Tire wear prediction system and tire wear prediction method

Abstract

A tire wear prediction system predicts wear of a tire mounted on a vehicle constituting a vehicle fleet. The tire wear prediction system sets a driver-related coefficient applied to a driver-related explanatory variable and a vehicle-related coefficient applied to a vehicle-related explanatory variable, based on actual values of the driver-related explanatory variable and actual values of the vehicle-related explanatory variable, and calculates wear of the tire using the driver-related explanatory variable and the vehicle-related explanatory variable as a target variable. The tire wear prediction system recalculates a predicted value of the wear in a case where at least either of attributes of the driver and attributes of the vehicle are changed, using the driver-related coefficient and the vehicle-related coefficient.

Description

Technical Field

[0001] This disclosure relates to a tire wear prediction system for predicting wear on tires mounted on vehicles constituting a fleet. Background Technology

[0002] Previously, there were known demand forecasting systems for consumables (e.g., tires) that target fleets of multiple vehicles (e.g., buses and trucks) owned by transportation companies, etc. (Patent Document 1).

[0003] The demand forecasting system described in Patent Document 1 statistically analyzes the deterioration rate of consumables and predicts the types and quantities of consumables needed in vehicle maintenance workshops. This can help reduce fleet maintenance costs.

[0004] Existing technical documents

[0005] Patent documents

[0006] Patent Document 1: International Publication No. 2016 / 071993 Summary of the Invention

[0007] In the case of a convoy of multiple vehicles, it is important to accurately predict the wear of the tires mounted on the vehicles in order to achieve efficient use.

[0008] Therefore, the following disclosure is made in view of the following situation, with the aim of providing a tire wear prediction system and tire wear prediction method that can more accurately predict the wear of tires mounted on vehicles constituting a fleet.

[0009] One aspect of this disclosure is a fuel consumption rate prediction system (fleet management system 10) for predicting the fuel consumption rate of vehicles constituting a fleet, comprising: a driver-related variable acquisition unit (driver-related variable acquisition unit 110) that acquires multiple driver-related explanatory variables related to the attributes of the driver driving the vehicle; a vehicle-related variable acquisition unit (vehicle-related variable acquisition unit 120) that acquires multiple vehicle-related explanatory variables related to the attributes of the vehicle; a coefficient setting unit (coefficient setting unit 140) that sets a driver-related coefficient applied to the driver-related explanatory variables and a vehicle-related coefficient applied to the vehicle-related explanatory variables based on the actual values ​​of the driver-related explanatory variables and the actual values ​​of the vehicle-related explanatory variables; and a target variable calculation unit (target variable calculation unit 150) that uses the driver-related explanatory variables and the vehicle-related explanatory variables to calculate the fuel consumption rate as a target variable, wherein the target variable calculation unit uses the driver-related coefficients and the vehicle-related coefficients to recalculate the predicted value of the fuel consumption rate under the condition that at least one of the attributes of the driver and the attributes of the vehicle has been changed.

[0010] One aspect of this disclosure is a fuel consumption rate prediction method for predicting the fuel consumption rate of vehicles constituting a fleet, comprising the following steps: obtaining a plurality of driver-related explanatory variables related to attributes of the driver driving the vehicle; obtaining a plurality of vehicle-related explanatory variables related to attributes of the vehicle; setting a driver correlation coefficient applied to the driver-related explanatory variables and a vehicle correlation coefficient applied to the vehicle-related explanatory variables based on the actual values ​​of the driver-related explanatory variables and the actual values ​​of the vehicle-related explanatory variables; and calculating the fuel consumption rate as a target variable using the driver-related explanatory variables and the vehicle-related explanatory variables, wherein, in the step of calculating the target variable, the predicted value of the fuel consumption rate is recalculated using the driver correlation coefficient and the vehicle correlation coefficient for cases where at least one of the attributes of the driver and the attributes of the vehicle has been changed.

[0011] One aspect of this disclosure is a tire wear prediction system (fleet management system 10) for predicting tire wear on vehicles constituting a fleet, comprising: a driver-related variable acquisition unit (driver-related variable acquisition unit 110) that acquires multiple driver-related explanatory variables related to the attributes of the driver driving the vehicle; a vehicle-related variable acquisition unit (vehicle-related variable acquisition unit 120) that acquires multiple vehicle-related explanatory variables related to the attributes of the vehicle; a coefficient setting unit (coefficient setting unit 140) that sets a driver-related coefficient applied to the driver-related explanatory variables and a vehicle-related coefficient applied to the vehicle-related explanatory variables based on the actual values ​​of the driver-related explanatory variables and the actual values ​​of the vehicle-related explanatory variables; and a target variable calculation unit (target variable calculation unit 150) that uses the driver-related explanatory variables and the vehicle-related explanatory variables to calculate the wear as a target variable, wherein the target variable calculation unit uses the driver-related coefficients and the vehicle-related coefficients to recalculate the predicted value of the wear in cases where at least one of the attributes of the driver and the attributes of the vehicle has been changed.

[0012] One aspect of this disclosure is a tire wear prediction method for predicting tire wear on vehicles constituting a fleet, comprising the following steps: obtaining a plurality of driver-related explanatory variables related to attributes of a driver driving the vehicle; obtaining a plurality of vehicle-related explanatory variables related to attributes of the vehicle; setting a driver correlation coefficient applied to the driver-related explanatory variables and a vehicle correlation coefficient applied to the vehicle-related explanatory variables based on actual values ​​of the driver-related explanatory variables and actual values ​​of the vehicle-related explanatory variables; and calculating the wear as a target variable using the driver-related explanatory variables and the vehicle-related explanatory variables, wherein, in the step of calculating the target variable, the predicted value of the wear is recalculated using the driver correlation coefficient and the vehicle correlation coefficient in cases where at least one of the attributes of the driver and the attributes of the vehicle has been changed. Attached Figure Description

[0013] Figure 1 This is an overall overview diagram of the fleet and fleet management system 10.

[0014] Figure 2 This is a functional block structure diagram of server computer 100.

[0015] Figure 3 This is a diagram showing the overall outline of the operation flow of the fleet management system 10.

[0016] Figure 4 This is a diagram showing the tire positions for each type of bus.

[0017] Figure 5 This is a diagram illustrating a summary example of the fleet's operation, tires, and driver characteristics.

[0018] Figure 6 It is a graph showing the relationships between the variables.

[0019] Figure 7 These are examples of representative descriptions of variables (data).

[0020] Figure 8 These are examples of representative descriptions of variables (data).

[0021] Figure 9 This is a graph showing the results of the 3SLS model.

[0022] Figure 10 This is a diagram illustrating an example of the simulated scheme.

[0023] Figure 11 This is a diagram that provides a descriptive overview of the maintenance cost dataset.

[0024] Figure 12 This is a graph showing the results of the maintenance cost model.

[0025] Figure 13 This is a diagram illustrating an example simulation scheme using the DSS tool with a 3SLS model. Detailed Implementation

[0026] The embodiments will now be described with reference to the accompanying drawings. Furthermore, the same or similar reference numerals will be used to denote the same functions and structures, and their descriptions will be omitted where appropriate.

[0027] (1) Overall structure of the fleet and fleet management system

[0028] Figure 1 This is an overall structural diagram of the fleet and fleet management system 10. The fleet management system 10 is used to manage a fleet, specifically, to manage multiple vehicles.

[0029] In this embodiment, the fleet management system 10 is used to manage multiple buses that constitute a fleet. For example... Figure 1 As shown, the fleet consists of multiple types of buses (vehicles).

[0030] Specifically, the fleet may include regular buses 20, double-decker buses 30, and articulated buses 40.

[0031] The ordinary bus 20 is a single-story rigid vehicle with a low floor or a high floor. In this embodiment, the ordinary bus 20 has one front axle and one rear axle.

[0032] The double-decker bus 30 has a double-decker structure, comprising a first deck section and a second deck section. The double-decker bus 30 is also referred to as a double-decker bus, etc. In this embodiment, the double-decker bus 30 has two front axles and one rear axle.

[0033] The articulated bus 40 has a structure that connects two bus bodies via joints. The articulated bus 40 is also referred to as an articulated or bent bus, etc. In this embodiment, the articulated bus 40 has a front axle and a rear axle of the foremost vehicle, and an axle of the rear vehicle.

[0034] In addition, the regular buses 20, double-decker buses 30, and articulated buses 40 can be provided as dedicated bus lines or as charter buses.

[0035] Regular buses 20, double-decker buses 30, and articulated buses 40 are driven by drivers from a bus operating company that manages the fleet. Specifically, these buses are driven by male drivers 50A and female drivers 50B. Within the bus operating company, multiple male drivers 50A and multiple female drivers 50B belong to that company. Furthermore, the age structure and years of driving experience of the drivers can vary.

[0036] The fleet management system 10 can provide multiple functions related to fleet management. In this embodiment, it can be configured as a fuel consumption rate prediction system for predicting the fuel consumption rate of vehicles (buses) constituting the fleet and a tire wear prediction system for predicting the wear of tires installed on the vehicles (buses) constituting the fleet.

[0037] Such fuel consumption rate prediction systems and tire wear prediction systems can also be provided as decision support systems (DSS) to support decision-making in businesses operating fleets (such as bus operating companies). Such DSSs can provide several guidelines regarding tire-related impacts on fleet fuel consumption and maintenance costs. Details regarding DSSs are described later.

[0038] The fleet management system 10 may include a server computer 100, a portable terminal 200, and a desktop terminal 300.

[0039] The server computer 100 can execute computer programs for implementing a fuel consumption rate prediction system and a tire wear prediction system. The server computer 100 can be a single server computer connected to a communication network or it can be virtually configured on a network cloud.

[0040] Typically, a network cloud includes the Internet, encompassing various information services (such as weather information), storage services, and application services provided on the Internet.

[0041] Portable terminal 200 and desktop terminal 300 can access server computer 100 via a communication network.

[0042] Typically, portable terminal 200 is a smartphone or tablet, but it can also be a laptop computer (PC). Desktop terminal 300 is a larger PC (including a monitor) set up in the office of a business entity operating a fleet.

[0043] Portable terminal 200 and desktop terminal 300 are used for inputting data to server computer 100 (fuel consumption rate prediction system and tire wear prediction system), displaying and outputting processing results in server computer 100, etc.

[0044] The data processed by fuel consumption prediction systems and tire wear prediction systems cover a wide range. When broadly distinguished, they can be based on data related to vehicles (such as bus types), tire-related measurements (tread depth, temperature, internal pressure, etc.), weather (temperature, etc.), driving (distance, fuel consumption, etc.), drivers (gender, age, years of driving experience, etc.), and roads (routes, number of traffic lights, number of stops, etc.).

[0045] In addition, some of this data may not be obtained through portable terminal 200 or desktop terminal 300. For example, tire-related measurement data may be obtained through a telemetry system that utilizes a tire pressure monitoring system (TPMS), and weather data may be obtained directly by server computer 100 via a communication network.

[0046] (2) Functional block structure of the fleet management system

[0047] Next, the functional block structure of the fleet management system 10 will be explained. Specifically, the functional block structure of the fuel consumption rate prediction system and the tire wear prediction system implemented by the server computer 100 will be explained.

[0048] Figure 2 This is a functional block structure diagram of server computer 100. (Example) Figure 2As shown, the server computer 100 includes a driver-related variable acquisition unit 110, a vehicle-related variable acquisition unit 120, a data storage unit 130, a coefficient setting unit 140, a target variable calculation unit 150, and an output unit 160.

[0049] These functional blocks are implemented by executing computer programs (software) on the hardware of the server computer 100.

[0050] Specifically, as hardware components, the server computer 100 may include a processor, memory, input devices, a display, and external interfaces. However, as mentioned above, the server computer 100 (fuel consumption prediction system and tire wear prediction system) can also be virtually configured on a network cloud.

[0051] The fuel consumption rate prediction system is used to predict the fuel consumption rate of the vehicles (buses) that make up the fleet. Additionally, the tire wear prediction system is used to predict the wear of the tires mounted on the vehicles (buses) that make up the fleet.

[0052] The driver-related variable acquisition unit 110 is capable of acquiring multiple driver-related explanatory variables. Driver-related explanatory variables can be defined as variables related to the attributes of the driver of the vehicle (bus).

[0053] In this embodiment, driver-related explanatory variables refer to variables related to the attributes of multiple male drivers 50A and female drivers 50B who drive ordinary buses 20, double-decker buses 30, or articulated buses 40.

[0054] This variable can be interpreted as an explanatory variable. An explanatory variable can be defined as a variable used to explain the target variable; it can also be called an independent variable. Explanatory variables can be selected from the data mentioned above.

[0055] The target variable can be defined as the variable that you want to predict, and it can also be called the dependent variable.

[0056] The driver-related variable acquisition unit 110 can acquire at least one of the following as driver-related explanatory variables: average age of drivers, gender ratio, average years of experience, and number of drivers per specified driving distance. The gender ratio can refer to the proportion of male (or female) drivers in the total population. Furthermore, the number of drivers per specified driving distance (Number of drivers per 1,000kms) refers to the number of drivers responsible for each specified distance (e.g., 1,000km).

[0057] Driver-related explanatory variables are not limited to these; statistical data related to driver attributes (such as accident history, blood type, etc.) can also be added.

[0058] The vehicle-related variable acquisition unit 120 is capable of acquiring multiple vehicle-related explanatory variables that are related to the attributes of the vehicle (bus). Vehicle-related explanatory variables can be defined as variables that are related to the attributes of the vehicle (bus).

[0059] In this embodiment, vehicle-related explanatory variables refer to variables related to the attributes of the ordinary bus 20, the double-decker bus 30, or the articulated bus 40.

[0060] This variable can be interpreted as an explanatory variable in the same way as the driver-related explanatory variables.

[0061] The vehicle-related variable acquisition unit 120 can acquire the wear rate of the tires installed on the vehicle (bus) as a vehicle-related explanatory variable. Wear rate can be defined as the amount of tread wear per unit distance traveled by the tires installed on each bus. Alternatively, wear rate can also be defined as the amount of tread wear per unit operating time of the bus. Furthermore, wear rate can also be defined as the residual tread depth (RTD).

[0062] In addition, the vehicle-related variable acquisition unit 120 can acquire the average speed of the vehicle (bus) as a vehicle-related explanatory variable. The average speed can be defined as the average speed of the buses on each route.

[0063] Alternatively, average speed can also be defined as the average speed of a bus per unit of travel time.

[0064] Furthermore, the vehicle-related variable acquisition unit 120 can acquire the average temperature of the tires installed on the vehicle (bus) (average tire temperature) as a vehicle-related explanatory variable. The average temperature can be defined as the average temperature obtained by taking multiple measurements of the tires installed on each bus. This temperature can be measured by a TPMS sensor installed on the inner surface of the tire, as described above. In addition, the timing of acquiring this temperature can be the start, end, or midway point of bus operation, but it is desirable to exclude periods when the bus is not running and is stopped.

[0065] The data storage unit 130 stores the data (variables) acquired by the driver-related variable acquisition unit 110 and the vehicle-related variable acquisition unit 120. Additionally, the data storage unit 130 stores data input via the portable terminal 200 or desktop terminal 300 and processed by the fuel consumption rate prediction system and the tire wear prediction system.

[0066] The coefficient setting unit 140 sets the coefficients applied to the driver-related explanatory variables obtained by the driver-related variable acquisition unit 110. Additionally, the coefficient setting unit 140 sets the coefficients for the vehicle-related explanatory variables obtained by the vehicle-related variable acquisition unit 120.

[0067] Specifically, the coefficient setting unit 140 sets the driver-related coefficients applied to the driver-related explanatory variables. Additionally, the coefficient setting unit 140 sets the vehicle-related coefficients applied to the vehicle-related explanatory variables.

[0068] More specifically, the coefficient setting unit 140 can set the driver correlation coefficient and the vehicle correlation coefficient based on the actual values ​​of the driver-related explanatory variables and the actual values ​​of the vehicle-related explanatory variables.

[0069] The actual values ​​of the driver-related explanatory variables and the actual values ​​of the vehicle-related explanatory variables refer to values ​​determined based on the past values ​​of the driver-related explanatory variables obtained by the driver-related variable acquisition unit 110 and the past values ​​of the vehicle-related explanatory variables obtained by the vehicle-related variable acquisition unit 120.

[0070] The driver correlation coefficient can be set for each driver-related explanatory variable. Similarly, the vehicle correlation coefficient can be set for each vehicle-related explanatory variable. Furthermore, a common (identical) coefficient can be set for a subset of all explanatory variables. These coefficients can also be replaced with terms such as weighted averages or parameters.

[0071] In this embodiment, the driver correlation coefficient and the vehicle correlation coefficient can be estimated using a three-stage least squares (3SLS) model. However, it is not necessary to use a 3SLS model; other SLS models can also be used.

[0072] Furthermore, specific examples of driver-related explanatory variables, vehicle-related explanatory variables, driver correlation coefficients, and vehicle correlation coefficients will be described later.

[0073] The target variable calculation unit 150 uses driver-related explanatory variables and vehicle-related explanatory variables to calculate the target variable.

[0074] Specifically, the target variable calculation unit 150 can use driver-related explanatory variables and vehicle-related explanatory variables to calculate the fuel consumption rate as the target variable. The fuel consumption rate refers to the rate (or amount) of fuel consumption that a vehicle, specifically a regular bus 20, a double-decker bus 30, or an articulated bus 40 is expected to consume when traveling under specified conditions on a specified route.

[0075] More specifically, the target variable calculation unit 150 can estimate the change in fuel consumption rate when the coefficient is changed by adjusting the coefficient of at least one of the driver-related explanatory variables and the vehicle-related explanatory variables that affect the fuel consumption rate (driver-related coefficient or vehicle-related coefficient).

[0076] In other words, the target variable calculation unit 150 can set the weights (parameters) of each explanatory variable based on the past actual situation of the fuel consumption rate as the target variable, and the past actual situation of driver attributes (average age, gender ratio, average years of experience, etc.) and vehicle information (years of use, vehicle type) as explanatory variables that affect the fuel consumption rate. Furthermore, the target variable calculation unit 150 can recalculate the predicted value of the target variable in the case of changes in driver attributes and / or vehicle information.

[0077] Furthermore, the target variable calculation unit 150 can use driver-related explanatory variables and vehicle-related explanatory variables to calculate tire wear as the target variable. Tire wear refers to the amount of wear (or RTD) of the tread pattern of tires mounted on ordinary buses 20, double-decker buses 30, or articulated buses 40. Alternatively, tire wear as described here can also refer to the wear rate.

[0078] Specifically, the target variable calculation unit 150 can estimate the wear that changes when the coefficient is changed by adjusting the coefficient of at least one of the driver-related explanatory variables and vehicle-related explanatory variables that affect wear (driver-related coefficient or vehicle-related coefficient).

[0079] In other words, the target variable calculation unit 150 can set the weights (parameters) of each explanatory variable based on the past actual conditions of the wear rate (wear amount per fixed driving distance) as the target variable, and the past actual conditions of the driver attributes (average age, gender ratio, average years of experience, etc.) and vehicle information (years of use, vehicle type) as explanatory variables that affect the wear rate. Furthermore, the target variable calculation unit 150 can recalculate the predicted value of the target variable in cases where driver attributes and / or vehicle information have changed.

[0080] In this way, the target variable calculation unit 150 can use the driver correlation coefficient and the vehicle correlation coefficient to recalculate the predicted value of fuel consumption rate under the condition that at least one of the driver's attributes and the vehicle's attributes has been changed.

[0081] In addition, the target variable calculation unit 150 can use the driver correlation coefficient and the vehicle correlation coefficient to recalculate the predicted value of tire wear in cases where at least one of the driver's attributes and the vehicle's attributes has been changed.

[0082] The output unit 160 can output the data (explanatory variables) obtained by the driver-related variable acquisition unit 110 and the vehicle-related variable acquisition unit 120.

[0083] In addition, the output unit 160 can output the driver correlation coefficient and vehicle correlation coefficient set by the coefficient setting unit 140.

[0084] Furthermore, the output unit 160 can output the predicted value of fuel consumption rate and the predicted value of tire wear calculated by the target variable calculation unit 150.

[0085] Specifically, the output unit 160 provides functions for outputting image data to a display or the like and for sending data to a portable terminal 200 or a desktop terminal 300.

[0086] Furthermore, the output unit 160 can also display, send, or output the aforementioned data through the platform of a decision support system (DSS).

[0087] (3) Operations of the fleet management system

[0088] Next, the operation of the fleet management system 10 will be explained. Specifically, the fuel consumption rate prediction of the vehicles (buses) constituting the fleet, performed by the fleet management system 10 (fuel consumption rate prediction system), and the wear prediction of the tires installed on the vehicles (buses) constituting the fleet, performed by the fleet management system 10 (tire wear prediction system), will be explained.

[0089] (3.1) Overall Summary of Actions

[0090] Figure 3 This illustrates the overall operational flow of the fleet management system 10. Specifically, Figure 3 The illustrated workflow is common to both the fuel consumption rate prediction system and the tire wear prediction system.

[0091] like Figure 3 As shown, the fleet management system 10 obtains driver-related explanatory variables and vehicle-related explanatory variables (S10).

[0092] Specifically, the fleet management system 10 obtains explanatory variables related to the attributes of the driver of the vehicle (bus) and explanatory variables related to the attributes of the vehicle (bus).

[0093] The fleet management system 10 calculates the coefficients applied to the acquired driver-related and vehicle-related explanatory variables (S20). Specifically, the fleet management system 10 calculates the driver-related coefficients applied to the driver-related explanatory variables and the vehicle-related coefficients applied to the vehicle-related explanatory variables.

[0094] More specifically, the fleet management system 10 uses a three-stage least squares (3SLS) model, as described above, to calculate (estimate) driver correlation coefficients and vehicle correlation coefficients.

[0095] The fleet management system 10 uses driver-related explanatory variables and vehicle-related explanatory variables to calculate the target variable (S30).

[0096] Specifically, the fleet management system 10 can use driver-related and vehicle-related explanatory variables to calculate fuel consumption rate as a target variable. Similarly, the fleet management system 10 can use driver-related and vehicle-related explanatory variables to calculate tire wear as a target variable.

[0097] Furthermore, the fleet management system 10 can use driver correlation coefficients and vehicle correlation coefficients to recalculate the predicted fuel consumption rate when at least one of the driver's attributes and vehicle attributes has been changed. Similarly, the fleet management system 10 can use driver correlation coefficients and vehicle correlation coefficients to recalculate the predicted tire wear when at least one of the driver's attributes and vehicle attributes has been changed.

[0098] The fleet management system 10 outputs the recalculated results of the predicted fuel consumption rate and / or the recalculated results of the predicted tire wear (S40). Specifically, the fleet management system 10 can display the results on a monitor or send them to a portable terminal 200 or a desktop terminal 300.

[0099] (3.2) Data collection

[0100] In the fleet management system 10 (fuel consumption rate prediction system and tire wear prediction system), in order to calculate the predicted values ​​of fuel consumption rate and tire wear, the data shown below are collected as driver-related explanatory variables and vehicle-related explanatory variables.

[0101] (3.2.1) Vehicle (bus)

[0102] As mentioned above, the type of bus can be any of the following: ordinary bus 20, double-decker bus 30, and articulated bus 40. There are no particular restrictions on the ratio of each type of bus, but in general, the ratio of ordinary bus 20 can be higher.

[0103] Figure 4 The tire positions for each type of bus are shown. (Example) Figure 4 As shown, the ordinary bus 20 has one front axle and one rear axle. The rear wheels of the ordinary bus 20 are equipped with a so-called dual-tire configuration, consisting of two tires mounted side by side, thus the ordinary bus 20 has six tires, numbered 1 to 6.

[0104] The double-decker bus 30 has two front axles and one rear axle. The rear wheels of the double-decker bus 30 also use dual tires, so the double-decker bus 30 is equipped with eight tires, numbered 1 to 8.

[0105] The articulated bus 40 not only has one front axle and one rear axle in the foremost vehicle, but also one axle in the rear vehicle. The articulated bus 40 also uses dual tires, with 10 tires installed from 1 to 10.

[0106] There are no particular restrictions on the number of buses and the number of tires, but in order to improve the accuracy of fuel consumption rate prediction and tire wear prediction, it is preferable to have a certain level of parameters. Preferably, the total number of buses is about 30 or more and the number of tires is about 200 or more.

[0107] In addition, the fleet management system 10 can obtain detailed driving records (based on unique vehicle IDs) of each bus from an external system (service) connected via a communication network.

[0108] This data can include the number of seats / passenger capacity of buses, the number of trips (runs) per day, the passenger volume of each trip, and the route corresponding to the direction at each time. For each trip of each vehicle, a unique driver ID corresponding to the driver's experience profile can be assigned.

[0109] This data can provide variables by vehicle ID level, including the number of trips, average passenger capacity per trip, bus seating capacity, average speed per trip, and time division (times of day, e.g., 6 trips per day).

[0110] Additionally, this data may also include the service (maintenance) history for each vehicle. The service history may also include the following items.

[0111] Service Team

[0112] • Date of service

[0113] Working hours

[0114] • Cost of each service

[0115] Service Types

[0116] • Distance meter display when the vehicle is receiving service

[0117] • Description of each service record

[0118] In addition, service types and costs can be counted and statistically analyzed based on the RTD of each vehicle.

[0119] A service group may include, for example, the following services.

[0120] Tire-related services: Repairing tire punctures, tire wear, tread retreading, internal pressure monitoring, tire nut and rim locking, and mechanical repairs of electrical safety valves, etc.

[0121] • Rack-related services: Braking, troubleshooting, electrical, engine, cooling / heating, steering, urea water (AdBlue) filters, etc.

[0122] Vehicle-related services: air conditioning, vehicle damage repair, maintenance, repair, ticket vending machines, etc.

[0123] Interior-related services: rearview mirrors, glass, windows, windshield, CCTV, steam cleaning, etc.

[0124] • Regular services: Regular inspections, pre-inspection checks (RMS), vehicle inspections (RMS)

[0125] (3.2.2) Tread groove allowance (RTD)

[0126] The tread depth (RTD) of a tire varies at different locations on the tire surface and may have different measured values ​​along the tire circumference.

[0127] However, the difference in these measured values ​​will not have a significant impact, so it is sufficient to measure the RTD at only one point along the tire circumference. On the other hand, for the tire width direction, considering that the outer edge is prone to wear and the inner edge sometimes experiences uneven wear, it is desirable to measure at multiple points (e.g., 3 to 4 points). The measurement can be performed using automatic tools or manually.

[0128] In addition, the wear of tire tread (groove allowance) can be defined as follows.

[0129] • TreadLoss = (Current tread depth) - (Previous tread depth)

[0130] Tread depth can be interpreted as the depth of the main grooves that form treads.

[0131] Tire temperature and pressure (internal pressure) can be monitored by the vehicle ID and tire position (reference) installed on the bus. Figure 4 The measurements are taken from sensors at each tire location. Average temperature and average internal pressure can be based on measurements collected at two consecutive RTD measurement times.

[0132] (3.2.3) Fuel Consumption

[0133] Fuel consumption (fuel consumption rate) can be calculated using a weighted average based on driving distance (wL 100km). Specifically, wL 100km can be calculated using "service kilometers" (operating distance) at measurement intervals of RTD.

[0134] Alternatively, fuel consumption data can be obtained on a monthly basis, and tread loss and other key variables can be statistically analyzed during the period in which a specific amount of tread loss occurred.

[0135] The weighted service kilometers can also be calculated for each sub-period obtained by dividing monthly fuel consumption data into measurement intervals based on the RTD.

[0136] (3.2.4) Roads and routes

[0137] Data regarding the roads along which buses travel on their routes may include the number of traffic lights and major roundabouts. The number of traffic lights may also include the number of individual traffic lights and mini-roundabouts. Additionally, intersections may include large roundabouts with a centerline diameter exceeding 15 meters.

[0138] In addition, road conditions can also be quantified based on segmented road datasets, using factors such as the length of various road functions, road surface, and number of lanes.

[0139] (3.2.5) Tire-related

[0140] In addition to the RTD mentioned above, tracking data (records) after tire removal or installation can also be included. For example, data related to a tire can be recorded when removing or replacing it. This data may include the vehicle ID (of the bus whose tires were replaced), the location of the replaced tire, the tire brand name, the date the tire was replaced, the retread depth, the tire ID (date + vehicle ID + tire location), the condition of each installed tire (new or retreaded), and the number of times the tire has been retreaded, etc.

[0141] (3.2.6) Driver Information Overview

[0142] Driver information can include data such as employee ID (driver ID), gender, date of birth, and date of employment. By combining this data with customer data, driver experience during the period when tire tread wear (deep wear) occurs can be obtained, thereby establishing a link between each vehicle and each tire.

[0143] In addition, the driving experience and driving profile calculated from this stage may also include the number of trips, service kilometers, average speed and passenger capacity for each RTD (time of tread depth loss of each tire).

[0144] (3.2.7) Meteorological data

[0145] Regarding meteorology (weather), it can include meteorological data for the area where the base (bus depot) of the business entity (bus operating company) operating the fleet is located. For example, it can include the highest temperature of each month or the average temperature for a period consistent with the RTD measurement interval. In addition, the data can include weather conditions (sunny, cloudy, rainy, snowy).

[0146] (3.3) Data Summary

[0147] One of the most important variables in the fleet management system 10 (fuel consumption prediction system and tire wear prediction system) is the tread depth loss, which is measured as the change in tread depth over two measured periods on a time axis. This period can also be called the tread depth measurement cycle (interval). There is no particular limitation on this measurement cycle, but it is usually around 10 to 130 days.

[0148] During this period, there are variables that do not cause problems (or remain roughly constant), such as the age of the buses, the type of buses, the brand of the tires, and the fuel consumption of the buses.

[0149] On the other hand, variables such as average speed, route, patronage, air temperature, tire temperature, and pressure (internal pressure) may vary during the tread depth measurement period.

[0150] Preferably, all variables are summarized (statistically) to represent a tire tread measurement cycle. Figure 5 This example shows a summary of the fleet's operation, tires, and driver characteristics.

[0151] like Figure 5 As shown, operational / route characteristics include average speed, distance traveled (every two weeks), and average maximum temperature. Furthermore, this data is aggregated using appropriate methods such as weighted averaging.

[0152] Tire characteristics include average temperature and average internal pressure. Additionally, driver characteristics include average age, average years of driving experience, and the percentage of male drivers.

[0153] Furthermore, these characteristics can be weighted more heavily for longer routes than for shorter routes. The distance traveled (kms) within the tread depth measurement cycle can be measured in two-week (14-day) units. This is because, for example, it is meaningless to directly compare the distance traveled by a bus in one measurement cycle of 15 days and another of 50 days.

[0154] (3.4) Speed, temperature, tread depth and fuel consumption model

[0155] Next, the model related to speed, temperature, tire tread depth, and fuel consumption will be explained.

[0156] (3.4.1) Simultaneous Equation Model System

[0157] The statistical method involved in this embodiment can use a system of equations of a jointly estimated three-stage least squares (3SLS) model (including the left-hand (LHS) and right-hand (RHS) endogenous variables).

[0158] As mentioned above, the tread depth loss between two measurement times (points) is important, but the number of measurements depends on the number of consistently identical data points that can be collected from the same tire.

[0159] The four models are related to average speed, average tire temperature, tread depth loss, and fuel consumption. Tread depth loss itself is highly dependent on the distance the vehicle travels during the measurement period. In addition, millimeter-level changes will not have the same effect on tires with different tread depths, and are therefore not very useful.

[0160] Therefore, a variable is obtained by transforming the tread depth using a ratio representing the ratio of tread depth loss per kilometer (tread wear per km) to the previous tread depth loss. This transformation is shown in Equation 1.

[0161] [Number 1]

[0162]

[0163] Here, the unit of this variable is as follows.

[0164] [Number 2]

[0165]

[0166] All attributes included in the formula below are calculated based on the tread depth measurement cycle. For example, distance traveled means the service distance traveled by each bus from the last measurement to the current time.

[0167] The system of equations is shown in (Equation 2) to (Equation 5).

[0168] [Number 3]

[0169]

[0170] [Number 4]

[0171]

[0172] [Number 5]

[0173]

[0174] [Number 6]

[0175]

[0176] Figure 6 The relationships between the variables are shown. Fuel consumption is explained partly by tire tread depth loss per km, partly by tire temperature, and partly by average speed.

[0177] Tire temperature is partly indicated by average speed.

[0178] Each of these endogenous variables is affected by Figure 6 The influence of numerous exogenous factors is represented by circles in the diagram. Figure 6 The exogenous factors included in the models shown are illustrative examples, as shown in the equations above. The combinations of interactions related to the variables are diverse.

[0179] like Figure 6As shown, when disturbances (error terms) are co-correlated, a feasible generalized least squares (FGLS) version of the two-stage least squares estimation is consistent and forms an asymptotically efficient estimate (Zellner and Theil 1962). Such an estimation process is called three-stage least squares (3SLS).

[0180] Zelner’s 3SLS estimator can be obtained by the following process: First, regress all variables on the right side of each equation with respect to all variables in the list of (external) instruments to maintain applicable values.

[0181] In this event, the regression in the initial stage produces a perfect fit with one parameter on that variable and zero for all other variables, so any variable that also appears in the list of instrumental variables and in the list of equations is accurately reproduced.

[0182] Following this estimation, the original variables are used instead of the matched variables to estimate the disturbance covariance matrix. This is because using non-iterative 3SLS does not provide a significant improvement in iteration and does not generate maximum likelihood estimation results.

[0183] (3.4.2) Description of the collected data

[0184] Figure 7 and Figure 8 These are examples of representative descriptive summaries of variables (data). For example... Figure 7 and Figure 8 As shown, the final data example of the statistical and descriptive analysis of the 3SLS model has 1,749 observations, each representing a tire.

[0185] Figure 7 and Figure 8 The values ​​shown are averages (standard deviations), differentiated by bus type (all-wheeled, high-floor, low-floor, articulated, double-decker).

[0186] In addition, the variables are divided into bus characteristics, route / operation characteristics, tire characteristics, and driver characteristics.

[0187] (3.4.3) Results and Interpretation of the 3SLS Model

[0188] As described above, in this embodiment, a 3SLS model is used, where each endogenous variable has the potential to have a statistically significant effect, namely, average speed, tread depth loss, average tire temperature, and fuel consumption.

[0189] The dependent (or left-hand) variable in (Equation 2) is the average speed, which is statistically explained based on several routes, operations, and the characteristics of buses.

[0190] The dependent variable in (Equation 3) is the average tire temperature, which is statistically explained based on average speed, peak-hour mileage ratio, patronage, air temperature, and tire position.

[0191] The dependent variable in (Equation 4) is the tread depth loss, which is statistically explained based on average speed, average tire temperature, tire pressure of the front tires, driving distance (km), route characteristics, passenger capacity, and tire position.

[0192] The dependent variable in (Equation 5) is fuel consumption, which is influenced by tire tread depth wear, bus age, driving distance, and bus type through a statistically meaningful method.

[0193] Figure 9 The results of the 3SLS model are shown. Figure 9 The table shows the coefficients and t-tests for each dependent variable.

[0194] according to Figure 9 The results show that the average speed of the double-decker bus 30 and the low-floor conventional bus 20 (Rigid) is higher than that of other types of buses (articulated bus 40 and high-floor conventional bus 20).

[0195] The rear tires showed greater tread depth loss compared to the expected front tires. Additionally, the inner rear tires exhibited a greater rate of tread depth loss per 1,000 km compared to the outer rear tires (3.7% vs. 1.3%).

[0196] Articulated buses (40) consume significantly more fuel than other types of buses, followed by double-decker buses (30).

[0197] Importantly, most of the statistically significant explanatory variables depend on context, representing the action environment, and are therefore not under the control of the operator (bus operating company). Of all the variables under the operator's control in the four interrelated equations (Equations 2–5) (i.e., variables with the ability to change values), the variables are drivers (years of experience, age, number of drivers), tire pressure, tire temperature, bus age, and bus type.

[0198] Furthermore, in the equation for average speed (Equation 2), all other conditions are the same, so it can be seen that there is a tendency for the average speed to be higher the distance traveled every two weeks. The more stops per kilometer, the lower the average speed.

[0199] Furthermore, as the maximum passenger capacity per trip increases, the speed decreases (the loading and unloading time becomes longer). The maximum load per trip has different effects depending on the type of bus. For low-floor conventional buses 20, the maximum load has a greater impact, followed by double-decker buses 30 and articulated buses 40.

[0200] In bus operating companies, long-serving drivers tend to provide higher average speeds. Conversely, older drivers tend to have lower average speeds, while male drivers tend to have higher average speeds. This is primarily related to the routes they drive.

[0201] Interestingly, most drivers tend to experience a decrease in average speed when driving a vehicle for an extended period (around 1,000 km). This may be related to familiarity with the vehicle and route, as the average speed decreases when unfamiliar with the terrain.

[0202] Additionally, there is a possibility of increased tire temperature during peak driving hours (due to congestion and increased braking), as well as during peak passenger times and periods of anticipated high temperatures.

[0203] Maximum load has a significant impact on the tire temperature of the articulated bus 40, followed by the high-floor conventional bus 20 and the double-decker bus 30, but no statistically significant impact is shown for the low-floor conventional bus 20.

[0204] Furthermore, tread depth loss increases with both higher average tire temperature and higher average speed. Interestingly, the optimal pressure level (reduced to 105-115 psi) appears to have a statistically significant effect on tread depth loss in front tires. This suggests that the optimal pressure level does not have a statistically significant effect on tread depth loss in rear tires. This may be due to the influence of other variables.

[0205] Clearly, improvements in tire tread depth can significantly contribute to fuel savings through improvements in fuel consumption. Opportunities to improve tire tread depth may exist, but generally, other variables such as average speed, aside from driving ability which varies depending on the driver's experience and age, are beyond the control of bus operators.

[0206] (3.5) Relationship between maintenance costs

[0207] The maintenance cost of a vehicle (bus) takes into account tire service, frame service, body service, interior service, and routine services, including all inventory (including the cost of tires) and labor operations.

[0208] Figure 11 This provides a descriptive overview of the maintenance cost dataset. Total maintenance costs per kilometer include personnel and inventory costs for each item. Furthermore, the unit is Australian Dollars (AUD).

[0209] The tread depth loss used in this model represents the average for all vehicles' tires. For example... Figure 11 As shown, maintenance costs arise from various cost components of total maintenance costs.

[0210] Therefore, all costs are statistically analyzed and expressed as the maintenance cost per kilometer as the dependent variable. The maintenance cost model is illustrated using other candidate operations, routes, and other variables. The linear regression equation is as follows.

[0211] [Number 7]

[0212]

[0213] Figure 12 The results of this maintenance cost model are shown. The main variables related to the impact on tire performance that explain the variation in total maintenance costs are as follows.

[0214] • Average tread depth loss of all tires installed on the vehicle

[0215] Average tire temperature

[0216] • Age of the bus

[0217] Number of drivers per route

[0218] • A dummy variable representing a regular bus with a high floor (20).

[0219] It can be confirmed that there is a correlation between tire tread depth loss and total maintenance costs.

[0220] (3.6) Decision Support System (DSS)

[0221] Decision support systems (DSS) are particularly effective in providing guidance on tire-related impacts, as well as several important impacts on fuel consumption and maintenance costs. When using a DSS, users can change the levels of explanatory variables (within permissible limits and within practical operational parameters), and the DSS automatically calculates the impact on all other affected variables.

[0222] DSS includes Figure 6All the relationships shown thus affect, for example, the average speed when the average age of drivers changes, thereby affecting the ratio between tire temperature and tread depth loss. Additionally, tread depth loss affects fuel consumption.

[0223] Therefore, even if a change in one explanatory variable (such as average speed) is not included in all four equations, it may still affect all four endogenous dependent variables (for example, average speed is not a direct explanatory variable in the fuel consumption model, but it has the effect of tire tread depth loss).

[0224] Furthermore, DSS adjusts the constant coefficients (see reference) without changing any other explanatory variables. Figure 9 This allows for changes in the dependent variable (i.e., average speed, average tire temperature, tread depth loss ratio, and fuel consumption).

[0225] In other words, the driver correlation coefficient and vehicle correlation coefficient can be adjusted without changing any other explanatory variables.

[0226] As mentioned above, several explanatory variables are included in the 3SLS model, but not all of them are controllable by the operator. Various scenarios are simulated to clarify the impact of these variables on fuel consumption and maintenance cost savings.

[0227] The scheme can be simulated by changing five explanatory variables that are considered to be controllable by the operator. Specifically, these are the ratio of tires with optimal internal pressure levels (105psi to 115psi), the age of the bus, the average age of the drivers, the average number of years of driving experience of the drivers, the ratio of male drivers with 50A, the number of drivers, and the number of drivers per route.

[0228] Figure 10 An example of the simulated scheme is shown. Specifically, Figure 10 Six different simulation scenarios and a baseline (current conditions) are shown, illustrating the predicted impact on four intrinsic dependent variables: average speed, average tire temperature, the rate of tread depth loss per 1,000 km, and fuel consumption.

[0229] In Option 1, the proportion of tires at the optimal internal pressure level (105psi–115psi) increases from 54.17% to 90.00%. The DSS can show the amount (amount) of fuel consumption savings in this option (the same applies to Options 2–7).

[0230] In Option 2, the age of buses is reduced from 6.73 years to 5 years.

[0231] In Option 3, the improved vehicle age of the buses from Option 2 is maintained, while the average age of drivers is increased to 55 years.

[0232] In Option 4, the improved explanatory variables from Option 3 are maintained, and the driver's experience is shortened to 3 years.

[0233] In Option 5, the level of Option 4 is maintained, but 50% of the driving is done by female drivers (50B).

[0234] Option 6 builds upon Option 5 by reducing the number of drivers per vehicle per 1,000 km to 6 (the basic option had 8.61 drivers).

[0235] In addition, from the perspective of maintenance costs, all changes can directly or indirectly affect maintenance costs by altering the ratio of tread depth loss to average tire temperature.

[0236] As illustrated in Scheme 2, the age of the buses, as an explanatory variable, has a significant impact on maintenance costs. Reducing the age of the buses from 6.73 years to 5 years, as in Scheme 2, can result in savings in annual maintenance costs. The same effect was confirmed in both the 3SLS model and the maintenance cost model for all simulated schemes.

[0237] Option 7 is a model that achieves significant savings in both fuel and maintenance costs.

[0238] Figure 13 An example simulation scheme using the DSS tool with a 3SLS model is shown. Specifically, Figure 13 This is an example screenshot showing a simulation scheme of a 3SLS model and maintenance costs using a DSS tool.

[0239] Savings in fuel consumption can be translated into monetary savings on fuel costs. For example, if fuel costs an average of AUD 1.61 per liter, and considering a fuel tax reduction of AUD 0.12003 per liter, the annual savings would be AUD 1.49 per liter.

[0240] Additionally, for example, the total annual mileage provided by the fleet as the object is assumed to be a fixed value, but when using the DSS tool, it is also possible to change the total annual mileage provided by the fleet as the object.

[0241] (4) Functions and effects

[0242] According to the above implementation method, the following effects can be achieved. Specifically, the fleet management system 10 (fuel consumption rate prediction system) can set driver correlation coefficients and vehicle correlation coefficients based on the actual values ​​of driver-related explanatory variables and vehicle-related explanatory variables. In addition, the fuel consumption rate prediction system (specifically, DSS) can use the driver correlation coefficients and vehicle correlation coefficients to recalculate the predicted fuel consumption rate under conditions where at least one of the driver's attributes and vehicle attributes has changed.

[0243] In other words, by focusing on the "driver attributes" and "vehicle attributes" that are variables that the operator (bus operating company) can control, and recalculating the fuel consumption rate when these variables are changed, a method for fleet management can be presented.

[0244] In addition, using tire wear rate, average vehicle speed, and average tire temperature as explanatory variables for fuel consumption rate can improve prediction accuracy.

[0245] In addition, the fleet management system 10 (tire wear prediction system) can use driver correlation coefficient and vehicle correlation coefficient to recalculate the predicted value of tire wear in cases where at least one of the driver's attributes and the vehicle (bus) attributes has been changed.

[0246] In other words, by focusing on the "driver attributes" and "vehicle attributes" that are variables that the operator (bus operating company) can control, and recalculating the wear rate (tread depth loss) when these variables are changed, a method for fleet management can be presented.

[0247] In addition, using the vehicle's average speed and average tire temperature as explanatory variables for wear rate can improve prediction accuracy.

[0248] In other words, the fuel consumption rate prediction system can help control the fuel consumption rate of the buses that make up the fleet. Additionally, the tire wear prediction system can more accurately predict the wear of the tires installed on the buses that make up the fleet.

[0249] In this embodiment, the fleet management system 10 can obtain at least one of the following as driver-related explanatory variables: average age of drivers, gender ratio, average years of experience, and number of drivers per specified driving distance. Therefore, it is possible to use driver-related explanatory variables that have a significant impact on fuel consumption rate or tire wear rate to predict fuel consumption rate and tire wear with higher accuracy.

[0250] In this embodiment, the fleet management system 10 can acquire the average speed of the bus, the wear rate of the tires mounted on the bus (in the case of a fuel consumption rate prediction system), and the average temperature of the tires mounted on the bus as vehicle-related explanatory variables. Therefore, it is possible to use vehicle-related explanatory variables that have a significant impact on fuel consumption rate or wear rate to predict fuel consumption rate and tire wear with higher accuracy.

[0251] (5) Other implementation methods

[0252] The above description of the implementation method is not limited to the described implementation method. Various modifications and improvements can be made, which is self-evident to those skilled in the art.

[0253] For example, in the above implementation, the main explanatory variables used for drivers are the average age of drivers, the gender ratio, the average number of years of experience, and the number of drivers per prescribed driving distance. However, driver-related explanatory variables are not limited to these variables and can use other variables. Figure 7 and Figure 8 Other explanatory variables are shown.

[0254] Furthermore, in the above embodiments, tire wear rate, average vehicle speed, and average tire temperature were used as vehicle-related explanatory variables. However, vehicle-related explanatory variables are not limited to these variables; other variables may be used. Figure 7 and Figure 8 Other explanatory variables are shown.

[0255] The present disclosure has been described in detail above, but it is not limited to the embodiments described herein, as will be apparent to those skilled in the art. The present disclosure can be implemented in modified and altered ways without departing from the spirit and scope of the disclosure as defined by the claims. Therefore, the purpose of this disclosure is illustrative and it is not intended to impose any limitations on the present disclosure.

[0256] Explanation of reference numerals in the attached figures

[0257] 10: Fleet Management System; 20: Regular Bus; 30: Double-decker Bus; 40: Articulated Bus; 50A: Male Driver; 50B: Female Driver; 100: Server Computer; 110: Driver-Related Variable Acquisition Unit; 120: Vehicle-Related Variable Acquisition Unit; 130: Data Storage Unit; 140: Coefficient Setting Unit; 150: Target Variable Calculation Unit; 160: Output Unit; 200: Portable Terminal; 300: Desktop Terminal.

Claims

1. A tire wear prediction system for predicting the wear of tires mounted on vehicles constituting a fleet, the tire wear prediction system comprising: The driver-related variable acquisition unit acquires multiple driver-related explanatory variables related to the attributes of the driver driving the vehicle. The vehicle-related variable acquisition unit acquires multiple vehicle-related explanatory variables related to the attributes of the vehicle. The coefficient setting unit sets the driver correlation coefficient applied to the driver-related explanatory variables and the vehicle correlation coefficient applied to the vehicle-related explanatory variables based on the actual values ​​of the driver-related explanatory variables and the actual values ​​of the vehicle-related explanatory variables. as well as The target variable calculation unit uses the driver-related explanatory variables and the vehicle-related explanatory variables to calculate the wear as the target variable. Specifically, the target variable calculation unit uses the driver correlation coefficient and the vehicle correlation coefficient to recalculate the predicted wear value under the condition that at least one of the driver's attributes and the vehicle's attributes has been changed. The driver-related variable acquisition unit acquires at least one of the following: the driver's average age, gender ratio, average years of experience, and the number of people driving for each specified distance, as the driver-related explanatory variables.

2. The tire wear prediction system according to claim 1, wherein, The vehicle-related variable acquisition unit obtains the average speed of the vehicle as the vehicle-related explanatory variable.

3. The tire wear prediction system according to claim 1 or 2, wherein, The vehicle-related variable acquisition unit acquires the average temperature of the tires installed on the vehicle as the vehicle-related explanatory variable.

4. A tire wear prediction method for predicting the wear of tires mounted on vehicles constituting a fleet, the tire wear prediction method comprising the following steps: Obtain multiple driver-related explanatory variables related to the attributes of the driver driving the vehicle; Obtain multiple vehicle-related explanatory variables associated with the attributes of the vehicle; Based on the actual values ​​of the driver-related explanatory variables and the actual values ​​of the vehicle-related explanatory variables, the driver correlation coefficient applied to the driver-related explanatory variables and the vehicle correlation coefficient applied to the vehicle-related explanatory variables are set. as well as The wear and tear was calculated as the target variable using the driver-related explanatory variables and the vehicle-related explanatory variables. In the step of calculating the target variable, the predicted value of wear is recalculated using the driver correlation coefficient and the vehicle correlation coefficient, considering changes to at least one of the driver's attributes and the vehicle's attributes. The average age, gender ratio, average years of experience, and number of drivers for each specified driving distance are obtained as explanatory variables related to the drivers.

Images