DE102023111515B4 - Method for identifying, locating and/or evaluating at least one defect in a motor vehicle, data processing device, system, computer program and computer-readable storage medium - Google Patents

Abstract

Method for identifying, locating and/or evaluating at least one defect (16) in a motor vehicle (10), the method comprising:
a) collecting and/or generating first damage data (22) on the basis of information acquired by means of at least one motor vehicle sensor (12) of the motor vehicle (10) relating to at least one defect (16) of the motor vehicle (10) suspected as a result of a damage and/or accident event (14) and/or relating to the damage and/or accident event (14);
b) collecting and/or generating second damage data (32) relating to the defect (16) suspected in step a) and/or the damage and/or accident event (14) by inspecting the motor vehicle (10) and/or an environment (26) thereof on the basis of the first damage data (22) after the damage and/or accident event (14) by means of an autonomously controlled, vehicle-external inspection device (24); and
c) identifying, locating and/or evaluating the at least one defect (16) suspected in step a) in the motor vehicle (10) on the basis of the first damage data (22) and the second damage data (32).

Description

The invention relates to a method for identifying, locating, and/or evaluating at least one defect in a motor vehicle. The invention further relates to a data processing device, a system, a computer program, and a computer-readable storage medium.

Today, a vehicle accident can be detected by activating the vehicle's inherent inertial sensors (crash sensors). This estimates the severity of damage at the component level and the collision point. Activating an emergency or breakdown call system in the vehicle transmits information about an accident, but usually not the detailed collision severity. Accident information typically only provides an indication of possible mechanical damage to the vehicle. Precise deduction of damage at the component level and the exact course of events may not be possible. Therefore, an expert and images of the vehicle and/or the scene of the accident are used for further damage processing. The description of the course of events based on an accident reconstruction is performed manually by the expert. However, the manual procedure may not guarantee process reliability, may also be dependent on the expert, and can therefore have disadvantages compared to an automated or semi-automated procedure. One such disadvantage could be that the damage must be verified by the expert. Furthermore, there may be no automatic comparison between the damage diagnosis from the internal sensor estimate and the actual damage before repair. Despite artificial intelligence (AI), a 100% accurate damage profile may not be determined based on vehicle-based inertial sensors. In addition, unnecessary costs and significant time expenditure may arise for the driver, vehicle owner, and their insurance company. Furthermore, the accident sequence usually has to be reconstructed at the scene by the police and/or accident assessors. However, a manual description of the accident sequence may not provide process reliability in some cases.

The DE 10 2019 133 172 A1 discloses a method for diagnosing damage to a vehicle, wherein the vehicle has a swivel arm or carries a drone that can inspect the components of the vehicle.

The DE 10 2020 103 903 A1 discloses a method for diagnosing paint damage on a vehicle, wherein photos of the body are taken by a drone for diagnosis.

The DE 11 2016 007 265 T5 discloses a method for checking a vehicle, wherein photos of the body are taken by a drone carried along.

The US 2021 / 0 334 767 A1 discloses a method for determining a defect in a vehicle, in which a video of the vehicle is recorded using a mobile phone and the video is analyzed for a defect by a data processing device using a machine learning model.

Furthermore, the US 10,891,694 B1 , the US 2019 / 0 266 715 A1 and the US 2021 / 0 142 590 A1 State of the art.

The object of the invention is to create a standardized method for identifying, locating and/or evaluating a defect in a motor vehicle.

To achieve this object, the invention provides a method for identifying, locating and/or evaluating at least one defect in a motor vehicle according to claim 1. A system, a computer program and a computer-readable storage medium are the subject of the independent claims.

Advantageous embodiments are the subject of the subordinate claims.

1. a) Sammeln und/oder Erzeugen von ersten Schadensdaten auf Grundlage von mittels wenigstens eines Kraftfahrzeugsensors des Kraftfahrzeugs erfassten Informationen betreffend wenigstens einen durch einen Schadens- und/oder Unfallvorgang vermuteten Defekt des Kraftfahrzeugs und/oder betreffend den Schadens- und/oder Unfallvorgang;
2. b) Sammeln und/oder Erzeugen von zweiten Schadensdaten betreffend den in Schritt a) vermuteten Defekt und/oder den Schadens- und/oder Unfallvorgang durch Inspizieren des Kraftfahrzeugs und/oder einer Umgebung davon auf Grundlage der ersten Schadensdaten nach dem Schadens- und/oder Unfallvorgang mittels einer autonom gesteuerten, kraftfahrzeugexternen Inspektionsvorrichtung; und
3. c) Identifizieren, Lokalisieren und/oder Bewerten des wenigstens einen in Schritt a) vermuteten Defektes in dem Kraftfahrzeug auf Grundlage der ersten Schadensdaten und der zweiten Schadensdaten.

In one aspect, the invention provides a method for identifying, locating and/or evaluating at least one defect in a motor vehicle, the method comprising:

1. a) collecting and/or generating first damage data on the basis of information acquired by at least one motor vehicle sensor of the motor vehicle relating to at least one defect of the motor vehicle suspected as a result of a damage and/or accident event and/or relating to the damage and/or accident event;
2. b) collecting and/or generating second damage data relating to the defect suspected in step a) and/or the damage and/or accident process by inspecting the motor vehicle and/or an environment thereof on the basis of the first damage data after the damage and/or accident process by means of an autonomously controlled, vehicle-external inspection device; and
3. c) identifying, locating and/or evaluating the at least one defect in the motor vehicle suspected in step a) on the basis of the first damage data and the second damage data.

One advantage of the method can be that two information sources are used according to a standardized, computer-implemented procedure to identify, localize and/or evaluate a defect. The information obtained from the two sources can be redundant or complementary. The information from the two sources can also be weighted against each other. The two sources here are, on the one hand, the first damage data, which is determined by the motor vehicle before, during or after a damage and/or accident event, and the second damage data, which is determined by an autonomously controlled, vehicle-external inspection device after the damage and/or accident event, preferably directly after and/or at the location of the damage and/or accident event. The method is therefore an automated process that enables objective and standardized identification, localization and/or evaluation of the defect.

The mechanical and functional damage of the vehicle can be assessed in detail at the component level. A backend system, for example, is capable of precisely quantifying the cost of the vehicle damage and creating a repair order without the need for an expert. Optionally, the accident sequence can also be automatically evaluated using data from the accident scene. This also enables a process-reliable description of the accident sequence with the option of certification, compared to manual analysis and evaluation.

Another advantage of the method may be that information can be transmitted to the inspection device after the damage and/or accident event. The inspection device can then preferentially inspect those components of the motor vehicle for which the initial damage data predict, suspect, and/or demonstrate possible damage. This allows the initial damage data to be verified, so that the method as a whole identifies, localizes, and/or evaluates a defect with greater probability or quality.

• c1) Identifizieren, Lokalisieren und/oder Bewerten des Defektes auf Grundlage eines maschinellen Lernmodells, einer Simulation des Schadens- und/oder Unfallvorgangs und/oder wenigstens eines vordefinierten Parameters;
• c2) Simulieren des Schadens- und/oder Unfallvorgangs und/oder Auswerten des Defektes auf Grundlage der ersten Schadensdaten, der zweiten Schadensdaten und/oder von Bewegungsgleichungen;
• c3) Simulieren des Schadens- und/oder Unfallvorgangs und/oder Auswerten des Defektes durch Analyse einer Situation, der Umgebung und/oder sonstiger spezifischer Merkmale nach dem und/oder an dem Ort des Schadens- und/oder Unfallvorgangs;
• c4) Automatisiertes oder halbautomatisiertes Simulieren des Schadens- und/oder Unfallvorgangs und/oder Auswerten des Defektes, bevorzugt durch die Inspektionsvorrichtung; und
• c5) Erzeugen eines Schadensberichts auf Grundlage einer Kombination von ersten Schadensdaten und zweiten Schadensdaten.

It is preferred that step c) comprises one or both of the following steps:

• c1) identifying, locating and/or evaluating the defect based on a machine learning model, a simulation of the damage and/or accident process and/or at least one predefined parameter;
• c2) Simulating the damage and/or accident process and/or evaluating the defect on the basis of the first damage data, the second damage data and/or equations of motion;
• c3) Simulating the damage and/or accident process and/or evaluating the defect by analyzing a situation, the environment and/or other specific characteristics after and/or at the location of the damage and/or accident process;
• c4) Automated or semi-automated simulation of the damage and/or accident process and/or evaluation of the defect, preferably by the inspection device; and
• c5) Generating a damage report based on a combination of first damage data and second damage data.

Preferably, the identification, localization, and/or evaluation of the defect is based on a machine learning model. The machine learning model can be trained using previous accident and/or damage events. Furthermore, the machine learning model can be continuously expanded by adding additional events to the training.

Both the initial damage data and the second damage data can be incorporated into the creation of a damage report. The initial damage data and the second damage data can also be weighted against each other in the damage report. The damage report can also address any discrepancies between the initial damage data and the second damage data.

It is preferred that the machine learning model is trained with categorized and/or classified damage and/or accident events.

It is preferred that the inspection device is an autonomously controlled vehicle, in particular an aircraft.

The inspection device can be any type of autonomously controlled, external device. For example, the inspection device can be a permanently installed device in a workshop. However, it is preferred that the autonomously controlled, external inspection device is a vehicle that can inspect the motor vehicle directly at the scene of the accident and/or damage event. In this case, the inspection device can also inspect the surroundings of the scene of the accident and/or damage event. An overview of the surroundings of the scene of the accident and/or damage event can be obtained particularly well by an aircraft, such as a drone. The vehicle of the Inspection device can also inspect on the ground.

It is preferred that the inspection device is controlled autonomously based on the first damage data.

Preferably, the autonomously controlled inspection device receives the first damage data collected and/or generated by the motor vehicle. In this case, the inspection device can then inspect the motor vehicle autonomously based on this first damage data. Thus, the method can be further automated and/or standardized.

It is preferred that a trajectory and/or orientation of the vehicle is calculated based on the first damage data.

If the autonomously controlled inspection device is not permanently installed but is designed as a mobile vehicle, the inspection device can also autonomously inspect suspect components, locations, and/or the surrounding area based on the initial damage data. This allows the process to be further automated and/or standardized.

It is preferred that the inspection device comprises a camera device, LIDAR device and/or a scanning device for, preferably three-dimensionally, scanning and/or recording the motor vehicle and/or the surroundings thereof.

• e) Übertragen der ersten Schadensdaten und/oder der zweiten Schadensdaten an eine Datenverarbeitungseinrichtung zum Identifizieren, Lokalisieren und/oder Bewerten des wenigstens einen Defektes auf Grundlage der ersten Schadensdaten und der zweiten Schadensdaten;
• f) Empfangen, durch die Inspektionsvorrichtung, der ersten Schadensdaten von dem Kraftfahrzeug und/oder von einer externen und/oder entfernten Datenverarbeitungseinrichtung; und
• g) Empfangen, durch das Kraftfahrzeug, der zweiten Schadensdaten von der Inspektionsvorrichtung und/oder von einer externen und/oder entfernten Datenverarbeitungseinrichtung.

It is preferred that the method further comprises one, two or all of the following steps:

• e) transmitting the first damage data and/or the second damage data to a data processing device for identifying, locating and/or evaluating the at least one defect on the basis of the first damage data and the second damage data;
• f) receiving, by the inspection device, the first damage data from the motor vehicle and/or from an external and/or remote data processing device; and
• g) Receiving, by the motor vehicle, the second damage data from the inspection device and/or from an external and/or remote data processing device.

The first damage data and/or the second damage data are preferably transmitted to a data processing device, e.g., an external and/or remote data processing device. The data processing device can store first and/or second damage data from a plurality of accident and/or damage events. The data processing device can, for example, use the machine learning model, which can be trained with the plurality of accident and/or damage events and continuously expanded. This allows the data from a plurality of accident and/or damage events to be stored centrally, so that the accident and/or damage events can all be evaluated according to the same criteria.

For use cases or application situations that may arise during the method and which are not explicitly described here, it may be provided that, in accordance with the method, an error message and/or a request to enter user feedback is issued and/or a default setting and/or a predetermined initial state is set.

• - von mittels wenigstens eines Kraftfahrzeugsensors des Kraftfahrzeugs erfassten Informationen betreffend wenigstens einen durch einen Schadens- und/oder Unfallvorgang vermuteten Defekt des Kraftfahrzeugs und/oder betreffend den Schadens- und/oder Unfallvorgang, gesammelten und/oder erzeugten ersten Schadensdaten und
• - von auf Grundlage der ersten Schadensdaten mittels einer autonom gesteuerten, kraftfahrzeugexternen Inspektionsvorrichtung zum Inspizieren des Kraftfahrzeugs und/oder einer Umgebung davon nach dem Schadens- und/oder Unfallvorgang gesammelten und/oder erzeugten zweiten Schadensdaten betreffend den vermuteten Defekt und/oder den Schadens- und/oder Unfallvorgang.

According to a further aspect, the invention provides a data processing device which is configured to identify, localize and/or evaluate at least one defect in a motor vehicle on the basis of:

• - first damage data collected and/or generated from information recorded by at least one motor vehicle sensor of the motor vehicle relating to at least one defect of the motor vehicle suspected as a result of a damage and/or accident process and/or relating to the damage and/or accident process, and
• - second damage data relating to the suspected defect and/or the damage and/or accident process, collected and/or generated on the basis of the first damage data by means of an autonomously controlled, vehicle-external inspection device for inspecting the motor vehicle and/or an environment thereof after the damage and/or accident process.

The invention also includes the data processing device or process device. The data processing device or processor device can be configured to at least partially carry out an embodiment of the method according to the invention. For this purpose, the processor device can have at least one microprocessor and/or at least one microcontroller and/or at least one FPGA (Field Programmable Gate Array) and/or at least one DSP (Digital Signal Processor). Furthermore, the processor device can have program code that is configured to at least partially carry out the embodiment of the method according to the invention when executed by the processor device. The pro The program code can be stored in a data memory of the processor device. The processor device can be based, for example, on at least one circuit board and/or on at least one SoC (System on Chip).

The invention also includes further developments of the data processing device according to the invention that have features already described in connection with the further developments of the method according to the invention. For this reason, the corresponding further developments of the data processing device according to the invention are not described again here.

According to a further aspect, the invention provides a system for identifying, locating and/or evaluating at least one defect in a motor vehicle, comprising means for carrying out the method according to one of the preceding embodiments.

The invention also includes the system configured to carry out an embodiment of the method according to the invention. The system may comprise the data processing device. Furthermore, the system may comprise the inspection device.

The invention also includes further developments of the system according to the invention that have features already described in connection with the further developments of the method according to the invention. For this reason, the corresponding further developments of the system according to the invention are not described again here.

According to a further aspect, the invention provides a computer program comprising instructions which cause the system to carry out the method according to any one of the preceding embodiments.

According to a further aspect, the invention provides a computer-readable storage medium on which the computer program is stored.

As a further solution, the invention also encompasses a computer-readable storage medium comprising program code which, when executed by a computer or computer network, causes the computer or computer network to execute an embodiment of the method according to the invention. The storage medium can be provided at least partially as a non-volatile data memory (e.g., as a flash memory and/or as an SSD - solid state drive) and/or at least partially as a volatile data memory (e.g., as a RAM - random access memory). The storage medium can be arranged in the computer or computer network. However, the storage medium can also be operated, for example, as a so-called app store server and/or cloud server on the Internet. The computer or computer network can provide a processor circuit with, for example, at least one microprocessor. The program code can be provided as binary code and/or as assembly code and/or as source code of a programming language (e.g., C) and/or as a program script (e.g., Python).

The invention also encompasses combinations of the features of the described embodiments. The invention therefore also encompasses implementations that each have a combination of the features of several of the described embodiments, unless the embodiments are described as mutually exclusive.

• 1 eine Ausführungsform eines Kraftfahrzeugs;
• 2 eine Darstellung einer Szene nach einem Schaden- und/oder Unfallvorgang;
• 3 eine Ausführungsform eines Verfahrens zum Identifizieren, Lokalisieren und/oder Bewerten wenigstens eines Defektes in dem Kraftfahrzeug; und
• 4 eine Ausführungsform eines Systems zum Identifizieren, Lokalisieren und/oder Bewerten des wenigstens einen Defektes.

Exemplary embodiments of the invention are described below. Shown are:

• 1 an embodiment of a motor vehicle;
• 2 a depiction of a scene following a damage and/or accident event;
• 3 an embodiment of a method for identifying, locating and/or evaluating at least one defect in the motor vehicle; and
• 4 an embodiment of a system for identifying, locating and/or evaluating the at least one defect.

The exemplary embodiments explained below are preferred embodiments of the invention. In the exemplary embodiments, the described components of the embodiments each represent individual features of the invention that can be considered independently of one another, each of which also develops the invention independently of one another. Therefore, the disclosure is intended to encompass combinations of the features of the embodiments other than those shown. Furthermore, the described embodiments can also be supplemented by further features of the invention already described.

In the figures, the same reference symbols denote elements with the same function.

1 shows an embodiment of a motor vehicle 10.

The motor vehicle 10 comprises a plurality of motor vehicle sensors 12. The motor vehicle sensors 12 are designed to detect information relating to a defect 16 of the motor vehicle caused by a damage and/or accident event 14 and/or relating to the damage and/or accident event 14.

For example, the motor vehicle sensors 12 are part of an airbag system. For this purpose, the motor vehicle sensors 12 can be configured, for example, as pressure sensors. However, the motor vehicle sensors 12 can also be of any other type. For example, the motor vehicle sensors 12 can each be configured as a speed, acceleration, or temperature sensor.

In 1 The motor vehicle sensors 12 are connected to a central unit 18. The central unit 18 is designed to collect and/or generate first damage data 22.

In the case of the damage and/or accident event 14, the information acquired by the motor vehicle sensors 12 may relate to the defect 16 of the motor vehicle 10 caused by the damage and/or accident event 14 and/or the damage and/or accident event 14 itself.

The central unit 18 can, for example, collect the acquired information as initial damage data 22. Based on the acquired information, the initial damage data 22 can, for example, indicate that a component of the motor vehicle 10, for example, a side door 23, may be damaged.

Alternatively or additionally, the central unit 18 can analyze the acquired information and generate the first damage data 22 based thereon. The central unit 18 can, for example, analyze the acquired information to directly identify, localize, and/or evaluate the defect 16. For the analysis, the central unit 18 can, for example, use a machine learning model, a simulation of the damage and/or accident process 14, and/or at least one predefined parameter. The machine learning model can be trained for this purpose with categorized and/or classified damage and/or accident processes 14.

The central unit 18 can transmit the collected and/or generated first damage data 22 to a remote data processing device 20. Alternatively or in addition to the central unit 18, the remote data processing device 20 can analyze the first damage data 22 to directly identify, localize, and/or evaluate the defect 16.

2 shows a representation of a scene after the damage and/or accident event 14.

In 2 an autonomously controlled, vehicle-external inspection device 24 inspects the motor vehicle 10 and/or an environment 26 thereof.

The autonomously controlled, vehicle-external inspection device 24 is in 2 as a vehicle 27, in particular an aircraft 28, more particularly a drone 30. However, the concept also encompasses embodiments in which the autonomously controlled, vehicle-external inspection device 24 is designed, for example, as a land vehicle, such as an autonomously controlled robot. The inspection device 24, here the drone 30, is designed to collect and/or generate second damage data 32.

For this purpose, the inspection device 24 can, for example, inspect the surroundings 26, such as a road layout 34 or road signage 36, for example, relating to a right-of-way regulation, or brake marks 38. The inspection device 24 can, for example, also inspect a position 40 and/or an orientation 42 of the motor vehicle 10 and/or other motor vehicles 10 in the surroundings 26. Furthermore, the inspection device 24 can, for example, also inspect specific features of the motor vehicle 10, such as a tire angle 44 or braking points 46.

For inspection, the inspection device 24 has, for example, a camera device 48, a LIDAR device 50 and/or a scanning device 52 for scanning and/or recording the motor vehicle 10 and/or the surroundings 26 thereof.

The inspection device 24 can collect the information acquired by the camera device 48, LIDAR device 50, and/or scanning device 52 as the second damage data 32. Based on the acquired information, the second damage data 32 can confirm that the component of the motor vehicle 10, for example, the side door 23, is damaged.

Alternatively or additionally, the inspection device 24 can analyze the information acquired by the camera device 48, LIDAR device 50 and/or scanning device 52 and generate the second damage data 32 based thereon. The inspection device 24 can, for example, analyze the acquired information to directly identify, localize and/or evaluate the defect 16. For the analysis, the For example, the inspection device 24 may use a machine learning model, a simulation of the damage and/or accident process 14, and/or at least one predefined parameter. For this purpose, the machine learning model may be trained with categorized and/or classified damage and/or accident processes 14.

The inspection device 24 can transmit the collected and/or generated second damage data 32 to the remote data processing device 20. Alternatively or in addition to the inspection device 24, the remote data processing device 20 can analyze the second damage data 22 to directly identify, localize, and/or evaluate the defect 16.

3 shows an embodiment of a method for identifying, locating and/or evaluating the at least one defect 16 in the motor vehicle 10.

• - Sammeln und/oder Erzeugen der ersten Schadensdaten 22 auf Grundlage von mittels des wenigstens einen Kraftfahrzeugsensors 12 des Kraftfahrzeugs 10 erfassten Informationen betreffend den durch den Schadens- und/oder Unfallvorgang 14 verursachten Defekt 16 des Kraftfahrzeugs 10 und/oder betreffend den Schadens- und/oder Unfallvorgang 14.

In a step S11, the method comprises:

• - Collecting and/or generating the first damage data 22 on the basis of information acquired by the at least one motor vehicle sensor 12 of the motor vehicle 10 relating to the defect 16 of the motor vehicle 10 caused by the damage and/or accident event 14 and/or relating to the damage and/or accident event 14.

• - Sammeln und/oder Erzeugen der zweiten Schadensdaten 32 durch Inspizieren des Kraftfahrzeugs 10 und/oder der Umgebung 26 davon mittels der autonom gesteuerten, kraftfahrzeugexternen Inspektionsvorrichtung 24 nach dem Schadens- und/oder Unfallvorgang 14.

In a step S12, the method comprises:

• - Collecting and/or generating the second damage data 32 by inspecting the motor vehicle 10 and/or the surroundings 26 thereof by means of the autonomously controlled, vehicle-external inspection device 24 after the damage and/or accident event 14.

• - Identifizieren, Lokalisieren und/oder Bewerten des wenigstens einen Defektes 16 auf Grundlage der ersten Schadensdaten 22 und der zweiten Schadensdaten 32.

In a step S13, the method comprises:

• - Identifying, locating and/or evaluating the at least one defect 16 based on the first damage data 22 and the second damage data 32.

For this purpose, the collected and/or generated first damage data 22 and second damage data 32 are transmitted, for example, to the data processing device 20. The data processing device 20 has been described above as a remote data processing device 20. However, the concept also encompasses embodiments in which the data processing device is part of the motor vehicle 10 or the inspection device 24.

The data processing device 20 can analyze the received first damage data 22 and second damage data 32 to identify, localize, and/or evaluate the defect 16. For the analysis, the data processing device 20 can use, for example, a machine learning model, a simulation of the damage and/or accident process 14, and/or at least one predefined parameter. For this purpose, the machine learning model can be trained with categorized and/or classified damage and/or accident processes 14.

To identify, localize, and/or evaluate the defect 16, the data processing device 20 combines or merges the first damage data 22 and the second damage data 32. Preferably, the data processing device 20 then generates a damage report 56 based on the first damage data 22 and the second damage data 32.

4 shows an embodiment of a system 58 for identifying, locating and/or evaluating the at least one defect 16 in the motor vehicle 10.

The system 58 comprises the motor vehicle 10, the inspection device 24 and the data processing device 20.

The central unit 18 of the motor vehicle 10 collects and/or generates the first damage data 22. The first damage data 22 are transmitted from the motor vehicle 10 to the inspection device 24 and the data processing device 20.

The inspection device 24 is controlled autonomously based on the first damage data 22. For example, the inspection device 24 can inspect motor vehicle parts to be inspected and/or locations in the surrounding area 26 based on the first damage data 22. For this purpose, the inspection device 24 calculates a trajectory and/or orientation based on the first damage data 22 in a step S21.

The inspection device 24 inspects the motor vehicle 10 and/or the surroundings 26 thereof by means of the camera device 48, LIDAR device 50 and/or scanning device 52 and collects and/or generates the second damage data 32. The second damage data 32 are transmitted to the data processing device 20.

The data processing device 20 combines or merges the first damage data 22 and the second damage data 32 in a step S22. Subsequently, the data processing device device 20 in a step S23 the damage report 56 based on the combination of first damage data 22 and second damage data 32.

The damage report 56 can describe the circumstances of the accident and the resulting damage. The damage report 56 can also assess the damage and include the costs of any necessary repairs. The damage report 56 is generated without subjective influences, but rather through an objective process. The damage report 56 can therefore also be used, for example, by a vehicle owner's insurance company.

The invention also includes the data processing device 20, which is configured to identify, localize, and/or evaluate the defect 16 in the motor vehicle 10 on the basis of the information acquired by the at least one motor vehicle sensor 12 of the motor vehicle 10 relating to the at least one defect 16 of the motor vehicle 10 caused by the damage and/or accident event 14 and/or relating to the damage and/or accident event 14, the first damage data 22 collected, and the second damage data 32 collected by the autonomously controlled, vehicle-external inspection device 24 for inspecting the motor vehicle 10 and/or the surroundings 26 thereof after the damage and/or accident event 14.

The invention also includes the system 58 for identifying, locating, and/or evaluating the at least one defect 16 in the motor vehicle 10. The invention further includes the computer program (not shown) comprising instructions that cause the system 58 to execute the described method. The invention further includes a computer-readable storage medium (not shown) on which the computer program is stored.

• Die einwandfreie Schadensdetektion aufgrund einer Panne oder eines Unfalls wird bevorzugt durch die Berücksichtigung einer Kombination aus fahrzeuginhärenter Inertialsensorik, externen mittels eines autonomen, fahrzeugunabhängigen Systems (Drohne - flugfähige oder bodengebundener Roboter) erfassten Bildern und künstlicher Intelligenz (z.B. künstliches, neuronales Netz) ermittelt.

A solution principle of preferred embodiments of the invention can thus be summarized as follows:

• The correct detection of damage due to a breakdown or an accident is preferably determined by taking into account a combination of vehicle-inherent inertial sensors, external images captured by an autonomous, vehicle-independent system (drone - flight-capable or ground-based robot) and artificial intelligence (e.g. artificial neural network).

The points to be assessed are controlled, documented, evaluated, and merged with the vehicle's inherent data by the autonomous system itself. Furthermore, the accident sequence can be reconstructed automatically or semi-automatically using a drone. The drone is preferably used for process-reliable analysis of the accident scene. Downstream data processing ensures reproducible evaluation and creates a description of the accident sequence or an accident report, which offers the possibility of certification.

The figures show one of the possible implementations of the method. A computer component processes the in-crash sensor signals from a collision in the vehicle. Damaged vehicle components are then identified using a model (e.g., a machine learning model trained with labeled FEM crash simulations). Based on the prediction of damage to components and vehicle regions, an independent autonomous system is started, which documents the vehicle and/or the accident scene using an image and time-of-flight (TOF) camera, as well as LIDAR.

Based on the (preliminary) estimate of the damaged vehicle components, the drone/robot's trajectory and flight path, as well as the positions for data acquisition by the sensors, are calculated. The recorded data from the autonomous system is then merged with the (preliminary) estimate (system overview, see 4 ). A detailed and accurate damage report at component level with costs is created in a backend.

If a drone is used at the scene of the accident, it can also record and evaluate data on the course of the accident (see 2 ). For example, road layout, tire positions, and skid marks can be recorded and used in the damage estimator.

The automated or semi-automated accident analysis preferably consists of at least the components of environmental analysis, situation analysis, and an analysis of specific characteristics. These analyses, along with the internal, vehicle-specific data described above, are incorporated into an automated accident reconstruction (e.g., in a cloud).

In the simplest case, an accident reconstruction is performed in the cloud based on known equations of motion, depending on acceleration, speed, vehicle positions relative to each other, and steering and braking points. The analysis is performed reliably in an automated or semi-automated manner (certification is possible) and finally approved by an expert.

1. 1) Fahrzeugschadensdetektion auf Bauteilebene mit In-Crash-Sensorsignalen und KI (künstliche Intelligenz; maschinelles Lernmodell). Die KI ist beispielsweise mittels FEM-Crashsimulationen und Fachwissen trainiert;
2. 2) Autonomes, unbemanntes System (bodengebunden oder fliegend), ausgestattet mit einer Kommunikationseinheit (mit Kraftfahrzeug und Cloud), einer bildgebenden Kamera und/oder einer 3D-Sensorik (TOF-Kamera oder LIDAR) mit Algorithmus zur Trajektorienberechnung bzw. mit Navigation basierend aus Schadensinformation des Fahrzeuges;
3. 3) Backendsystem zur Fusionierung der Information und Auswertung zur Schadensberichterstellung auf Bauteilebene;
4. 4) Automatisierung der Unfallhergangsanalyse durch Flugdrohne;
5. 5) Automatisierte, methodische Unfallhergangsanalyse (Umweltverständnis, Situationsverständnis, spezifische Merkmale) durch die Drohne;
6. 6) Automatisierung der Unfallhergangsauswertung durch eine definierte Toolchain in der Cloud, mit bekannten Bewegungsgleichungen basierend auf den ermittelten Drohnenparametern einschließlich den fahrzeuginhärenten Daten; und
7. 7) Möglichkeit der Zertifizierung durch die automatisierte bzw. halbautomatisierte Auswertung.

Embodiments of the inventions may have the following features:

1. 1) Component-level vehicle damage detection using in-crash sensor signals and AI (artificial intelligence; machine learning model). The AI is trained, for example, using FEM crash simulations and specialist knowledge;
2. 2) Autonomous, unmanned system (ground-based or flying), equipped with a communication unit (with vehicle and cloud), an imaging camera and/or a 3D sensor (TOF camera or LIDAR) with an algorithm for trajectory calculation or with navigation based on vehicle damage information;
3. 3) Backend system for merging information and evaluation for damage reporting at component level;
4. 4) Automation of accident analysis by drone;
5. 5) Automated, methodical accident analysis (understanding of the environment, understanding of the situation, specific characteristics) by the drone;
6. 6) Automation of the accident analysis through a defined toolchain in the cloud, with known equations of motion based on the determined drone parameters including the vehicle-inherent data; and
7. 7) Possibility of certification through automated or semi-automated evaluation.

Overall, the examples demonstrate how a method for verifying vehicle damage can be provided by merging information from a vehicle diagnostic system and an external, autonomous system assessment system.

List of reference symbols:

10

motor vehicle

12

Automotive sensor

14

Damage and/or accident process

16

defect

18

central unit

20

Data processing facility

22

initial damage data

23

side door

24

Inspection device

26

Vicinity

27

vehicle

28

aircraft

30

flying drone

32

second damage data

34

Road layout

36

Road signs

38

skid marks

40

position

42

Orientation

44

Tire angle

46

braking point

48

Camera setup

50

LIDAR device

52

Scanning device

56

Damage report

58

system

Claims (12)

A method for identifying, locating, and/or evaluating at least one defect (16) in a motor vehicle (10), the method comprising: a) collecting and/or generating first damage data (22) based on information acquired by at least one motor vehicle sensor (12) of the motor vehicle (10) relating to at least one defect (16) of the motor vehicle (10) suspected as a result of a damage and/or accident event (14) and/or relating to the damage and/or accident event (14); b) collecting and/or generating second damage data (32) relating to the defect (16) suspected in step a) and/or the damage and/or accident event (14) by inspecting the motor vehicle (10) and/or an environment (26) thereof based on the first damage data (22) after the damage and/or accident event (14) using an autonomously controlled, vehicle-external inspection device (24); and c) identifying, locating, and/or evaluating the at least one defect (16) suspected in step a) in the motor vehicle (10) based on the first damage data (22) and the second damage data (32). Procedure according to Claim 1 , characterized in that step c) comprises one, several or all of the following steps: c1) identifying, localizing and/or evaluating the defect (16) based on a machine learning model, a simulation of the damage and/or accident process (14) and/or at least one predefined parameter; c2) simulating the damage and/or accident process (14) and/or evaluating the defect (16) based on the first damage data (22), the second damage data (32) and/or equations of motion; c3) simulating the damage and/or accident process (14) and/or evaluating the defect (16) by analyzing a situation, the environment (26) and/or other specific features after and/or at the location of the damage and/or accident process (14); c4) automated or semi-automated simulation of the damage and/or accident process (14) and/or evaluating the defect (16), preferably by the inspection device (24); and c5) generating a damage report (56) based on a combination of first damage data (22) and second damage data (32). Procedure according to Claim 2 , characterized in that the machine learning model is trained with categorized and/or classified damage and/or accident processes (14). Method according to one of the preceding claims, characterized in that the inspection device (24) is an autonomously controlled vehicle (27), in particular an aircraft (28). Method according to one of the preceding claims, characterized in that the inspection device (24) is controlled autonomously on the basis of the first damage data (22). Procedure according to the Claims 4 and 5 , characterized in that a trajectory and/or orientation of the vehicle (27) is calculated on the basis of the first damage data (22). Method according to one of the preceding claims, characterized in that the inspection device (24) comprises a camera device (48), LIDAR device (50) and/or a scanning device (52) for, preferably three-dimensionally, scanning and/or recording the motor vehicle (10) and/or the surroundings (26) thereof. Method according to one of the preceding claims, further comprising one, two, or all of the following steps: e) transmitting the first damage data (22) and/or the second damage data (32) to a data processing device (20) for identifying, locating, and/or evaluating the at least one defect (16) based on the first damage data (22) and the second damage data (32); f) receiving, by the inspection device (24), the first damage data (22) from the motor vehicle (10) and/or from an external and/or remote data processing device (20); and g) receiving, by the motor vehicle (10), the second damage data (32) from the inspection device (24) and/or from an external and/or remote data processing device (20). A data processing device (20) configured to identify, localize, and/or evaluate at least one defect (16) in a motor vehicle (10) based on: - first damage data (22) collected and/or generated, acquired by at least one motor vehicle sensor (12) of the motor vehicle (10), relating to at least one defect (16) of the motor vehicle (10) suspected as a result of a damage and/or accident event (14) and/or relating to the damage and/or accident event (14), and - second damage data (32) collected and/or generated based on the first damage data (22) by means of an autonomously controlled, vehicle-external inspection device (24) for inspecting the motor vehicle (10) and/or an environment (26) thereof after the damage and/or accident event (14). System (58) for identifying, locating and/or evaluating at least one defect (16) in a motor vehicle (10), comprising means for carrying out the method according to one of the Claims 1 until 8 . Computer program comprising instructions that cause the system to Claim 10 the procedure according to one of the Claims 1 until 8 executes. Computer-readable storage medium on which the computer program is Claim 11 is stored.

Images