Tire Load Monitoring

Abstract

1. A computer-implemented method (100) for monitoring a current load (201) on a tire (3) mounted on a vehicle (2), the method (100) including: obtaining (e.g., by a server device (7)) sensor readings (23) from one or more tire monitoring system (TMS) sensors (5) mounted on the tire (5); accessing (e.g., by the server device (7)) characteristic information (25 a) related to the tire (5); selecting (e.g., by the server device (7)) an appropriate one of a plurality of computer models (15) based on the characteristic information (25 a); and using (e.g., by the server device (7)) the selected computer model (15) to estimate a current tire load value (201) based on the sensor readings (23).

Description

[Technical Field]

[0001] The present invention relates to a method for monitoring the current load on a tire mounted on a vehicle, and further to a system for monitoring the current load on a tire. [Background technology]

[0002] Overloading a tire can lead to potential failure with the risk of an accident, while loading with incorrect inflation pressure can increase tire distortion, reduce the durability of the casing, and cause irregular wear. Overloading a vehicle can also lead to driving instability and legal violations.

[0003] Sensors mounted inside the tire are commonly called tire monitoring system (TMS) sensors. TMS sensors are used to monitor several parameters of the tire itself, such as inflation pressure, internal air temperature, tire rotation time, etc., as well as to extract information about the interaction of the tire with its surrounding environment (e.g., road and vehicles).

[0004] Current methods of monitoring tire load use telematics information such as vehicle speed, acceleration, engine RPM, engine load, and shock absorber data. Such methods can be inaccurate and may require the collection of a significant amount of data, meaning that accurate load estimation calculations can take a significant amount of time.

[0005] It is known to use acceleration sensors located within the tire to measure the amount of deflection when the tire makes contact with the ground (the so-called "contact patch").

[0006] There remains a need for a fast and reliable load assessment that can be used to monitor tire usage. Summary of the Invention

[0007] According to a first aspect of the present invention there is provided a computer implemented method of monitoring a current load on a tire fitted to a vehicle, the method comprising: obtaining (e.g., by a server device) sensor readings from one or more tire monitoring system (TMS) sensors attached to the tires; accessing (e.g., by a server device) characteristic information regarding the tire; selecting (e.g., by a server device) an appropriate one of a plurality of computer models based on the characteristic information; and using (eg, by a server device) the selected computer model to estimate a current tire load value based on the sensor readings.

[0008] It will be understood that the term "current" as used herein means real-time or near real-time, so that monitoring the current load means that the load is estimated in real-time or with a small delay. Similarly, current sensor readings are sensor readings taken at or near the time that the load estimation is performed.

[0009] Those skilled in the art will appreciate that current sensor readings can be more effectively used to estimate current tire loads by selecting an appropriate (e.g., adjusted) model based on the characteristic information. Because multiple different models are provided and an appropriate model is selected based on the tire characteristic information, the model itself does not need to consider the tire characteristic information as a variable. Therefore, the selected model can be simpler and therefore less computationally demanding. This has the added advantage that the model can be run on a device with less processing power and / or can be run more quickly. Clearly, quickly estimating the current tire load is important, especially if the tire is overloaded, as this can lead to dangerous tire failure. Furthermore, the reduction in variables within each model leads to a more accurate estimation of the current tire load.

[0010] In some embodiments, the characteristic information about the tire includes one or more of the tire manufacturer, tread pattern, tire specification, tire compound, tire stiffness, tire size, tire mounting position, and retread information.

[0011] In some preferred embodiments, the plurality of models includes separate models that are specifically optimized for known tire products from a given manufacturer. By providing separate models for specific tire products, the complexity of each model is significantly reduced, and thus the models can be used to more quickly and accurately calculate current tire loads while being less computationally demanding.

[0012] In some embodiments, the method includes accessing characteristic information related to the vehicle. When the characteristic information related to the vehicle is accessed, an appropriate one of the plurality of models may be selected based on both the characteristic information related to the tire and the characteristic information related to the vehicle.

[0013] In some embodiments, the characteristic information about the vehicle includes one or more of the vehicle manufacturer, vehicle chassis type, vehicle usage, number of axles, product type, number of tires, and number of tires mounted per axle.

[0014] Such vehicle characteristic information can contribute to the accuracy of load estimation. Vehicle usage can be defined as a road type driving profile, which can include any information that can be derived from the amount of driving performed per road type. For example, the road type driving profile can include a breakdown of each road type and the distance driven per road type. It can also optionally include the average time driven per road type for a particular vehicle or vehicle type. It can further optionally include a percentage based on the distance and / or time driven per road type. For example, it can be known that 60% of a vehicle's total distance is driven on highways, 20% is driven on suburban roads, and 10% is driven on urban roads.

[0015] A product type may be defined as a particular vehicle product from a given manufacturer.

[0016] In some embodiments, the plurality of models includes separate models specifically optimized for known tire products from a given manufacturer and adapted to known vehicle products from a given manufacturer.

[0017] In some embodiments, each of the multiple models is a fitting algorithm developed by studying the correlation between each parameter from the sensor readings and the tire load of a tire (and optionally also a vehicle) with particular characteristic information.

[0018] In some embodiments, the method includes obtaining current telematics information from the vehicle. In such embodiments, the telematics information may include one or more of GPS location, vehicle speed, vehicle lateral / longitudinal acceleration, road type, engine load, gear shift, engine RPM, wheel speed, throttle / brake pedal position, tire temperature, exterior temperature, and / or steering wheel angle.

[0019] In some embodiments, the method includes detecting a favorable load measurement condition using current telematics information and obtaining TMS sensor readings when the favorable load measurement condition is detected. One or more of engine load information, throttle input, brake input, steering input, lateral vehicle acceleration, and longitudinal vehicle acceleration can be used to detect the favorable load measurement condition. The load on the tires changes as a result of dynamic vehicle movement; for example, front tire load increases under braking. Detecting a favorable load measurement condition can reduce the likelihood of erroneous load estimation.

[0020] In some embodiments, the sensor readings comprise one or more of a radial acceleration value, an inflation pressure, and an internal air temperature. For example, one or more tire monitoring system (TMS) sensors attached to the tire may include an acceleration sensor and / or a pressure sensor and / or a temperature sensor.

[0021] It is now understood that radial acceleration values ​​can be particularly useful for load estimation because they provide an accurate measurement of a tire's rolling dynamics. This approach can be used in addition to, or separately from, more traditional sensor readings of pressure and / or temperature. Thus, in some embodiments, a method further includes obtaining sensor readings including radial acceleration values, analyzing the radial acceleration values ​​to calculate one or more tire parameters selected from tire contact patch length, tire wear estimation, tire rotation time (t_rev), automatic mounting position identification, and tire mileage estimation, and estimating a current tire load value based on one or more of the sensor readings and the tire parameters using a selected model. Tire contact patch length may be defined as the total time a single portion of the tire remains on the contact patch (t_patch), or may be defined as a contact patch ratio, which is the ratio between the total time a single portion of the tire remains on the contact patch (t_patch) and the total tire rotation time (t_rev). Automatic mounting position identification may be defined as identifying the location of the tire on an axle / chassis (e.g., left side, front axle).

[0022] In some embodiments, the method includes providing feedback to the user related to the current tire load value, hi some embodiments, the feedback is provided in real time.

[0023] In some embodiments, the feedback includes one or more of a notification that a tire load index has been exceeded, a notification that a tire is running at an incorrect inflation pressure for the current load, a notification that the vehicle is exceeding its maximum load capacity, and a notification that the vehicle has an uneven weight distribution across its tires. Early notification of a load problem may enable a user, who may be a driver or fleet manager, to take steps to correct the problem and reduce the likelihood of a failure such as a tire blowout.

[0024] In some embodiments, the model is used to detect the current tire load based on sensor readings and telematics information. Leveraging both TMS sensor readings and telematics information increases the amount of data input into the model, which may result in more accurate load estimation.

[0025] In some preferred embodiments, the sensor readings include inflation pressure and radial acceleration values, which are analyzed (e.g., by a TMS sensor) to provide information about tire contact patch length and tire rotation time (t_rev), and the characteristic information about the tire includes tire compound or tire stiffness and tire size. While the radial acceleration and inflation pressure are raw sensor readings, the footprint length and tire rotation time (t_rev) can be calculated using the measured raw values ​​for radial acceleration. In embodiments, the rotation time may be measured directly. In embodiments, the rotation time (t_rev) (whether measured directly or calculated using the radial acceleration values) can be used to calculate the tangential velocity of the tire using the tire radius. The inflation pressure may be a non-normalized pressure reading or a normalized pressure reading, where the normalized pressure reading is the internal air pressure divided by the internal temperature.

[0026] In an embodiment, the method includes estimating the tire contact patch length and tire rotation time (t_rev) by evaluating peaks that appear in a waveform obtained by differentiating a time series waveform of tire radial acceleration detected by an acceleration sensor attached to the tire. Such a technique is known from WO2020071249(A1), which is incorporated herein by reference.

[0027] In some embodiments, the tangential tire velocity v is:

[0028]

number

[0029] In another embodiment, the tangential tire velocity v is:

[0030]

number

[0031] In some embodiments, the characteristic information about the tire includes the tire compound or tire stiffness, and the tire size.

[0032] In some embodiments, the method includes suggesting an action to the user, optionally the action being load rebalancing and / or pressure adjustment.

[0033] In some embodiments, the current tire load value is calculated only once per trip made by the vehicle. Because a vehicle is unlikely to be loaded or unloaded while in motion, tire loads are unlikely to change significantly during a trip. Reducing load estimation to once per trip can save energy, which is particularly advantageous when using battery-powered TMS sensors.

[0034] In some embodiments, a method includes detecting the start of a journey using tire monitoring system (TMS) data and / or telematics data from one or more TMS sensors, and estimating a current tire load value in response to detecting the start of the journey.

[0035] In some embodiments, the start of run is detected by determining the presence of a non-zero radial acceleration reading in the TMS data after a period of inactivity.

[0036] In an embodiment, the length of the period of inactivity is compared to a threshold, and if the time exceeds the threshold, a decision is made to start a new run.

[0037] By comparing the period with a threshold, the possibility of false start of journey detection, for example after the vehicle has briefly stopped at a traffic light and then started off.

[0038] In a preferred set of embodiments, a time history set of the vehicle's GPS location is collected from telematics data, and the start of a journey is determined based on the GPS location changing after a stationary period. In embodiments, the time the vehicle has been stationary is compared to a threshold, and if the time exceeds the threshold, a new journey start is determined.

[0039] Possibility of false start of journey detection, for example after the vehicle has been stationary for a short time at a traffic light and then started off, by comparing the time the vehicle has been stationary with a threshold.

[0040] In other embodiments, the start of a journey may be detected by identifying an "ignition on" signal from the telematics data.

[0041] In an embodiment, after start-of-motion is detected, the method includes periodically acquiring data for a predetermined number of tire revolutions over a period of time after start-of-motion is detected. Data acquisition can begin as soon as start-of-motion is detected or shortly thereafter, for example, as soon as an optimal load estimation condition is detected after start-of-motion is detected.

[0042] In some embodiments, the current tire load value is calculated at fixed time intervals or fixed distance traveled intervals. This can reduce energy consumption by reducing the number of load estimations performed. Therefore, the methods described herein may be repeated at fixed time intervals or fixed distance traveled intervals.

[0043] In some embodiments, the method includes storing the estimated current tire load value. The estimated current tire load value may be stored on a server, a user output device, or another device.

[0044] In some embodiments, the method includes collecting estimated current tire load values ​​for a given tire over a period of time (e.g., the life of the tire installed on a particular vehicle).

[0045] In some embodiments, the method includes outputting tire usage data indicating the current load on the tire as a function of time over a period of time. Such an output may enable a user, such as a fleet manager or tire manufacturer, to see if the tire was routinely overloaded and therefore understand how this has affected the tire's lifespan.

[0046] In some embodiments, each of the multiple models is pre-trained on historical data acquired for tires and / or vehicles with characteristic information. This means that each model is optimized to correlate (e.g., linearly and / or non-linearly) the characteristic information with observed measurements of tire load. Such modeling may be based, for example, on linear regression, random forests, support vector regressors, neural networks, or any combination thereof.

[0047] According to a second aspect of the present invention, applying a method as described herein (according to any embodiment of the first aspect) to each tire mounted on a vehicle; accessing information regarding the placement of tires on each axle of a vehicle (e.g., vehicle characteristic information); A computer-implemented method is provided for monitoring the current load of a vehicle by calculating the total vehicle load or the load per axle.

[0048] It will be understood that this aspect may (and preferably does) include, for example, one or more (e.g., all) of the preferred and optional features disclosed herein in relation to other aspects and embodiments of the invention, where applicable.

[0049] According to a third aspect of the present invention, there is provided a system for monitoring a current load on a tire, configured to carry out a method according to the first aspect, the system comprising: one or more TMS sensors mounted on the tire and configured to obtain sensor readings; a processor configured to process the sensor readings; and a memory configured to store the sensor readings and / or parameters based on the sensor readings.

[0050] It will be understood that this aspect may (and preferably does) include, for example, one or more (e.g., all) of the preferred and optional features disclosed herein in relation to other aspects and embodiments of the invention, where applicable.

[0051] In some embodiments, the processor is embodied as a tire-mounted processor, for example, in the same tire-mounted device as the one or more TMS sensors. In some embodiments, the processor is included in a receiver of the sensor readings, for example, a receiver attached to the vehicle, such as a suitable network communication device (e.g., a dongle). In some embodiments, the processor is embodied as a server remote from the vehicle, for example, a cloud server. In some embodiments, the sensor readings are processed (e.g., analyzed) by one or more of these processors.

[0052] In some embodiments, the system includes one or more user output devices for displaying estimated tire load values ​​and / or notifications and / or suggested actions to a user, and the above description applies equally to such systems.

[0053] In some embodiments, at least one of the tire monitoring system (TMS) sensors attached to the tire is an acceleration sensor, preferably a radial acceleration sensor, and the sensor readings may include a radial acceleration reading (e.g., a waveform) as the radial acceleration value.

[0054] In some embodiments, at least one of the tire monitoring system (TMS) sensors mounted on the tire is an air pressure sensor, preferably mounted inside the tire to measure the internal air pressure.

[0055] In some embodiments, at least one of the tire monitoring system (TMS) sensors mounted on the tire is preferably an air temperature sensor mounted inside the tire to measure the internal air temperature.

[0056] In some embodiments, each sensor (or a group of sensors) is located on the inside surface of the tire, for example, attached to the surface of the inner liner facing the tread, or on the inside surface of the tire on the sidewall side, although the sensors do not necessarily have to be attached to the inside surface of the tire, for example, the sensors may be partially or completely embedded within the tire.

[0057] In at least some embodiments, one or more tire monitoring system (TMS) sensors mounted on a tire include a transmitter for transmitting sensor readings, optionally a processor configured to pre-process the sensor readings, and a battery for powering the transmitter at regular detection intervals (e.g., 10 seconds, 30 seconds, 60 seconds, etc.).

[0058] It will be understood, of course, that the term “server,” as used herein, refers to a computer or machine (e.g., a server device) connected to a network such that the server transmits and / or receives data from other devices (e.g., computers or other machines) on the network. Additionally or alternatively, the server may provide resources and / or services to other devices on the network. The network may be the Internet or any other suitable network. The server may be embodied in any suitable server type or device, e.g., a file server, an application server, a communication server, a computing server, a web server, a proxy server, etc. The server may be a single computing device or may be a distributed system, i.e., the functionality of the server may be divided across multiple computing devices. For example, the server may be a cloud-based server, i.e., its functionality may be divided across many computers “on demand.” In such a configuration, server resources may be obtained from one or more data centers, which may be located in different physical locations.

[0059] It will therefore be appreciated that the processes described above as being performed by a server may be performed by a single computing device, i.e., a single server, or by multiple separate computing devices, i.e., multiple servers. For example, all of the processes may be performed by a single server that has access to all the relevant information needed. [Brief explanation of the drawings]

[0060] One or more non-limiting examples will now be described, by way of example only, with reference to the accompanying drawings. [Figure 1] 1 shows a schematic diagram of a load monitoring system installed in a heavy goods vehicle. [Figure 2] FIG. 1 is a schematic diagram of a load monitoring system. [Figure 3] 1 is a flow chart illustrating a load monitoring method. [Figure 4] 3D plot showing the correlation between pressure, load and contact patch. [Figure 5] 4 is a flow chart illustrating a portion of the method of FIG. 3 in more detail. [Figure 6] 6 is a flow diagram illustrating an alternative method to that shown in FIG. 5. [Figure 7] 1 shows a scatter plot of predicted load (using a model according to an embodiment of the present invention) against true load. [Figure 8] 1 is a histogram showing the distribution of the variable "true load - predicted load". [Figure 9] Indicates the output provided to the user. DETAILED DESCRIPTION OF THE INVENTION

[0061] It will be understood that in the figures, features and method steps shown with dashed lines are optional and may be omitted at will in embodiments of the present invention.

[0062] FIG. 1 illustrates a system 1 for monitoring loads on multiple vehicle tires 3 mounted on a vehicle 2. The system 1 includes multiple tire monitoring system (TMS) sensors 5, each mounted within a tire 3, and a network communication device 6 provided on the vehicle 2. In an embodiment, the network communication device 6 may be a dongle that plugs into a port on the vehicle 2, such as an OBD port, an FMS port, or the like. In an alternative embodiment, the network communication device 6 may be a permanently installed transceiver box. The TMS sensors 5 are configured to communicate with a remote server 7 via the network communication device 6 within the vehicle 2. The remote server 7 may comprise an online database / cloud / platform. In this example, the TMS sensors 5 communicate with the network communication device 6 via a Bluetooth® connection, although it will be understood that any suitable form of short-range wireless communication or wired connection may be used. The network communication device 6 is networked and communicates with the remote server 7 via a wireless network connection (e.g., a cellular network). The remote server 7 is connected to a user output device 10 via the wireless network. In the illustrated embodiment, several user output devices 10 are provided: a user terminal 10a provided in the cabin of the vehicle 2, a user mobile device 10b that may be carried by a user, and a fleet manager's computer 10c.

[0063] FIG. 2 is a schematic diagram illustrating inputs to a model 15 implemented by a remote server 7. First, characteristic information 25, including characteristic information about a tire 25a and, optionally, characteristic information about a vehicle 25b, is received from memory 11. In an embodiment, memory 11 may be implemented on the remote server 7 itself. Alternatively, the memory may be included in the TMS sensor 5, the network communication device 6, or a further server (not shown) operatively connected to the remote server 7. Because models are built and trained for use with specific tire products, the characteristic information 25a is used to select a model 15 from a plurality of models 14, as will be described below in connection with FIG. 3. The selected model 15 receives input data. TMS information 23 is obtained from a plurality of TMS sensors 5 attached to tires 3 of the vehicle 2 and is input to the selected model 15. In this embodiment, telematics information 21 is received from the vehicle 2 and, optionally, is also input to the selected model 15, as indicated by the dashed arrow in FIG. 2. Telematics information 21 and TMS information 23 may be input into the selected model 15 to improve the accuracy of the model 15. This telematics information 21 may be obtained by sensors on the vehicle 2 in any known manner.

[0064] The operation of the system 1 will be explained with reference to the flow chart of FIG. 3, which illustrates a load monitoring method 100.

[0065] Steps 101-107 outline a process for monitoring the load on a single tire 3 of a vehicle 2. In step 101, sensor readings are obtained from the TMS sensors 5 attached to the tire. Each sensor 5 acquires data for n tire revolutions every m seconds at the beginning of each run for several minutes. In an embodiment, each sensor 5 may acquire data for 10 tire revolutions every 15 seconds for the first 5 minutes of the run. The sensor readings 23 include radial acceleration, inflation pressure, contact patch length estimate, and tire rotation time (t_rev) / tangential tire velocity. The radial acceleration readings come in two forms. First, they are raw radial acceleration readings. These are used to generate a waveform (acceleration plotted against time) from which contact patch length can be estimated. Second, an average radial acceleration is calculated, considering only acceleration readings taken outside the contact patch. This average radial acceleration can be used to estimate tangential tire velocity. These estimates are not measured directly, but the calculation of the estimates is performed by a processor of the TMS sensor 5 itself, as is known, and therefore still considered as sensor readings received from the TMS sensor. The contact patch length is estimated by evaluating peaks appearing in a waveform obtained by differentiating a time series waveform of tire radial acceleration detected by an acceleration sensor attached to the tire. Such a technique is known from WO2020071249(A1), which is incorporated herein by reference. The tire contact patch length is estimated as the ratio of contact patch time to rotation time. Meanwhile, the tire tangential velocity can be calculated according to Equation 1,

[0066]

number

[0067] where v is the tangential tire velocity, a is the average radial acceleration outside the contact patch, and r is the radius of the tire. In another embodiment, the tangential tire velocity v is

[0068]

number

[0069] In step 103, characteristic information 25 is accessed from memory 11. The characteristic information includes vehicle characteristic information 25b, tire characteristic information 25a, or both. Tire characteristic information 25a may include one or more of tire manufacturer, tread pattern, specification, size, mounting location, retread information, and additional remaining tread depth (RTD) measured during tire inspection. In a preferred embodiment, tire characteristic information 25a may correspond to a particular tire product from a particular manufacturer. Vehicle characteristic information 25b may include one or more of vehicle manufacturer, chassis (number of tires / axles), vehicle usage, and tractor load.

[0070] In step 105, one of a plurality of pre-trained models 14 is selected based on the characteristic information 25. The model 15 includes an algorithm pre-trained on historical data, which includes data from three sources: Indoor test rig A drum or similar device capable of simulating tire behavior at different speeds, pressures, and load conditions Outdoor testing Testing vehicles in a controlled environment, such as a test ground or public road, running under specific tire and vehicle conditions. Load variations can be applied during outdoor testing by weighing the vehicle or by using a dedicated vehicle capable of applying variable tire loads / forces. Test vehicle Vehicles on public roads in an uncontrolled state where the vehicle / tyre condition may be unknown or only partially known.

[0071] Because vehicle characteristic information 25b and tire characteristic information 25a can have a significant impact on tire load monitoring, separate models are provided for different vehicle and / or tire characteristics. This is because a model that can handle all vehicle and tire characteristics is likely to be computationally limited. In a preferred embodiment, a separate model 15 is built for each tire product based on characteristic information 25a related to that tire. These models 15 are vehicle-independent and therefore work with any vehicle on which the tire product can be installed. This improves model accuracy and reduces model 15 complexity. As a result, building many separate models can be more demanding than building a single model, but simple models are more suitable for implementation on devices with limited processing power, such as the TMS sensor device 5 or the network communication device 6 in the vehicle 2. In a preferred embodiment, different tire product-specific tire compound and stiffness, as well as tire size, are used as tire characteristic information 25a on which model selection is based.

[0072] In step 107, the TMS information 23 is input into the selected model 15. The model 15 correlates the input TMS information with the characteristic information to determine the load on the tire 3. This correlation may be linear or non-linear. The algorithm used to create the model 15 is linear regression, random forest, support vector regressor, neural network, or an ensemble of all or a subset of the above. During the training phase of the model 15, historical data is analyzed to study the correlation between each parameter from the TMS data 23 and the tire load for each specific tire product. From this correlation, a fitting model 15 can be created for that specific tire product.

[0073] FIG. 4 shows a 3D scatter plot created using historical (training) data. The plot shows load plotted against pressure and contact patch ratio. This particular training data includes measurements taken at three different pressures, and therefore includes three different lines in the data. However, it can be clearly seen that there is a correlation between these three variables, and therefore, any known correlation algorithm can be used to generate a plane of best fit based on the plotted training data. This best fit plane represents a model 15 that can be used to estimate load when the contact patch and pressure are known.

[0074] The model 15 is optimized using a cross-validation procedure with multiple folds. The historical data is divided into multiple sub-datasets (folds), and different combinations of model variations and historical data are sequentially tested to determine which model variations best solve the problem. Thus, the model 15 is optimized around the best model variations.

[0075] While steps 101-107 in Figure 3 relate to load estimation itself, Figure 3 also includes other steps related to evaluating when to perform load estimation. In step 111, a "start of trip" is detected. Methods for detecting the start of trip are described in more detail in conjunction with Figures 5 and 6. The objective of the present invention is to monitor the current tire load and provide feedback to the user about the current tire / vehicle load condition. However, because most TMS sensors are battery-powered, it is advantageous to minimize the number of TMS transmissions to reduce data costs and maximize the battery life of the TMS sensors. A possible solution to achieving these two objectives is to reduce the number of load detection calculations to once per vehicle trip to detect possible weight changes. This solution operates under the assumption that the load is unlikely to change during a trip, since the most likely cause of a load change is the addition, removal, or movement of cargo, which requires vehicle 2 to stop. By implementing a start of trip detection procedure in step 111, the start of a new trip can be detected. Because a new run has begun, the load conditions may have changed between runs, and therefore updated tire loads need to be detected in response to run start detection. As shown in Figure 3, once run start is detected, method 100 can proceed directly to step 101 to begin load estimation, although alternatively, method 100 can proceed to step 113, as described below.

[0076] In embodiments, the number of TMS transmissions may be controlled by performing data measurements during the run at fixed time / distance intervals. Such embodiments may be implemented separately from or in combination with detecting the start of a run.

[0077] The load on the vehicle tires 3 is affected by dynamic conditions that change during driving due to weight transfer. For example, when the vehicle 2 is braking, the front tires 3 of the vehicle 2 experience an increased load, while the rear tires 3 of the vehicle 2 experience a decreased load. Similarly, when the vehicle 2 is accelerating, the rear tires 3 of the vehicle 2 experience an increased load, while the front tires 3 of the vehicle 2 experience a decreased load. The same applies to lateral acceleration. When the vehicle 2 is turning, the tires 3 on the outside of the turn experience a greater load than the tires 3 on the inside of the turn. Obviously, loads calculated based on TMS sensor readings obtained during such load transfer events do not have high accuracy. Therefore, to improve the accuracy of the load monitoring method 100, it is advantageous to detect favorable conditions for load monitoring evaluation. This can be done by acquiring telematics data in step 113 and analyzing the telematics data in step 115 to determine whether the current dynamic conditions are favorable for load monitoring evaluation.

[0078] Telematics data 21 is acquired by sensors within vehicle 2 and collected by a network communication device 6 connected to an OBD, FMS, or similar port on vehicle 2. In the illustrated embodiment, a single network communication device 6 is provided to communicate with all of the TMS sensors 5, although it will be understood that there may be multiple network communication devices 6. For example, multiple network communication devices 6 may be required in large vehicles where the range of the wireless signals from the TMS sensors 5 is not sufficient for all TMS sensors 5 to be within range of a single network communication device 6. Telematics information 21 may include one or more of acceleration, speed, GPS coordinates, road type, engine load, gear shift, engine RPM, wheel speed, throttle / brake, pedal position, tire temperature, exterior temperature, and steering angle.

[0079] For example, the vehicle's longitudinal and lateral accelerations may each be compared to a threshold value, and if the vehicle's acceleration is lower than that value, the condition is determined to be favorable for load monitoring evaluation. The best dynamic condition for load monitoring evaluation may exist when the vehicle is stationary or traveling in a straight line at a constant speed, such as on a highway. If, in step 115, the condition is determined to be favorable for load monitoring estimation, method 100 proceeds to step 101. If the condition is not favorable, method 100 returns to step 113, where updated telematics data 21 is obtained.

[0080] In the illustrated embodiment, the network communication device 6 streams the telematics data 21 to a remote server 7 .

[0081] It will be understood that steps 111, 113 and 115 are not inseparably linked and the method according to the invention can detect the start of travel without analyzing the dynamic conditions, and similarly can analyze the dynamic conditions without first detecting the start of travel.

[0082] In embodiments, dynamic conditions may be assessed in other ways that do not require the use of telematics data 21. For example, in embodiments, TMS information 23 may be used to determine favorable conditions. In such embodiments, a relatively constant value of radial wheel acceleration readings may indicate a constant speed and therefore a favorable condition. Alternatively, multiple load monitoring estimates may be performed and subsequently analyzed. High fluctuations in estimated load over a short period of time may indicate that conditions were unfavorable, and therefore, such results should be ignored.

[0083] Method 100, as described with respect to steps 101-103, estimates the current load on a single tire. In embodiments, it may be useful to calculate the load on the entire vehicle, as overloading the vehicle may result in driving instability and / or legal violations.

[0084] In step 117, steps 101-107 are repeated for each tire of the vehicle 2 to obtain a current load estimate for each tire 3. While FIG. 3 is formatted as a flow diagram, it will be understood that steps 101-107 are performed simultaneously for each tire 3. Because the load values ​​are intended to indicate the current load, the load value estimates for each tire 3 should be aligned in time. Each tire 3 is equipped with at least one TMS sensor 5. Each TMS sensor 5 communicates with the same network communication device 6.

[0085] In step 119, the current total vehicle load is calculated simply by summing the individual tire load values ​​for each tire 3 of the vehicle 2. In order to perform this summation, it is necessary to know the vehicle configuration, in particular the number of axles and the number of tires 3. If the number of tires 3 per axle is known, the load per axle can be calculated in the same way according to equation 2.

[0086]

number

[0087] During the ceremony, w i is the individual tire load and N is the number of tires on the axle or vehicle, depending on whether total axle load or total vehicle load is being calculated.

[0088] The vehicle configuration is included in the vehicle characteristics information 25b obtained from the memory 11 as described in relation to FIG.

[0089] In step 109, the load values ​​are transmitted to the output device 10 for display to the user. As described in connection with FIG. 1, the output device 10 may be a user terminal 10a in the vehicle, a user's mobile device 10b, and / or a fleet manager's computer 10c. Because the method 100 of the present invention is designed to monitor current tire loads, feedback is advantageously provided in real time. While in simpler embodiments, only raw load values ​​(tire load values ​​or total vehicle load values, or both) may be displayed to the user, in embodiments, the load values ​​may be analyzed and specific warnings may be provided to the user. For example, the user may be notified when a tire load index is exceeded, or when a tire is running at an incorrect inflation pressure for the current load, or when the vehicle exceeds its maximum load capacity, or when the vehicle features an uneven weight distribution (left / right and / or between axles). This analysis may be performed in the TMS sensor 5, the vehicle 2, the network communication device 6, the remote server 7, or the output device 10. Specifically, the analysis may be performed by an application installed on the user's mobile device 10b.

[0090] 5 shows a flow diagram illustrating a method for detecting the start of a journey in step 111 of method 100 of FIG. 3. In step 121, a time history set of GPS positions of vehicle 2 is collected from telematics information 21. However, it will be understood that the GPS information can come from any suitable source. The GPS information can come from a GPS sensor integrated into vehicle 2, or network communication device 6 may include a GPS sensor, or a separate GPS sensor may be provided in vehicle 2. In step 123, the GPS coordinate number resolution is reduced to a certain spatial accuracy. This reduction may be performed by truncation, rounding, or similar methods. In an embodiment, the resolution may be 100 m.

[0091] In step 125, a time instance at which the GPS position changes is identified, and the time difference between that time instance and the minimum time of the previous GPS coordinate is calculated. It will be appreciated that the minimum time of the previous GPS coordinate is the first time that the GPS coordinate was recorded, and therefore the time difference between the first instance that the GPS coordinate was recorded and the instance at which the GPS coordinate changes is equal to the amount of time that the vehicle 2 was stationary. In step 127, the time difference, i.e., the time that the vehicle 2 was stationary, is compared to a predetermined threshold, and if the time exceeds the threshold, a new travel event is triggered in step 139. If the time is less than the threshold, the method is restarted.

[0092] The threshold is set as a compromise between the need to reduce the likelihood that a new driving event is falsely triggered and the need to ensure that a new driving event is not missed. For example, setting the threshold too low may lead to a new driving event being triggered after vehicle 2 is stopped at a traffic light or in traffic. On the other hand, the threshold should not be set so high as to potentially miss a stop long enough for the vehicle load condition to change, for example, a stop at a depot where cargo is added to vehicle 2. In an embodiment, the predetermined threshold is set to 10 minutes.

[0093] In step 123, the precision of the spatial coordinates is reduced. This is to reduce the possibility of false trip initiation. For example, if a vehicle is moving around the depot, the GPS coordinates will change, but a new trip has not yet started. By reducing the spatial precision of the GPS coordinates to, for example, 100 m, a new trip will only be triggered when the vehicle's position has changed significantly, as that significant movement indicates a potential new trip has begun.

[0094] FIG. 6 shows a flow diagram illustrating an alternative method for detecting the start of a new journey using TMS information 23. In step 131, sensor readings are obtained from the TMS sensors 5 attached to the tires 3. These readings may include radial acceleration values. In step 133, periods of inactivity are identified, which may be periods during which the radial acceleration value is equal to zero. In step 135, time instances during which the sensor readings change are identified, and the time difference from the start of the inactivity period is calculated. This allows the length of time the vehicle has been stationary to be determined in the same manner as described above with respect to the method of FIG. 4. In step 137, the time difference is compared to a predetermined threshold, and if it exceeds the threshold, a new journey event is triggered. If the difference is less than the threshold, the method resumes. The threshold is set according to the same criteria as described with respect to the method of FIG. 5.

[0095] 7 shows a scatter plot of true load versus predicted load using the TMS load prediction algorithm in accordance with an embodiment of the present invention. The model used was specifically optimized for Bridgestone's Duravis Drive tire product. As can be clearly seen from the plot, for all three pressures tested, the data points fall on or near the Y=X line, indicating that the predictions made by the model are accurate.

[0096] Furthermore, Figure 8 shows a histogram showing that the distribution of true load minus predicted load values ​​is concentrated around the bin corresponding to low values ​​(meaning that true load and predicted load are of similar values).

[0097] 9 shows an output screen 200 that is displayed to a user on output device 10. Individual current tire loads 201 are shown for each tire 3. Output screen 200 also displays the axle loads 203 for each axle, the total vehicle load 205, and the payload 207, which is equal to the total vehicle load 205 minus the weight of the vehicle itself. In the illustrated embodiment, output screen 200 also displays other information such as the current vehicle speed 209 and the current vehicle coordinates 211.

[0098] In the described embodiment, processing steps such as implementation of model 15 are performed on a remote server 7. However, it is envisioned that some or all of the processing steps of method 100 may be performed on the edge, by the network communication device 6, by the vehicle 2 itself, or by the TMS sensor 5. Start of journey detection 111 may similarly be implemented on the edge, on the network communication device 6, or directly on the vehicle 2, or in embodiments that do not require telematics data, directly on the TMS sensor firmware.

[0099] In the above-described embodiment, a single current load value is calculated. It is envisioned that the method may be configured to detect multiple load values ​​during travel and combine these values ​​into a running average, a total measurement average, or combine the values ​​using other suitable aggregation functions. Such an embodiment may further improve load detection accuracy.

Claims

1. 1. A computer-implemented method for monitoring a current load on a tire mounted on a vehicle, the method comprising: obtaining sensor readings from one or more tire monitoring system (TMS) sensors mounted on the tire; accessing characteristic information about the tire; selecting an appropriate model from among a plurality of models based on the characteristic information; and using the selected model to estimate a current tire load value based on the sensor readings.

2. The method of claim 1 , wherein the characteristic information about the tire includes one or more of a tire manufacturer, a tread pattern, a tire specification, a tire compound, a tire stiffness, a tire size, a tire mounting position, and retread information.

3. The method of claim 1 or 2, wherein the plurality of models includes separate models specifically optimized for known tire products from a given manufacturer.

4. The method of any one of claims 1 to 3, further comprising accessing characteristic information about the vehicle.

5. 5. The method of claim 4, wherein the characteristic information about the vehicle includes one or more of a vehicle manufacturer, a chassis type of the vehicle, a vehicle usage status, a number of axles, a product type, a number of tires, and a number of tires mounted per axle.

6. 6. The method of any one of claims 1 to 5, comprising obtaining current telematics information from the vehicle, optionally the telematics information comprising one or more of GPS location, vehicle speed, vehicle lateral / longitudinal acceleration, road type, engine load, gear shift, engine RPM, wheel speed, throttle / brake pedal position, tire temperature, exterior temperature, and / or steering wheel angle.

7. The method of claim 6 , comprising using the current telematics information to detect a favorable load measurement condition, and obtaining the sensor reading when a favorable load measurement condition is detected.

8. The method of any one of claims 1 to 7, wherein the sensor readings include one or more of a radial acceleration value, an inflation pressure, and an internal air temperature.

9. obtaining sensor readings including radial acceleration values; analyzing the radial acceleration values ​​to calculate one or more tire parameters selected from tire contact patch length, tire wear estimation, tire rotation time, automatic mounting position determination, and tire mileage estimation; 9. The method of claim 1, further comprising: using the selected model to estimate a current tire load value based on the sensor readings and one or more of the tire parameters.

10. 10. The method of any one of claims 1 to 9, comprising providing a user with real-time feedback related to the current tire load value, optionally the feedback comprising one or more of a notification that a tire load index has been exceeded, a notification that the tire is running at an incorrect inflation pressure for the current load, a notification that the vehicle has exceeded its maximum load capability, and a notification that the vehicle has an uneven weight distribution across its tires.

11. The method according to any one of claims 1 to 10, wherein the current tire load value is calculated only once in a journey made by the vehicle.

12. 12. The method of claim 11, comprising detecting the start of a journey using data from the one or more tire monitoring system (TMS) sensors and / or telematics data; and estimating the current tire load value in response to detecting the start of a journey.

13. The method according to any one of claims 1 to 10, wherein the current tire load value is calculated at regular time intervals or regular mileage intervals.

14. The method of any one of claims 1 to 13, wherein each of the plurality of models is pre-trained on historical data acquired for tires having the same characteristic information.

15. A system for monitoring the current load of a tire, configured to carry out the method according to any one of claims 1 to 14, said system comprising: one or more TMS sensors mounted on the tire and configured to obtain sensor readings; a processor configured to process the sensor readings; a memory configured to store the sensor readings and / or parameters based on the sensor readings.

Images