Methods for recording fault data of a vehicle and methods for predicting the failure of a component of a vehicle

Abstract

The invention relates to a method for acquiring fault data from a vehicle, wherein the fault data is acquired from diagnostic data and claim data. The invention further relates to a method for predicting the failure of a vehicle component, wherein the prediction is made from fault data of the vehicle.

Description

State of the art

[0001] Vehicles send fault or diagnostic data, for example in the form of so-called Diagnostic Trouble Codes (DTCs), to backend servers, such as those operated by vehicle manufacturers. Diagnostic Trouble Codes (DTCs) are standardized codes generated by vehicles when a problem or fault is detected in the vehicle system. These codes are generated by the vehicle's on-board diagnostic (OBD) systems and serve to help mechanics and diagnostic equipment identify and diagnose problems. Each code corresponds to a specific problem or fault in the vehicle, which facilitates diagnosis and repair. The code "P0300," for example, represents a general cylinder fault in the engine. Diagnostic Trouble Codes can indicate serious problems, but are often also transient, meaning they only occur temporarily.After a certain period of time, these codes are no longer issued because the underlying problem has resolved itself, or they are even actively deleted if the cause no longer exists. Diagnostic Trouble Codes related to the mass airflow sensor or exhaust gas recirculation are particularly common and transient. Examples include "P0101," a DTC indicating a dirty mass airflow sensor or a loose connection, and "P0135," which can be caused by an aging oxygen sensor. Diagnostic Trouble Codes can be transmitted to a backend system via the vehicle's communication device or read out in workshops using diagnostic testers.

[0002] When a vehicle is repaired in a workshop, data about the repair is often entered into a database. This is particularly true when the repair involves replacing a vehicle component. This data is also available in a backend system, but is often manually generated and therefore of varying quality. Information about the replaced component is typically supplemented by metadata such as the vehicle's current mileage and the repair date. This repair data is also referred to as claim data. Diagnostic Trouble Codes (TDCs) are only linked to claim data if the workshop also enters the TDCs along with the claim data.

[0003] Automated analysis of claim data together with diagnostic trouble codes is therefore not possible, as there is no correlation between the data. In particular, weak but statistically detectable correlations and causes cannot be systematically identified. Disclosure of the invention

[0004] The inventive method for acquiring vehicle fault data has the advantage that the fault data is acquired from diagnostic data and claim data. The diagnostic data advantageously consists of fault codes, in particular Diagnostic Trouble Codes. Claim data advantageously consists of data relating to a replaced vehicle component, as well as associated metadata, such as the vehicle's mileage or the time of the component's replacement. The claim data can, in particular, contain an identifier by which a replaced vehicle component can be uniquely identified. The identifier can, for example, be a manufacturer-specific part number. Fault data within the meaning of the present invention are therefore, in particular, combined claim and diagnostic data.

[0005] It is advantageous to combine diagnostic and claim data using temporal correlation. Combining, in this context, refers specifically to the process of matching the data. This step makes it advantageous to establish a causal relationship between the claim data and the diagnostic data. For example, it can be determined which Diagnostic Trouble Code (DTC) occurred in connection with the replacement of a component, or which DTC led a vehicle driver to a workshop where the component was then replaced due to the underlying problem. Combining data using temporal correlation increases the probability of a genuine causal relationship existing between claim data and the correlated diagnostic data.

[0006] It is advantageous to correlate diagnostic data with claim data over time by merging diagnostic data whose recording time falls within a predefined time interval around a claim time assigned to the claim data. The claim time is, in particular, a timestamp that is entered into a database along with the other claim data, for example, as metadata when a component is replaced. Claim data is usually entered manually and is therefore sometimes quite imprecise, especially with regard to the metadata. A workshop employee might, for example, enter the date of the replacement itself as the time of the component replacement. However, it is also possible that the timestamp is automatically extracted from order planning software, so it could also be the date the order was accepted.A timestamp associated with the claim data is also referred to as the claim time. The acquisition times of the diagnostic data are advantageously recorded and stored automatically by vehicle software along with the diagnostic data. Therefore, the acquisition times of the diagnostic data are usually reliable.

[0007] It is advantageous if the time interval begins 30 days before the claim date and ends 30 days after the claim date. Empirical observations have shown that a genuine causal relationship between claim data and correlated diagnostic data is particularly evident when only diagnostic data within a ±30-day period around the claim date are assigned to the claim data. This relatively long period can be explained by the fact that diagnostic trouble codes often appear some time before a serious vehicle problem necessitates a workshop visit. Additionally, for the reasons mentioned above, there is always a certain degree of inaccuracy in the claim date. Furthermore, due to organizational reasons, such as limited workshop capacity, a few days always elapse between the occurrence of a serious fault and the actual workshop visit.Diagnostic data occurring up to 30 days after the claim date (which may be a calendar date) are less frequent than data occurring before the claim date, but are nevertheless attributable to the same event if they do occur. The probability of independently occurring Diagnostic Trouble Codes occurring up to 30 days after the claim date has been shown to be quite low.

[0008] It is advantageous if the temporal correlation of diagnostic data to claim data is performed additionally or alternatively by correlating diagnostic data whose associated mileage lies within a predefined mileage interval around a claim mileage assigned to the claim data. The claim mileage is specifically the mileage of the vehicle whose component is being replaced, and this is entered as metadata along with the remaining claim data. Since a mileage can be uniquely assigned to a (at least approximate) point in a vehicle's life, it also represents temporal information, making it suitable for temporal correlation. Using the mileage in addition to the timestamp for correlating claim data and diagnostic data increases the robustness of the method.In particular, merging can also be omitted if a plausible claim date exists, but the mileage is implausible, as this suggests an incorrect data entry when creating the claim data.

[0009] A further advantage is a method for predicting the failure of a vehicle component, wherein the prediction is based on vehicle fault data, the fault data being acquired according to the inventive method. In particular, an impending component failure in a large number of vehicles of a certain type can be assumed if some vehicles of this type exhibit a striking pattern between claim data and, for example, previous diagnostic data. This can be an indicator of a fundamental problem with the vehicle type. This embodiment is therefore particularly suitable for economically optimized recall campaigns, which can thus be carried out with significantly greater precision than blanket recall campaigns.This allows the recall of vehicles of a certain type with a recognized, conspicuous correlation between certain prior Diagnostic Trouble Codes and subsequent component replacements to be postponed until the Diagnostic Trouble Codes are actually present in the affected vehicles. This avoids unnecessary or premature component replacements.

[0010] It is advantageous if the claim data is recorded in connection with a workshop visit.

[0011] Advantageously, data quality is improved before the claim data and diagnostic data are merged. For this purpose, raw diagnostic and claim data are validated using rule sets. These rule sets can be manufacturer-specific. Since raw diagnostic and claim data are often incomplete, incomplete data is discarded during validation. Only data deemed complete after validation is processed into claim and / or diagnostic data. Alternatively, the validation process can also include supplementing the raw claim and / or diagnostic data. Missing data fields can be filled with default values.

[0012] An embodiment of the present invention will now be explained in more detail with reference to the accompanying drawings. These show: Brief description of the drawing Fig. 1 A schematic representation of a temporal correlation between diagnostic data and claim data of a vehicle. embodiment of the invention

[0013] Fig. Figure 1 shows a schematic representation of a temporal correlation of diagnostic data (20, 22, 24) and claim data of a vehicle to a fault data set (40). Fig.Figure 1 shows a timeline (10) on which a first diagnostic session is plotted, resulting in a first set of diagnostic data (20). In the example shown, the first set of diagnostic data includes the Diagnostic Trouble Codes "A", "B", and "C". A second diagnostic session results in a second set of diagnostic data (22). In the example shown, the second set of diagnostic data includes the Diagnostic Trouble Codes "A", "B", and "D". A third diagnostic session results in a third set of diagnostic data (24). In the example shown, the third set of diagnostic data includes the Diagnostic Trouble Codes "A", "B", and "E". A set of diagnostic data can include one or more Diagnostic Trouble Codes.

[0014] The point in time when a vehicle visits a workshop and a vehicle component is replaced is referred to below as the claim time (30). The claim time can, for example, be the date the component is replaced. Data, referred to below as claim data, is collected during the component replacement. A time interval encompassing 30 days before and 30 days after the claim time is defined by a lower boundary (12) and an upper boundary (14). The second set of diagnostic data (22) and the third set of diagnostic data (24) fall within this time interval. Therefore, the second set of diagnostic data (22) and the third set of diagnostic data (24) are combined with the claim data to form a fault data record (40). The first set of diagnostic data (20) lies outside the interval and is therefore not considered when forming the fault data record (40).

[0015] The fault data set advantageously includes the entries of the second set of diagnostic data (22), those of the third set of diagnostic data (24) and the claim data.

[0016] In particular, the fault data set (40) therefore includes the Diagnostic Trouble Codes “A”, “B”, “D” and “E” as well as the claim data.

[0017] The fault data set can then be used for further analyses and / or predictions of failure probabilities of components of other, especially identical, vehicles.

Claims

[1] Method for recording fault data of a vehicle, characterized by that the error data is recorded based on diagnostic data and claim data. [2] Method according to claim 1, characterized by that the diagnostic data and the claim data are combined via a temporal correlation. [3] Method according to claim 2, characterized by , that the temporal correlation of the diagnostic data to the claim data is achieved by correlating diagnostic data whose recording time lies within a predefinable time interval around a claim time assigned to the claim data with the claim data. [4] Method according to claim 3, characterized by that the time interval begins 30 days before the claim date and ends 30 days after the claim date. [5] Method according to any one of claims 2 to 4, characterized by, that the temporal correlation of the diagnostic data to the claim data is achieved by correlating diagnostic data whose assigned mileage lies within a predefinable mileage interval around a claim mileage assigned to the claim data with the claim data. [6] Methods for predicting the failure of a vehicle component, characterized by that the prediction is based on fault data of the vehicle, wherein the fault data are recorded according to the method according to one of claims 1 to 5. [7] Method according to any of the preceding claims, characterized by that the diagnostic data is recorded during normal operation of the vehicle. [8] Method according to any of the preceding claims, characterized by that the claim data is recorded in connection with a workshop visit. [9] Apparatus configured to carry out the method according to any one of claims 1 to 8.

Images