System and method for diagnosing vehicle components

Abstract

The system includes a vehicle component installed within the vehicle and a sensor coupled to the vehicle component. The sensor is configured to acquire operational data and output a signal including the operational data. The system also includes a communication device configured to receive the signal and wirelessly transmit the operational data to an electronic processor located on a server external to the vehicle. The electronic processor is configured to analyze the operational data via a machine learning algorithm, generate a health state for the component, and transmit the health state for the component from the electronic processor to an external device in response to the generated health state.

Description

[Technical Field]

[0001] Aspects, features, and embodiments described herein relate to a system for vehicle component health detection. Summary of the Invention [Problem to be solved by the invention]

[0002] Many vehicles include components, such as air filters or batteries, that are designed to be replaced after use or over time. In addition, many vehicles include tools for monitoring the overall health of those components. Some vehicles include on-board computer systems designed to diagnose the health of components and provide an indication that replacement or repair is necessary. It may be desirable to improve the accuracy of such diagnoses. Accordingly, embodiments described herein provide, among other things, systems and methods for detecting the health of vehicle components. [Means for solving the problem]

[0003] Some examples provide a system for diagnosing the health of a vehicle component. In one example, the system includes a vehicle component installed, for example, within the vehicle, and a sensor coupled to the vehicle component. The sensor is configured to acquire operational data and output a signal including the operational data. The system also includes a communication device configured to receive the signal and wirelessly transmit the operational data to an electronic processor located on a server external to the vehicle. The electronic processor is configured to analyze the operational data via a machine learning algorithm, such as a linear regression algorithm, to generate a health state of the component, and to transmit the health state of the component from the electronic processor to an external device in response to the generated health state.

[0004] Another example provides a method for diagnosing a vehicle component. In one example, the method includes, via an electronic processor, acquiring operational data corresponding to operational parameters of the vehicle component and, via the electronic processor, determining, using the operational data, a plurality of machine learning model features, such as classification model features. The method also includes, via the electronic processor, matching the plurality of classification model features to a plurality of ordered rows and inputting the plurality of ordered rows to the machine learning model. The method also includes, via a linear regression model, determining a classification score for each of the plurality of ordered rows and determining an average classification score for the plurality of ordered rows. The method also includes determining a health state of the vehicle component based on the average classification score and, in response to determining the health state of the vehicle component, transmitting the health state of the vehicle component to an external device. In some examples, alternatives to the linear regression model may be used when determining the classification scores.

[0005] Another example provides a method for diagnosing the health of a motorcycle battery, the method including: acquiring, via an electronic processor, operational data corresponding to operating parameters of the motorcycle battery; and determining, via the electronic processor, a plurality of classification model features using the operational data. The method also includes, via the electronic processor, matching the plurality of classification model features to a plurality of ordered rows and inputting the plurality of ordered rows to an algorithm. The method also includes, via the algorithm, determining a classification score for each of the plurality of ordered rows and calculating, via the electronic processor, an average classification score for the plurality of ordered rows. The method also includes, via the electronic processor, determining a vehicle component health state based on the average classification score and, in response to determining the motorcycle battery health state, transmitting the motorcycle battery health state to an external device.

[0006] Other aspects, features, and embodiments will become apparent by consideration of the detailed description and accompanying drawings. [Brief explanation of the drawings]

[0007] [Figure 1] FIG. 1 is a diagram of a system for vehicle component health detection, according to some aspects. [Figure 2] 1 is a flowchart of a process for detecting the health of a vehicle component, according to some aspects. [Figure 3] FIG. 1 is a diagram of data flow to feature computation and through a regression model, according to some aspects. [Figure 4] 1 is a graph of the flow coefficient of a vehicle air filter, according to some embodiments. [Figure 5] FIG. 1 is an illustration of a process for detecting the health of a vehicle component, according to some aspects. [Figure 6] 1 is a graph of a starting response of a vehicle battery, in accordance with some embodiments. DETAILED DESCRIPTION OF THE INVENTION

[0008] The patent or application file contains at least one drawing executed in color. Copies of this patent or patent application publication with color drawing(s) will be provided by the Office upon request and payment of the necessary fee.

[0009] FIG. 1 is a diagram of a system for detecting the health of vehicle components, according to some examples and embodiments. The system 100 includes a vehicle 105, such as a motorcycle (or bike). The vehicle 105 includes replaceable or serviceable components, such as an air filter, a battery, tires, one or more brake pads, a chain, and a belt. The vehicle 105 also includes one or more sensors configured to monitor the vehicle components and generate vehicle operating parameters. For example, the vehicle 105 includes a sensor 102 configured to detect air flow through an air filter and generate a signal. Operational data 107 corresponding to the operating parameters of the air filter can be derived from the signal. In one example, the operating parameters of the air filter include the amount of air passing through the air filter or the percentage of the total amount of air filter clogged by trapped foreign matter. Other vehicle components may include additional sensors, such as a sensor for detecting battery voltage, a sensor for detecting brake pad wear, or a sensor for detecting tire pressure. In the following description, operational data 107 refers to operational data derived from information or signals provided by one or more sensors (e.g., the sensors described above) or timing devices of the vehicle. Thus, operational data 107 may include information about general vehicle operation, such as the time of day that the vehicle 105 is traveling, the duration that the vehicle 105 is traveling, the load on the vehicle 105, the vehicle speed (e.g., a speed log), engine revolutions per minute (RPM), engine speed, and throttle position (or a throttle log), etc.

[0010] The operational data 107 is transmitted or otherwise provided to an embedded controller, such as the controller 110 in the vehicle 105. In one example, the controller 110 includes an input / output interface 120, an electronic processor 125, and a memory 130. In some examples, the electronic processor 125 is implemented as a microprocessor with a separate memory, e.g., the memory 130. In other examples, the electronic processor 125 may be implemented as a microcontroller (with the memory 130 on the same chip). In other examples, the electronic processor 125 may be implemented using multiple processors. Additionally, the electronic processor 125 may be implemented partially or fully as, for example, a field programmable gate array (FPGA), an application specific integrated circuit (ASIC), etc., and the memory 130 may be unnecessary or modified accordingly. In some examples, the memory 130 includes a non-transitory computer-readable memory that stores instructions received and executed by the electronic processor 125 to perform the methods described herein. The memory 130 may include, for example, program storage and data storage. The program storage area and the data storage area may include a combination of different types of memory, such as, for example, read-only memory and random-access memory. The input / output interface 120 may include one or more input mechanisms and one or more output mechanisms (e.g., a general-purpose input / output (GPIO), a controller area network bus (CAN) bus interface, an analog input, and a digital input, etc.). The controller 110 is configured to wirelessly communicate with a cloud-based server, such as a first server 140, via a wireless link or connection 135. The wireless connection 135 may be a wireless access network connection, a cellular connection, a satellite connection, etc. In some examples, the first server 140 is owned or operated by an original equipment manufacturer (OEM). The OEM may provide services based on the operational data 107, such as component repair, replacement, or other services.

[0011] In another example, the operational data 107 is communicated to an external device, such as a dongle 115, configured to communicate with the vehicle 105. The dongle 115 includes internal input / output interfaces, an electronic processor, and memory similar to the components described for the controller 110. The dongle 115 is configured to communicate with an external device 145, such as a smartphone, tablet, or laptop. The dongle 115 preferably communicates with the external device 145 wirelessly via a near-field communication (NFC) connection, a short-range wireless connection (e.g., a Bluetooth® connection), or an IEEE 802.11 protocol connection (e.g., Wi-Fi). In another example, the dongle 115 communicates with the external device 145 via a wired connection. The external device 145 is configured to wirelessly communicate with a cloud-based server. For example, the external device 145 wirelessly communicates with the first server 140 via the wireless connection 135, which is functionally similar to the wireless communication between the controller 110 and the first server 140.

[0012] The first server 140 is configured to receive the operational data 107 from the controller 110 or the external device 145. The first server 140 is also configured to communicate with other cloud-based servers, such as the second server 150. In some examples, the first server 140 acts as a relay communication service to collect the operational data 107 from the vehicle 105 before transmitting the operational data 107 to the second server 150. In other examples, the first server 140 processes, compresses, collates, or classifies the operational data 107 before transmitting the operational data 107 to the second server 150. In some examples, the transfer of data is facilitated by an application programming interface (API) 155. For example, the first server 140 may transmit the operational data 107 to the second server 150 via the application programming interface (API) 155. For example, in some examples, the second server 150 generates a call requesting the operational data 107 from the first server 140. In another example, the operational data 107 is batch processed by the first server 140 before being pushed to the second server 150. The second server 150 processes the operational data 107 using an algorithm to determine the health status of the vehicle components associated with the operational data. The algorithm and process for determining the health status of the vehicle components is described below in conjunction with Figures 2-6.

[0013] In one example, after the second server 150 generates the health status of the vehicle components, the second server 150 transmits insights 160 to the first server 140. The insights 160 include actionable advice based on the component health status. For example, the insights 160 may provide actionable advice including replacing an air filter that the second server 150 has determined to be in a low health state. In another example, the insights 160 may include actionable advice including repairing an air filter that the second server 150 has determined to be in a medium health state. Alternatively, the insights 160 may include information that the air filter does not require repair or replacement and is in a good health state. The second server 150 is configured to output these insights 160 to any combination of a vehicle operator, such as a rider 165, an OEM 170, or a service technician 175, via the interface 180. The interface 180 may be a wireless or wired interface. In some examples, aspects of the algorithms, first server 140, or second server 150 may be combined or divided into separate software, firmware, and / or hardware. For example, instead of being located within and executed by a single electronic processor, logic and processing may be distributed across multiple electronic processors. Regardless of how they are combined or divided, the hardware and software components may be located on the same computing device or distributed among different computing devices connected by one or more networks or other suitable communication links.

[0014] FIG. 2 is a diagram of an example process 200 for vehicle component health detection, according to some embodiments. In some cases, process 200 is executed on a second server 150. In such an example, second server 150 obtains operational data 107 (step 205). Next, second server 150 determines whether data loss of the operational data is within acceptable limits (step 210). If the data loss is not within acceptable limits, the process stops (step 255). If the data loss is within acceptable limits, second server 150 determines relevant features of a classification model using a machine learning algorithm (step 215). In this example, the classification model is a linear regression algorithm. In some examples, the relevant features of the classification model are also referred to as multiple classification model features. The relevant features of the classification model include weights assigned to individual data in operational data 107. In some examples, different weights are assigned to different data depending on the particular vehicle component. For example, when calculating the state of health of an air filter, particular weight values ​​are assigned to data relating to intake air temperature, throttle opening, air flow rate, or air pressure, etc. In another example, different weight values ​​are assigned to data when calculating the state of health of a vehicle battery. In some cases, the weights of the operational data depend on the particular vehicle and its operating conditions, with the impact of each factor varying between vehicles. In some cases, the classification model uses alternative or additional regression models.

[0015] Process 200 includes matching operational data 107 into a plurality of ordered rows. The plurality of ordered rows includes individual rows of data, which are matched by second server 150 (step 220). Once the operational data has been matched into the individual rows, second server 150 determines whether a measurement entry condition for the individual row is met (step 225). In some examples, the entry condition includes a predetermined threshold, such as engine throttle position or engine RPM. Second server 150 determines whether the measurement entry condition for the individual row is not met. If so, second server 150 continues matching the data (the process returns to step 220). If second server 150 determines that the measurement entry condition for the individual row is met, process 200 proceeds (step 230), where associated features are input into a classification model to calculate a score for the individual row. The classification model then generates a classification score for each individual row (step 235). Next, the second server 150 determines whether a minimum number of data points has been achieved (step 240). The minimum number of data points is required to calculate a final health state score. If the second server 150 determines that the minimum number of data points has not been achieved, the process 200 stops (step 255). If the second server 150 determines that the minimum number of data points has been achieved, the classification model calculates an average of the scores of the individual rows, also referred to as an average classification score of multiple ordered rows (step 245). The classification model then calculates a final health state score (step 250) and stops at step 255. In some examples, the final health state is further output to another device, such as an external device. For example, in some cases, the final health state of the vehicle component is transmitted to the rider 165, the OEM 170, or a service technician 175. In this example, the process 200 is executed by the second server 150. However, in other examples, the process 200 is performed by another server, such as the first server 140 .

[0016] 3 is a diagram of data flow to feature calculation and data flow through a regression model, also referred to as process 300, according to some embodiments. In some examples, the regression model is executed by an electronic processor of second server 150. Process 300 is applied to an air filter of vehicle 105 and is similar to process 200. Process 300 includes step 305 in which second server 150 obtains operational data 107 including entry conditions such as the vehicle's engine throttle opening 310, engine speed 315 (also known as engine revolutions per minute), engine manifold pressure level 320, and intake air temperature 325. Second server 150 then uses the operational data 107 in step 330 to assign mathematical weights to the operational data during feature calculation. Process 300 includes step 335, in which second server 150 uses the weights generated during step 330 to generate weighted data, including a weighted flow coefficient 340, a weighted engine RPM 345, a weighted engine throttle opening 350, and a differential pressure value 355. In step 360, the weighted data is used to calculate the air filter health state via a classification model or a linear regression model. For example, the air filter health state may include a clogged air filter percentage. This clogged percentage is then output by the linear regression model in step 365. Optionally, other vehicle components are analyzed via a similar process using the same or other operational data.

[0017] FIG. 4 is a graph 400 of a vehicle air filter's flow coefficient, according to some embodiments. In some examples, graph 400 corresponds to process 300 described above. Graph 400 includes a flow coefficient trace 405. In this example, flow coefficient trace 405 is generated by a classification model of process 200 or process 300 described above. In some cases, the airflow classification model is a linear regression model. The linear regression model uses a quadratic equation that includes coefficient weights for various vehicle components, as described above. For example, in some examples, each coefficient of a variable in the quadratic equation is associated with a different weight. As described above, the weights differ for various components and vary from vehicle to vehicle. Additionally, in some examples, the quadratic equation includes constants that also vary based on component and vehicle variations.

[0018] FIG. 5 illustrates an example process 500 for vehicle battery health detection, according to some embodiments. Process 500 may be performed, for example, by second server 150 and may include monitoring vehicle 105 running and engine starting conditions. In some examples, the battery is monitored by sensor 102 or a similar sensor configured to detect the battery's voltage level. For example, the battery is monitored during startup or cranking of vehicle 105. In particular, the battery is monitored for the coup-de-fouet effect after cranking begins until the engine speed exceeds a certain threshold, as shown in FIG. 6. In some cases, the threshold occurs within the first 500 milliseconds of cranking. In some cases, the coup-de-fouet effect coincides with the initial moment of electrical connection between the battery and the load. For example, during engine startup, the battery experiences the coup-de-fouet effect as it activates the engine starter and initiates cranking of the engine crankshaft. Additionally, process 500 includes monitoring and recording engine RPM and engine temperature, as described above. In some cases, process 500 monitors the coup de feuillete effect and correlates operational data 107 with the battery's state of health. Process 500 includes step 505, in which a vehicle start is activated. In step 510, when the vehicle is started (e.g., on the first run of the day), a logging program running on electronic processor 125 of controller 110 begins logging operational data 107 and saving the data to a first log file 517. The logging program includes a timer configured to increment only if engine RPM is greater than a predetermined RPM threshold and engine throttle opening is greater than a predetermined throttle opening threshold. When the timer exceeds the predetermined timer threshold and the battery voltage after engine shutdown exceeds the battery threshold, monitoring for the next engine start or engine crank is enabled. Process 500 includes step 515, in which a log file for the first run of the day is generated. The first log file 517 contains the operational data 107 collected during the first run of the day.The engine is shut off by the rider and the log file is completed in step 520. The log file is then sent to the first server 140 in step 525.

[0019] Process 500 also monitors additional or subsequent vehicle runs, such as when vehicle 105 is started in step 530. A second log file 532 for the subsequent vehicle run is generated similarly to first log file 517, as described above. During the subsequent vehicle run, the engine may experience an engine-on phase 535 and an engine-off phase 540. If engine monitoring is enabled as described above and the engine-off time is greater than a predetermined threshold, monitoring is disabled. Process 500 includes step 545, in which vehicle 105 is stopped, ending the subsequent vehicle run. Second log file 532 is then sent to first server 140, as described above. Process 500 includes a final run of the day, in which the vehicle is started in step 550 and stopped in step 555. A third log file 557 for the final run of the day is generated similarly to the previous log files in step 560. The log file includes operating data 107 used to generate the motorcycle's battery health status, as described above. In some examples, the health status is the total battery life remaining or the approximate number of engine cranks remaining. In some cases, the health status includes other information such as the available current, the total voltage, the approximate power output, or the estimated charge time.

[0020] 6 is a graph 600 of a starting response of a vehicle battery, according to some embodiments. Graph 600 includes a first trace 605 representing a battery in a medium state of health. Graph 600 also includes a second trace 610 representing a battery in a good state of health. Graph 600 also shows a 13.5 volt trace 615 as a reference for first trace 605 and second trace 610. Graph 600 also shows a coup de feu effect 620 corresponding to engine starting, as previously described.

[0021] Accordingly, various implementations of the systems and methods described herein provide, among other things, techniques for vehicle component health detection. Other features and advantages of the invention are set forth in the following claims.

[0022] In the foregoing specification, particular examples have been described. However, those skilled in the art will recognize that various changes and modifications can be made without departing from the scope of the invention as set forth in the following claims. Accordingly, the specification and figures are to be regarded in an illustrative rather than a restrictive sense, and all such modifications are intended to be included within the scope of the present teachings.

[0023] Advantages, advantages, solutions to problems, and any elements that may cause or make more pronounced any advantage, advantage, or solution should not be construed as critical, necessary, or essential features or elements of any or all claims. The present invention is defined solely by the appended claims, including any amendments made during the pendency of this application and all equivalents of those claims as issued.

[0024] Additionally, relational terms such as first and second, and upper and lower, may be used herein solely to distinguish one entity or operation from another, without necessarily requiring or implying an actual relationship or order between such entities or operations. The terms "comprises," "comprising," "has," "having," "includes," "including," "contains," "containing," or any other variation thereof, are intended to cover a non-exclusive inclusion, so that a process, method, article, or apparatus that comprises, has, includes, or contains a list of elements does not include only those elements, but may include other elements not expressly listed or inherent to such process, method, article, or apparatus.

[0025] An element preceded by "comprises," "has," "includes," or "includes," without further constraints, does not exclude the presence of additional identical elements in the process, method, article, or apparatus that comprises, has, includes, or encompasses the element. The terms "a" and "an" are defined herein as one or more, unless expressly stated otherwise. The terms "substantially," "essentially," "approximately," "about," or any other variation thereof, are defined as close as understood by one of ordinary skill in the art, and in one non-limiting example, the terms are defined as within 10%, within 5%, within 1%, and within 0.5%. The term "coupled," as used herein, is defined as connected, although not necessarily directly, and not necessarily mechanically. A device or structure that is "configured" in a certain way is configured in at least that way, but may also be configured in ways not listed. [Explanation of symbols]

[0026] 100 systems 102 Sensors 105 vehicles 107 Operational Data 110 Controller 115 Dongle 120 Input / Output Interface 125 Electronic Processor 130 memory 135 Wireless Connection 140 First Server 145 External Devices 150 Second Server 155 Application Programming Interface 160 Insight 165 Rider 170 OEM 175 Service Technicians 180 Interface 300 processes 310 Engine throttle opening 315 engine RPM 320 Engine Manifold Pressure Level 325 Intake temperature 340 Weighted Discharge Coefficient 345 weighted engine RPM 350 weighted engine throttle opening 355 differential pressure value 400 graphs 405 Flow Coefficient Trace 500 processes 517 First Log File 532 Secondary Log File 535 Engine Operation Phase 540 Engine Shutdown Phase 557 Third Log File 600 graphs 605 First Trace 610 Second Trace 615 13.5 volt trace 620 Coupe de Hue Effect

Claims

1. 1. A system for diagnosing the health of a vehicle component, the system comprising: Vehicle components and a sensor coupled to the component of the vehicle, the sensor configured to acquire operational data and output a signal including the operational data; a communications device configured to receive the signal and wirelessly transmit the operational data to an electronic processor located on a server external to the vehicle; Equipped with the electronic processor is configured to analyze the operational data via a machine learning algorithm to generate a health state of the component; transmitting the health status of the component from the electronic processor to an external device in response to the generated health status; system.

2. The system of claim 1 , wherein the component of the vehicle is an air filter.

3. The system of claim 1 , wherein the communication device is a dongle device installed in the vehicle and configured to wirelessly communicate with the server via cellular communication.

4. The system of claim 1 , wherein the communication device is a vehicle central control unit configured to wirelessly communicate with the server via cellular communication.

5. The system of claim 1 , wherein the vehicle is a motorcycle and the component of the vehicle is a motorcycle battery.

6. 6. The system of claim 5, wherein the sensor is configured to measure the voltage level of the motorcycle battery until the motorcycle engine RPM exceeds a threshold value.

7. The system of claim 5 , wherein the operational data includes a measured coup-de-fouet effect.

8. The system of claim 7 , wherein the health status of the component is based on the measured coup de feu effect.

9. 1. A method of diagnosing a vehicle component, the method comprising: obtaining, via an electronic processor, operational data corresponding to operational parameters of the vehicle component; determining, via the electronic processor, features for a plurality of classification models using the motion data; collation, via the electronic processor, of the plurality of classification model features into a plurality of ordered rows; inputting the plurality of ordered rows into a machine learning model; determining a classification score for each of the plurality of ordered rows via the machine learning model; and determining a mean classification score for the plurality of ordered rows; determining a state of health of the vehicle component based on the average classification score; In response to determining the health status of the vehicle component, transmitting the health status of the vehicle component to an external device; A method comprising:

10. The method of claim 9 , wherein the plurality of classification model features include one selected from the group consisting of air filter flow coefficient, vehicle speed, vehicle throttle opening, and rate of change of air flow velocity pressure.

11. 10. The method of claim 9, wherein the vehicle component is a motorcycle air filter.

12. The method of claim 9 , wherein the plurality of classification model features comprises mathematical weights assigned to the vehicle components.

13. The method comprises: acquiring second operational data corresponding to a second operational parameter of a second vehicle component; calculating features of a second plurality of classification models using the second motion data; matching the second plurality of classification model features to the plurality of ordered rows; inputting the plurality of ordered rows into the machine learning model; 10. The method of claim 9, further comprising:

14. 10. The method of claim 9, wherein the vehicle component is a motorcycle air filter and the operational data includes at least one of the group consisting of engine throttle opening, engine RPM, engine manifold pressure, and intake air temperature.

15. 1. A method for diagnosing the health of a motorcycle battery, said method comprising: obtaining, via an electronic processor, operational data corresponding to operating parameters of a battery of the motorcycle; determining, via the electronic processor, features for a plurality of classification models using the motion data; collation, via the electronic processor, of the plurality of classification model features into a plurality of ordered rows; inputting said plurality of ordered rows into an algorithm; determining, via the algorithm, a classification score for each of the plurality of ordered rows; calculating, via said electronic processor, an average classification score for said plurality of ordered rows; determining, via the electronic processor, a health status of the vehicle component based on the classification score; responsive to determining the state of health of the motorcycle battery, transmitting the state of health of the motorcycle battery to an external device; A method comprising:

16. 16. The method of claim 15, wherein the operating parameters of the motorcycle battery include a measured coup de feuille effect.

17. The method of claim 15 , wherein the algorithm is a machine learning model.

18. 16. The method of claim 15, wherein the operational data includes voltage measurements recorded during the first 50 milliseconds of operation of the motorcycle battery.

19. The method of claim 15, wherein the state of health of the motorcycle battery comprises a total battery life remaining.

20. 20. The method of claim 19, wherein the total remaining battery life is transmitted to a rider of the motorcycle.

Images