Wear prediction of chassis using machine learning model

Abstract

A system may include an apparatus configured to receive sensor data related to wear of one or more components of a chassis of the machine from one or more sensor devices of the machine and to predict an amount of wear of the one or more components based on a wear rate of the one or more components using the machine learning model and the sensor data. The machine learning model is trained using training data to predict a wear rate of the one or more components. The training data includes two or more of past sensor data, past inspection data, or simulation data from a simulation model of one or more third devices. The apparatus may perform an action based on the amount of wear.

Description

Technical Field

[0001] The present disclosure generally relates to monitoring wear of a machine chassis, for example, predicting chassis wear using a machine learning model.

Background Art

[0002] Parts of a machine chassis (such as track links, bushings, and / or pins) wear over time. Techniques for detecting wear of parts include obtaining manual measurements of the part dimensions of such parts. The manual measurements can be compared to the specified dimensions of the parts. To obtain the manual measurements, the machine needs to pause the execution of tasks at the work site. To obtain the manual measurements, the machine needs to interrupt the execution of tasks, which is a time-consuming process (e.g., due to travel time to obtain the manual measurements and / or time to obtain the manual measurements), and obtaining the manual measurements can negatively impact the productivity at the work site. In this regard, the task (the task performed by the machine) may be interrupted for a long period (e.g., the period to obtain the manual measurements).

[0003] Also, such manual measurements may not be accurate. Conversely, inaccurate measurement of the part dimensions may lead to inaccurate prediction of the wear amount of the parts. As a result of such inaccurate predictions, the parts may fail prematurely, or may need to be repaired or replaced (e.g., the parts may not be worn enough to require replacement or repair). Such premature failure of parts, premature replacement or repair of parts may also negatively impact the productivity at the work site. Therefore, in order to prevent or reduce downtime at the work site (e.g., downtime to obtain manual measurements of part dimensions, downtime related to premature part failure, downtime related to premature part repair, downtime related to premature part replacement, and / or similar situations), it is necessary to improve the above techniques for detecting part wear.

[0004] German Patent Application Publication No. DE10257793 (“‘793 Publication”) discloses a model-based life observer for calculating the remaining life of selected components. In the ‘793 Publication, it is disclosed that the model-based life observer associates measured values of operating loads with a model-based preparation of the measured loads via available sensor devices.

[0005] The ‘793 Publication discloses a model-based life observer, but the ‘793 Publication does not disclose that data (from available sensor devices) takes into account external factors that may affect the wear of the selected components. Thus, the model-based life observer of the ‘793 Publication may not accurately predict the remaining life of the components.

[0006] The wear detection device according to the present disclosure solves one or more of the above problems and / or other problems in the art.

SUMMARY OF THE INVENTION

[0007] A method executed by a first device, comprising receiving, from one or more second devices, past sensor data related to wear of one or more components of a chassis of a machine; receiving, from one or more third devices, past inspection data related to wear of the one or more components; training a machine learning model using the past sensor data and the past inspection data to predict the remaining life of the one or more components; receiving sensor data related to wear of the one or more components from the one or more sensor devices of the machine; predicting the remaining life of the one or more components based on the sensor data using the machine learning model; and executing an action based on the remaining life of the one or more components.

[0008] A machine comprising one or more memories and one or more processors, wherein the one or more processors receive sensor data related to wear of the one or more components of the chassis of the machine from one or more sensor devices of the machine, and based on the wear rate of the one or more components, use the machine learning model and the sensor data to predict the wear amount of the one or more components, and are configured to execute the action based on the wear amount of the one or more components, and the machine learning model is trained to predict the wear rate of the one or more components using training data including two or more of past sensor data, past inspection data, or simulation data from one or more third devices, and two or more of the past sensor data, the past inspection data, or the simulation data are related to the wear of the one or more components.

[0009] A system comprising an apparatus configured to receive sensor data related to wear of one or more components of the chassis of the machine from one or more sensor devices of the machine, use the machine learning model to predict the remaining life of the one or more components based on the sensor data, and execute an action based on the remaining life of the one or more components, wherein the machine learning model is trained to predict the remaining life of the one or more components using training data including two or more of the past sensor data, the past inspection data, or the simulation data from the simulation model, and execute an action based on the remaining life of the one or more components, and two or more of the sensor data, the past inspection data, or the simulation data are related to the wear of the one or more components.

Brief Description of the Drawings

[0010]

Figure 1

Figure 2

Figure 3

DETAILED DESCRIPTION OF THE INVENTION

[0011] The present disclosure relates to an apparatus for predicting the remaining life of one or more components of a chassis of a machine using a machine learning model. The term "machine" can mean any machine that performs operations related to industries such as, for example, mining, construction, agriculture, transportation, or other industries. Additionally, one or more tools can be connected to the machine.

[0012] FIG. 1 is a diagram of an exemplary implementation form 100 described in this specification. As shown in FIG. 1, the exemplary implementation form 100 includes a machine 105 and a wear detection device 190. The machine 105 is embodied as an earthmoving machine such as a dozer. Alternatively, the machine 105 may be another type of track-type machine such as an excavator.

[0013] As shown in FIG. 1, the machine 105 includes an engine 110, a sensor system 120, an operator cabin 130, a controller 140, a rear attachment portion 150, a front attachment portion 160, and a ground engaging member 170.

[0014] The engine 110 can include an internal combustion engine such as a compression ignition engine, a spark ignition engine, a laser ignition engine, a plasma ignition engine, etc. The engine 110 supplies power to the machine 105 and / or a set of loads related to the machine 105 (e.g., components that absorb power and / or operate using power). For example, the engine 110 can supply power to one or more control systems (e.g., the controller 140), the sensor system 120, the operator cabin 130, and / or the ground engaging member 170.

[0015] The engine 110 can supply power to the tools of the machine 105 used in mining, construction, agriculture, transportation, or other industries. For example, the engine 110 can provide power to components (such as one or more hydraulic pumps, one or more actuators, and / or one or more electric motors) to facilitate the control of the rear attachment portion 150 and / or the front attachment portion 160 of the machine 105.

[0016] The sensor system 120 can include sensor devices that can generate signals regarding the wear amount of one or more components of the chassis of the machine 105 (as will be described in more detail below). The type of sensor devices of the sensor system 120 will be described in more detail below in relation to FIG. 2.

[0017] The operator cabin 130 includes an integrated display (not shown) and operator controls (not shown). The operator controls can include one or more input components (such as an integrated joystick, buttons, control levers, and / or a steering wheel) for controlling the operation of the machine 105. In the case of an autonomous machine, the operator controls can be designed to operate independently of the operator rather than being designed for the operator to use. In this case, for example, the operator controls can include one or more input components that provide input signals used by other components without operator input.

[0018] The controller 140 (such as an electronic control module (ECM)) can control and / or monitor the operation of the machine 105. For example, the controller 140 can control and / or monitor the operation of the machine 105 based on signals from the operator controls, the sensor system 120, and / or the wear detection device 190. In some cases, the controller 140 can predict the wear amount of one or more components of the chassis based on signals from the sensor system 120 and the wear detection device 190, as will be described in more detail below.

[0019] The front attachment portion 150 can include a blade assembly. The rear attachment portion 150 can include a ripper assembly, a winch assembly, and / or a drawbar assembly.

[0020] The ground engaging member 170 may be configured to propel the machine 105. The ground engaging member 170 can include wheels, tracks, rollers, etc. of the propulsion machine 105. In some cases, the ground engaging member 170 can include a chassis including a track (as shown in FIG. 1). The track can include track links. The track links can include track link bushings and track link pins. As an example, the track can include a first track link 172 and a second track link 174. The first track link 172 includes a first track link bushing 176 and a first track link pin 178. The second track link 174 includes a second track link pin 180.

[0021] The sprocket 182 can include one or more sections 184 (referred to herein simply as "section 184" and collectively as "section 184"). The sprocket 182 can be configured to engage the ground engaging member 170 to drive the ground engaging member 170. For example, the segment 184 can be configured to engage a track link bushing (e.g., the track of the ground engaging member 170) and rotate such that the track propels the machine 105.

[0022] The wear detection device 190 can include one or more devices that can predict the wear amount of one or more components of the chassis (for example, one or more trucks, truck links, one or more truck link bushings, one or more truck link pins, one or more sprockets 182, and / or one or more segments 184). The wear detection device 190 can predict the remaining life of one or more components based on the wear amount. In some examples, the wear detection device 190 can predict the wear rate of one or more components and predict the wear amount based on the wear rate. The wear detection device 190 can predict the wear amount of one or more components using a machine learning model, as described in more detail below. The wear detection device 190 may be disposed within the machine 105, may be disposed outside the machine 105, may be partially disposed within the machine 105, or may be partially disposed outside the machine 105.

[0023] As described above, FIG. 1 is provided as an example. Other examples may be different from the examples described with reference to FIG. 1.

[0024] FIG. 2 is a diagram of an exemplary system 200 described herein. As shown in FIG. 2, the system 200 includes a sensor system 120, a controller 140, a wear detection device 190, an inspection device 210, a simulation device 220, and a machine learning model 230. In some examples, the wear detection device 190, the inspection device 210, and / or the simulation device 220 may be part of a site management system (for example, a work site associated with the machine 105).

[0025] Alternatively, the wear detection device 190, the inspection device 210, and / or the simulation device 220 may be part of a background system. The wear detection device 190, the inspection device 210, and / or the simulation device 220 may be included in the same device. Alternatively, the wear detection device 190, the inspection device 210, and / or the simulation device 220 may be separate devices.

[0026] The sensor system 120 can include a sensor device that generates sensor data related to the wear amount of one or more components of the chassis (e.g., one or more tracks, track links, one or more track link bushings, one or more track link pins, one or more sprockets 182, and / or one or more segments). The sensor data may be used to estimate the wear amount of the one or more components. The sensor data can include information identifying the time and / or date at which the sensor data was generated.

[0027] The sensor data can include past sensor data for training a machine learning model 230 to predict the wear amount of one or more components of the chassis. For example, the sensor system 120 can provide past sensor data to a wear detection device 190 for training the machine learning model 230, as described in more detail below in connection with the training of the machine learning model 230.

[0028] For example, the sensor system 120 can provide past sensor data to the wear detection device 190 (e.g., to train the machine learning model 230) periodically (e.g., every hour, at time intervals, and / or for each work shift). Additionally or alternatively, the sensor system 120 can provide past sensor data to the wear detection device 190 (e.g., to train the machine learning model 230) based on a trigger event (e.g., a request from the wear detection device 190, a request from the controller 140, and / or a request from an operator of the machine 105 (e.g., via an integrated display and / or operator control)).

[0029] After the machine learning model 230 is trained, the sensor system 120 can provide sensor data as an input to the machine learning model 230 to predict the wear amount of the one or more components. The sensor system 120 can provide sensor data as an input to the machine learning model 230 periodically and / or based on a trigger event.

[0030] The sensor device can include a vibration sensor device, a sound sensor device, a track link wear sensor device, a position sensor device, a speed sensor device, a motion sensor device, a load sensor device, a pressure sensor device, a flow rate sensor device, and / or a temperature sensor device.

[0031] The vibration sensor device can include one or more devices that sense the vibration of the machine 105 and generate machine vibration data based on the vibration. As an example, the vibration sensor device can include one or more inertial measurement units (IMUs). The machine vibration data can indicate the vibration measurement of the machine 105.

[0032] The sound sensor device can include one or more devices that sense the sound (or noise) emitted from the machine 105 and generate machine sound data based on the sound. The sound data can identify the measurement of the sound related to the machine 105. The track link wear sensor device can include one or more devices that sense the wear of the track link of the machine 105 and generate track link wear data. The track link wear data can identify the measurement of the track link wear.

[0033] The position sensor device can include one or more devices that sense the position of the machine 105 and generate position data for identifying the position. As an example, the position sensor device can include a global positioning system (GPS) receiver and / or a GPS sensor. The position data can identify the position of the machine. This position can include the work location where the machine 105 performs tasks.

[0034] The motion sensor device can include one or more devices that sense the speed associated with the machine 105 and generate speed data that identifies the speed associated with the machine 105. The motion sensor device can include an accelerometer, a tachometer, a speedometer, and / or an IMU. In some implementations, the motion sensor device may further sense the distance the machine 105 moves and generate distance data that identifies the distance the machine 105 moves. This distance can correspond to the distance the machine 105 moves during the execution of a task. Additionally or alternatively, the distance can correspond to the distance moved since one or more components have been repaired and / or replaced. The motion sensor device can monitor the amount of time the machine 105 is in use (e.g., while a task is being executed) and generate machine time data that identifies the amount of time the machine 105 is in use.

[0035] The load sensor device can include one or more devices that sense the load on the engine 110 and generate load data that identifies the load on the engine 110. The pressure sensor device can include one or more sensor devices that sense the fluid pressure of a hydraulic system that facilitates the control of the rear attachment 150 and / or the front attachment 160 of the machine 105 and generate pressure data that identifies the fluid pressure of the hydraulic system. The pressure sensor device can include a pressure sensor and / or a pressure sensor.

[0036] The flow sensor device can include one or more sensor devices that sense the flow rate of the fluid in the hydraulic system and generate flow rate data that identifies the flow rate of the fluid in the hydraulic system. The flow sensor device can include a flow sensor, a flow monitor, and / or a pump flow rate.

[0037] The temperature sensor can include one or more sensor devices that sense the temperature of different components of the machine 105 (e.g., the temperature of the hydraulic system and / or the temperature of the engine 110) and generate temperature data that identifies the temperature of different components of the machine 105.

[0038] The controller 140 can include one or more processors 240 (alternatively referred to herein as the "processor 240" and collectively referred to as the "processor 240") and one or more memories 250 (alternatively referred to herein as the "memory 250" and collectively referred to as the "memory 250"). The processor 240 can be implemented in hardware, firmware, and / or a combination of hardware and software. The processor 240 includes a central processing unit (CPU), a graphics processing unit (GPU), an accelerated processing unit (APU), a microprocessor, a microcontroller, a digital signal processor (DSP), a field-programmable gate array (FPGA), an application-specific integrated circuit (ASIC), or other types of processing components. The processor 240 can be programmed to perform functions.

[0039] The memory 250 includes random access memory (RAM), read-only memory (ROM), and / or other types of dynamic or static storage devices (e.g., flash memory, magnetic memory, and / or optical memory) that store information and / or instructions used by the processor 240 to perform functions. For example, when this function is performed, the controller 140 can obtain sensor data (e.g., from the sensor system 120), and the wear detection device 190 can predict the wear amount of one or more components based on the sensor data (e.g., using the machine learning model 230).

[0040] The wear detection device 190 can include one or more devices (e.g., a server device or a set of server devices) configured to train the machine learning model 230 to predict the wear amount of one or more components of the chassis, as described in more detail below. In some implementations, the wear detection device 190 can be implemented by one or more computing resources in a cloud computing environment. The wear detection device 190 can be hosted, for example, in a cloud computing environment. Alternatively, the wear detection device 190 can be non-cloud-based or partially cloud-based.

[0041] The inspection device 210 can include one or more devices that can provide past inspection data regarding past inspections of the machine 105. The past inspection data can include information identifying that it is related to the time and / or date when the past inspection was performed (e.g., the time and / or date of the past inspection). The past inspection may be performed (e.g., manually) at one or more locations (e.g., one or more workplaces) of the machine 105. In some examples, when the inspection data is provided, the inspection device 210 can provide data from a past inspection report regarding past inspections of the machine 105. As an example, the past inspection report can include information identifying one or more inspections performed and information identifying the time and / or date associated with the inspection report (e.g., the time and / or date of the inspection of the machine 105). The information identifying one or more inspections can include measurements of wear of one or more parts, an overall assessment of the condition of one or more parts, measurements of the wearability of tasks performed by the machine 105, the track tension of the machine 105, the environmental conditions of locations associated with the machine 105, and / or other types of inspections related to the amount of wear of one or more parts.

[0042] The past inspection data can be used to train the machine learning model 230 to predict the amount of wear of one or more parts of the chassis. For example, the inspection device 210 can provide the past inspection data to the wear detection device 190 to train the machine learning model 230, as will be described later in connection with the training of the machine learning model 230.

[0043] Past inspection data can include information obtained based on measurement values related to machine 105 (e.g., manual measurement values related to past inspections). For example, past inspection data can include wear data that identifies a measurement of the wearability of a task performed by machine 105 at one location, environmental data that identifies environmental conditions at that location during the execution of the task, and track tension data that identifies the track tension of machine 105 (e.g., as a result of performing the task at that location). The measurement of wearability can indicate the amount of wear of one or more components as a result of performing the task at this location. The environmental data can of the machine include humidity data that identifies a measurement of humidity at the location (e.g., soil humidity) and / or dryness data that identifies a measurement of dryness at the location (e.g., soil dryness).

[0044] Also, past inspection data can include operator behavior data. For example, operator behavior data can include information that identifies the speed of machine 105 during task execution, information that identifies the load on engine 110 of machine 105 during task execution, information that identifies the travel distance of machine 105 during task execution, the amount of time of machine 105 during task execution, information that identifies the fluid pressure of the hydraulic system of machine 105 during task execution, and / or information that identifies the fluid flow rate of the hydraulic system of machine 105 during task execution. Some or all of the operator behavior data can be determined based on sensor data. In some cases, the operator behavior data can include information that identifies the type of task being performed by machine 105.

[0045] The simulation device 220 can include one or more devices that can include a simulation model that simulates the operation of machine 105 (e.g., to achieve a specific measurement of wear of one or more components). The simulation device 220 can generate simulation data by simulating the operation of machine 105 and the wear of one or more components.

[0046] The simulation data can be used to train the machine learning model 230 to predict the wear amount of one or more components of the chassis. For example, the simulation device 220 can provide the simulation data for training the machine learning model 230 to the wear detection device 190 as will be described later in connection with training the machine learning model 230. In some implementations, the simulation data can indicate the correlation between the vibration measurement of the machine 105 and the wear amount of one or more components.

[0047] As shown in FIG. 2, the wear detection device 190 can receive past sensor data from the sensor system 120, past inspection data from the inspection device 210, and simulation data from the simulation device 220. The past sensor data, past inspection data, and / or simulation data may be included in the training data used to train the machine learning model 230 to predict the wear amount of one or more components. The machine learning model 230 can predict the remaining life of one or more components based on the wear amount. In some examples, the machine learning model 230 may be trained to predict the wear rate of one or more components and predict the wear amount of one or more components based on the wear rate.

[0048] In some implementations, the machine learning model 230 may be trained to predict the wear amount of one or more components of the machine 105. Further, the machine learning model 230 may be trained to predict the wear amount of one or more components of the chassis of a group of machines similar to the machine 105 (e.g., similar or identical machines, similar or identical specifications, similar or identical components, and / or similar or identical tasks being performed). In this regard, the training data can include past sensor data, past inspection data, and / or simulation data related to a set of machines.

[0049] Based on the trained machine learning model 230, the wear detection device 190 can identify factors that affect the wear rate (and / or the wear amount of one or more components) of one or more components (e.g., using the machine learning model 230). These factors can include measurements of wear, position, humidity, the behavior of the operator (which can be recognized by the tasks performed by the machine 105), travel distance, speed related to the travel distance, track tension, drawbar force, measurements of the vibration of the machine 105, and / or measurements of the sound of the machine 105.

[0050] In some examples, based on the trained machine learning model 230, the wear detection device 190 can determine the correlation (e.g., based on past inspection data) between the measurement of wear and the position of the machine 105 (e.g., using the machine learning model 230). For example, the wear detection device 190 (e.g., the machine learning model 230) can determine the measurement of wear based on the position of the machine 105. For example, the wear detection device 190 can determine that the first measurement of wear at the first position (e.g., the first workplace) is higher than the second measurement of wear at the second position (e.g., the second workplace). Therefore, the wear detection device 190 can determine (e.g., using the machine learning model 230) that the wear rate (and the corresponding wear amount) of one or more components at the first position is higher than the wear rate (and the corresponding wear amount) of one or more components.

[0051] Further, the wear detection device 190 can determine the correlation between the environmental conditions at a location and the wear rate of one or more components (e.g., using the machine learning model 230) (e.g., based on past inspection data). For example, the wear detection device 190 can determine that the humidity measurement at the first location is higher than the humidity measurement at the second location. Therefore, the wear detection device 190 can determine that the wear rate of one or more components increases as the humidity measurement increases (e.g., using the machine learning model 230). In some examples, if the wear detection device 190 determines that the machine 105 is performing the same task at the first and second locations, the wear detection device 190 can confirm that the wear rate of one or more components increases with the increase in humidity measurement.

[0052] The wear detection device 190 can determine the correlation between the tasks performed (e.g., using the machine learning model 230) and the wear rate of one or more components (e.g., based on past inspection data and / or past sensor data). For example, the wear detection device 190 can identify a first type of task (e.g., a task related to a moving body) based on the analysis of the first operator behavior data (e.g., using the machine learning model 230), and can identify a second type of task (e.g., a task not related to a moving body, such as driving to a work site) based on the analysis of the second operator behavior data. The wear detection device 190 can determine that the wear rate related to the first type of task is higher than the wear rate related to the second type of task.

[0053] In some implementations, the wear detection device 190 can determine the correlation between the tasks performed and the track tension in order to determine the correlation between the tasks performed (e.g., using the machine learning model 230) and the wear rate of one or more components. For example, the wear detection device 190 can analyze past track tension data and determine that the first track tension of the machine 105 (generated by performing the first type of task) is smaller than the second track tension of the machine 105 (generated by performing the second type of task).

[0054] The wear detection device 190 can determine the correlation between the traction lever force for performing a task (e.g., using the machine learning model 230) and the wear rate of one or more components (e.g., based on past inspection data and / or past sensor data). In some examples, the wear detection device 190 can determine traction lever force data that identifies the amount of traction lever force used by the machine during the execution of the task. The wear detection device 190 can determine the traction lever force data based on past load data, past speed data, past distance data, and / or past temperature data. The wear detection device 190 can determine that the wear rate of one or more components increases with an increase in the traction lever force (e.g., using the machine learning model 230). For example, the wear detection device 190 can determine that the wear rate of one or more components increases with the load of the engine 110.

[0055] The wear detection device 190 can determine the correlation between the wear measurement of one or more components and the vibration measurement of the machine 105 (e.g., using simulation data, past sensor data, and / or past inspection data). For example, the wear detection device 190 can determine simulation data that shows the correlation between the wear amount of one or more components and the vibration measurement of the machine 105.

[0056] The wear detection device 190 can analyze past sensor data (e.g., past mechanical vibration data) and past inspection data (e.g., past wearability data and / or past track tension) to associate vibration measurement with wearability measurement or track tension measurement. For example, the wear detection device 190 can determine that a first vibration measurement (e.g., based on first past vibration data) corresponding to a first wearability measurement (e.g., first wearability data) exceeds a second vibration measurement (e.g., based on second past vibration data) corresponding to a second wearability measurement (e.g., second wearability data) (e.g., using the machine learning model 230). The wear detection device 190 can determine that the first measurement of wearability exceeds the second measurement of wearability and the first measurement of vibration exceeds the second measurement value of vibration. The wear detection device 190 can confirm the correlation between the wear amount of one or more components and the vibration measurement of the machine 105 based on the analysis of past sensor data and past inspection data. For example, the wear detection device 190 can determine (e.g., using the machine learning model 230) that the vibration measurement increases with the increase in wear of one or more components.

[0057] Similarly, the wear detection device 190 can analyze past data of mechanical vibration and past data of track link wear to determine that the vibration measurement increases with the increase in wear of one or more components. Similarly, the wear detection device 190 can determine (e.g., using the machine learning model 230) the correlation between the wear measurement of one or more components and the sound measurement of the machine 105 (e.g., using simulation data, past sensor data, and / or past inspection data). For example, the wear detection device 190 can determine (e.g., using the machine learning model 230) that the sound measurement increases with the increase in wear of one or more components.

[0058] Based on the above, the machine learning model 230 may be trained to predict the wear rate and / or wear amount of one or more components based on position, humidity measurement, operator behavior (e.g., related to task execution), travel distance (e.g., when machine 105 executes a task), speed related to the travel distance, track tension, drawbar force (e.g., when machine 105 executes a task), vibration measurement, and / or sound measurement. The machine learning model 230 can predict the date and / or time when one or more components should be replaced and / or repaired based on the predicted wear rate and / or wear of one or more components. The predicted wear rate of one or more components, the predicted wear amount of one or more components, and / or the predicted date and time may hereinafter be referred to as "predicted component wear information".

[0059] When training the machine learning model 230, the wear detection device 190 can divide the training data into a training set (e.g., a data set for training the machine learning model 230), a validation set (e.g., a data set for evaluating the fitting of the machine learning model 230 and / or a data set for fine-tuning the machine learning model 230), a test set (e.g., a data set for evaluating the final fitting of the machine learning model 230), and / or the like. The wear detection device 190 can perform preprocessing and / or execute dimensionality reduction to reduce the training data to a minimum feature set. Since the wear detection device 190 can train the machine learning model 230 on this minimum feature set, the process for training the machine learning model 230 can be reduced, and classification techniques can be applied to the minimum feature set.

[0060] The wear detection device 190 can determine a classification result (e.g., the wear amount of one or more components) using classification techniques such as logistic regression classification technology, random forest classification technology, gradient boosting machine learning (GBM) technology, and / or similar technologies. The wear detection device 190 can use naive Bayes classifier technology in addition to or instead of using the classification technique. In this case, the wear detection device 190 can perform binary recursive partitioning to divide the training data of the minimum feature set into regions and / or branches, and use the partitions and / or branches to perform predictions (e.g., the wear rate and / or wear amount of one or more components). Based on the use of recursive partitioning, the wear detection device 190 can reduce the use of computational resources for manual linear classification and analysis of data items, enable training of the model using thousands, millions, or billions of data items, and result in a more accurate model than using fewer data items.

[0061] The wear detection device 190 can train the machine learning model 230 using a supervised training process that includes receiving inputs from experts in the target field (e.g., the machine 105 and / or one or more operators associated with one or more machines) into the machine learning model 230, which can reduce the amount of time, processing resources, and / or similar amounts for training the machine learning model 230 compared to an unsupervised training process. The wear detection device 190 can use one or more other model training technologies such as neural network technology, potential semantic indexing technology, and / or similar technologies.

[0062] For example, the wear detection device 190 can perform pattern recognition regarding different wear amount patterns of one or more components by executing artificial neural network processing technology (such as using a two-layer feedforward neural network architecture, a three-layer feedforward neural network architecture, etc.). In this case, by using artificial neural network processing technology, it is more robust against noise, inaccurate or incomplete data, and enables the wear detection device 190 to detect patterns and / or trends that cannot be detected by a human analyst or system or by a less complex technology, thereby improving the accuracy of the machine learning model 230 generated by the wear detection device 190.

[0063] After training, the predicted component wear information can be determined (or predicted) using the machine learning model 230. In other words, after training the machine learning model 230 that can output data regarding the wear rate and / or wear amount of one or more components, the wear detection device 190 can receive sensor data from the machine 105 and input the received sensor data into the machine learning model 230. The received sensor data can include position data, operator behavior data, distance data (e.g., related to the task), speed data (e.g., related to the task), traction lever force data (e.g., related to the task), vibration data, and / or sound data. The output of the machine learning model 230 can include a score of the predicted component wear information. In the case of the predicted component wear information, the score can represent a measure of the reliability of the predicted component wear information.

[0064] Different devices, such as server devices, can generate and train the machine learning model 230. The different devices can provide the machine learning model 230 for use by the wear detection device 190. The different devices can update the machine learning model 230 (e.g., based on scheduling, on-demand, trigger, periodic, and / or the like) and provide the machine learning model 230 to the wear detection device 190. In some cases, the wear detection device 190 can receive additional training data (e.g., additional past sensor data, additional past inspection data, and / or additional simulation data) and retrain the machine learning model 230. Alternatively, the wear detection device 190 can provide additional training data to another device for training the machine learning model 230. The machine learning model 230 can be retrained periodically and / or based on trigger events.

[0065] In some implementations, the wear detection device 190 can provide the machine learning model 230 to the controller 140 to enable the controller 140 to determine the predicted component wear information. Alternatively, the wear detection device 190 may receive a request from the controller 140 for determining the predicted component wear information. The request can include the sensor data of the machine 105.

[0066] The wear detection device 190 (and / or the controller 140) can perform an action based on the predicted component wear information. For example, the action can include the wear detection device 190 adjusting the operation of the machine 105 based on the predicted wear amount of one or more components (e.g., when the predicted wear amount meets a threshold wear amount). For example, the wear detection device 190 can cause a decrease in the speed of the machine 105, a decrease in the load of the engine 110, a decrease in the pressure of the hydraulic system, a decrease in the flow rate of the hydraulic system, a decrease in the temperature of the hydraulic system, and / or other operations that can reduce the wear rate of one or more components and extend the time until one or more components need to be repaired or replaced.

[0067] The wear detection device 190 can include navigating the machine 105 to different work locations and performing one or more tasks at the different work locations to extend the life of one or more parts. For example, the different work locations may be associated with a wear rate of one or more parts that is less than the wear rate of one or more parts associated with the work location where the machine 105 is currently located. Additionally or alternatively, the wear detection device 190 can cause the machine 105 to perform different tasks to extend the life of one or more parts. For example, the different tasks may be associated with a wear rate of one or more parts that is less than the wear rate of one or more parts associated with the task that the machine 105 is currently performing.

[0068] This action can include the wear detection device 190 sending remaining life information to one or more devices that monitor the wear amount of parts of multiple machines (e.g., including the machine 105). In some examples, the wear detection device 190 can send the remaining life information when the wear amount of one or more parts meets a threshold wear amount. The remaining life information can indicate the wear amount of one or more parts, the wear rate of one or more parts, the remaining life of one or more parts, and / or a proposal related to the repair and / or replacement of one or more parts. The one or more devices can include devices of a location management system, devices of a back office system, devices related to an operator of the machine 105, devices related to a technician, and / or the controller 140.

[0069] In some examples, the wear detection device 190 may send the remaining life information to cause one or more devices to order one or more replacement parts. In some cases, the remaining life information can include information identifying one or more parts and / or one or more replacement parts.

[0070] The wear detection device 190 can send remaining life information to one or more devices (e.g., the controller 140) to autonomously navigate the machine 105 to a repair facility. Additionally or alternatively, the wear detection device 190 can send the remaining life information so that one or more devices can populate the technician's calendar with calendar events for inspecting and / or repairing one or more parts. Additionally or alternatively, the wear detection device 190 can send the remaining life information to activate an alarm in one or more devices (e.g., the controller 140). The alarm may suggest that one or more parts be repaired or replaced.

[0071] In some cases, the wear detection device 190 can send the remaining life information to cause one or more devices to generate a service request for repairing and / or replacing one or more parts. As part of generating the service request, one or more devices can perform one or more actions described herein.

[0072] In some examples, the action can include causing the wear detection device 190 to deliver one or more replacement parts to a location associated with the machine 105 by a first autonomous device. This location can include the current location of the machine 105, the location of the workplace where the machine 105 performs multiple tasks, the location where the machine 105 is located when it is not performing a task, and / or the location where the machine 105 is located when it is being repaired and / or replaced. In some cases, the remaining life information can include information identifying a location associated with the machine 105.

[0073] In some examples, the action can include causing the wear detection device 190 to navigate a second autonomous device to a location associated with the machine 105 to verify predicted part wear information. The second autonomous device can generate verification information based on the verification of the part wear information and send the verification information to the wear detection device 190. The wear detection device 190 can use the verification information to retrain the machine learning model 230.

[0074] In some cases, the wear detection device 190 can determine whether a failure of one or more components will occur soon (e.g., based on predicted component wear information). If the wear detection device 190 determines that a failure is about to occur, the wear detection device 190 can execute one or more of the above-described actions. If the wear detection device 190 determines that a failure is not imminent, it may refrain from taking an action.

[0075] An example of the number and arrangement of the devices and network shown in FIG. 2 is provided. In practice, additional devices, fewer devices, different devices, or devices with different arrangements may exist compared to those shown in FIG. 2. Further, two or more of the devices shown in FIG. 2 may be implemented within a single device, or the single device shown in FIG. 2 may be implemented as a plurality of distributed devices. Additionally or alternatively, a set of devices (e.g., one or more devices) of the system 200 can perform one or more functions described as being performed by another set of devices of the system 200.

[0076] FIG. 3 is a flowchart of an exemplary process 300 related to chassis wear prediction using a machine learning model. One or more of the processing blocks in FIG. 3 can be executed by a first device (e.g., the wear detection device 190). One or more of the processing blocks in FIG. 3 can be executed by another device or set of devices, such as a controller (e.g., controller 140) that is separate from or includes the wear detection device.

[0077] As shown in FIG. 3, the process 300 can include receiving past sensor data related to wear of one or more components of a machine's chassis from one or more second devices (block 310). For example, as described above, the first device can receive past sensor data related to wear of one or more components of a machine's chassis from one or more second devices.

[0078] As further shown in FIG. 3, process 300 can include receiving, from one or more third devices, past inspection data related to wear of one or more components (block 320). For example, as described above, the first device can receive past inspection data related to wear of one or more components from one or more third devices.

[0079] As further shown in FIG. 3, process 300 can include training a machine learning model using past sensor data and past inspection data to predict a remaining life of one or more components (block 330). For example, as described above, the first device can train a machine learning model using past sensor data and past inspection data to predict a remaining life of one or more components.

[0080] The past sensor data and past inspection data are included in training data, and the training data includes two or more position data for identifying a position of the machine, distance data for identifying a distance the machine has moved while performing a task at the position, speed data for identifying a speed related to the distance the machine has moved, machine time data for identifying an amount of time the machine has performed the task, machine vibration data for identifying a measurement of vibration of the machine, machine sound data for identifying a measurement of sound related to the machine, traction lever force data for identifying an amount of traction lever force used by the machine while performing a task at the position, wearability data for identifying a measurement of wearability of the task performed by the machine at the position, and track tension data for identifying a track tension of the one or more components as a result of performing the task at the position, and training the machine learning model includes training the machine learning model using two or more of the position data, distance data, speed data, machine time data, machine vibration data, machine sound data, traction lever force data, wearability data, or track tension data.

[0081] The training data is during the execution of the task of the machineTraining a machine learning model, including environmental data that identifies environmental conditions at a location, can include using two or more of position data, distance data, speed data or machine time data, machine vibration data, machine sound data, traction lever force data, wearability data, track tension data, or environmental data to train the machine learning model.

[0082] Training a machine learning model includes training the machine learning model to predict the wear rate of one or more components and predicting the remaining life of one or more components based on the wear rate. Determining the remaining life of one or more components includes determining the amount of wear of one or more components based on the wear rate and determining the remaining life of one or more components based on the amount of wear.

[0083] As further shown in FIG. 3, process 300 can include receiving, from one or more sensor devices of the machine, sensor data related to wear of one or more components (block 340). For example, as described above, the first device can receive, from one or more sensor devices of the machine, sensor data related to wear of one or more components.

[0084] As further shown in FIG. 3, process 300 can include predicting, using the machine learning model, the remaining life of one or more components based on the sensor data (block 350). For example, as described above, the first device can predict, using the machine learning model, the remaining life of one or more components based on the sensor data.

[0085] As further shown in FIG. 3, process 300 can include causing an action to be performed based on the remaining life of one or more components (block 360). For example, as described above, the first device can perform an action based on the remaining life of one or more components.

[0086] Process 300 includes receiving simulation data from one or more fourth devices that are generated by simulating the operation of a machine and are related to the wear of one or more components. Training the machine learning model includes training the machine learning model using sensor data, past inspection data, and simulation data.

[0087] The sensor data includes machine vibration data related to the vibration measurement of the machine. The simulation data shows the correlation between the vibration measurement of the machine and the wear of one or more components. Training the machine learning model includes training the machine learning model using the machine vibration data and the simulation data.

[0088] Causing an action to be performed includes adjusting the operation of the machine based on the predicted remaining life of one or more components, transmitting the remaining life information to a device to cause the device to generate a service request for repairing or replacing at least one of the one or more components based on the predicted remaining life indicating the predicted remaining life of the one or more components, and transmitting the remaining life information to a device associated with the operator of the machine to cause the operator to adjust the operation of the machine based on the remaining life information or to transmit a service request using the device associated with the operator, including at least one of these.

[0089] Figure 3 shows an example block of process 300. However, in some implementations, process 300 can include additional blocks, fewer blocks, different blocks, or blocks in a different arrangement than those shown in Figure 3. Additionally or alternatively, two or more blocks of process 300 can be executed in parallel.

Industrial Applicability

[0090] The present disclosure relates to a method for predicting the remaining useful life of one or more components of a machine's chassis using a machine learning model. The method for predicting the remaining useful life of one or more components can prevent problems associated with manual measurements of a machine truck (for determining the wear of the truck) and inaccurate predictions of the remaining useful life of the truck.

[0091] When obtaining manual measurements, the manual measurements of the truck can waste machine resources for preventing the movement of the machine. Further, incorrect manual measurements of the truck and / or incorrect predictions of the remaining useful life of the truck can waste computing resources for improving problems associated with the incorrect manual measurements and / or incorrect predictions of the remaining useful life (e.g., premature failure of the truck, premature repair of the truck, and / or premature replacement of the truck).

[0092] The method for predicting the remaining useful life of one or more components of the chassis using a machine learning model can solve the problems associated with the above-mentioned manual measurements and inaccurate predictions of the remaining useful life. Some advantages may be associated with the disclosed method. For example, by using a machine learning model to predict the remaining useful life of one or more components, this process can prevent manual measurements of the truck (which can be inaccurate) and prevent inaccurate predictions of the remaining useful life of the truck.

[0093] This process can prevent (or limit) interruptions in the operation of the machine by preventing manual measurements, prevent the machine from being fixed when manual measurements are obtained, and prevent inaccurate predictions of the remaining useful life. This process can hold computing or machine resources that would otherwise be used to improve problems associated with the inaccuracy of manual measurements and inaccurate predictions of the remaining useful life of the truck (e.g., premature failure of the truck, premature repair of the truck, and / or premature replacement of the truck) by preventing manual measurements and preventing inaccurate predictions of the remaining useful life of the truck.

[0094] The foregoing disclosure provides illustration and description, but is not intended to limit or restrict implementation to the exact forms disclosed. Modifications and changes may be made based on the above disclosure, or obtained from the practice of the embodiments. Further, any of the embodiments described herein may be combined unless there is a clear reason in the above disclosure precluding such combination. Specific combinations of features, even if recited in the claims and / or disclosed in the specification, are not intended to limit the disclosure of the various embodiments. Each of the dependent claims listed below can only directly depend on one claim, but the disclosure of the various embodiments includes each dependent claim in combination with all the other claims in the claim set.

[0095] As used herein, "a", "one", and "a set" are intended to include one or more and may be used interchangeably with "one or more". Further, as used herein, the article "the" is intended to include one or more when combined with the article "the" and may be used interchangeably with "one or more". Further, the phrase "based on" is intended to mean "at least partially based on" unless otherwise specified. Further, as used herein, the term "or" is intended to be inclusive when used herein and may be used interchangeably with "and / or" unless otherwise expressly stated (e.g., when used in combination with "one of" or "only one of").

Claims

1. A method executed by a first device (190), comprising: receiving, from one or more second devices (120), past sensor data related to wear of one or more components of a machine chassis; receiving, from one or more third devices (210), past inspection data related to wear of the one or more components; training a machine learning model (230) using the past sensor data and the past inspection data to predict a remaining life of the one or more components; receiving, from the one or more second devices (120) of the machine, sensor data related to wear of the one or more components; using the machine learning model (230) to predict the remaining life of the one or more components based on the sensor data; and executing an action based on the remaining life of the one or more components.

2. The method of claim 1, further comprising receiving, from one or more fourth devices (220), simulation data generated by simulating an operation of the machine and related to wear of the one or more components, wherein training the machine learning model (230) comprises training the machine learning model (230) using the sensor data, the past inspection data, and the simulation data.

3. The past sensor data and the past inspection data are included in training data, and the training data includes two or more of: position data identifying a position of the machine; distance data identifying a distance the machine has moved while performing a task at the position; speed data identifying a speed related to the distance the machine has moved; machine time data identifying an amount of time the machine has been performing a task; machine vibration data identifying a measurement of vibration of the machine; machine sound data identifying a measurement of sound related to the machine; traction lever force data identifying an amount of traction lever force used by the machine while performing a task at the position; abrasiveness data identifying a measurement of abrasiveness of the task performed by the machine at the position; and track tension data identifying a track tension of the one or more components as a result of performing the task at the position. ​ ​ The step of training the machine learning model (230) includes training the machine learning model (230) using two or more of position data, distance data, speed data, machine time data, machine vibration data, machine sound data, traction lever force data, wearability data, or track tension data. The method according to any one of claims 1 and 2.

4. The past sensor data and the past inspection data are included in the training data, The training data includes environmental data for identifying environmental conditions at the position of the machine during the execution of the task, The step of training the machine learning model (230) includes training the machine learning model (230) using two or more of position data, distance data, speed data or machine time data, machine vibration data, machine sound data, traction lever force data, wearability data, track tension data, or environmental data. The method according to any one of claims 1 to 3.

5. The step of training the machine learning model (230) includes training the machine learning model (230) to predict the wear rate of the one or more parts and predicting the remaining life of the one or more parts based on the wear rate, The step of determining the remaining life of the one or more parts is, Determining the amount of wear of the one or more parts based on the wear rate; Determining the remaining life of the one or more parts based on the amount of wear. The method according to any one of claims 1 to 4.

6. A system, Receiving sensor data related to wear of one or more parts of a machine chassis from one or more second devices of the machine, Including a first device (190) configured to predict the remaining life of the one or more parts based on the sensor data using a machine learning model (230), The machine learning model (230) is, Past sensor data Past inspection data, or Using training data including two or more of simulation data of a simulation model to predict the remaining life of the one or more parts, Trained to execute an action based on the remaining life of the one or more parts, Two or more of the sensor data, the past inspection data, or the simulation data are related to wear of the one or more parts. A system.

7. When the first device (190) executes the action, To reduce the wear rate of the one or more components, the operation of the machine is adjusted based on the remaining life information of the one or more components indicating the remaining life of the one or more components. The remaining life information is transmitted to a device to cause the device to generate a service request for repairing or replacing at least one of the one or more components based on the remaining life information. Or The system according to claim 6, wherein the remaining life information is transmitted to a device associated with an operator of the machine and is configured to cause the operator to adjust the operation of the machine to reduce the wear rate of the one or more components.

8. The training data includes Position data for identifying the position of the machine, Distance data for identifying the distance the machine has moved since the one or more components were repaired or replaced, Speed data for identifying the speed associated with the distance the machine has moved, Machine vibration data for identifying a measurement of the vibration of the machine, Machine sound data for identifying a measurement of the sound generated by the machine, Traction lever force data for identifying the amount of traction lever force used by the machine while performing a task at the position, Wearability data for identifying a measurement of the wearability of the task performed by the machine at the position, The system according to any one of claims 6 and 7, including two or more of the track tension data for identifying the track tension of the one or more components as a result of the task being performed at the position.

9. The first device (190) is configured to retrain the machine learning model (230) using two or more of position data, distance data, speed data, machine vibration data, machine sound data, traction lever force data, wearability data, or track tension data. The system according to any one of claims 6 to 8.

10. The training data further includes humidity data for identifying a humidity measurement at the position of the machine. The first device (190) is configured to retrain the machine learning model (230) using the humidity data and two or more of the position data, distance data, speed data, Machine vibration data, machine sound data, traction lever force data, wearability data, track tension data. The system according to any one of claims 6 to 9.

Images