, the second user device 1022 can also include sensors, such as, but are not limited to, cameras, audio sensors, motion sensors, accelerometers, gyroscopes, pressure sensors, temperature sensors, light sensors, humidity sensors, any type of sensors, or a combination thereof. In certain embodiments, the second user 1021 can also utilize additional user devices as well.

[0086] In certain embodiments, the second user device 1022, electric machines 100, sensor devices 220, and additional user devices belonging to the communications network 1034 can share and exchange data with each other via the communications network 1034. For example, the user devices can share information relating to the various components of the devices, information identifying and / or associated with the remaining useful lifetimes for components of the electric machine 100 and / or the electric machine 100 itself, information associated with content accessed and / or attempting to be accessed by the second user 1021 of the devices, information identifying the locations of the devices, information indicating the types of sensors that are contained in and / or on the devices, information identifying the applications being utilized on the devices, information identifying how the user devices are being utilized, information identifying user profiles for users of the devices, information identifying device profiles for the devices, information identifying the number of devices in the communications network 1034, information identifying devices being added to or removed from the communications network 1034, any other information, or any combination thereof.

[0087] In certain embodiments, the devices described herein can have any number of software functions, applications and / or application services stored and / or accessible thereon. For example, the user devices can include applications for controlling and / or accessing the operative features and functionality of the system 1000, applications for controlling and / or accessing any device of the system 1000, artificial intelligence and / or machine learning applications, applications for controlling, generating, and / or training the artificial intelligence and / or machine learning models, cloud-based applications, electric machine health condition detection applications, electric machine health condition mitigation applications, communication applications, business applications, e-commerce applications, media streaming applications, content-based applications, media applications, database applications, gaming applications, internet-based applications, browser applications, mobile applications, service-based applications, productivity applications, video applications, music applications, social media applications, any other type of applications, any types of application services, or a combination thereof. In certain embodiments, the software applications can support the functionality provided by the system 1000 and methods described in the present disclosure. In certain embodiments, the software applications and services can include one or more graphical user interfaces so as to enable the first and / or second users 1001, 1021 to readily interact with the software applications. The software applications and services can also beutilized by the first and / or second users 1001, 1021 to interact with any device in the system 1000, any network in the system 1000, or any combination thereof. In certain embodiments, devices can include associated telephone numbers, device identities, network identifiers (e.g., IP addresses, etc.), and / or any other identifiers to uniquely identify the user devices. In certain embodiments the software can be open-source and users 1001 and 1021 may be able to create their own software.

[0088] The system 1000 can also include a communications network 1035. The communications network 1035 can include resources (e.g., data, web pages, content, documents, computing resources, applications, and / or any other resources) that can be accessible to the first user 1001, second user 1021, and / or any of the devices in the system 1000, such as the electric machines 100 and sensor devices 220. The communications network 1035 of the system 1000 can be configured to link any number of the devices in the system 1000 to one another. For example, the communications network 1035 can be utilized by the second user device 1022 to connect with other devices within or outside communications network 1035. Additionally, the communications network 1035 can be configured to transmit, generate, and receive any information and data traversing the system 1000. In certain embodiments, the communications network 1035 can include any number of servers, databases, or other componentry. The communications network 1035 can also include and be connected to a neural network, a mesh network, a local network, a cloud-computing system or network, an IMS network, a VoIP network, a security network, a VoLTE network, a wireless network, an Ethernet network, a satellite network, a broadband network, a cellular network, a private network, a cable network, the Internet, an internet protocol network, MPLS network, a content distribution network, any network, or any combination thereof. Illustratively, servers 1040, 1045, and 1050 are shown as being included within communications network 1035. In certain embodiments, the communications network 1035 can be part of a single autonomous system that is located in a particular geographic region, or be part of multiple autonomous systems that span several geographic regions.

[0089] Notably, the functionality of the system 1000 can be supported and executed by using any combination of the servers 1040, 1045, 1050, and 1060. The servers 1040, 1045, and 1050 can reside in communications network 1035, however, in certain embodiments, the servers 1040, 1045, 1050 can reside outside communications network 1035. The servers 1040, 1045, and 1050 can provide and serve as a service that performs the various operations and functions provided by the system 1000. In certain embodiments, the server 1040 can include a memory 1041 thatincludes instructions, and a processor 1042 that executes the instructions from the memory 1041 to perform various operations that are performed by the server 1040. The processor 1042 can be hardware, software, or a combination thereof. Similarly, the server 1045 can include a memory 1046 that includes instructions, and a processor 1047 that executes the instructions from the memory 1046 to perform the various operations that are performed by the server 1045. Furthermore, the server 1050 can include a memory 1051 that includes instructions, and a processor 1052 that executes the instructions from the memory 1051 to perform the various operations that are performed by the server 1050. In certain embodiments, the servers 1040, 1045, 1050, and 1060 can be network servers, routers, gateways, switches, media distribution hubs, signal transfer points, service control points, service switching points, firewalls, routers, edge devices, nodes, computers, mobile devices, or any other suitable computing device, or any combination thereof. In certain embodiments, the servers 1040, 1045, 1050 can be communicatively linked to the communications network 1035, any network, any device in the system 1000, or any combination thereof.

[0090] The database 1055 of the system 1000 can be utilized to store and relay information that traverses the system 1000, cache content that traverses the system 1000, store data about each of the devices in the system 1000 and perform any other typical functions of a database. In certain embodiments, the database 1055 can be connected to or reside within the communications network 1035, any other network, or a combination thereof. In certain embodiments, the database 1055 can serve as a central repository for any information associated with any of the devices and information associated with the system 1000. Furthermore, the database 1055 can include a processor and memory or can be connected to a processor and memory to perform the various operations associated with the database 1055. In certain embodiments, the database 1055 can be connected to the servers 1040, 1045, 1050, 1060, the first user device 1002, a second user device 1022, the communications network 1033, the communications network 1034, the communications network 1035, a server 1040, a server 1045, a server 1050, a server 1060, and a database 1055, the additional user devices, any devices in the system 1000, any process of the system 1000, any program of the system 1000, any other device, any network, or any combination thereof.

[0091] The database 1055 can also store information and metadata obtained from the system 1000, stores sensor data measured by the sensor devices 220, store remaining useful lifetime predictions and / or determinations, store information indicating an accuracy of predictions and / ordeterminations, store information for replacement components and / or information for repairing components of the electric machine 100, store electric machine health condition information determined by the sensor devices 220, store information associated with the electric machine 100 and components 202 of the electric machine 100, store information associated with measures that can be taken to counteract an electric machine-health condition, store any sensor data, store baseline sensor data indicating normal operating conditions for components 202 of the electric machine 100, store specifications for normal operation conditions for components 202 of the electric machine 100, store information identifying measures to take depending on the type of condition detected by the sensor devices 220, store metadata and other information associated with the first and second users 1001, 1021, store profiles for the networks of the system, information identifying the networks of the system 1000, store configuration information for the networks and / or devices of the system 1000, store user profiles associated with the first and second users 1001, 1021, store device profiles associated with any device in the system 1000, store communications traversing the system 1000, store user preferences, store information associated with any device or signal in the system 1000, store information relating to patterns of usage relating to the devices, store any information obtained from any of the networks in the system 1000, store historical data associated with the first and second users 1001, 1021, store device characteristics, store information relating to any devices associated with the first and second users 1001, 1021, store information associated with the communications network 1035, store any information generated and / or processed by the system 1000, store any of the information disclosed for any of the operations and functions disclosed for the system 1000 herewith, store any information traversing the system 1000, or any combination thereof. Furthermore, the database 1055 can be configured to process queries sent to it by any device in the system 1000.

[0092] Notably, as shown in Figure 10, the system 1000 can perform any of the operative functions disclosed herein by utilizing the processing capabilities of server 1060, the storage capacity of the database 1055, or any other component of the system 1000 to perform the operative functions disclosed herein. The server 1060 can include one or more processors 1062 that can be configured to process any of the various functions of the system 1000. The processors 1062 can be software, hardware, or a combination of hardware and software. Additionally, the server 1060 can also include a memory 1061, which stores instructions that the processors 1062 can execute to perform various operations of the system 1000. For example, the server 1060 can assist inprocessing processes and / or loads handled by the various devices in the system 1000, such as, but not limited to, receiving signals from sensor devices including sensor data; determining, such as by utilizing physics-based models and / or artificial intelligence models, the remaining useful lifetimes of components of electric machines 100 and / or the overall electric machines 100 themselves; outputting the determining remaining useful lifetimes via interfaces; providing rewards to artificial intelligence models based on the accuracy of the predictions and / or determinations, such as to facilitate reinforcement learning; initiating the repair and / or replacement of components of the electric machines 100, training the artificial intelligence / machine learning models; receiving additional sensor data via signals; generating predictions for next evaluation intervals for the remaining useful lifetimes for components and / or the electric machine 100 itself; and performing any other suitable operations conducted in the system 1000 or otherwise. In certain embodiments, multiple servers 1060 can be utilized to process the functions of the system 1000. The server 1060 and other devices in the system 1000, can utilize the database 1055 for storing data about the devices in the system 1000 or any other information that is associated with the system 1000. In one embodiment, multiple databases 1055 can be utilized to store data in the system 1000.

[0093] In certain embodiments, the system 1000 can be utilized by the electric machines 100, the sensor devices 220, and / or other components of the system 1000 to interact and share information with each other. In certain embodiments, for example, the sensor data from each of the sensor devices 220 can be aggregated in the communications network 1035 and can be analyzed to determine trends in the sensor data among multiple electric machines 100, such as whether certain types of electric machines 100 have components that have shorter or longer remaining lifetimes in comparison to comparable electric machines 100, whether repairs or replacement of components lead to superior long-term overall electric-machine health, whether certain training data results in superior predictive capabilities for estimating remaining useful lifetime for components of electric machines 100, whether certain types of sensor data should have greater weight in determining or predicting remaining useful lifetime of components of electric machines 100 in comparison to other types of sensor data, whether certain types of sensor data should have greater weight when being utilized to train artificial intelligence models to perform predictions relating to remaining useful lifetime for components of an electric machine, along with any other information associated with the electric machines 100 and / or sensor devices 220. Incertain embodiments, the sensor devices 220 of each electric machine 100 can form a communication network with other sensor devices 220 of other electric machines 100. In certain embodiments, a first sensor device 220 associated with a first electric machine 100 can be configured to initiate actions on a second electric machine 100, such as if a second sensor device 220 of the second electric machine 100 shares information with the with first electric machine 100 and / or first sensor device 220. In certain embodiments, the system 1000 can also be configured to assist with the performance of any of the operations and / or functionality described herein and can also be utilized to execute the methods described in the present disclosure.

[0094] Referring now also to Figure 11, an exemplary method 1100 for determining and / or estimating remaining useful lifetime for electric machines according to embodiments of the present disclosure is provided. In certain embodiments, the method 1100 can be implemented by utilizing the electric machine 100 and / or sensor device 220. In certain embodiments, the method of Figure 10 can be implemented in the system 1000 of Figure 10 and / or any other systems, devices, and / or componentry illustrated in the Figures. In certain embodiments, the method 1100 can be implemented by utilizing a combination of the electric machine 100, the system 1000, and / or any other systems. In certain embodiments, the method of Figure 11 can be performed by processing logic that can include hardware (e.g., processing device, circuitry, dedicated logic, programmable logic, microcode, hardware of a device, integrated circuit, etc.), software (e.g., instructions run or executed on a processing device), or a combination thereof. In certain embodiments, the method of Figure 11 can be performed at least in part by one or more processing devices (e.g., processor 1042, processor 1047, processor 1052, and processor 1062 of Figure 10) and / or other devices, systems, components, or a combination thereof, of Figures 1-12. Although shown in a particular sequence or order, unless otherwise specified, the order of the operations in the method 1100 can be modified and / or changed depending on implementation and objectives. Thus, the illustrated embodiments should be understood only as examples, and the illustrated processes can be performed in a different order, and some processes can be performed in parallel. Additionally, one or more processes can be omitted in various embodiments. Thus, not all processes are required in every embodiment. Other process flows are possible. For example, in certain embodiments, the method 1100 can be combined with other methods. In certain embodiments, a greater or fewer number of operations as illustrated in Figure 11 can be incorporated into method 1100. In certain embodiments, the method 1100 can be modified to incorporate any of the functionality describedin the present disclosure.

[0095] Generally, the method 1100 can include operations, actions, or blocks for determining, prediction, and / or estimation remaining useful lifetimes for components of electric machines 100 and / or the overall electric machine 100 itself according to embodiments of the present disclosure. For example, in certain embodiments, the method 1100 can include measuring sensor data by utilizing one or more sensors that can be embedded and / or in proximity to an electric machine 100 being monitored. In certain embodiments, the method 1100 can include receiving signals, such as during a first evaluation interval, including sensor data from the sensors. Based on the sensor data, the method 1100 can include determining and / or predicting the remaining useful lifetime of components of the electric machine 100 and / or the overall electric machine 100 itself. The method 1100 can include outputting the remaining useful lifetime, such as via an interface (e.g., an interface of the electric machine 100 and / or an interface of a user device, such as first user device 1002). In certain embodiments, the method 1100 can include providing a reward to the artificial intelligence model in accordance with an accuracy of the remaining useful lifetime determination and / or prediction. In certain embodiments, the method 1100 can include initiating or scheduling a repair and / or replacement of the at least one component based on the remaining useful lifetime. In certain embodiments, the method 1100 can include training the artificial intelligence model based on the determinations, predictions, and / or accuracy. In certain embodiments, the method 1100 can include receiving additional signals including sensor data associated with the electric machine 100, such as during a next evaluation interval. In certain embodiments, the method 1100 can include predicting (or determining), for the next evaluation interval, a next remaining useful lifetime for the components of the electric machine 100 by utilizing the prediction capability of the artificial intelligence machines and / or physics-based models.

[0096] At block 1102, the method 1100 can include receiving, such as during an initial evaluation interval, one or more signals from one or more sensors that include sensor data associated with an electrical machine 100. In certain embodiments, for example, the initial evaluation interval can be a time period during which the sensor data is to be obtained and / or the remaining useful lifetime for the electric machine 100 is to be determined. In certain embodiments, the sensors can be embedded within the electric machine 100, in proximity to the electric machine 100, at other locations, or a combination thereof. In certain embodiments, the measuring of the sensor data by the one or more sensors can be triggered and / or activated by the first user 1001,such as via the first user device 1002, and / or by input devices of the electric machine 100 itself. In certain embodiments, the sensors can include any types of sensors, such as, but not limited to, pressure sensors, light sensors, acoustic sensors, temperature sensors, torque output sensors, accelerometers, gyroscopes, vibration sensors, voltage sensors, current-detection sensors, air sensors, cameras, humidity sensors, altitude sensors, any other types of sensors, or a combination thereof. In certain embodiments, the sensor data can include electric machine usage parameters, such as, but not limited to, current measurements, voltage measurements, vibration measurements, speed measurements, temperature measurements, output torque measurements, orientation measurements, content (e.g., video taken by a camera), light measurements, pressure measurements, sound measurements, any other types of measurements, air quality measurements, humidity measurements, altitude measurements or a combination thereof. In certain embodiments, the receiving of the one or more signals can be performed and / or facilitated by utilizing the electric machine 100, any other component of the sensor device 220 or system, any component of the system 1000, or a combination thereof.

[0097] At block 1104, the method 1100 can include determining the damage for one or more components of the electric machine 100, such as by utilizing the physics-based model and / or by utilizing an artificial intelligence model. Additionally, at block 1104, the method 1100 can include determining, such as by utilizing the physics-based model, a remaining useful lifetime for one or more components of the electric machine 100 based on the determined damage for the one or more components of the electric machine 100. In certain embodiments, at block 1104, the method 1100 can also include generating a prediction for the remaining useful lifetime for one or more components of the electric machine 100. In certain embodiments, the remaining useful lifetime can be the expected amount of time remaining before a particular component fails, needs to be replaced, needs to be repaired, and / or has certain operative capabilities. In certain embodiments, the physics-based models can constitute equations related to the materials used for the various components of the electric machine 100 (e.g., stator and rotor materials), empirical relations from manufacturers of components of the electric machines 100, component specifications related to expected performance and / or characteristics of properly-functioning components, or a combination thereof. For physics-based model approaches, the sensor data can be analyzed by the electric machine 100 and / or system 1000 and utilized for values in variables in the equations, utilized to compare to empirical relations, and / or utilized to compare to expected performanceand / or characteristics of property-functioning components to determine the remaining useful lifetime of each component and / or the overall electric machine 100 itself. In certain embodiments, at step 1104, the method 1100 can include predicting the remaining useful lifetime by utilizing one or more artificial intelligence / machine learning models. The artificial intelligence models can be trained to make the predictions for remaining useful lifetime using training data associated with comparable electric machines 100 and / or components that are the same or similar (or have a threshold correlation with) the electric machine 100 being evaluated. Once trained, the artificial intelligence model(s) can generate a prediction, based on the sensor data, for the remaining useful lifetime for one or more components of the electric machine 100 and / or the overall electric machine 100 itself. In certain embodiments, the determining and / or predicting can be performed and / or facilitated by utilizing the electric machine 100, any other component of the sensor device 220 or system, any component of the system 1000, or a combination thereof.

[0098] At step 1106, the method 1100 can include outputting the remaining useful lifetime prediction and / or determination via an interface. For example, in certain embodiments, the remaining useful lifetime for each component of the electric machine 100 under evaluation can be displayed on an interface 228 of the electric machine 100, on an interface of a user device, or a combination thereof. In certain embodiments, the outputting can be visual, auditory, haptic, virtual reality, augmented reality, any other type of output, or a combination thereof. In certain embodiments, the outputting of the remaining useful lifetime can include rendering each of the components of the electric machine 100 on the interface along with information identifying the remaining useful lifetime for each of the components. Additionally, an overall remaining lifetime can also be displayed for the entire electric machine 100. In certain embodiments, the outputting of the remaining useful lifetime can be performed and / or facilitated by utilizing the electric machine 100, any other component of the sensor device 220 or system, any component of the system 1000, or a combination thereof.

[0099] At step 1108, the method 1100 can include providing a reward to the artificial intelligence model in accordance with an accuracy of the remaining useful lifetime prediction, which can be utilized to modify the prediction capability of the artificial intelligence model. For example, if the model is trained via reinforcement learning, the providing of the reward as feedback to the artificial intelligence model can indicate how good or bad the prediction capability was for the model. Over time, the predictive capability will improve as rewards are given based on goodactions taken versus bad actions taken by the model. As a result, the artificial intelligence model can learn to make predictions and can learn from the consequences of its actions and can receive feedback in the form of rewards or penalties to help guide the learning process for the artificial intelligence model. In certain embodiments, the providing of the reward can be performed and / or facilitated by utilizing the electric machine 100, any other component of the sensor device 220 or system, any component of the system 1000, or a combination thereof.

[0100] At step 1110, the method 1100 can include initiating or scheduling a repair of the one or more components based on the remaining useful time determined and / or predicted for each of the components. For example, if a particular rotor of an electric machine 100 is determined to have a remaining useful life of 10 days, the system 1000 can initiate a repair or replacement of the rotor prior to the expiration of the 10-day period. In certain embodiments, the initiating or scheduling can be performed and / or facilitated by utilizing the electric machine 100, any other component of the sensor device 220 or system, any component of the system 1000, or a combination thereof. At step 1112, the method 1100 can include training the artificial intelligence model(s) based on the determinations made, predictions made, and / or accuracy of the predictions. In certain embodiments, at step 1112, the method 1100 can include training the artificial intelligence model(s) based on updated historical electric machine usage information (e.g., information for comparable and / or identical electric machines 100 relating to component useful life, etc ). In certain embodiments, the training can be performed and / or facilitated by utilizing the electric machine 100, any other component of the sensor device 220 or system, any component of the system 1000, or a combination thereof.

[0101] At step 1114, the method 1100 can include receiving, such as during a next evaluation internal of the electric machine 100, one or more additional signals from the one or more sensors, which can include sensor data associated with the electrical machine during the next evaluation interval. In certain embodiments, the receiving of the signals can be performed and / or facilitated by utilizing the electric machine 100, any other component of the sensor device 220 or system, any component of the system 1000, or a combination thereof.. At step 1116, the method 1100 can include predicting, for the next evaluation interval, the next remaining useful lifetime for the one or more components of the electrical machine by utilizing the prediction capability of the artificial intelligence model(s) and / or determining the remaining useful lifetime using the physics-based model.

[0102] In certain embodiments, the method 1 100 can be repeated as desired, which can be on a continuous basis, periodic basis, or at designated times. Notably, the method 1100 can incorporate any of the other functionality as described herein and can be adapted to support the functionality of the system 1000, the electric machine 100, the sensor device 220, and / or other components described in the present disclosure. In certain embodiments, functionality of the method 1100 can be combined with other methods and / or functionality described in the present disclosure. In certain embodiments, certain operations of the method 1100 can be replaced with other functionality of the present disclosure and the sequence of operations can be adjusted as desired.

[0103] Referring now also to Figure 12, at least a portion of the methodologies and techniques described with respect to the exemplary embodiments of the electric machine 100, sensor device 220, system 1000, and / or methods 300, 400, 500, 700, 800, 900, 1200 can incorporate a machine, such as, but not limited to, computer system 1200, or other computing device within which a set of instructions, when executed, can cause the machine to perform any one or more of the methodologies or functions discussed above. The machine can be configured to facilitate various operations conducted by the electric machine 100, sensor device 220, system 1000, and / or methods 300, 400, 500, 700, 800, 900, 1200. For example, the machine can be configured to, but is not limited to, assist the system 1000 by providing processing power to assist with processing loads experienced in the system 1000, by providing storage capacity for storing instructions or data traversing the system 1000, or by assisting with any other operations conducted by or within the system 1000. As another example, in certain embodiments, the computer system 1200 can assist in obtaining signals, such as during an evaluation interval, from sensors that include sensor data associated with an electric machine 100; determining and / or predicting the remaining useful lifetime of components of the electric machine 100 and / or the electric machine 100 as a whole based on the sensor data, such as by utilizing physics-based models and / or machine learning models; outputting the determined remaining useful lifetime via an interface; providing rewards to the artificial intelligence model(s) in accordance with an accuracy of the remaining useful lifetime prediction to modify the prediction capability of the model; initiating or scheduling a repair or replacement of a component of the electric machine 100 and / or components of the electric machine 100; training artificial intelligence / machine learning models based on previous determinations, predictions, and / or the accuracy; receiving additional sensor data during a nextevaluation interval; predicting, for the next evaluation interval, the next remaining useful lifetime for the electric machine and / or components of the electric machine 100; and / or performing any other operations of the system 1200, the sensor device 220, the electric machine 100, or a combination thereof. In certain embodiments, the computer system 1200 can be configured to assist in facilitating communications between electric machines 100, sensor devices 220, and / or communication networks 1035, performing any other operations, or a combination thereof.

[0104] In some embodiments, the machine can operate as a standalone device. In some embodiments, the machine can be connected (e.g., using communications network 1035, another network, or a combination thereof) to and assist with operations performed by other machines and systems, such as, but not limited to, the communications network 1033, the communications network 1034, the communications network 1035, the server 1040, the server 1045, the server 1050, the server 1060, the database 1055, the electric machines 100, the sensor devices 220, any other system, program, and / or device, or any combination thereof. The machine can be connected with any component in the system 1000. In a networked deployment, the machine can operate in the capacity of a server or a client user machine in a server-client user network environment, or as a peer machine in a peer-to-peer (or distributed) network environment. The machine can comprise a server computer, a client user computer, a personal computer (PC), a tablet PC, a laptop computer, a desktop computer, a control system, a network router, switch or bridge, or any machine capable of executing a set of instructions (sequential or otherwise) that specify actions to be taken by that machine. Further, while a single machine is illustrated, the term “machine” shall also be taken to include any collection of machines that individually or jointly execute a set (or multiple sets) of instructions to perform any one or more of the methodologies discussed herein.

[0105] The computer system 1200 can include a processor 1202 (e.g., a central processing unit (CPU), a graphics processing unit (GPU, or both), a main memory 1204 and a static memory 1206, which communicate with each other via a bus 1208. The computer system 1200 can further include a video display unit 1210, which can be, but is not limited to, a liquid crystal display (LCD), a flat panel, a solid-state display, or a cathode ray tube (CRT). The computer system 1200 can include an input device 1212, such as, but not limited to, a keyboard, a cursor control device 1214, such as, but not limited to, a mouse, a disk drive unit 1216, a signal generation device 1218, such as, but not limited to, a speaker or remote control, and a network interface device 1220.

[0106] The disk drive unit 1216 can include a machine-readable medium 1222 on which isstored one or more sets of instructions 1224, such as, but not limited to, software embodying any one or more of the methodologies or functions described herein, including those methods illustrated above. The instructions 1224 can also reside, completely or at least partially, within the main memory 1204, the static memory 1206, or within the processor 1202, or a combination thereof, during execution thereof by the computer system 1200. The main memory 1204 and the processor 1202 also can constitute machine-readable media.

[0107] Dedicated hardware implementations including, but not limited to, application specific integrated circuits, programmable logic arrays and other hardware devices can likewise be constructed to implement the methods described herein. Applications that can include the apparatus and systems of various embodiments broadly include a variety of electronic and computer systems. Some embodiments implement functions in two or more specific interconnected hardware modules or devices with related control and data signals communicated between and through the modules, or as portions of an application-specific integrated circuit. Thus, the example system is applicable to software, firmware, and hardware implementations.

[0108] In accordance with various embodiments of the present disclosure, the methods described herein are intended for operation as software programs running on a computer processor. Furthermore, software implementations can include, but not limited to, distributed processing or component / object distributed processing, parallel processing, or virtual machine processing can also be constructed to implement the methods described herein.

[0109] The present disclosure contemplates a machine-readable medium 1222 containing instructions 1224 so that a device connected to the communications network 1033, the communications network 1034, the communications network 1035, another network, or a combination thereof, can send or receive voice, video or data, and communicate over the communications network 1035, another network, or a combination thereof, using the instructions. The instructions 1224 can further be transmitted or received over the communications network 1033, the communications network 1034, the communications network 1035, another network, or a combination thereof, via the network interface device 1220.

[0110] While the machine-readable medium 1222 is shown in an example embodiment to be a single medium, the term "machine-readable medium" should be taken to include a single medium or multiple media (e.g., a centralized or distributed database, and / or associated caches and servers) that store the one or more sets of instructions. The term "machine-readable medium" shall also betaken to include any medium that is capable of storing, encoding or carrying a set of instructions for execution by the machine and that causes the machine to perform any one or more of the methodologies of the present disclosure.

[0111] The terms "machine-readable medium," "machine-readable device," or "computer- readable device" shall accordingly be taken to include, but not be limited to: memory devices, solid-state memories such as a memory card or other package that houses one or more read-only (non-volatile) memories, random access memories, or other re-writable (volatile) memories; magneto-optical or optical medium such as a disk or tape; or other self-contained information archive or set of archives is considered a distribution medium equivalent to a tangible storage medium. The "machine-readable medium," "machine-readable device," or "computer-readable device" can be non-transitory, and, in certain embodiments, cannot include a wave or signal per se. Accordingly, the disclosure is considered to include any one or more of a machine-readable medium or a distribution medium, as listed herein and including art-recognized equivalents and successor media, in which the software implementations herein are stored.The illustrations of arrangements described herein are intended to provide a general understanding of the structure of various embodiments, and they are not intended to serve as a complete description of all the elements and features of apparatus and systems that might make use of the structures described herein. Other arrangements can be utilized and derived therefrom, such that structural and logical substitutions and changes can be made without departing from the scope of this disclosure. Figures are also merely representational and cannot be drawn to scale. Certain proportions thereof can be exaggerated, while others can be minimized. Accordingly, the specification and drawings are to be regarded in an illustrative rather than a restrictive sense.

[0112] Thus, although specific arrangements have been illustrated and described herein, it should be appreciated that any arrangement calculated to achieve the same purpose can be substituted for the specific arrangement shown. This disclosure is intended to cover any and all adaptations or variations of various embodiments and arrangements of the invention. Combinations of the above arrangements, and other arrangements not specifically described herein, will be apparent to those of skill in the art upon reviewing the above description. Therefore, it is intended that the disclosure is not limited to the particular arrangement(s) disclosed as the best mode contemplated for carrying out this invention, but that the invention will include all embodiments and arrangements falling within the scope of the appended claims.

[0113] The foregoing is provided for purposes of illustrating, explaining, and describing embodiments of this invention. Modifications and adaptations to these embodiments will be apparent to those skilled in the art and can be made without departing from the scope or spirit of this invention. Upon reviewing the aforementioned embodiments, it would be evident to an artisan with ordinary skill in the art that said embodiments can be modified, reduced, or enhanced without departing from the scope and spirit of the claims described below.

[0114] At least some aspects of the present disclosure will now be described with reference to the following numbered clauses.

[0115] Clause 1 : A system, may include an electric machine, and may further include; at least one sensor configured to capture sensor data associated with at least one component of the electrical machine; and at least one processor configured to: receive, during an evaluation interval, at least one signal from the at least one sensor, wherein the at least one signal comprises the sensor data; determine, based on the sensor data and by utilizing a physics-based model, at least one input parameter for use in calculating a remaining useful lifetime for the at least one component of the electrical machine; calculate, by utilizing the physics-based model, the remaining useful lifetime for the at least one component of the electrical machine, wherein the remaining useful lifetime for the at least one component is calculated based on the at least one input parameter, at least one assumption for at least one future operating load and environmental condition, and at least one electrical machine specification specified for the electrical machine; and output the remaining useful lifetime via an interface.

[0116] Clause 2: The system of clause 1, wherein the assumption for the at least one future operating load and environmental condition are an extrapolation of measured load and environmental condition histograms; and wherein the at least one assumption for the at least one future operating load and environmental condition comprises an extrapolation of the measured load and environmental condition histograms.

[0117] Clause 3: The system of clause 1 or 2, wherein the at least one processor is further configured to initiate an evaluation interval for evaluating the remaining useful lifetime of the electrical machine, the at least one component of the electrical machine, or a combination thereof; and wherein the at least one processor is further configured to assume electrical machine load histograms and environmental condition histograms for at least one period of time during which the evaluation interval was not running.

[0118] Clause 4: The system of clause 1, 2, or 3, wherein the at least one processor is further configured to determine a shortest remaining useful lifetime for a component of the at least one component by comparing each remaining useful lifetime for each component of the at least one component.

[0119] Clause 5: The system of clause 1, 2, 3, or 4, output the shortest remaining useful lifetime as the remaining useful lifetime of the electrical machine via the interface; and provide a signal to indicate whether a component of the at least one component having the shortest remaining lifetime is to be repaired or replaced.

[0120] Clause 6: The system of clause 1, 2, 3, 4, or 5, wherein the sensor data comprise operating parameters comprising a current draw associated with the at least one component, a phase voltage and imbalance associated with the at least one component, a temperature associated with the at least one component, an acoustic pattern associated with the at least one component, a temperature pattern associated with the at least one component, an air gap flux associated with the electrical machine, stray flux associated with the electrical machine, a housing vibration associated with the electrical machine, strain and stress on a body of the electrical machine, or a combination thereof.

[0121] Clause 7: The system of clause 1, 2, 3, 4, 5, or 6, wherein the at least one processor is further configured to: initiate a next evaluation interval; receive at least one additional signal from the at least one sensor comprising additional sensor data; determine, based on the additional sensor data and by utilizing the physics-based model, at least one new input parameter for use in calculating a current remaining useful lifetime for the at least one component of the electrical machine; and calculate the current remaining useful lifetime for the at least one component of the electrical machine based on the at least one new input parameter, the remaining useful lifetime from the evaluation interval, and the at least one electrical machine specification.

[0122] Clause 8: The system of clause 1, 2, 3, 4, 5, 6, or 7, wherein the at least one processor is further configured to train an artificial intelligence model to generate a remaining useful lifetime prediction for the electrical machine, the at least one component, or a combination thereof, wherein the artificial intelligence model is trained based on the sensor data, the at least one input parameters, the at least one electrical machine specification, at least one electrical machine specification for another electrical machine having a correlation to the electrical machine, or a combination thereof.

[0123] Clause 9: The system of clause 1, 2, 3, 4, 5, 6, 7, or 8, wherein the at least one processor is further configured to adjust the remaining useful lifetime prediction generated by the artificial intelligence model based on reward function inputs and the remaining useful lifetime determined using the physics-based model.

[0124] Clause 10: A system may include: an electrical machine; at least one sensor configured to capture sensor data associated with at least one component of the electrical machine; and at least one processor configured to: receive, during an evaluation interval, at least one signal from the at least one sensor, wherein the at least one signal comprises the sensor data; determine, by utilizing an artificial intelligence model and based on at least one manufacturing parameter associated with the electrical machine, a comparable electrical machine that is comparable to the electrical machine, wherein historical data associated with the comparable electrical machine is utilized to train the artificial intelligence model to predict a remaining useful lifetime for the at least one component of the electrical machine; predict, by utilizing the artificial intelligence model, the remaining useful lifetime for the at least one component of the electrical machine, wherein the remaining useful lifetime for the at least one component is calculated based on the sensor data being compared with the historical data associated with the comparable electrical machine; and output the remaining useful lifetime predicted by the artificial intelligence model via an interface.

[0125] Clause 11 : The system of clause 10, wherein the at least one processor is further configured to cumulate prior damage to the at least one component of the electrical machine determined from a prior evaluation interval with current damage determined during the evaluation interval when determining the remaining useful lifetime for the at least one component of the electrical machine.

[0126] Clause 12: The system of clause 10 or 11, wherein the at least one processor is further configured to select a strategy from a plurality of strategies for determining the remaining useful lifetime of the at least one component of the electrical machine.

[0127] Clause 13: The system of clause 10, 11, or 12, wherein the at least one processor is further configured to weight a first portion of the sensor data over a second portion of the sensor data when determining the remaining useful lifetime of the at least one component.

[0128] Clause 14: The system of clause 10, 11, 12, or 13, wherein the at least one processor is further configured to determine an accuracy of the remaining useful lifetime predicted by the artificial intelligence model based on a comparison with a result of an inspection of the at least onecomponent of the electrical machine.

[0129] Clause 15: The system of clause 10, 11, 12, 13, or 14, wherein the at least one processor is further configured to provide a reward to the artificial intelligence model in accordance with the accuracy.

[0130] Clause 16: The system of clause 10, 11, 12, 13, 14, or 15, wherein the at least one processor is further configured to utilize at least one industry standard associated with the electrical machine, the at least one component, or a combination thereof, to determine the remaining useful lifetime.

[0131] Clause 17: The system of clause 10, 11, 12, 13, 14, 15, or 16, wherein the at least one processor is further configured to optimize an electrical machine maintenance schedule for the electrical machine, the at least one component, or a combination thereof, based on the remaining useful lifetime.

[0132] Clause 18: The system of clause 10, 11, 12, 13, 14, 15, 16, or 17, wherein the at least one processor is further configured to provide an electrical machine health indication associated with the electrical machine, the at least one component, or a combination thereof, via the interface and based on the remaining useful lifetime.

[0133] Clause 19: A method comprising: receiving, during an evaluation interval, at least one signal from at least one sensor, wherein the at least one signal comprises sensor data associated with an electrical machine; predicting, by utilizing an artificial intelligence model, a remaining useful lifetime for at least one component of the electrical machine, wherein the remaining useful lifetime for the at least one component is calculated based on the sensor data; outputting the remaining useful lifetime predicted by the artificial intelligence model via an interface; providing a reward to the artificial intelligence model in accordance with an accuracy of the remaining useful lifetime prediction to modify a prediction capability of the artificial intelligence model; and predicting, for a next evaluation interval, a next remaining useful lifetime for the at least one component of the electrical machine by utilizing the prediction capability.

[0134] Clause 20: The method of clause 19, further comprising scheduling a repair or replacement of the at least one component based on the remaining useful lifetime indicating onset of a failure of the at least one component.

Claims

CLAIMSWhat is claimed is:

1. A system, comprising: an electrical machine; at least one sensor configured to capture sensor data associated with at least one component of the electrical machine; and at least one processor configured to: receive, during an evaluation interval, at least one signal from the at least one sensor, wherein the at least one signal comprises the sensor data; determine, based on the sensor data and by utilizing a physics-based model, at least one input parameter for use in calculating a remaining useful lifetime for the at least one component of the electrical machine; calculate, by utilizing the physics-based model, the remaining useful lifetime for the at least one component of the electrical machine, wherein the remaining useful lifetime for the at least one component is calculated based on the at least one input parameter, at least one assumption for at least one future operating load and environmental condition, and at least one electrical machine specification specified for the electrical machine; and output the remaining useful lifetime via an interface.

2. The system of claim 1, wherein the assumption for the at least one future operating load and environmental condition are an extrapolation of measured load and environmental condition histograms; and wherein the at least one assumption for the at least one future operating load and environmental condition comprises an extrapolation of the measured load and environmental condition histograms.

3. The system of claim 1, wherein the at least one processor is further configured to initiate an evaluation interval for evaluating the remaining useful lifetime of the electrical machine,the at least one component of the electrical machine, or a combination thereof; and wherein the at least one processor is further configured to assume electrical machine load histograms and environmental condition histograms for at least one period of time during which the evaluation interval was not running.

4. The system of claim 1, wherein the at least one processor is further configured to calculate an updated remaining useful lifetime for the at least one component based on calculating incremental damage to the at least one component based on the at least one input parameter and at least one prior damage value calculated for the at least one component during a prior evaluation interval.

5. The system of claim 1 , wherein the at least one processor is further configured to determine a shortest remaining useful lifetime for a component of the at least one component by comparing each remaining useful lifetime for each component of the at least one component.

6. The system of claim 5, wherein the at least one processor is configured to: output the shortest remaining useful lifetime as the remaining useful lifetime of the electrical machine via the interface; and provide a signal to indicate whether a component of the at least one component having the shortest remaining useful lifetime is to be repaired or replaced.

7. The system of claim 1, wherein the sensor data comprise operating parameters comprising a current draw associated with the at least one component, a phase voltage and imbalance associated with the at least one component, a temperature associated with the at least one component, an acoustic pattern associated with the at least one component, a temperature pattern associated with the at least one component, an air gap flux associated with the electrical machine, stray flux associated with the electrical machine, a housing vibration associated with the electrical machine, strain and stress on a body of the electrical machine, or a combination thereof.

8. The system of claim 1, wherein the at least one processor is further configured to: initiate a next evaluation interval;receive at least one additional signal from the at least one sensor comprising additional sensor data; determine, based on the additional sensor data and by utilizing the physics-based model, at least one new input parameter for use in calculating a current remaining useful lifetime for the at least one component of the electrical machine; and calculate the current remaining useful lifetime for the at least one component of the electrical machine based on the at least one new input parameter, the remaining useful lifetime from the evaluation interval, and the at least one electrical machine specification.

9. The system of claim 1, wherein the at least one processor is further configured to train an artificial intelligence model to generate a remaining useful lifetime prediction for the electrical machine, the at least one component, or a combination thereof, wherein the artificial intelligence model is trained based on the sensor data, the at least one input parameter, the at least one electrical machine specification, at least one electrical machine specification for another electrical machine having a correlation to the electrical machine, or a combination thereof.

10. The system of claim 9, wherein the at least one processor is further configured to adjust the remaining useful lifetime prediction generated by the artificial intelligence model based on reward function inputs and the remaining useful lifetime determined using the physics-based model.

11. A system comprising: an electrical machine; at least one sensor configured to capture sensor data associated with at least one component of the electrical machine; and at least one processor configured to: receive, during an evaluation interval, at least one signal from the at least one sensor, wherein the at least one signal comprises the sensor data; determine, by utilizing an artificial intelligence model and based on at least one manufacturing parameter associated with the electrical machine, a comparable electrical machine that is comparable to the electrical machine, wherein historical data associatedwith the comparable electrical machine is utilized to train the artificial intelligence model to predict a remaining useful lifetime for the at least one component of the electrical machine; predict, by utilizing the artificial intelligence model, the remaining useful lifetime for the at least one component of the electrical machine, wherein the remaining useful lifetime for the at least one component is calculated based on the sensor data being compared with the historical data associated with the comparable electrical machine; and output the remaining useful lifetime predicted by the artificial intelligence model via an interface.

12. The system of claim 11, wherein the at least one processor is further configured to cumulate prior damage to the at least one component of the electrical machine determined from a prior evaluation interval with current damage determined during the evaluation interval when determining the remaining useful lifetime for the at least one component of the electrical machine.

13. The system of claim 11, wherein the at least one processor is further configured to select a strategy from a plurality of strategies for determining the remaining useful lifetime of the at least one component of the electrical machine.

14. The system of claim 11, wherein the at least one processor is further configured to weight a first portion of the sensor data over a second portion of the sensor data when determining the remaining useful lifetime of the at least one component.

15. The system of claim 11, wherein the at least one processor is further configured to determine an accuracy of the remaining useful lifetime predicted by the artificial intelligence model based on a comparison with a result of an inspection of the at least one component of the electrical machine.

16. The system of claim 15, wherein the at least one processor is further configured to provide a reward to the artificial intelligence model in accordance with the accuracy.

17. The system of claim 1 1 , wherein the at least one processor is further configured to utilize at least one industry standard associated with the electrical machine, the at least one component, or a combination thereof, to determine the remaining useful lifetime.

18. The system of claim 11, wherein the at least one processor is further configured to optimize an electrical machine maintenance schedule for the electrical machine, the at least one component, or a combination thereof, based on the remaining useful lifetime.

19. The system of claim 11, wherein the at least one processor is further configured to provide an electrical machine health indication associated with the electrical machine, the at least one component, or a combination thereof, via the interface and based on the remaining useful lifetime.

20. A method, comprising: receiving, during an evaluation interval, at least one signal from at least one sensor, wherein the at least one signal comprises sensor data associated with an electrical machine; predicting, by utilizing an artificial intelligence model, a remaining useful lifetime for at least one component of the electrical machine, wherein the remaining useful lifetime for the at least one component is calculated based on the sensor data; outputting the remaining useful lifetime predicted by the artificial intelligence model via an interface; providing a reward to the artificial intelligence model in accordance with an accuracy of the remaining useful lifetime predicted to modify a prediction capability of the artificial intelligence model; and predicting, for a next evaluation interval, a next remaining useful lifetime for the at least one component of the electrical machine by utilizing the prediction capability.

Images