Tire identification gate and method for long multi-wheeled vehicles such as trucks, semi-trailer trucks, road trains, and buses
The method effectively identifies tire configurations on long multi-wheeled vehicles by clustering and sequencing RFID data using K-means, Davies-Bouldin index, and edit distance, addressing data loss and noise issues for precise tire monitoring.
Patent Information
- Application Number
- JP2025552077
- Authority / Receiving Office
- JP · JP
- Patent Type
- Applications
- Current Assignee / Owner
- Priority Date
- 2023-03-09
- Filing Date
- 2024-03-07
- Publication Date
- 2026-03-06
AI Technical Summary
Efficient tire identification for long multi-wheeled vehicles with dual-type wheels is challenging due to proximity of wheels, leading to data loss, noise interference, and misassociation, affecting monitoring tasks like tire pressure and tread wear.
A computer-implemented method using K-means clustering, Davies-Bouldin index, and edit distance to sort and evaluate RFID data from dual-type wheels, determining tire configuration by grouping, sorting, and iteratively refining clusters based on unique identifiers.
Achieves accurate and efficient tire identification with low computational load, robustness to missing readings, and adaptability to various vehicle types, making it suitable for low-performance devices.
Smart Images
Figure 2026507899000001_ABST
Abstract
Description
[Technical Field]
[0001] The present invention relates generally to tire identification for motor vehicles, particularly long multi-wheeled vehicles such as trucks, semi-trailer trucks, road trains, buses, etc. More particularly, the present invention relates to a tire identification system and related tire identification method that is particularly advantageous for use on long multi-wheeled vehicles equipped with dual-type wheels. [Background technology]
[0002] As is well known, radio frequency identification (RFID) technology is widely used in the automotive field for identification and monitoring purposes. In particular, nowadays, the incorporation of RFID devices in / on tires for tire identification and monitoring purposes is widespread for the following purposes: -Optimizing manufacturing processes, logistics, and assembly line operations for original equipment manufacturers (OEMs); -Ensuring tracking and traceability of tires; -improving work efficiency in vehicle management (e.g., maintenance work); -Speed up inventory management and audits with greater data accuracy; -Enabling services for mobility solutions.
[0003] For example, EP 3074917(A1) discloses a system for dynamic reading of data from vehicle tire transponders having a linearly polarized antenna, the system comprising at least one antenna capable of receiving data transmitted by the transponder, and at least one reader coupled to the antenna and capable of reading and storing data from the transponder, the antenna being a linearly polarized antenna with a vertical electric field.
[0004] EP 2829659 A1 discloses a stand and antenna assembly for receiving data transmissions from an electronic transmitting device mounted on a vehicle. The vehicle has at least one tire assembly including a tire. The stand and antenna assembly comprises a stand having an upper or substantially convex upper surface, the upper surface including approach and exit ramps aligned and operative to capture and engage a tire as it traverses the upper surface of the stand. The stand further comprises a base having at least one pocket disposed therein and configured to receive an antenna assembly, and at least a first antenna assembly within one pocket of the base. The antenna assembly is seated in the pocket at an oblique angle relative to a bottom surface of the base and is operative to direct a read field in the direction of the vehicle's approach path to the stand.
[0005] U.S. Patent No. 10,935,466 (B2) discloses an integrated tire sensor and reader system including at least one sensor unit mounted on a tire or wheel. The reader is located remotely from the sensor unit. The sensor unit includes at least one sensor for measuring parameters of the tire or wheel and an antenna for communicating with the reader. The sensor unit is configured to receive radio frequency power signals from the reader and transmit data to the reader. The reader includes an antenna for transmitting radio frequency power signals to one sensor unit to activate the sensor unit. When the sensor unit is activated, the sensor measures parameters of the tire or wheel, and data from the sensor is transmitted from the sensor unit to the reader.
[0006] U.S. Patent No. 8,009,027 (B2) discloses a tire monitoring system and related method. The tire monitoring system includes a plurality of tire sensor modules configured to transmit tire data at predetermined time intervals using non-contact sensors. The system also includes a central control unit configured to receive tire data from the tire sensor modules, and further includes an external control unit capable of communicating with the tire sensor modules, the central control unit, and an external source.
[0007] In this context, it is worth noting that tire identification can be an extremely challenging task for trucks, semi-trailer trucks, road trains, buses, and more generally, long multi-wheeled vehicles due to the large number of wheels, often dual-type wheels, located in close proximity. Indeed, in these cases, identification data from some tires may be lost, affected by too much noise to be usable, or misassociated or confused / combined with data from other tires. Therefore, in these scenarios, tire identification performance is significantly degraded, thereby affecting the performance of any other monitoring tasks (e.g., tire pressure monitoring, tread wear monitoring, etc.) that are performed in conjunction with and based on tire identification.
[0008] Therefore, today there is a strong need for efficient tire identification techniques in the automotive sector, especially in all cases where known tire identification techniques are unreliable and inefficient, e.g. in the case of trucks, semi-trailer trucks, road trains, buses and more generally in the case of long multi-wheeled vehicles with dual-type wheels. Summary of the Invention
[0009] In view of the above, the applicant felt the need to develop an efficient tire identification technology that would enable efficient identification of tires on motor vehicles such as long multi-wheeled vehicles with dual-type wheels (e.g., trucks, semi-trailer trucks, road trains, buses, etc.), and thus devised the present invention.
[0010] It is therefore an object of the present invention to provide a tire identification technique that generally overcomes the technical drawbacks of prior art solutions and, more specifically, provides efficient tire identification for motor vehicles such as long multi-wheeled vehicles with dual-type wheels, trucks, semi-trailer trucks, road trains, buses, etc.
[0011] This and other objects are achieved by the present invention, which relates to a computer-implemented method for efficiently identifying the tire configuration of a vehicle, and an associated tire identification gate and method, as defined in the appended claims.
[0012] In particular, the present invention relates to a computer-implemented method for efficiently identifying the tire configuration of a motor vehicle having tires each mounted with a respective radio frequency identification device, the method comprising: receiving identification data indicative of a unique identifier of a tire located on the left side of a given one of said vehicles; receiving identification data indicative of the unique identifier of a tire located on the right side of a given motor vehicle; - grouping the received identification data into a left data group and a right data group; - Sorting the identification data of the left group and the identification data of the right group in order of acquisition time; -For either the left or right group, - iteratively applying a K-means clustering method to the sorted identification data, thereby obtaining, at each iteration, a respective identification data clustering representing a respective candidate tire configuration for a given motor vehicle; - evaluating the obtained discriminative data clustering by the Davies-Bouldin index metric, thereby determining an estimated tire configuration for a given vehicle; - determining the actual tire configuration of a given vehicle by evaluating the estimated tire configurations by an edit distance metric; - determining a respective position of each tire on the side associated with the determined group in the determined actual tire configuration and assigning a respective unique identifier to said positions based on said determined actual tire configuration and the sorted identification data; - determining a respective position of each tire on opposite sides of a given vehicle in the determined actual tire configuration, and assigning a respective unique identifier to said positions based on said determined actual tire configuration and the sorted identification data. [Brief explanation of the drawings]
[0013] For an understanding of the invention, an embodiment thereof will now be described, purely by way of non-limiting and non-binding example, with reference to the accompanying drawings, in which: [Figure 1] 1 illustrates a schematic diagram of a high-level functional architecture of an RFID gate for a motor vehicle, such as a truck, according to an embodiment of the present invention. [Figure 2] 2 illustrates an exemplary implementation of the RFID gate of FIG. 1. [Figure 3] Figure 2 shows a schematic bottom view of a semi-trailer truck. [Figure 4] 1 illustrates a schematic diagram of a tire identification method according to an embodiment of the present invention. [Figure 5] An example of the results obtained by performing some steps of the tire identification method illustrated in FIG. 4 is shown. [Figure 6] An example of the results obtained by performing some steps of the tire identification method illustrated in FIG. 4 is shown. [Figure 7]An example of the results obtained by performing some steps of the tire identification method illustrated in FIG. 4 is shown. [Figure 8] An example of the results obtained by performing some steps of the tire identification method illustrated in FIG. 4 is shown. [Figure 9] An example of the results obtained by performing some steps of the tire identification method illustrated in FIG. 4 is shown. DETAILED DESCRIPTION OF THE INVENTION
[0014] The present invention will now be described in detail with reference to the drawings so that those skilled in the art can understand, reproduce, and use the present invention. Various modifications to the described embodiments will be readily apparent to those skilled in the art, and the general principles described can be applied to other embodiments and applications without departing from the scope of protection of the present invention as defined in the appended claims. Therefore, the present invention should not be considered limited to the described and illustrated embodiments, but should be accorded the widest scope of protection in accordance with the features described and claimed.
[0015] FIG. 1 shows a schematic diagram of a high-level functional architecture of an RFID gate (generally designated 1) for a vehicle such as a truck (e.g., a semi-trailer truck, a road train, a bus, or more broadly, a long multi-wheeled vehicle) according to an embodiment of the present invention, and FIG. 2 shows an exemplary implementation of RFID gate 1.
[0016] In particular, the RFID gate 1 is designed to be passed by a vehicle such as a truck (e.g., a semi-trailer truck as shown in Figure 2 and generally designated 2 therein) and comprises a pair of RFID antennas 11, 12 arranged on either side of the path passed by the truck. In this regard, it is worth noting that in Figure 2 only the RFID antenna 11 on the right side (the right side relative to the direction of travel of the semi-trailer truck 2) is shown, while the RFID antenna 12 on the left side is not visible.
[0017] Furthermore, the RFID gate 1 further includes an RFID reader 13 and a tire monitoring system 14 configured to monitor one or more tire characteristics while the semi-trailer truck 2 passes through the RFID gate 1, and in the example of Figure 2 is implemented by a ramp designed to determine (e.g., by laser measurement) tire tread wear on the tires of the truck traveling on the ramp.
[0018] The semi-trailer truck 2 includes a tractor unit 21 and a semi-trailer 22 attached to and towed by the tractor unit 21. The tractor units 21 are conveniently equipped with respective RFID tags storing respective identification data indicative of a respective unique identifier (UID) of the tractor units 21. In an exemplary embodiment, the RFID tags may be conveniently attached / located on or near the front panel 210 of the tractor units 21 so as to be easily readable by the RFID antennas 11, 12. The semi-trailer 22 is also conveniently equipped with respective RFID tags storing respective identification data indicative of a respective UID of the semi-trailer 22. In an exemplary embodiment, the RFID tags may be conveniently attached / located on the underside 220 of the semi-trailer 22 so as to be easily readable by the RFID antennas 11, 12. Furthermore, each tire of the semi-trailer truck 2 (i.e. the tractor unit 21 and the semi-trailer 22) is fitted with a respective RFID tag that stores respective identification data indicating the respective UID of the tire.
[0019] In this regard, Figure 3 shows a schematic bottom view of the semi-trailer truck 2, and it is worth noting the following: The tractor unit 21 comprises a pair of front steering wheels 211, 212 and two rear dual drive wheels 213, 214; and The semi-trailer 22 is provided with three pairs of trailer wheels 221, 222, 223.
[0020] While the semi-trailer truck 2 passes through the RFID gate 1, the RFID antennas 11 and 12 - Tractor unit 21, - tyres for the steering wheels 211, 212 and the driving wheels 213, 214, - Semi-trailer 22, and -Tires for trailer wheels 221, 222, 223 and reading therefrom the respective identification data provided by the UID.
[0021] The RFID antennas 11 , 12 are further configured to provide read identification data to an RFID reader 13 .
[0022] The RFID reader 13 is a processing device comprising at least one processor and memory. It may be located near the RFID antennas 11, 12 and connected thereto in a wired manner. Alternatively, the RFID reader 13 may be a remote device / system remotely connected to the RFID antennas 11, 12 in a wired and / or wireless manner. In an exemplary embodiment, the RFID reader may be implemented in a cloud computing system located remotely from the RFID gate 1.
[0023] The RFID reader 13 is configured to identify and locate the mounting position of each tire of the semi-trailer truck 2 relative to the trailer truck 2 (i.e., determine the position of each tire relative to the semi-trailer truck 2) based on the read identification data received from the RFID antennas 11, 12.
[0024] The respective UID and location of each identified and located tire may then be stored in data lake 15 in association with the respective tire monitoring data provided by monitoring system 14 .
[0025] To identify and locate each tire of the semi-trailer truck 2, the RFID reader 13 is programmed to execute a predefined tire identification method comprising two main stages: a clustering stage and a sequencing stage.
[0026] In fact, generally speaking, with respect to a vehicle that can be classified as a truck, the actual position of each tire of the truck, and therefore its type, strictly depends on the configuration of the truck (more specifically, the tire / wheel configuration / layout of the truck considered). Therefore, to determine the tire configuration of the truck, it is necessary to distinguish tire clusters, if any. That is, a cluster represents a group of tires according to the type of truck considered (e.g., tires on steering wheels, driving wheels, or trailer wheels). In most cases, each cluster contains multiple tires, and therefore a technique for determining the cluster hierarchy is required (sequencing stage).
[0027] The predefined tire identification method executed by the RFID reader 13 focuses on each acquired cluster of tires, rather than all tires of the vehicle considered as a whole. In this way, read errors are not propagated throughout the system and can be efficiently processed within a single cluster in a faster and more accurate manner.
[0028] More specifically, the predefined tire identification method comprises, as mentioned above, two main stages: a clustering stage in which groups of axles are identified, and then a sequencing stage in which the positions of the tires within the determined tire configuration are identified. Furthermore, the method can advantageously comprise some additional specific steps (such as twin tire recognition) to further improve accuracy.
[0029] It is worth noting that the predefined tire identification method is independent of the way in which the identification data is acquired by the reader. It allows determining the tire configuration of a vehicle (e.g., truck) by using only the readings of the tire-embedded RFID tags of the vehicle's tires. That is, the method is applicable even in the absence of any specific RFID tags applied / installed on the vehicle itself (e.g., those identifying the tractor unit 21 and semi-trailer unit 22 of a semi-trailer truck 2), and is even independent of the specific RFID gate architecture being used, as long as an appropriate reader acquires the read identification data.
[0030] For a better understanding of the present invention, Figure 4 illustrates schematically a tire identification method, generally designated 4, which is carried out in use by an RFID reader 13 according to one embodiment of the present invention.
[0031] As shown in FIG. 4, the tire identification method 4 includes a pre-processing step 41 followed by the clustering and sequencing stages (denoted 42 and 43 respectively).
[0032] The tire identification method 4 will now be described in detail using a semi-trailer truck as an example, as shown in Figure 2. However, it will be apparent to those skilled in the art in light of this disclosure that the method is equally applicable to any type of wheeled vehicle.
[0033] 1. Pretreatment The clustering stage 42 is based on the idea of clustering the read identification data (i.e. UIDs) according to their "collection" over time by the RFID antennas 11, 12. The pre-processing step 41 therefore comprises sorting the read identification data received from the RFID antennas 11, 12 according to the chronological order of their reading, as shown diagrammatically in Figure 5. Furthermore, to make the algorithm more robust, the read identification data are separated at this stage into two groups for which the RFID antennas 11, 12 provided said identification data, and therefore according to the side of the track to which the read identification data relate.
[0034] 2. Clustering Once the pre-processing step 41 is completed, the clustering stage 42 is performed.
[0035] The separation into two data groups (left and right groups) during the pre-processing stage allows the algorithm to be more robust against missing readings. Indeed, it may happen that RFID tags are not read during the collection stage. To overcome this drawback, the clustering stage 42 is advantageously applied first to the group (associated with the corresponding side of the track) containing the largest number of read RFID tags.
[0036] More particularly, the clustering step 42 comprises: a K-means based clustering step 421 in which an initial clustering is achieved representing an initial estimated tire configuration of the truck; an evaluation step 422 based on the Davies-Bouldin index, during which the K-means based clustering step 421 is iterated, and the quality of the clusters obtained at each iteration of the K-means based clustering step 421 is evaluated by the Davies-Bouldin index metric to select the best cluster and therefore the best tire configuration for the truck; an edit distance based evaluation step 423 for determining the actual tire configuration of the truck based on an edit distance metric and a given dictionary of predefined tire configurations; an optional twin tire recognition step 424 for detecting twin tires on dual type wheels.
[0037] 2.1 K average In K-means based clustering step 421, a K-means algorithm is performed, the inputs of which are the read identification data (conveniently along with the median read timestamp associated / assigned to that read identification data) and the expected number of clusters.
[0038] K-means evaluates the distance between a sample and a reference point (centroid). This distance is evaluated for all centroids, the number of which is equal to the number of clusters. The closest centroid for a given sample determines the cluster to which the given sample belongs. Centroids are initially selected randomly, and for each iteration, they are evaluated as the mean value of the cluster, and the distance for each sample is evaluated again. K-means converges when a sample no longer changes its cluster.
[0039] In this regard, Figures 6 and 7 show two examples of data clusters before and after K-means processing.
[0040] Since the initial selection of centroids is random, the K-means algorithm is run N times (N is an integer greater than zero, e.g., N=10) by evaluating the metric (i.e., within-cluster sum of squares - WCSS) for each iteration.
[0041]
number
[0042] This metric simply evaluates the variance of the data within each cluster. The smallest value therefore represents the best clustering for the given dataset and is used for the next step (i.e., evaluation step 422 based on the Davies-Bouldin index), as shown in Figure 8, where the lowest WCSS corresponds to the fifth iteration, which is therefore selected as the one to be used for evaluation step 422 based on the Davies-Bouldin index.
[0043] In summary, in the K-means based clustering step 421, q (number of clusters) centroids are randomly selected (initialized). For each read data item, the distance from each centroid is evaluated. The shortest distance identifies the centroid of the i-th read data item. A new centroid is selected and new iteration(s) are performed until convergence occurs.
[0044] 2.2 Davies-Bouldin Index It is worth noting that the expected number of clusters, q, varies according to the track being considered. Thus, the K-means algorithm is run n times (n is an integer greater than 0) by changing the number of clusters for each iteration (e.g., n=2, q=2, 3).
[0045] To determine the best clustering for a given dataset, the Davies-Bouldin Index (DBI) is used. This metric simply evaluates the ratio between the within-cluster variance and the between-cluster separation.
[0046]
number
[0047] The minimum value of DBI is obtained by the ideal number of clusters, q, as shown in Figure 9. Ideal , the minimum value of DBI is 3, and therefore the ideal number of clusters is 3 (i.e., q Ideal =3).
[0048] At this stage, we extract the estimated tire configuration of the truck by counting the number of distinct UIDs for each cluster. For example, if one UID is detected in the first cluster, two UIDs in the second cluster, and three UIDs in the third cluster, the truck's tire configuration is 1-2-3 (i.e., one steering tire / wheel, two driving tires / wheels, and three trailer tires / wheels).
[0049] 2.3 Edit Distance Next, in an edit distance-based evaluation step 423, the estimated tire configuration is searched in a given dictionary (which may be stored on the RFID reader 13 or on any other device connected to the RFID gate 1 in a wired or wireless manner) containing several predefined tire configurations of trucks. It may happen that the estimated tire configuration does not correspond to any predefined tire configuration in the given dictionary. In this case, the estimated tire configuration is then compared with all predefined tire configurations having the same total number of tires. For each comparison, an edit distance metric is evaluated, which represents the number of operations required to convert the estimated tire configuration into the considered predefined tire configuration. The smallest edit distance determines the most likely tire configuration.
[0050]
number
[0051] Thus, the output of the edit distance based evaluation step 423 is the tire / wheel configuration of the truck and the number of axles (i.e., the number of UIDs in each cluster) on the side of the truck with the greatest number of read UIDs, with a similar configuration possible on the other side of the truck, of course.
[0052] In view of the above, the clustering stage 42 jointly and synergistically utilizes three statistical methodologies (i.e., K-means, Davies-Bouldin index, and edit distance) to efficiently achieve the best clustering results.
[0053] 2.4 Twin tire recognition The twin tire recognition substep 424 is an optional step that is performed only if it is clear from the tire configuration identified at the end of the previous step that twin tires are expected to be on the vehicle. When the edit distance based evaluation step 423 is performed, a threshold based analysis of the relative time distance between the read RFID tags is performed to detect twin tires on dual type wheels of the truck.
[0054] Advantageously, a track-length-related time difference is calculated between the first and last RFID tag reads, and each read timestamp is normalized to the calculated track-length-related time difference. Then, for each UID, a respective median timestamp is determined, and a respective relative time distance between that respective median timestamp and the median timestamps of the other read UIDs is calculated. Thus, if the calculated relative time distance is lower than a given threshold (e.g., equal to a given percentage of the track-length-related time difference), a twin tire is detected.
[0055] 3. Sequencing Once the clustering step 42 has been performed, a sequencing step 43 is performed to correctly assign UIDs to the actual tires.
[0056] The best-side selection is based on the maximum number of read RFID tags, and since at the end of the clustering stage tire configurations have been identified for all real tires (even those that may have missing RFID reads), no "missing tires" are expected at this stage. Therefore, UIDs of RFID tags are assigned according to their cluster and their temporal position within it.
[0057] If at least a twin tire configuration is detected and sub-step 424 is performed, it is also possible to specifically identify which tire is in the inner position and which tire is in the outer position within the twin tire configuration. -It is assumed that there is at least one RFID reading for each tire in a twin tire configuration, and that if the number of readings for one tire differs from the number of readings for the other tire, the UID with the highest number of readings is associated with the outer tire and the UID with the lowest number is associated with the inner tire. In practice, it is expected that reading the tags of the inner tire by the antennas 11, 12 will be more prone to errors due to noise and interference. If there is at least one RFID reading for only one UID, then that UID is assumed to correspond to the outer tire. - In the unlikely event that (a) there is a zero reading for both tires in a twin tire configuration, or (b) the number of readings for each tire in a twin tire configuration is the same, it may be necessary to repeat the reading of the tags of tires in a twin tire configuration.
[0058] The tire allocation on the other side of the track (ie the side corresponding to the data group to which the clustering step has not been applied) can be divided into two different cases. - all expected RFID tags have been read, and therefore the read UIDs are assigned to the tires according to the chronological order of the readings as described above, or - there is at least one missing RFID tag. In this case, for each UID, the median timestamp is evaluated, and then the time distance between each timestamp associated with the side of the truck with fewer readings and the timestamp associated with the other side is evaluated. In this way, tires on the side of the truck with fewer readings are associated with the corresponding tires on the other side of the truck to determine their cluster. Once the cluster is determined, the read UIDs are assigned to the tires according to the chronological order of the readings, as described above. ******
[0059] In view of the above, the technical advantages and innovative features of the present invention will be readily apparent to those skilled in the art.
[0060] It is particularly noteworthy that the tire identification method according to the present invention can achieve excellent tire identification performance in terms of successful tire identification and computational load, which makes it possible to use this lightweight and energy-efficient method even on low-performance processing devices.
[0061] As mentioned above, the tire identification method according to the present invention is primarily based on three statistical methods, namely, K-means, Davies-Bouldin index, and edit distance; all three statistical methods are known in the art, but their specific combination, sequence, and synergistic use according to the present invention is completely unknown in the art and brings significant benefits to the overall performance of tire identification.
[0062] Additionally, the tire identification method of the present invention can identify the configuration of the truck using the axle arrangement. At the end of the process, a dictionary check is performed to find a correlation with one known configuration in a predefined dictionary containing the previously identified configuration. This approach actually makes the methodology independent not only of different types of trucks, but also of the type of vehicle itself. In fact, this method is applicable to other types of vehicles, unlike known methods that require a priori fixation of the structure / vehicle type in order to utilize a priori knowledge of possible RFID tag locations / shapes.
[0063] This concept is strengthened by the fact that the method according to the invention uses only one dimension, namely time, which makes the algorithm pipeline potentially adaptable to other acquisition systems than those relying on RFID, and therefore usable for new applications (even non-automotive) that require configuration sequencing and identification.
[0064] In summary, the following advantages of the present invention can be highlighted: 1) The tire identification method can be considered lightweight due to its low computational load, effectively making the system energy efficient. 2) The algorithm does not require a training phase, which makes the implementation of the solution independent of the application context. 3) The dictionary containing the list of track types can be created with predefined configurations, so the system is customizable, and the algorithm uses the predefined configurations to find an exact match between the found configuration and an existing configuration, or finds the most likely configuration based on an edit distance-based comparison. 4) The system is robust to the loss of some RFID tag readings or the loss of all readings for a particular RFID tag. The algorithm identifies the best side of the vehicle and uses it to find the actual tire configuration; the determined tire configuration is then used for tire identification and placement on the other side of the truck. In either case, the edit distance helps to identify the correct configuration, to a good approximation, even in the case of inaccurate readings and / or noisy data. 5) Only a time metric is used, which effectively makes the tire identification method independent of the RFID gate 1, as it is irrelevant what technology or method is used to acquire the data. 6) Problems similar to the problem of identifying truck configurations may potentially be solved by the present invention. As anticipated above, the proposed tire identification method is applicable to any type of vehicle, such as a train, car, or bus.
[0065] In conclusion, it will be apparent that numerous modifications and variations can be made to the present invention, all of which are encompassed within the scope of the invention as defined in the appended claims.
[0066] In this regard, it is important to note that although the present invention has been described above with specific reference to the use of RFID technology, it is absolutely clear to those skilled in the art that any other radio frequency identification technology may be used, such as, for example, Internet of Things (IoT) type radio frequency identification technology.
[0067] Furthermore, although the description refers to the specific example of tire tread wear monitoring, it will be apparent to those skilled in the art that the present invention may be advantageously utilized in conjunction with monitoring any type of tire characteristic (e.g., tire pressure monitoring, tire damage detection and monitoring, etc.).
Claims
1. 1. A computer-implemented method (4) for efficiently identifying the tire configuration of a motor vehicle (2) having tires each mounted with a respective radio frequency identification device, said method (4) comprising: - receiving identification data indicative of the unique identifier of a tire located on the left side of a given one of said vehicles (2); - receiving identification data indicative of the unique identifier of a tire located on the right side of said given vehicle (2); - grouping the received identification data into a left data group and a right data group; - sorting the identification data of the left group and the identification data of the right group in order of acquisition time; for either the left group or the right group, - iteratively applying a K-means clustering method to the sorted identification data, thereby obtaining, at each iteration, a respective identification data clustering representative of each candidate tire configuration of the given motor vehicle (2); - evaluating the obtained discriminative data clustering by the Davies-Bouldin index metric, thereby determining an estimated tire configuration of the given vehicle (2); - determining the actual tire configuration of said given vehicle (2) by evaluating said estimated tire configuration by an edit distance metric; - determining a respective position of each tire of the side associated with the determined group in the determined actual tire configuration and assigning a respective unique identifier to said positions based on the determined actual tire configuration and the sorted identification data; - determining a respective position of each tire on opposite sides of the given vehicle (2) in the determined actual tire configuration and assigning a respective unique identifier to said positions based on the determined actual tire configuration and the sorted identification data.
2. 2. The method of claim 1, wherein the actual tire configuration of the given vehicle (2) is determined based on a given dictionary of predefined tire configurations.
3. the actual tire configuration of the given vehicle (2) is determined by looking up the estimated tire configuration in the given dictionary; - if the estimated tire configuration corresponds to a given predefined tire configuration of the predefined tire configurations, the actual tire configuration is determined to be the given predefined tire configuration, 3. The method of claim 2, wherein if the estimated tire configuration does not match any of the predefined tire configurations, a respective edit distance between the estimated tire configuration and each predefined tire configuration in the given dictionary is calculated, whereby the actual tire configuration is the predefined tire configuration that has the shortest edit distance from the estimated tire configuration.
4. 4. The method of claim 1, wherein a respective position of each tire on opposite sides of the given vehicle (2) in the determined actual tire configuration is determined, and wherein the respective unique identifiers are assigned to said positions also based on the determined position of the tire on the side associated with the determined group in the determined actual tire configuration.
5. 5. The method according to any one of claims 1 to 4, further comprising performing a threshold-based analysis of relative time distances between the sorted identification data to detect twin tires of dual wheels of the given motor vehicle (2).
6. The method according to any one of claims 1 to 5, wherein said given motor vehicle is a truck, a semi-trailer truck (2), a road train, a car, a bus, etc.
7. A tire identification gate (1) designed to be passed by a motor vehicle (2) having tires fitted with respective radio frequency identification devices and to identify the tires of said motor vehicle (2), said tire identification gate (1) comprising at least two antennas (11, 12) and a reader (13); The antennas (11, 12) - located on both sides of the path traversed by the vehicle (2) passing through the tire identification gate (1); - configured to obtain, while a given one of the motor vehicles (2) passes through the tyre identification gate (1), respective identification data from the respective radio frequency identification devices of each tyre of said given motor vehicle (2), indicative of a respective unique identifier of said tyres; A tire identification gate (1), wherein the reader (13) is configured to perform the method according to any one of claims 1 to 6.
8. a tire monitoring system (14) configured to monitor one or more tire characteristics while the given vehicle (2) passes through the tire identification gate (1); 8. The tire identification gate of claim 7, wherein each identified tire is associated with respective tire monitoring data provided by the monitoring system (14) and relating to the identified tire.
9. A tire identification method, comprising: a) providing a tire identification gate (1) through which a vehicle (2) having tires each fitted with a respective wireless identification device passes, said tire identification gate (1) comprising: - at least two antennas (11, 12) arranged on either side of the path traversed by the vehicle (2) passing through the tire identification gate (1); - providing a tire identification gate (1) comprising a reader (13); b) obtaining, while a given one of the vehicles (2) passes through the tire identification gate (1), respective identification data from the respective radio frequency identification devices of each tire of the given vehicle (2) by means of the antennas (11, 12), the respective identification data being indicative of a unique identifier of each of the tires; c) carrying out by said reader (13) the method (4) according to any one of claims 1 to 6.
10. The tire identification gate (1) further comprises a tire monitoring system (14), and the tire identification method includes: - monitoring one or more tire characteristics by said tire monitoring system (14) while said given vehicle (2) passes through said tire identification gate (1); 10. The tire identification method of claim 9, further comprising: associating each identified tire with respective tire monitoring data provided by said monitoring system (14) and relating to said identified tire.
11. A computer program comprising instructions which, when executed by a computer, cause the computer to carry out the method according to any one of claims 1 to 6.
12. A computer readable storage medium comprising instructions which, when executed by a computer, cause the computer to perform the method of any one of claims 1 to 6.
13. 1. A software or firmware program product comprising a software / firmware code portion, said software / firmware code portion comprising: - each radio frequency identification device is designed to be passed by a motor vehicle (2) having a respective tire attached thereto, and also comprises at least two antennas (11, 12) arranged on either side of the path passed by the motor vehicle (2) passing through the tire identification gate (1), said radio frequency identification device being readable / storable in and executable by a reader (13) of said tire identification gate (1), said antennas (11, 12) being: - obtaining, while a given one of the motor vehicles (2) passes through the tire identification gate (1), respective identification data from the respective radio frequency identification devices of each tire of said given motor vehicle (2), indicative of a respective unique identifier of said tire; - providing said obtained identification data to said reader (13); - so that, when executed by said reader (13), A software or firmware program product adapted to cause and to carry out the method according to any one of claims 1 to 6.