Method for identifying defective vehicles in a vehicle fleet, computer-readable medium, system, and vehicle

Abstract

The invention relates to a method for identifying defective vehicles in a vehicle fleet, the method comprising: determining an expected communication pattern of a vehicle in the vehicle fleet based on communication events between the vehicle in the vehicle fleet and one or more external servers by the external server; classifying a deviation from the expected communication pattern of the vehicle in the vehicle fleet using the communication events between the vehicle in the vehicle fleet and the external servers by an external server; and identifying vehicles in the vehicle fleet as defective vehicles in the vehicle fleet by an external server whose respective communication patterns correspond to the classified deviation from the expected communication pattern of the vehicle in the vehicle fleet.

Description

[0001] The invention relates to a method for identifying faulty vehicles in a vehicle fleet. The invention further relates to a computer-readable medium for identifying faulty vehicles in a vehicle fleet and a system for identifying faulty vehicles in a vehicle fleet.

[0002] Monitoring a connected application in a vehicle is known. For example, a single backend system of the connected application can be monitored at the system level. An error in the connected application can only be detected for the connected application. However, monitoring a vehicle at the vehicle level is not known in the prior art.

[0003] US 2021 / 0 144 026 A1 describes a computer-implemented method for identifying an abnormal event in an operating environment of a vehicle.

[0004] EP 3 798 090 A1 describes a method for monitoring spoken communication in train traffic.

[0005] DE 10 2017 214 661 A1 describes a method for detecting manipulation of at least one of several control units of a motor vehicle.

[0006] It is therefore an object of the invention to more efficiently detect faulty vehicles in a vehicle fleet. In particular, an object of the invention is to efficiently detect communication and / or functional errors in vehicles in a vehicle fleet.

[0007] This object is achieved by the features of the independent claims. Advantageous embodiments and further developments of the invention emerge from the dependent claims.

[0008] According to a first aspect, the invention is characterized by a method for determining faulty vehicles in a vehicle fleet. The method can be a computer-implemented method and / or a control unit-implemented method. The vehicle fleet can be a vehicle fleet of a vehicle manufacturer. A faulty vehicle can have one or more communication errors and / or one or more functional errors in one or more applications of the vehicle. Additionally or alternatively, a faulty vehicle can have one or more communication errors in one or more communication middlewares and / or one or more applications. One or more communication middlewares and / or one or more applications of the vehicle can be executed on one or more control units of the vehicle.For example, one vehicle control unit can run an application and another control unit can run a communication middleware.

[0009] The method comprises determining, by the external-vehicle server, an expected communication pattern of a vehicle in the vehicle fleet as a function of communication events between the vehicle in the vehicle fleet and a server external to the vehicle. The expected communication pattern can be an error-free communication pattern of the vehicle. The vehicle can have one or more error-free communication patterns between itself and an external-vehicle server that are representative of typical, recurring communication patterns of a vehicle in the vehicle fleet with the external-vehicle server. For example, a typical, recurring communication pattern can be representative of a typical communication pattern during the trip with a vehicle in the vehicle fleet, at the beginning of each trip with a vehicle in the vehicle fleet, and / or during the trip with a vehicle in the vehicle fleet.A communication pattern, in particular a typical, recurring communication pattern, of a vehicle in a vehicle fleet can be country-specific and / or specific to a geographical region. The first communication pattern can comprise at least one communication event of a first communication data stream and at least one communication event of a second communication data stream between the vehicle and a server external to the vehicle.

[0010] The method further comprises classifying a deviation from the expected communication pattern of the vehicle in the vehicle fleet using the communication events between the vehicle in the vehicle fleet and the off-vehicle server by the off-vehicle server. Furthermore, the method comprises identifying vehicles in the vehicle fleet as faulty vehicles in the vehicle fleet by the off-vehicle server whose respective communication patterns correspond to the classified deviation from the expected communication pattern of the vehicle in the vehicle fleet.

[0011] Advantageously, the method can identify a faulty vehicle within the fleet using a communication pattern from a single vehicle. This allows communication and / or functional errors in vehicles within a fleet to be efficiently identified.

[0012] According to a further advantageous embodiment, the expected communication pattern can comprise a communication event of a first communication data stream from the vehicle to the off-vehicle server and a communication event of a second communication data stream from the vehicle to the off-vehicle server. This allows the method to efficiently determine deviations from communication patterns that transmit communication results on different communication data streams between a vehicle and an off-vehicle server.

[0013] According to a further advantageous embodiment, the expected communication pattern can be representative of a cross-correlation between the communication event of a first communication data stream from the vehicle to the vehicle-external server and the communication event of a second communication data stream from the vehicle to the vehicle-external server. This allows an expected communication pattern between multiple communication events from multiple communication data streams to be efficiently determined. By using the cross-correlation, deviations from the communication pattern with respect to an individual vehicle can be determined. Including communication events from other vehicles is not absolutely necessary for this. This makes it possible to efficiently reduce the number of false-positive vehicles identified as faulty within the vehicle fleet.

[0014] According to a further advantageous embodiment, the communication event of the first communication data stream can be a communication event of a first application via a first communication channel between the vehicle and the vehicle-external server, and the first application can preferably be executed on a first control unit on the vehicle. Furthermore, the communication event of the second communication data stream can be a communication event of a second application via a second communication channel between the vehicle and the vehicle-external server, and the second application can preferably be executed on a second control unit of the vehicle.Furthermore, the first application can preferably communicate via the first communication channel using a communication middleware, and the second application can preferably communicate via the second communication channel using the communication middleware or another communication middleware. The method can thus use communication patterns of one or more applications and / or one or more communication middleware to identify faulty vehicles.

[0015] According to a further advantageous embodiment, the classification of a deviation from the expected communication pattern of the vehicle in the vehicle fleet using the communication events between the vehicle in the vehicle fleet and the off-vehicle server by the off-vehicle server can include classifying the deviation from the expected communication pattern of the vehicle depending on a cross-correlation between the communication events of the expected communication pattern of the vehicle. A cross-correlation between communication events can thus be used to detect a deviation from an expected communication pattern of the vehicle.

[0016] According to a further advantageous embodiment, the method may further comprise determining a faulty software component of the further vehicle based on one or more cross-correlations between communication events of communication data streams, wherein the faulty software component is preferably an application or a communication middleware of the vehicle, and wherein the faulty software component is preferably the cause of the deviation of the communication pattern of the further vehicle from the first communication pattern of the first vehicle of the vehicle fleet. This allows a faulty software component of the vehicle to be efficiently determined.

[0017] According to a further advantageous embodiment, the identification of faulty vehicles enabled by the method can potentially be used to send a fault correction command to a repair software component on the faulty vehicle(s). This allows faulty vehicles to be efficiently restored to a fault-free state.

[0018] According to a further aspect, the invention is characterized by a computer-readable medium for determining faulty vehicles in a vehicle fleet, wherein the computer-readable medium comprises instructions which, when executed on a computer or a control unit, carry out the method described above.

[0019] According to a further aspect, the invention is characterized by a system for determining faulty vehicles in a vehicle fleet, wherein the system is designed to carry out the method described above.

[0020] Further features of the invention emerge from the claims, the figures, and the description of the figures. All features and combinations of features mentioned above in the description, as well as the features and combinations of features mentioned below in the description of the figures and / or shown alone in the figures, can be used not only in the respective specified combination, but also in other combinations or even on their own.

[0021] In the following, an embodiment of the invention is described with reference to the accompanying drawings. Further details, preferred embodiments and developments of the invention emerge from these. In detail, schematically Fig. 1 an exemplary method for determining one or more faulty vehicles in a vehicle fleet, Fig. 2 an exemplary system for detecting one or more faulty vehicles in a vehicle fleet, and Fig. 3 an example diagram for analyzing faulty vehicles in a vehicle fleet.

[0022] In detail, Fig. 1 shows an exemplary method 100 for determining one or more faulty vehicles in a vehicle fleet. The method may use communication events between a vehicle and one or more off-vehicle servers to determine faulty vehicles in a vehicle fleet. A faulty vehicle may, for example, have a communication and / or functional error, which may be determined using the communication events.

[0023] In detail, the method 100 can determine 102 an expected communication pattern of a vehicle in the vehicle fleet based on communication events between the vehicle in the vehicle fleet and an external server by the external server. An external server, in particular each external server, can store each received communication event between a vehicle in the vehicle fleet and the respective external server. For example, each external server can store a communication event between a vehicle in the vehicle fleet and the respective external server in a log file or a log database. Additionally or alternatively, each vehicle can store an expected communication pattern with an external server.The expected communication pattern can be received for a single vehicle in the fleet and for a given time window, for example, one day. The expected communication pattern can include multiple communication events from one or more communication data streams between a vehicle in the fleet and one or more off-board servers. Examples of communication events include: Number of vehicle status messages from a vehicle to an off-board server, connection setup events of a communication middleware, connection termination events of a communication middleware, timeout events of a communication middleware, a number of mobile radio sessions for a given time window, and / or application-specific messages between a vehicle and an off-board server. Additionally or alternatively, the expected communication pattern may include the content of one or more communication events between a vehicle and an off-board server.

[0024] Furthermore, the expected communication pattern can comprise one or more relationships between communication events. A relationship between at least two communication events of one or more communication data streams can be temporal and / or causal. Furthermore, a relationship between at least two communication events can be determined using a machine learning method. For this purpose, the machine learning method can be trained using the communication events of one or more communication data streams between fault-free vehicles and one or more external servers. The trained machine learning method can finally be used to determine relationships between at least two communication events. The relationships determined using the machine learning method can, for example, be representative of a correlation or a cross-correlation between at least two communication events.Preferably, the method uses a cross-correlation between at least two communication events of one or more communication data streams of an individual vehicle of a vehicle fleet to determine one or more faulty vehicles of the vehicle fleet.

[0025] An exemplary vehicle may include an application A, wherein application A of the exemplary vehicle transmits a communication event to an off-board server every 5 minutes when the exemplary vehicle is awake. The exemplary vehicle may include an application B, wherein application B of the exemplary vehicle transmits a communication event to the off-board server at the start of a trip. Furthermore, the expected communication pattern of the exemplary vehicle may include a correlation that specifies that when a communication event of application B occurs, a communication event of application A must occur within a predetermined time interval.

[0026] Additionally or alternatively, the expected communication pattern may include one or more pieces of indirect information about the respective vehicle. For example, the indirect information about the respective vehicle may include a software version of an application and / or a communication middleware, a geographical location of the vehicle, e.g., a country of the vehicle, and / or a vehicle type.

[0027] The method 100 may classify 104 a deviation from the expected communication pattern of the fleet vehicle using the communication events between the fleet vehicle and the off-vehicle server by the off-vehicle server. In the above-mentioned exemplary vehicle with applications A and B, the method 100 may classify a deviation from the expected communication pattern if there is a deviation in the time interval between the communication event of application A and the communication event of application B with respect to the established correlation of the expected communication pattern.

[0028] Finally, the method 100 can identify 106 vehicles in the vehicle fleet as faulty vehicles in the vehicle fleet through the vehicle-external server, whose respective communication patterns correspond to the classified deviation from the expected communication pattern of the vehicle in the vehicle fleet. The method 100 can receive one or more communication patterns and determine deviations from them. This may allow a conclusion to be drawn about an original error. For example, it can be determined whether an application of the vehicle or a communication middleware of the vehicle is the cause of the error.

[0029] In detail, Fig. 2 shows an exemplary system 200 for determining one or more faulty vehicles in a vehicle fleet. A vehicle in the vehicle fleet may include a control unit 202, for example, a communications control unit, and a control unit 204, for example, an infotainment control unit, often also called a head unit. Within a predetermined period of time 206, for example, one day, the vehicle is driven for a period of time 208. An error-free communication pattern, also called the first communication pattern, of the vehicle may include a communication event 210 of a connection setup message from a communication middleware of the control unit 204 at the beginning of a trip, a communication event 212 of a message from an application of the control unit 204, and a communication event 214 of a connection termination message from a communication middleware of the control unit 204 at the end of a trip.Furthermore, the error-free communication pattern of the vehicle can include a communication event 216 of a connection setup message from a communication middleware of the control unit 202 at the beginning of a trip, three communication events 218 of a message for transmitting a vehicle status from a vehicle status application of the control unit 202, and a communication event 220 of a connection termination message from a communication middleware of the control unit 202 at the end of a trip. The error-free communication pattern can include a cross-correlation between the communication event 218, which is transmitted with a communication data stream of the control unit 202 from the vehicle to a server external to the vehicle, and the communication event 212, which is transmitted with the communication data stream of the control unit 204 from the vehicle to a server external to the vehicle. In the specific example of . Fig. 2 can be specified: if a communication event 212 has been transmitted from the vehicle to the off-board server, then one or more communication events 218 should also have been transmitted from the vehicle to the off-board server. Furthermore, in the example of Fig. 2: If no communication event 218 and no communication event 216 were transmitted from the vehicle to the off-vehicle server, there is an error in the communication middleware of the vehicle's control unit 202. In other words, if a communication pattern of the same vehicle at a later time and / or a communication pattern of another vehicle in the vehicle fleet does not have communication events 218 and 216, then there is a deviation in the communication pattern, and these vehicles are identified as faulty vehicles.

[0030] In detail, Fig.3 shows an exemplary diagram 300 for analyzing the number of faulty vehicles in a vehicle fleet. The diagram shows identified faulty vehicles in a vehicle fleet over time. The x-axis 302 shows a temporal progression for a predefined period of time, for example, one day. The y-axis 304 shows a number of faulty vehicles or a proportion of faulty vehicles in the vehicle fleet. While a certain number of false-positive detections cannot be ruled out, an increase 306, 308 in faulty vehicles over several time units, for example, several days, can indicate vehicles that have communication errors and / or functional errors with a low false-positive rate. The analysis can also be used to efficiently test the effectiveness of remedial measures.

[0031] Advantageously, the method can use communication events and / or communication patterns representative of a vehicle's connection status with one or more off-board servers to identify faulty vehicles with a low false positive rate. Communication patterns are considered at the vehicle level rather than at the system level, and cross-relations between various applications and communication middleware of the vehicle are used to identify faulty vehicles. Potentially faulty vehicles can thus be quickly identified, and remedial measures can potentially be applied to bring the faulty vehicles into a faulty state as quickly as possible. List of reference symbols 100 procedures 102 Determining an expected communication pattern 104 Classifying a Deviation 106 Detecting vehicles 200 systems 202 control unit 204 Control unit 206 Period 208 Period 210 Communication event 212 Communication event 214 Communication event 216 Communication event 218 Communication event 220 communication event 300 diagram 302 Time 304 Number 306 increase over period 308 increase over period

Claims

[1] Method (100) for determining faulty vehicles of a vehicle fleet, the method (100) comprising: Determining (102) an expected communication pattern of a vehicle of the vehicle fleet as a function of communication events (210, 212, 214, 216, 218) between the vehicle of the vehicle fleet and a vehicle-external server by the vehicle-external server, wherein the expected communication pattern comprises a communication event (210, 212, 214, 216, 218) of a first communication data stream from the vehicle to the off-vehicle server and a communication event (210, 212, 214, 216, 218) of a second communication data stream from the vehicle to the off-vehicle server; Classifying (104) a deviation from the expected communication pattern of the vehicle of the vehicle fleet using the communication events (210, 212, 214, 216, 218) between the vehicle of the vehicle fleet and the off-vehicle server by the off-vehicle server; and Determining (106) vehicles of the vehicle fleet as faulty vehicles of the vehicle fleet by the vehicle-external server, whose respective communication patterns correspond to the classified deviation from the expected communication pattern of the vehicle of the vehicle fleet. [2] Method (100) according to the preceding claim, wherein the expected communication pattern is representative of a cross-correlation between the communication event (210, 212, 214, 216, 218) of a first communication data stream from the vehicle to the off-vehicle server and the communication event (210, 212, 214, 216, 218) of a second communication data stream from the vehicle to the off-vehicle server. [3] Method (100) according to one of the preceding claims, wherein the communication event of the first communication data stream is a communication event (210, 212, 214, 216, 218) of a first application via a first communication channel between the vehicle and the vehicle-external server, and wherein the first application is preferably executed on a first control unit on the vehicle; and wherein the communication event (210, 212, 214, 216, 218) of the second communication data stream is a communication event of a second application via a second communication channel between the vehicle and the vehicle-external server, and wherein preferably the second application is executed on a second control unit of the vehicle, and wherein preferably the first application communicates via the first communication channel by means of a communication middleware; and wherein preferably the second application communicates via the second communication channel by means of the communication middleware or another communication middleware. [4] The method (100) of any preceding claim, wherein classifying a deviation from the expected communication pattern of the vehicle of the vehicle fleet using the communication events (210, 212, 214, 216, 218) between the vehicle of the vehicle fleet and the off-vehicle server by the off-vehicle server comprises: Classifying the deviation from the expected communication pattern of the vehicle depending on a cross-correlation between the communication events of the expected communication pattern of the vehicle. [5] Method (100) according to one of the preceding claims, the method further comprising: Determining a faulty software component of the further vehicle depending on the cross-correlation between the communication events (210, 212, 214, 216, 218) of the first communication data stream and the second communication data stream, wherein the faulty software component is preferably an application or a communication middleware of the vehicle, and wherein the faulty software component is preferably the cause of the deviation of the communication pattern of the further vehicle from the first communication pattern of the first vehicle of the vehicle fleet. [6] Method (100) according to one of the preceding claims, the method further comprising: Executing a fault correction command of a repair software component on the faulty vehicle(s), wherein the repair software component is linked to the deviation or the respective deviation of the first communication pattern of the further vehicle of the vehicle fleet from the first communication pattern of the first vehicle of the vehicle fleet. [7] A computer-readable medium for detecting faulty vehicles of a vehicle fleet, the computer-readable medium comprising instructions which, when executed on a server, carry out the method according to any one of claims 1 to 6. [8] System (200) for determining faulty vehicles of a vehicle fleet, wherein the system is designed to carry out the method according to one of claims 1 to 6.

Images