System and method for predicting correlated failure of machine components

Abstract

A system (10) for predicting failure of one or more components (110) of a machine (106) is disclosed. The system (10) includes at least one interface (112, 120) configured to input current repair data (128) of a first component (110), a database (126) configured to record the current repair data (128) of the first component (110), and a processor (118) operably connected to the at least one interface (112, 120) and the database (126). The processor (118) analyzes the current repair data (128) of the first component (110) based on historical repair data stored in the database (126), wherein the historical repair data includes identifications of multiple components of the machine (106), including the first component (110) and a second component (110). The processor (118) generates a recommendation for repairing the second component (110) based on the historical repair data stored in the database (126).

Description

Technical Field

[0001] The present invention relates to the field of parts repair and replacement, and more particularly to systems and methods for predicting associated failures of machine components. Background Art

[0002] Diagnosing, maintaining, and repairing complex products, such as vehicles, appliances, industrial equipment, and other complex products, can be difficult and time consuming. Specialized knowledge and / or expensive diagnostic equipment may be required to ensure that the product can be properly diagnosed, maintained, or repaired. Personnel involved in repairing and maintaining these complex machines understand that when a complex machine is taken out of service for repair or maintenance, unexpected defects may be encountered.

[0003] Various tools have been developed to assist in these tasks. One such tool is described in U.S. Patent Application Publication No. 2005 / 0144183 (the '183 Publication) to McQuown et al. The '183 Publication describes a handheld portable unit that can be used in the field by a locomotive technician to access information needed to repair, diagnose and troubleshoot locomotive problems, and to make necessary repairs. For example, the technician can download schematics, repair manuals, repair suggestions, and other resources to help complete the task at hand. In addition, the technician can use the portable unit to order needed parts from suppliers.

[0004] Although the portable unit disclosed in '183 can help technicians diagnose, maintain and repair locomotives, this alone may not be enough. A field technician can identify a specific part of the locomotive that needs to be replaced and therefore order the part using the portable unit. However, the portable unit may not be able to identify other related parts that should be ordered with the part to ensure that the technician can complete any associated repairs before any associated failures that may occur unexpectedly within a specified time window in the near future. Therefore, the technician needs to have the knowledge and foresight to identify these related parts when ordering. Due to the complexity of the machine and the difficulty of being able to identify the related parts, defects may be overlooked or not noticed. The unit disclosed in '183 also does not provide an indication of possible associated repairs of different components, subsystems or systems that should be addressed during the inspection. Therefore, the field technician may need to have several years of experience and be able to relate the occurrence of various related or seemingly unrelated early failures, but this may be difficult or impossible for some complex machines. This may lead to operating conditions that are about to fail.

[0005] Complex machines (including but not limited to off-highway mining trucks, hydraulic excavators, crawler tractors, and wheel loaders) can represent a large capital investment and can have significant productivity when operated. Therefore, it is important to predict component, subsystem and / or system failures so that maintenance can be scheduled during a period when productivity will be less affected and so that any minor repairs can be performed before a minor failure leads to a potentially catastrophic failure.

[0006] The present invention is directed to overcoming one or more of the above-mentioned problems and / or other shortcomings of the prior art. Summary of the invention

[0007] One aspect of the present invention relates to and discloses a system for predicting failure of one or more components of a machine. The system may include at least one interface configured to input current repair data of a first component, a database configured to record the current repair data of the first component, and a processor operably connected to the at least one interface and the database. The processor may analyze the current repair data of the first component based on historical repair data stored in the database, wherein the historical repair data includes identifications of multiple components of the machine (including the first component and the second component). The processor may generate a recommendation for repairing the second component based on the historical repair data stored in the database.

[0008] Another aspect of the invention relates to a method for predicting a failure of a component of a machine. The method may include inputting current repair data for a first component of the machine into a database, and processing the repair data. The processing may include analyzing the current repair data for the first component based on historical repair data stored in the database, wherein the historical repair data includes identifications of multiple components of the machine (including the first component and the second component). The processing may also include generating a recommendation for repairing the second component based on the historical repair data stored in the database, and outputting a recommended repair checklist.

[0009] Yet another aspect of the present invention relates to a computer-readable medium having computer-readable instructions stored thereon, which instructions, when executed by a processor, cause the processor to perform a method of predicting a failure of one or more components of a machine. The method may include inputting current repair data for a first component of the machine into a database, and processing the repair data. The processing may include analyzing the current repair data for the first component based on historical repair data stored in the database, wherein the historical repair data includes identifications of multiple components of the machine (including the first component and the second component). The processing may also include generating a recommendation for repairing the second component based on the historical repair data stored in the database, and outputting a recommended repair checklist. BRIEF DESCRIPTION OF THE DRAWINGS

[0010] Figure 1is a block diagram of an exemplary system for predicting failure of a machine component;

[0011] Figure 2 is stored in Figure 1 A representation of exemplary data in a database of a system;

[0012] Figure 3 is a flow chart of an exemplary embodiment of a method of predicting failure of a machine component;

[0013] Figure 4 is shown as input and stored in Figure 1 an exemplary graph of an exemplary set of repair data within a database of a system; and

[0014] Figure 5 is shown based on the storage Figure 1 An exemplary diagram of an exemplary collection of suggestions for data within a database of a system. DETAILED DESCRIPTION

[0015] Reference will now be made in detail to the exemplary embodiments illustrated in the accompanying drawings. Wherever possible, the same reference numerals will be used throughout the drawings to refer to the same or like parts.

[0016] Figure 1 A block diagram of an exemplary system for predicting failure of a machine component is depicted, wherein the system is generally designated 10. For purposes of the present invention, Figure 3 As shown in , the present system and method for predicting associated failures in a machine are described in conjunction with a remotely located machine (including machines such as off-highway mining trucks, hydraulic excavators, crawler tractors, wheel loaders, etc.). However, the disclosed system and method are equally applicable to various other devices or machines. In addition, the present invention can involve analyzing information collected from a part, subsystem, or system of a machine. However, data can be collected and analyzed from multiple machines.

[0017] Figure 1The system 10 shown in the figure includes an on-site system 100 and a remote system 102, which can be operably connected via a communication network 104. The on-site system 100 and its system elements can be located on-site of the machine currently being repaired, while the remote system 102 and its system elements can be located remotely from or off-site of the machine currently being repaired. The system 10 and / or the on-site system 100 and / or the remote system 102 can be a server, a client, a mainframe, a desktop, a laptop, a network computer, a workstation, a personal digital assistant (PDA), a tablet PC, etc. The communication network 104 can be, for example, a telephone-based network (e.g., a PBX or POTS), a local area network (LAN), a wide area network (WAN), the Internet or another packet-switched network, a private intranet, a workstation peer-to-peer network, a direct link network, a wireless network, or another suitable network.

[0018] The field system 100 may include at least one machine 106. Figure 1 In the embodiment of the present invention, the field system 100 is depicted as including a single machine 106; however, the systems and methods of the present invention are equally applicable to field systems 100 having more than one machine 106. Furthermore, when the field system 100 includes more than one machine 106, the machine may be the same machine, a group of related or similar machines, or in some cases a plurality of different machines. Multiple machines 106 from various work sites, such as different mining locations, may be included.

[0019] like Figure 1 As shown in , the machine 106 may include one or more sensors 108 and / or components 110. The component 110 may be a single part of the machine 106, or a system or subsystem of the machine 106. The single-part component 110 may be, for example, a seal, a tube, a valve, a bellows, etc., while the subsystem or system may be a cooler or heat exchanger, an intake and / or exhaust manifold, a brake group, etc. The sensor 108 may include one or more sensors 108 to sense data from one or more components 110 of the machine 106. The sensor 108 may be a sensor of a type known in the art for generating an electrical signal in response to a level of an operating parameter, and may sense data from the machine 106 and its components 110, including pulse width modulated sensor data, frequency-based data, five-volt analog sensor data, and switch data that has been effectively debounced. The sensor 108 may also be connected to an electronic module (not shown) of the machine 106.

[0020] Figure 1The field system 100 also includes an interface 112 that is operably connected to the communication network 104 and the machine 106 (including the sensor 108 and the component 110). The interface 112 is capable of communicating with the machine 106 and is capable of communicating with the remote system 102 via the communication network 104. The interface 112 may include a display 114 and an input device 116. The display 114 may be an electronic display, including but not limited to an LCD, a CRT, a plasma display, etc., and may include a graphical user interface (GUI) (not shown). The input device 116 may be any known device for inputting information, including but not limited to a keyboard. Although the input device 116 is shown as an element separate from the display 114, in some embodiments, the input device 116 may be formed as part of the display 114. In addition, other types of interface devices such as, for example, handheld computing devices, voice recognition devices, touch screens, etc. may be used to interface with the machine 106 and the remote system 102.

[0021] Figure 1 The remote system 102 shown in FIG. 1 includes a processor 118, an interface 120, and a database 126. The processor 118, shown as being operably connected to the communication network 104, can communicate with the field system 100 and the database 126 to perform an analysis of the current repair data, as described below with respect to FIG. Figure 3 Processor 118 may include one or more known processing devices, such as those from Intel TM Pentium manufactured TM or Xeon TM family of microprocessors, from AMD TM Made by Turion TM family or any other type of processor.

[0022] The interface 120 may be operably connected to the processor 118. In some cases, the interface 120 may also be directly operably connected to the database 126, as opposed to being operably connected through the processor 118. The interface 120 may include a display 122 and an input device 124. The display 122 may be an LCD, a CRT, a plasma display, etc., and may include a graphical user interface (GUI) (not shown). The input device 124 may be any known device for inputting information, including but not limited to a keyboard. Although the input device 124 is shown as a separate element from the display 122, in some embodiments, the input device 124 may be formed as part of the display 122. In addition, other types of interface devices such as, for example, handheld computing devices, voice recognition devices, touch screens, etc. may be used to interface with the processor 118, the database 126, and the field system 100.

[0023] Data from one or more machines 106 can be collected and stored in a database 126 for use in the embodiments disclosed herein. Data can be collected over the course of hours, days, weeks, months, or years and stored and recorded in the database 126.

[0024] like Figure 2 As shown in , the database 126 can be configured to store various types of data, including repair data 128 and operational data 130. "Repair data" can refer to "current repair data" from a current repair process on a machine, and can include data such as identification of components, subsystems, and / or systems of the machine. "Repair data" can also refer to data from previous repair processes on a machine, and in this context can be referred to as "historical repair data", which can include data such as identification of components, subsystems, and / or systems. Thus, the identification of a first component and the identification of a second component or associated components can be stored in the database 126 along with the identification of other machine components, subsystems, and / or systems. "Historical repair data" can refer to data collected over a period of time (e.g., months or years) that can be stored in the database 126 for use by the disclosed system 10. "Repair data" can also be referred to herein as "work order data". Historical repair data can be data collected from a single machine over a period of time, or collected from a group of machines or multiple machines that are the same or similar. For example, historical repair data may be collected over a period of months or years and stored in a single database 126 for a single haul truck, a group of haul trucks, or multiple haul trucks, and similar but not identical mining trucks. Although repair data may be described herein for a given machine, it may also refer to data collected for several identical or similar machines that may be used to analyze the performance of a group of machines.

[0025] "Operational data" and "historical operational data" may include, for example, engine RPM, oil pressure, water temperature, boost pressure, oil contamination, motor current, hydraulic pressure, system voltage, etc. "Operational data" may also include data related to other conditions of the machine 106, including but not limited to payload, tire performance, etc. The processor may analyze the current repair data of the current (first) component based on the historical operational data as part of generating recommendations for repairing the associated (second) component. The data entered and stored in the database 126 may each include a time and / or date stamp. For example, the data may be entered with a stamp of January 1, 2000. The data stored in the database 126 may originate from, for example, a technician, a machine manufacturer, a dealer, and / or a service provider. The repair data and / or operational data may be collected and recorded in the database 126 manually or by one or more sensors 108.

[0026] In some embodiments, the database 126 may include one or more storage devices configured to store information or data (such as the repair data and operational data described above), which may be used by the processor 118 to perform certain functions related to the disclosed embodiments. The database 126 may include volatile or non-volatile, magnetic, semiconductor, tape, optical, removable, non-removable, or other types of storage devices or computer-readable media. The database 126 or another storage device (not shown) operably connected to the processor 118 may store programs and / or other information, such as information related to processing data. In an exemplary embodiment, the remote system 102 includes a memory (not shown), which may include one or more programs or subroutines loaded from a storage device or elsewhere, which, when executed by the processor 118, perform various processes, operations, or procedures consistent with the disclosed embodiments. For example, the memory may include one or more programs that enable the processor 118 to analyze current repair data based on historical repair data (as well as perform other tasks), as described below with respect to Figure 3 Discussed in detail.

[0027] Figure 3 is a flow chart of an exemplary embodiment of a method for predicting failure of a machine component, the method comprising Figure 1 118. In general, the system 10 is configured to identify recurring patterns in which historically different events (including component failures) have occurred within a specified time window and apply value prioritization, as discussed in more detail below. The system can be used to predict associated component failures that are likely to occur based on statistically significant historical evidence triggered by current repair data. In addition, the system includes functionality to generate or output recommendations (also referred to as checklists) to recommend preemptive repair and / or maintenance of one or more associated components predicted to fail.

[0028] An associated component refers to a component that often needs to be serviced (e.g., repaired or replaced) at approximately the same time that a different component is being serviced. The component being serviced may be referred to as a "first component" or "current component," and the component that may need service may be referred to as a "second component," "at least one second component," or "associated component." In some cases, multiple other components may need service at approximately the same time that one component (the first component or current component) is being serviced, in which case the multiple components that need service may be referred to as "additional components" or "associated components." The term "associated" does not require that a component be related to another component. In fact, as described in the present invention, a component may be an associated component even though it is logically unrelated to other components being serviced.

[0029] Now refer to Figure 3The details of the method for predicting failures in machine components are discussed with reference to the flowchart in FIG.

[0030] Figure 3 The method 200 shown in FIG. 2 comprises step 202, wherein according to Figure 1 and its related description, collects part data from one or more machines 106. The part data may be collected using, for example, Figure 1 1 , is manually collected by a technician or other entity that enters the part data. Alternatively, one or more machine sensors 108 may be used to automatically collect the part data and send it to the database 126.

[0031] In step 204, component data is entered as current repair data. This can be done manually or automatically using interface 112. The entered current repair data may include, for example, component information, such as a part number, for the component currently being repaired that is identified. In other cases, interface 120 may be used to enter component data as current repair data. For example, a field technician may contact a remote technician who has access to remote system 102 and provide the remote technician with component data to enter as current repair data.

[0032] In step 206, the current repair data is recorded in the database 126. The current repair data may be recorded by, for example, identifying component information, such as a part number of the component currently being repaired, and a date and / or time stamp indicating when the component was repaired.

[0033] A data cleaning step (not shown) may be included as part of step 204 and / or step 206, which may also be referred to as a data cleaning or data washing step. The raw data may be input and recorded as uncleaned data. Commonly employed data cleaning techniques may be used to detect and correct errors in the data input into the database. With respect to the current repair data input and recorded into the database as described herein, known data cleaning techniques may be used to identify incomplete, incorrect, erroneous, or irrelevant aspects of the repair data, and replace, modify, or delete the data so that it is consistent with other repair data stored in the database.

[0034] In steps 208, 210, and 214, an analysis of the current repair data is performed based on the historical repair data stored in the database 126. The processor 118 of the remote system 102 may be configured to perform the analysis. Specifically, one or more algorithms may be applied to perform the analysis, which may be accessed by the processor 118 or in some cases stored on the processor 118, as described in more detail below. One or more algorithms may be used to extract data (such as repair data and / or operational data) from the database to determine when a machine component should be repaired or replaced, and to recommend repair or replacement of the component.

[0035] In step 208, the current repair data recorded in step 206 is analyzed based on the historical repair data. In order to perform the analysis in step 208, a decision can be made according to step 210. Specifically, in step 210, Figure 3 The operation method shown in can determine whether one or more associated events occurred within a specified time of a previous component failure. As discussed herein, the associated event can be, for example, a failure of an associated component (a second component) other than the component currently being repaired (a first component). And "previous component failure" can refer to a previous failure of the component currently being repaired (a first component), wherein data of the previous component failure is stored in the database 126.

[0036] An algorithm accessible by or stored on the processor 118 may be utilized to determine whether one or more associated events have occurred within a specified time of a previous component failure. Specifically, the algorithm applied in step 210 may be a version of an Apriori algorithm, which is generally understood to be a general type of algorithm useful for data mining and determining the frequency of occurrence of items in a data set. Data mining may generally refer to the process of identifying patterns in a data set and extracting useful information from the identified patterns. The terms "event," "transaction," or "item" as used herein may be used interchangeably to refer to a component, subsystem, or system failure or repair that occurs outside of normal operating conditions of a given machine.

[0037] The present systems and methods may use a modified version of the a priori algorithm to determine whether an associated event occurred within a specified time of a previous component failure. Using the modified a priori algorithm, each repair data recorded in the database 126 as historical repair data may be recorded as a timestamp record. For example, a repair data entry or item may be "Change cooler / heat exchanger seal, January 1, 2000." The a priori algorithm of the present invention may be applied by the processor 118 of the remote system 102 to detect whether one or more associated events occurred within a specified time of a previous component failure. For example, the modified a priori algorithm may include a time limit (also referred to herein as a time period or time window) between 7 and 15 days. In this example, the analysis using the modified a priori algorithm will select from possible suggestions ( Figure 3 , step 218; Figure 5 ) excludes any associated events that occurred outside of a 7 to 15 day period of a previous failure of the component currently being repaired from the a priori algorithm. In another example, the a priori algorithm may include a 10 day time limit, thereby excluding from possible suggestions any associated events that occurred more than 10 days after a previous failure of the component currently being repaired. The time limit may be entered using an input device 124 of the interface 120. Alternatively, the time limit may be entered using another input device of the remote system 102 or another input device of a system separate from the remote system 102 but operably connected to the remote system 102.

[0038] As yet another example, if a failure of a first component (e.g., a cooler or heat exchanger seal) occurs at a first time, the time at which a failure of a second component (e.g., a cooler or heat exchanger hose or a transmission pump hose) occurs is measured as a second time after the first time. These time measurements are compiled in a database 126, which can be accessed by the modified a priori algorithm via the processor 118 of the remote system 102. It can then be determined that the failure of the second component is likely to occur within a time N (where N=second time-first time) after the failure of the first component.

[0039] Thus, a modified a priori algorithm is an algorithm that can limit possible suggestions by providing a time limit, as described above. The modified a priori algorithm described herein may be referred to as an Apriori-like algorithm, a modified a priori algorithm, or a simple a priori algorithm. The present invention will specify whether and when to refer to a general a priori algorithm rather than the modified a priori algorithm described above.

[0040] If it is determined that one or more associated events have not occurred within the specified time of the previous component failure, then the process ends, as shown in step 212. However, if one or more associated events have occurred within the specified time of the previous component failure, then it can be determined whether the one or more associated events meet the priority threshold, as shown in step 214.

[0041] In step 214, in order to determine whether the one or more associated events meet the priority threshold, another algorithm accessible by or stored on the processor 118 can be utilized. Specifically, the algorithm applied in step 214 can be a Pareto algorithm. The Pareto algorithm is generally understood to be an algorithm that uses stored data to determine which events contribute the majority (usually 80%) of certain effects. The principle of the Pareto algorithm is generally considered in business, that is, 80% of a company's sales may come from 20% of the company's consumers or customers. In the context of the present invention, the Pareto algorithm applied in step 214 can be used to determine which repair events tend to contribute to a specific percentage of the cost of machine downtime. The Pareto algorithm can also be used to determine whether one or more associated events is one of the repair events that tend to contribute to a high percentage of the cost, so that the one or more associated events will be included in the recommendation of step 218.

[0042] In one example, a threshold of approximately 80% may be set in the Pareto algorithm. Applying this threshold, the analysis will determine which machine repairs have historically contributed 80% or more of the total downtime repair costs. Those repairs that have not contributed will not be included in the recommendations generated in step 218.

[0043] For any single machine, there may be thousands (even tens of thousands or hundreds of thousands) of components, most or all of which may be grouped into subsystems, which may be grouped into separate systems of the machine. Database 126 may store the cost of separate repair events, including one or more associated events and the cost of the systems, subsystems, and components involved in each repair event. The identification of the systems, subsystems, and / or components involved in each repair event may constitute the historical repair data discussed in the present invention. For example, assume that in a year, a given machine has 1,000 repair events, and the 1,000 repair events result in a total repair cost of $1,000,000 for the machine. And assume that a total of 10,000 different components are involved in these 1,000 repair events. 1,000 repair events involving 10,000 components and resulting in a total repair cost of $1,000,000 may be referred to as historical repair data. The disclosed Pareto algorithm, and in particular the priority threshold of the Pareto algorithm, can determine which of the 10,000 parts are involved in the most expensive repair events totaling at least $800,000 (80% of the total cost). The Pareto algorithm can therefore be applied to calculate the sum of the repair events, starting with the most expensive repair event and continuing to add the second most expensive repair event until the combined cost equals at least $800,000. In many cases, a small number of repair events can account for a large amount of the total downtime repair cost of the machine compared to the total number of repair events. As a simple example, out of 1,000 repair events, 20 repair events may account for 80% of the total downtime repair cost. Although a time period of one year is discussed above, the time period on which the total repair cost of a given machine is based can be a time period of days, weeks, months, or years. The time period can be set in the Pareto algorithm, for example, using the input device 124.

[0044] The 80% threshold is only one typical example of a Pareto algorithm. In other examples, the threshold may be set to a value higher than 80% (e.g., about 90% or about 95%) to provide a more selective and exclusive analysis, or lower than 80% (e.g., about 70% or about 75%) to provide a more inclusive analysis. For example, if a large number (e.g., twenty) of events occurred within a specified time of a previous component failure after the prior algorithm has been run in step 210, the analysis using the Pareto algorithm in step 214 may further limit the number of events according to the threshold as discussed herein. However, if only a small number of events (e.g., five) occurred after the prior algorithm has been run, none of these events may be excluded when generating these events as recommendations in step 218. The maximum number of events may be predetermined and set in the Pareto algorithm, for example, using the input device 124. For example, if the Pareto algorithm is applied after the prior algorithm, and the Pareto algorithm includes a maximum limit of five events, no more than five recommendations may be generated in step 218. When there is a maximum limit on events, these events can be the most expensive events as discussed above. The maximum number of events allowed by the Pareto algorithm can be more or less than five, depending on factors including but not limited to the machine being repaired, and the number of technicians available to provide repairs.

[0045] Due to machine complexity, when one seemingly unrelated component is being repaired, there may be many other components that need repair. When repairing a machine, a professional technician may try to apply domain knowledge expertise, which can be described as expertise in the field, to predict whether different components are on the verge of failure and whether they should also be repaired or replaced. However, using such domain knowledge expertise may result in repairing too many potentially related components. An analysis including applying the Pareto algorithm in step 214 can determine the priority of repairing associated components during any single instance of machine maintenance. This process may also be referred to as value prioritization.

[0046] Data representing the downtime costs of various machine components (e.g., as threshold percentage values) may be stored in the database 126 for use in Figure 3The data may be input into the Pareto algorithm manually or automatically during the analysis of the remote system 102. The data may be input using the input device 124 of the interface 120, or alternatively using another input device of the remote system 102 or another input device of a system separate from but operatively connected to the remote system 102. Additionally, the data of the downtime costs of the various machine components may be stored on a separate database or storage device that is part of the remote system 102 or a system separate from but operatively connected to the remote system 102.

[0047] Therefore, analysis using the a priori algorithm and the Pareto algorithm described herein can be used to discover sequences of frequent item sets using data mining. As used herein, the term "frequent item sets" can refer to component failures that occur together. Whether a component failure can be identified as a frequent item set can be determined based on the timing input into the a priori algorithm, and the threshold value of the value prioritization input into the Pareto algorithm, as discussed above.

[0048] If one or more of the associated events do not satisfy the priority threshold of the Pareto algorithm, the process ends, as shown in step 216. However, if the one or more associated events satisfy the priority threshold of the Pareto algorithm, the process proceeds to step 218, and the processor 118 generates (outputs) recommendations for which associated components should be repaired based on the historical repair data stored in the database 126. The recommendations may include one or more recommendations for repairing at least one of the associated components, which may be displayed on the display 114. In some cases, the recommendations may be displayed on the display 122. After generating the recommendations at step 218, the method may end at step 220.

[0049] Thus, the systems and methods described herein can allow for preemptive repair of a machine component that is not currently being repaired based on a determination that the component is likely to fail, where the determination is made based on the association of an identification associated with a failure of the component currently being repaired. For example, if the component currently being repaired is a heat exchanger seal, and historically the heat exchanger hose has failed within seven days of replacing the heat exchanger seal, the disclosed systems and methods for predicting failure of a machine component can provide instructions (recommendations) to a technician or other entity to inspect the hose when replacing the seal. If the hose needs to be repaired or replaced, the heat exchanger seal and the hose can be repaired at the same time.

[0050] although Figure 3 The flowchart refers to "repair data", but the same method can be applied using "operation data", as referenced above Figure 2 discussed. For example, Figure 3The method 200 may be used with operational data 130, including but not limited to engine RPM, oil pressure, water temperature, boost pressure, oil contamination, motor current, hydraulic pressure, and system voltage. As described above, the operational data 130 may also be stored in the database 126. In addition, although the algorithm is described as being stored in and accessible by the processor 118, in other embodiments, either or both algorithms may be stored in other processing units and / or storage devices of the field system 100 or the remote system 102. In addition, in other embodiments, steps 210 and 214 may be reversed so that the a priori algorithm is applied after the Pareto algorithm. That is, the method 200 may be performed so that it is first determined whether one or more associated events meet a priority threshold, and if so, it is determined whether the one or more associated events occur within a specified time of a previous component failure. This method of operation may filter out low-cost repair events from possible suggestions before filtering out other associated repairs based on the time limit of the a priori algorithm. Therefore, the a priori algorithm and the Pareto algorithm may not necessarily be dependent on each other.

[0051] Figure 4 Depicted shows the input and storage Figure 1 400 of an exemplary collection of repair data within the database 126 of the system 10 of FIG. "Previous Items" refers to repair data for a previous event (e.g., a failure) for a given (first) component. "Later Items" refers to repair data for an event (an associated event) subsequent to the previous event, which event is determined to be associated with the previous event according to the systems and methods described herein. That is, the later item may relate to an associated (second) component that is different from the first component. The "ID Number" column may include a number for a given component, subsystem, or system, such as a part number and / or model number.

[0052] The "Confidence (%)" column indicates the confidence that the antecedent and the consequent are related events. The confidence percentage can be statistically derived by applying the analysis of step 208, step 210, and / or step 214, such as Figure 3 For example, the a priori algorithm applied in step 210 of the present invention can be used to determine a confidence percentage. For a given antecedent event, the a priori algorithm can use the data stored in the database 126 to calculate the likelihood that the consequent will also occur within the time limit specified by the a priori algorithm, where the likelihood can be expressed as a percentage. For example, Figure 4 It shows that when the cooler / heat exchanger seal is serviced, there is a 100.000% probability that the cooler / heat exchanger hose should also be serviced. Figure 4 Involves depicting repair data, but the chart can also depict Figure 2 "Operational data" of the type shown and discussed.

[0053] Figure 5 Depicted is a diagram showing a method based on storing Figure 1 An exemplary chart 500 of an exemplary collection of recommendations for data within a database of system 10 is provided. The recommended chart 500 is a checklist showing which events are relevant and should be addressed by a technician. The chart 500 may include a description column describing the recommendation, which may include a recommendation to preemptively repair an associated (i.e., second) component. For example, if historically repairs to cooler / heat exchanger seals were often followed by repairs to cooler / heat exchanger hoses, a recommendation may be generated with the following description: Historically, repair events for cooler / heat exchanger seals were often followed by repair events for cooler / heat exchanger hoses. The description informs the technician that the cooler / heat exchanger hoses should be repaired at the same time as the cooler / heat exchanger seals. Additional possible descriptions are provided in the accompanying description. Figure 5 , and may also include various other descriptions depending on the component being repaired.

[0054] like Figure 5 As shown in , a single chart such as chart 500 may include recommendations for various machine components. In some cases, a recommendation chart providing recommendations may be generated based on repairing a single machine component. Figure 5 In addition to the "Description" column shown in FIG. 5 , the chart 500 may include additional columns specifying the part serial number, repair date, model code, and / or manufacturer for a specific machine part currently being repaired. The suggested chart 500 may be accessed in the field system 100. For example, the display 114 may display the chart 500. Alternatively, Figure 1 Another display not shown in the figure may be used to display chart 500 .

[0055] Those skilled in the art will appreciate that all or part of the systems and methods consistent with the present invention may be stored on or read from other computer-readable media. Figure 1 , system 10 may include a computer-readable medium having stored thereon computer-readable machine instructions that, when executed by processor 118, may cause processor 118 to perform the methods disclosed herein, including methods for predicting failures in a machine, and other tasks. Exemplary computer-readable media may include secondary storage devices, such as hard disks, floppy disks, and CD-ROMs; or other forms of computer-readable memory, such as read-only memory (ROM) or random access memory (RAM). Such computer-readable media may be implemented by one or more components of system 10, such as processor 118, database 126, interface 112, interface 120, machine 106, server system, or a combination of these and other components.

[0056] Industrial Applicability

[0057] Although described with respect to machines including trucks, hydraulic excavators, crawler tractors, and wheel loaders, the systems and methods of the present invention may be applied to other industries that rely on machine equipment. For example, in addition to the automotive industry, the airline or shipping industries may apply the described systems and methods, and heavy equipment manufacturers may seek to provide data utilization services to customers seeking to minimize unplanned machine downtime.

[0058] The disclosed system and method for predicting associated failures of machine components is a predictive tool triggered by repair data or operational data that can be used to reliably preemptively repair associated component failures. The system and method can leverage and mine large amounts of historical data, typically spanning months or years, to explore patterns and relationships when repairing machine components. The system and method can also provide specific, actionable recommendations whenever a machine component is inspected, repaired, replaced, or otherwise serviced.

[0059] Additionally, because failure of certain related components can be predicted, an inventory of parts for repair and replacement can be maintained in a cost-effective manner by using the systems and methods disclosed herein. For example, if the system determines that there is a relationship between an inlet exhaust manifold tube failure and an inlet exhaust manifold bellows failure, an inventory can be maintained for both conditions so that they can be addressed at once.

[0060] Thus, the disclosed systems and methods for predicting associated failures of machine components can minimize unplanned and expensive downtime for servicing machines and their components, subsystems, and systems by preemptively repairing failures of associated components. The systems and methods described herein are particularly useful for providing actionable intelligence in the form of recommendations to preemptively repair associated failures of components in machines that are often too complex for expert technicians to accurately diagnose, particularly when the associated components may not be logically connected within a given machine.

[0061] Although the systems and methods described herein relate to determining failure modes between machine components by predicting failure of another component based on repair of one component, they can also be used to determine failure modes between other machine subsystems or systems based on repairs to a subsystem or system of a machine, where the subsystems and systems may include multiple components.

[0062] It is obvious to those skilled in the art that various modifications and variations may be made to the disclosed system and method for predicting associated failures of machine components. Other embodiments of the invention will be apparent to those skilled in the art after consideration of the specification and practice of the invention. Therefore, the specification and examples are to be considered as exemplary only, and the true scope of the invention is indicated by the appended claims and their equivalents.

Claims

1. A system (10) for predicting failure of one or more components (110) of a machine (106), the system comprising: at least one interface (112, 120) configured to input current repair data (128) of the first component; a database (126) configured to record the current repair data of the first component; as well as a processor (118) operably connected to the at least one interface and the database, wherein the processor: analyzing the current repair data for the first component based on historical repair data stored in the database, wherein the historical repair data includes identification of a plurality of components of the machine, the plurality of components including the first component and a second component; and generating a recommendation for repairing the second component based on the historical repair data stored in the database and the current repair data for the first component, In order to analyze the current repair data, the processor: applying an a priori algorithm with a time limit to determine whether one or more associated events occur within the time limit; as well as A Pareto algorithm with a priority threshold is applied to determine whether the one or more associated events satisfy the priority threshold.

2. The system of claim 1, wherein: The database stores costs of individual repair events including repair events included in the one or more associated events; The priority threshold of the Pareto algorithm is a percentage of the total repair cost of the machine over a period of time; and The Pareto algorithm determines which machine components of the historical repair data were involved in repair events that totaled a percentage of the total repair cost that constituted the priority threshold.

3. The system of claim 2, wherein the Pareto algorithm calculates the sum of the costs of the individual repair events, starting with the most expensive repair event and continuing to add the second most expensive repair event until the combined cost is at least equal to the priority threshold. 4 . The system of claim 1 , wherein the a priori algorithm is used to determine a confidence percentage indicating a likelihood that the one or more associated events will occur within the time limit.

5. A method (200) for predicting failure of a component of a machine, the method comprising: Entering current repair data of a first component of the machine into a database (204); processing the repair data, wherein the processing comprises: analyzing the current repair data for the first component based on historical repair data stored in the database (208), wherein the historical repair data includes identification of a plurality of components of the machine, the plurality of components including the first component and a second component; and generating a recommendation for repairing the second component based on the historical repair data stored in the database and the current repair data of the first component (218); and Output a recommended repair checklist, The analysis includes: applying an a priori algorithm with a time limit to determine whether one or more associated events occur within the time limit; and A Pareto algorithm with a priority threshold is applied to determine whether the one or more associated events satisfy the priority threshold.

6. The method of claim 5, wherein the a priori algorithm is used to determine a confidence percentage indicating a likelihood that the one or more associated events will occur within the time limit.

7. The method according to claim 5, wherein The database stores costs of individual repair events including repair events included in the one or more associated events; The priority threshold of the Pareto algorithm is a percentage of the total repair cost of the machine over a period of time; and The Pareto algorithm determines which machine components of the historical repair data were involved in repair events that totaled a percentage of the total repair cost that constituted the priority threshold.

8. The method of claim 7, wherein the Pareto algorithm calculates the sum of the costs of the individual repair events, starting with the most expensive repair event and continuing to add the second most expensive repair event until the combined cost is at least equal to the priority threshold.

Images