Failure rate estimates for powertrain / machine components based on usage

Abstract

Failure rate estimates for power system / machine components based on usage. Models are created to predict failure rates for power system / machine components based on operational data. Risk scores are generated based on how often the component has experienced one or more damaging duty cycles. The risk scores can be used in mathematical models to determine which component or components are expected to fail prior to warranty expiration so that the components can be proactively replaced to save warranty costs and reduce downtime.

Description

Background Art

[0001] The present application generally relates to a method for estimating the failure rate of a power system / machine component based on customer usage. Some systems use algorithms that obtain health status information from one or more components and determine the functional status of the one or more components and / or the overall system. Some systems estimate the health status by creating an expected operating signature of one or more operating parameters of the component and comparing the actual operating signature to the expected operating signature to determine whether the component deviates from normal. Other systems use comprehensive data to determine residual value based on mileage and then apply the residual value to the health status consideration.

[0002] However, these systems do not fully consider actual component usage to identify one or more destructive duty cycles and determine the state of machine / powertrain components or predict component failure by the frequency of operation of a particular component in a destructive duty cycle. Therefore, there is still a great need for improvement in this technical field.

[0003] Disclosure of Illustrative Embodiments

[0004] For the purpose of clearly, concisely and accurately describing the illustrative embodiments of the disclosure, the manner and process of making and using the disclosure, and to enable the practice, manufacture and use of the disclosure, reference will now be made to certain exemplary embodiments, including those illustrated in the drawings, and specific language will be used to describe the disclosure. However, it should be understood that no limitation of the scope of the invention is thereby created, and the present invention includes and protects such changes, modifications and further applications of the exemplary embodiments as will occur to those skilled in the art. Summary of the invention

[0005] A system method for identifying destructive duty cycles for one or more power system / machine components from operational data is disclosed herein. Certain exemplary embodiments include one or more unique processes for determining the effects of one or more component parameters on one or more components in a large sample through multivariate regression analysis.

[0006] Certain exemplary embodiments include one or more unique processes for creating a risk score transfer function based on regression analysis and identifying a risk score based on how often a component has experienced a type of damaging duty cycle. Certain exemplary embodiments include one or more unique processes for generating a failure model based on the risk score and the component age or mileage associated with component failure. Certain exemplary embodiments include one or more unique processes for monitoring component failure rates through telematics or repair tools, and determining whether certain powertrain / machine components are expected to fail before the warranty period expires. Recalls or other proactive measures can be initiated before failures occur to save warranty costs and reduce inconvenience.

[0007] Other embodiments, forms, objects, features, advantages, aspects and benefits will become apparent from the following description and accompanying drawings. BRIEF DESCRIPTION OF THE DRAWINGS

[0008] Figure 1 is a schematic diagram illustrating certain aspects of an example system according to the present disclosure.

[0009] Figure 2 is a flow chart illustrating certain aspects of an exemplary process according to the present disclosure.

[0010] Figure 3 is a flow chart illustrating certain aspects of an exemplary process according to the present disclosure.

[0011] Figure 4 are diagrams illustrating certain aspects of exemplary embodiments of the present disclosure. DETAILED DESCRIPTION

[0012] The following discussion relates to an exemplary embodiment of a system process for determining a model and / or predicting a failure rate of a powertrain / machine component based on operational data and failure data. The operational data may be collected via a telematics device or during a maintenance event, such as during maintenance of a vehicle in which the component resides. A multivariate regression analysis may be performed on the operational data to determine the primary impact of each component parameter on a large population of machines / powertrains.

[0013] Destructive operating cycles can be defined based on the identified critical operating parameters. Subsequently, a fault model can be generated based on the destructive cycle information and a database of component failures to identify expected failures of components. After the failure model is determined, the risk of component failure can be monitored in real time through telematics services. If such services are not available, the same functionality can be achieved by extracting component data during maintenance events. If a component is expected to fail before the warranty period, the component can be proactively replaced to save warranty costs and reduce downtime.

[0014] One or more systems and subsystems of the powertrain / machine may include a database for collecting data associated with the operation of one or more parameters of a component. The data may be used in a system approach such as multivariate analysis to determine one or more destructive duty cycles for one or more components at which the one or more components are susceptible to failure.

[0015] The data can be further used, for example, in a regression analysis to create a risk score transfer function. Subsequently, a risk score for one or more components can be generated based on the type of destructive duty cycle. The risk score determination can also include the frequency with which one or more components undergo destructive cycles. The risk score and age / use can be used to determine a mathematical model associated with the failure of one or more components. Subsequently, the use / age of one or more components can also be monitored via telematics or service to support field action plans. For example, for field actions, instead of replacing an entire group, the risk score can be used to determine which components within the group to replace. Without the ability to monitor and assess the risk of failure, field actions to replace an entire group of components may be required, which can significantly increase costs. Using risk scores to selectively replace components at risk of failure reduces costs compared to replacing the entire group.

[0016] Atypical failure behavior of powertrain / machine components can be predicted by identifying destructive duty cycles with multivariate analysis using telematics data and warranty records. A risk score transfer function can be created based on regression analysis, and a risk score is generated based on the type and frequency of destructive cycles. A mathematical model can be determined based on the risk score and the age, usage or mileage of the failed component. Component usage can be monitored and analyzed based on the mathematical model from usage data collected through telematics or repair tools, and high-risk components can be recalled before catastrophic failure of the monitored component.

[0017] An exemplary process includes a controller configured to perform certain operations and receive and interpret signals from any component of the power system / machine and any sensor, which may or may not be associated with the component. It should be understood that the controller can be provided in various forms and configurations, including one or more computing devices forming the whole or part of a processing subsystem having non-transitory memory storing computer executable instructions, processing and communication hardware. The controller can be a single device or a distributed device, and the functions of the controller can be performed by hardware or software. The controller communicates with any actuator, sensor, data link, computing device, wireless connection, or other device capable of performing any of the described operations.

[0018] Processing logic can be implemented as a module, which can be implemented in the operating logic as an operation performed by software, hardware, artificial intelligence, fuzzy logic, or any combination thereof, or at least partially by a user or operator. In certain embodiments, the module represents the software element as a computer program encoded on a computer-readable medium, and the computer performs the described operation when executing the computer program. A module can be a single device, distributed across multiple devices, and / or a module can be grouped with other modules or devices in whole or in part. The operation of any module can be performed in whole or in part in hardware / software or by other modules.

[0019] The controller includes stored data values, constants and functions, and operating instructions stored on a computer readable medium. Any operation of the exemplary process described herein may be performed at least in part by the controller. It should be understood that other groupings that perform similar overall operations are within the scope of this application. A more specific description of certain embodiments of the controller operation is discussed herein. The operations shown are understood to be exemplary only, and the operations may be combined or divided, and added or removed, and reordered in whole or in part.

[0020] The exemplary process may include various powertrain / machine sensors and sensor systems, each operable to provide a signal to the controller indicating a parameter of a given component. It should be appreciated that the sensors and sensor arrangements of the various embodiments are applicable to the disclosed principles and techniques. It should be further appreciated that the sensors and sensor systems utilized may be physical sensors, virtual sensors, and / or combinations thereof.

[0021] The controller is operably coupled to the memory and is configured to store instructions in the memory, which can be read and executed by the controller to control the operation of the exemplary process as described herein. Certain operations described herein include operations for determining one or more parameters. Determination as used herein includes operation or calculation of values, obtaining values ​​from a lookup table or using a lookup operation, receiving values ​​from a data link or network communication, receiving electronic signals (e.g., voltage, frequency, current, or pulse width modulation (PWM) signals) indicating values, receiving software parameters indicating values, reading values ​​from memory locations on computer-readable media, by any means known in the art, and / or by receiving values ​​that can be interpreted as parameters by its operation, and / or by receiving values ​​as runtime parameters by reference to default values ​​interpreted as parameter values.

[0022] The controller can be an independent computer processing system, or an instance of a component of an integrated circuit-based electronic control system (ECS), which can be configured to control various operational aspects of the exemplary processes as described in further detail herein. The ECS according to the present disclosure can be implemented in various forms and can include a plurality of different elements and configurations of elements. In some forms, the ECS can be incorporated with one or more microprocessor-based or microcontroller-based electronic control units, which are sometimes referred to as electronic control modules. The ECS according to the present disclosure can be provided in a form having a single processing or computing component, or in a form including a plurality of operably coupled processing or computing components; and can include digital circuits, analog circuits, or a hybrid combination of these two types. The integrated circuits of the ECS and / or any of its constituent processors / controllers or other components may include one or more signal conditioners, modulators, demodulators, arithmetic processing units (ALUs), central processing units (CPUs), limiters, oscillators, control clocks, amplifiers, signal conditioners, filters, format converters, communication ports, clamps, delay devices, memory devices, analog-to-digital (A / D) converters, digital-to-analog (D / A) converters, and / or different circuits or functional components that will occur to those skilled in the art for providing and performing the communication and control aspects disclosed herein.

[0023] refer to Figure 1 , illustrates a schematic diagram showing certain aspects of an exemplary power system / machine system 100. The system 100 includes a vehicle and / or machine 102 having a power system 104 and a controller 110. The power system 104 may be used in a stationary application or vehicle, for example, having a generator or a machine or vehicle. The system 100 collects operational data 122 for one or more machine components 108 and / or power system components 106 during a first time period. The controller 110, or a separate computer processor or server analyzes the operational data 122 to identify one or more parameters 112 associated with the components 106, 108. For example, the controller 110 or other computer processor determines a threshold 116 associated with the speed and / or torque of the engine, or other criteria, an associated destructive duty cycle 118 of the one or more components 106, 108 based on the identified one or more parameters 112. The controller 110 or other computer processor determines a failure behavior 120 of the one or more components 106, 108 through a model based on the destructive duty cycle 118 and age of the components 106, 108. The controller 110 or other processor compares the actual usage / duty cycle of the components 106, 108 in one or more machines and / or vehicles 102 during the second time period with the failure behavior model 120 to predict component failure based on the usage data. The user can use this comparison to select which components 106, 108 to replace before failure occurs.

[0024] The operation data 122 of the machine and / or vehicle 102 is collected by one of the telematics device 130 or the repair tool module 132, for example, in communication with the remote server 134. The controller 110 or the remote server 134 performs a suitable analysis, such as a multivariate regression analysis, to identify one or more parameters 112 indicating a failure of the components 106, 108 in a sample population from one or more machines and / or vehicles. The one or more parameters 112 identified may include, for example, braking frequency, speed changes of powertrain components exceeding a threshold amount, and torque of powertrain components exceeding a threshold amount. The controller 110 or other processor may compare the actual use of the components 106, 108 with the current mileage of the machine and / or vehicle 102 and the end-of-warranty mileage of the components 106, 108. During a maintenance event from the maintenance event module 132, the actual use data may be collected in real time by the telematics device 130 or from the controller 110. Comparing the actual use may include determining a risk score 114 of the components 106, 108 based on one or more of the parameters 112. The risk score may also be based on the frequency with which the components 106 , 108 experience or operate within one or more destructive duty cycles.

[0025] refer to Figure 2 , illustrates a flow chart showing certain aspects of an exemplary process 200. Process 200 begins at operation 202, which includes collecting powertrain component data and vehicle / machine component data (collectively and also individually referred to as component data) from a vehicle / machine controller or service port module during a maintenance event, or providing such as via telematics of powertrain or vehicle / machine component history. Component data may be collected via signals from sensors that may be used to provide operational data of the component. Component data may be stored to and retrieved from a memory of the controller 110 or other onboard storage device, or transmitted for storage in a remote server.

[0026] Process 200 continues from operation 202 at operation 204 to define key parameters for multivariate analysis that may include component data. Sensitivity analysis may also be used to filter irrelevant parameters. Multivariate and sensitivity analysis may determine how changes in process output may be allocated to different input parameters, and how changes in input parameters affect changes in output. There are multiple applications of this type of analysis to utilize inherent knowledge of the data to quantify process uncertainty, optimize the design of the process, and rank the impact of various parameters of the process. This analysis may provide a comprehensive understanding of the impact of different input parameters and their changes on process results.

[0027] Process 200 continues from operation 204 at operation 206 to create a risk score transfer function by utilizing regression analysis for risk score generation. Regression analysis is a set of statistical procedures used to estimate the relationship between variables, such as the parameters identified as key parameters in operation 204, and failure of one or more components. Process 200 continues from operation 206 at operation 208, which includes building a mathematical model to model the failure behavior of the component by using the risk score and the component age and / or use.

[0028] Process 200 continues from operation 208 at operation 210, which includes receiving powertrain data collected at operation 202 during a period of time after the mathematical model is established. Operation 210 uses the mathematical model from operation 208 to predict the probability of failure of one or more components from one or more vehicles and / or machines based on actual use of the components. Process 200 provides a method for defining a destructive duty cycle of one or more components using a large amount of data, and estimating the unreliability or potential failure of the components based on actual customer use and / or age of the components. The prediction can be used to identify potential failure areas.

[0029] Process 200 continues from operation 210 at operation 212 where a field action plan is determined in response to assigning a risk score and failure rate to the entire population. The field action plan may include repairing, replacing, servicing, or otherwise responding to conditions that predict failure of one or more components based on actual use and / or age of the components and the mathematical model. The field action plan may be based on the risk of component failure and / or predicted failure timing.

[0030] refer to Figure 3 , a block diagram depicting certain aspects of an exemplary process 300 is shown, which may be implemented in one or more control components, such as the electronic control system described above, a remote server, or a combination thereof. The process 300 illustrates an embodiment method for determining a risk score 324 for a component. The risk score 324 is determined from a sum 322 of three transfer functions associated with the standard deviation of the mobile eRPM 312, the high torque output 314, and the braking frequency 316, respectively. The standard deviation of the mobile eRPM 312 and the high torque output 314 are both based on inputs 302 from a duty cycle map of the component. The braking frequency 316 is determined from an input 304 of bicycle brake actuation and an input 308 from an engine control module (ECM) distance. An estimate of current mileage 318 and an estimate of end-of-warranty mileage 320 may also be determined from the ECM distance input 308 and an input 310 regarding a start-of-service date. Figure 3The process 300 in is merely an example, and the identification of key parameters is not limited to standard deviation of moving eRPM, high torque, and braking frequency.

[0031] refer to Figure 4 , a diagram 400 is shown, which provides a method for using Figure 2 process and Figure 3 An example of determining the need and priority of replacing one or more components based on a risk score of. Graph 400 depicts usage in kilometers over time (t) on its horizontal axis and failure rate or probability as a function of unreliability as F(t)=100-R(t)(%) on its vertical axis. Graph 400 depicts the probability or likelihood of determining a powertrain component replacement with unreliability as a failure, the estimated end-of-warranty mileage 402, the estimated current mileage 404, and the risk score 406. The process of determining the need and priority of component replacement may include: 1) selecting a group based on the risk score; 2) estimating the current and end-of-warranty failure rates; 3) determining the need for component replacement based on the end-of-warranty mileage; and 4) determining the timing of component replacement based on the current mileage.

[0032] An exemplary embodiment of a process for determining a prediction of a powertrain component failure based on a duty cycle analysis is disclosed. The process includes predicting a failure rate using a risk score identified by the duty cycle analysis. Each component may include a risk score calculated from, for example, the percentage of high torque time, brake density, and standard deviation of moving engine rpm. The probability of component failure may be a function of mileage and risk score. Different groups with different risk scores may have different failure rates.

[0033] Various aspects of the present disclosure are considered. According to one aspect, a method for estimating failure of a machine / power system component includes: determining a destructive duty cycle for the component based on one or more parameters identified for the component of the machine / power system; modeling a failure behavior of the component based on the destructive duty cycle of the component and at least one of age and use; comparing at least one of an actual age or actual use of a second component to the model of the failure behavior; and replacing the second component based on the comparison.

[0034] In one embodiment, the one or more identified parameters include one or more of a braking frequency, a speed change of the machine component exceeding a threshold amount, and a torque of the machine component exceeding a threshold amount. In another embodiment, the comparison further includes comparing the actual usage of the second component with the current mileage and the end-of-warranty mileage of the component.

[0035] In another embodiment, actual usage data is collected in real time during the maintenance event via telematics or from an onboard computer system of the vehicle. In another embodiment, the comparison includes determining a risk score for the second component based on the destructive duty cycle, the risk score including one or more of a braking frequency, a change in speed of the second component exceeding a threshold amount, and a torque of the second component exceeding a threshold amount, and the risk score is further based on a frequency with which the second component experiences the destructive duty cycle.

[0036] In another aspect, a method for estimating failure of a machine / power system component includes: collecting operating data of a component of the machine / power system during a first time period; analyzing the operating data to identify one or more parameters associated with failure of the component; determining a destructive duty cycle of the component based on the identified one or more parameters; modeling the failure behavior of the component based on the destructive duty cycle and usage of the component; comparing actual usage of a second component during a second time period with the model of the failure behavior; and replacing the second component based on the comparison.

[0037] In one embodiment, the operational data is collected via one of a telematics device and a service portal module. In another embodiment, the analyzing includes performing a multivariate regression analysis to identify one or more parameters indicative of a failure of a component in the sample population.

[0038] In another embodiment, the one or more parameters identified include one or more of a braking frequency, a speed change of the machine component exceeding a threshold amount, and a torque of the machine component exceeding a threshold amount. In another embodiment, the comparing further includes comparing the actual usage of the second component to the current mileage and the end-of-warranty mileage of the component.

[0039] In another embodiment, the actual usage data is collected in real time via telematics. In another embodiment, the actual usage data is collected from an onboard computer system of the vehicle during a maintenance event. In another embodiment, replacing the second component includes replacing the second component prior to an end of warranty period for the second component.

[0040] In another embodiment, the comparing includes determining a risk score for the second component based on the destructive duty cycle, the risk score including one or more of a braking frequency, a speed change of the second component exceeding a threshold amount, and a torque of the second component exceeding a threshold amount. In a refinement of this embodiment, the risk score is based on a frequency with which the second component experiences the destructive duty cycle.

[0041] In another aspect of the present disclosure, a device for estimating failure of a machine / power system component is disclosed. The device includes at least one controller for collecting operating data of the machine / power system during a first time period; and at least one processor. The at least one processor is configured to analyze the operating data to identify one or more parameters associated with the component; determine a destructive duty cycle of the component based on the identified one or more parameters; and model the failure behavior of the component based on the destructive duty cycle and at least one of the age and use of the component. In addition, at least one controller and processor are configured to compare the actual use of the component during a second time period with the model of the failure behavior, and determine the failure probability of the component in response to the comparison.

[0042] In one embodiment, the apparatus includes a telematics device for transmitting the operational data to the at least one processor. In another embodiment, the operational data is collected from the controller during a maintenance event.

[0043] In another embodiment, the one or more parameters identified include one or more of a braking frequency, a speed change of the component exceeding a threshold amount, and a torque of the component exceeding a threshold amount. In another embodiment, one of the controller and the computer processor is configured to compare the actual usage of the component with the current mileage and the end of warranty mileage of the component.

[0044] Although the illustrative embodiments of the present disclosure have been shown and described in detail in the accompanying drawings and the foregoing description, they should be regarded as illustrative rather than restrictive in nature, and it should be understood that only certain exemplary embodiments are shown and described, and all changes and modifications falling within the spirit of the claimed invention need to be protected. It should be understood that although the words such as "preferred", "preferably", "preferred" or "more preferably" used in the above description indicate that the features described so may be more desirable, it may not be necessary and the embodiments lacking the features can be expected to be within the scope of the present invention, which is limited by the appended claims. When reading the claims, it is expected that when using words such as "one", "one", "at least one" or "at least a portion", it is not intended to limit the claims to only one item, unless otherwise specifically stated in the claims. When using the language "at least a portion" and / or "portion", the project can include a part and / or the entire project, unless otherwise specifically stated.

Claims

1. A method for estimating a failure of a machine / power system component, the method comprising: determining a destructive duty cycle for the machine / power system based on one or more parameters identified for a component of the machine / power system and correlating the destructive duty cycle with a failure of the component; modeling a failure behavior of the component based on the destructive duty cycle and at least one of an age and usage of the component; comparing at least one of an actual age or an actual usage of a second component to said model of said failure behavior; as well as The second component is replaced based on the comparison. 2 . The method of claim 1 , wherein the one or more identified parameters include one or more of a braking frequency, a speed change of a machine component exceeding a threshold amount, and a torque of the machine component exceeding a threshold amount.

3. The method of claim 1, wherein said comparing further comprises comparing said actual usage of said second component to a current mileage and an end-of-warranty mileage of said component.

4. The method of claim 1, wherein the actual usage data is collected in real time during a maintenance event via telematics or from an onboard computer system of the vehicle.

5. The method of claim 1 , wherein the comparing comprises determining a risk score for the second component based on the destructive duty cycle, the risk score comprising one or more of a braking frequency, a speed change of the second component exceeding a threshold amount, and a torque of the second component exceeding a threshold amount, and the risk score is further based on a frequency at which the second component experiences the destructive duty cycle.

6. A method for estimating a failure of a machine / power system component, the method comprising: collecting operational data of components of the machine / power system during a first time period; analyzing the operational data to identify one or more parameters associated with a failure of the component; determining a destructive duty cycle for the machine / power system based on the identified one or more parameters and correlating the destructive duty cycle with a failure of the component; modeling a failure behavior of the component based on the destructive duty cycle and usage of the component; comparing actual usage of a second component during a second time period to said model of said fault behavior; as well as The second component is replaced based on the comparison. 7 . The method of claim 6 , wherein the operational data is collected via one of a telematics device and a service portal module.

8. The method of claim 6, wherein the analyzing comprises performing a multivariate regression analysis to identify the one or more parameters indicative of failure of the component in a sample population.

9. The method of claim 6, wherein the identified one or more parameters include one or more of a braking frequency, a change in speed of a machine component exceeding a threshold amount, and a torque of the machine component exceeding a threshold amount.

10. The method of claim 6, wherein said comparing further comprises comparing said actual usage of said second component to a current mileage and an end-of-warranty mileage of said component.

11. The method of claim 6, wherein the actual usage data is collected in real time via telematics.

12. The method of claim 6, wherein the actual usage data is collected from an onboard computer system of the vehicle during a maintenance event.

13. The method of claim 6, wherein replacing the second component comprises replacing the second component before an end-of-warranty period for the second component.

14. The method of claim 6, wherein the comparing comprises determining a risk score for the second component based on a destructive duty cycle, the risk score comprising one or more of a braking frequency, a speed change of the second component exceeding a threshold amount, and a torque of the second component exceeding a threshold amount.

15. The method of claim 14, wherein the risk score is based on a frequency with which the second component experiences the destructive duty cycle.

16. An apparatus for estimating failure of a machine / power system component, the apparatus comprising: at least one controller for collecting operational data of the machine / power system during a first time period, and the apparatus further comprising at least one processor configured to: analyzing the operational data to identify one or more parameters associated with the component; determining a destructive duty cycle for the machine / power system based on the identified one or more parameters and correlating the destructive duty cycle with a failure of the component; as well as modeling a failure behavior of the component based on the destructive duty cycle and at least one of an age and usage of the component; Wherein one of the at least one controller and processor is configured to compare actual usage of the component during a second time period to the model of the failure behavior and determine a probability of failure of the component in response to the comparison.

17. The apparatus of claim 16, further comprising a telematics device for transmitting the operational data to the at least one processor.

18. The apparatus of claim 16, wherein the operational data is collected from the controller during a maintenance event.

19. The apparatus of claim 16, wherein the identified one or more parameters include one or more of a braking frequency, a change in speed of the component exceeding a threshold amount, and a torque of the component exceeding a threshold amount.

20. The apparatus of claim 16, wherein said one of said controller and computer processor is configured to compare said actual usage of said component to a current mileage and an end-of-warranty mileage of said component.

Images