Method for operating a vehicle, computer program product, control unit and vehicle

Abstract

The invention relates to a method for operating a vehicle (F1), in particular an automated or autonomously driving vehicle, comprising: - Detection of at least one other vehicle (F2) in the vicinity of the vehicle (F1), in particular in front of, behind and / or beside the vehicle (F1), - Evaluating the driving behavior of at least one other vehicle (F2), - Determining a level of automation (L2, L3, L4, L5) in the operation of at least one other vehicle (F2), - Providing feedback on the level of automation (L2, L3, L4, L5) in the operation of at least one other vehicle (F2) to a user (B) of the vehicle (F1).

Description

[0001] The invention relates to a method for operating a vehicle. Furthermore, the invention relates to a corresponding computer program product, a corresponding control unit, and a corresponding vehicle for carrying out this method.

[0002] Nowadays, powerful machine learning algorithms can be used to recognize complex patterns in images and videos, while computer vision algorithms can aid in scene understanding. On the other hand, there is a growing number of drivers using a certain degree of automated driving.

[0003] When vehicles use machine learning and / or machine vision during operation, the safety aspects for these vehicles are determined individually. However, the operation of such vehicles does affect the safety of other road users. Road users, however, have no control over whether other vehicles use machine learning and / or machine vision. Nor do road users have any way of verifying the safety aspects resulting from other vehicles using machine learning and / or machine vision, or of determining whether their own safety is compromised as a result.

[0004] It is therefore an object of the present invention to at least partially overcome at least one of the disadvantages described above. In particular, it is an object of the present invention to provide an improved method for operating a vehicle that increases road safety, raises awareness of the use of machine learning and / or machine vision by other vehicles, and provides a way to check whether the use of machine learning and / or machine vision by other vehicles endangers one's own safety. Furthermore, it is an object of the invention to provide a corresponding computer program, a corresponding control unit, and a corresponding vehicle for carrying out such a method.

[0005] The aforementioned problem is solved by the patent claims. Accordingly, the problem is solved by a method for operating a vehicle with the features of the independent method claim. Furthermore, the problem is solved by a corresponding computer program product, a corresponding control unit, and a corresponding vehicle for carrying out a corresponding method with the features of the dependent claims. Further features and advantages of the invention will become apparent from the dependent claims, the description, and the drawings. Features and details described in connection with the different embodiments and / or aspects of the invention naturally also apply in connection with the other embodiments and / or aspects, and vice versa, so that the disclosure relating to the individual embodiments and / or aspects always includes, or can include, reciprocal references.

[0006] The present invention provides for: a method for operating a vehicle, which itself may be designed as an automated or autonomous driving vehicle, and which may also be referred to as an EGO vehicle, exhibiting the following procedural steps: - Detection of at least one other vehicle (which can also be described as a foreign vehicle) in the vicinity of the vehicle, in particular in front of, behind and / or next to the vehicle, - Evaluating the driving behavior of at least one other vehicle (especially depending on the detection), - Determining the degree of automation (or degree for short) in the operation of at least one other vehicle (especially depending on the evaluation), - Providing feedback on the degree of automation in the operation of at least one other vehicle to a user (the user can be, for example, a driver) of the vehicle (especially depending on the determination).

[0007] The invention recognizes that nowadays more and more vehicles use powerful machine learning algorithms to process images and videos, while computer vision algorithms are used for scene understanding.

[0008] If other vehicles use machine learning and / or machine vision during operation, this has an impact on the safety of other road users, such as the safety of the EGO vehicle.

[0009] Using this method, the EGO vehicle can detect whether other vehicles are using machine learning and / or machine vision.

[0010] The EGO vehicle thus gains the ability to check whether its own safety is endangered by the use of machine learning and / or machine vision by other vehicles.

[0011] The procedure helps to raise user awareness that other vehicles use machine learning and / or machine vision, so that the user can be warned and take better care.

[0012] This procedure can therefore contribute to increasing road safety.

[0013] While driving, camera data, video data and / or data from the vehicle's LIDAR and / or radar sensors can be used to check, possibly with the help of machine learning methods (e.g., using an artificial neural network), whether nearby cars (in front and / or behind and / or to the side) are using a certain degree of driving automation, e.g.: - Grade 2.5, including so-called Adaptive Cruise Control, - advanced levels of automation, e.g., level 3, level 4 or even - a so-called autopilot below grade 5).

[0014] This allows the driver to be informed that other cars near the Ego vehicle have a certain degree of driving automation and that the other drivers do not have full control over the vehicle commands.

[0015] This can attract the attention of the EGO vehicle's user. Consequently, the user can be prompted to pay more attention.

[0016] Furthermore, the user may be prompted to activate safety measures in their own vehicle, e.g., to refrain from automatically orienting themselves towards such vehicles, for example, in the case of distance and speed control.

[0017] Furthermore, it may be provided that at least one of the following vehicle sensors is used to detect at least one other vehicle and / or to evaluate the driving behavior of at least one other vehicle: - optical sensor, - Camera, - Radar, - Lidar and / or - acoustic sensor.

[0018] In this way, the vehicle's surroundings, particularly in front of, behind, and / or beside the vehicle, can be advantageously captured. Sensor data from different sensors can be advantageously checked and / or compared against each other to detect at least one other vehicle, to evaluate its driving behavior, and / or to determine the level of automation used.

[0019] In principle, an area in front of the vehicle, an area behind the vehicle, and / or an area beside the vehicle can be monitored. In this way, the vehicle's surroundings can be fully covered.

[0020] Advantageously, at least one machine learning method, at least one machine image processing method, and / or at least one artificial neural network can be used to evaluate the driving behavior of at least one other vehicle and / or to determine the degree of automation in the operation of at least one other vehicle. In this way, the method can utilize modern machine learning techniques to process sensor data, to detect at least one other vehicle, to evaluate its driving behavior, and / or to determine the level of automation employed.

[0021] The following levels of automation can be determined using this method: Grade 2: Partially automated driving, in particular level 2.5: which can provide adaptive distance and speed control, Grade 3: Highly automated driving, Grade 4: Fully automated driving, Grade 5: Autonomous driving.

[0022] Level 2: Semi-automated: The user can be assisted with both lateral and longitudinal vehicle control for a certain period of time or in specific situations, such as overtaking on the motorway. The user must continuously monitor the traffic situation and be able to take over control of the vehicle immediately at any time.

[0023] Level 3: Conditionally automated: As with level 2 automation, the vehicle takes over lateral and longitudinal control for a certain period or in specific situations. Continuous monitoring of the traffic situation by the user is no longer necessary. However, the user must be able to regain control of the vehicle with sufficient time if requested by the vehicle.

[0024] Level 4: Highly Automated: At this level of automation, the vehicle can automatically handle all situations in a defined use case, such as driving on highways. The user must be prompted to take over the driving task before the use case is completed. If the user does not comply with the prompt, the vehicle must assume a low-risk state (e.g., come to a stop on a highway shoulder).

[0025] Level 5: Fully automated: At this level of automation, no user is required from start to finish. The vehicle takes over the driving task completely, i.e., on all road types, at all speeds, and in all environmental conditions.

[0026] Firstly, it is conceivable that the feedback to the user could include a warning that at least one other vehicle is driving automatically and / or autonomously. This alone may be sufficient to attract the user's attention.

[0027] Furthermore, it is conceivable that the feedback to the user could include a specific level of automation. For example, if the user learns that a vehicle ahead is using adaptive cruise control, they can be alerted that they may need to pay attention to the vehicle's longitudinal control. If, for instance, a vehicle ahead uses level 3 automation or higher, the user can be alerted that they may need to pay attention to the vehicle's longitudinal and / or lateral control. Thus, such specific feedback can provide the user with improved information.

[0028] In principle, it is conceivable that feedback to the user could be provided acoustically, e.g., via voice input, and / or visually, e.g., via a display. In this way, the vehicle's existing output units can be used to inform the user.

[0029] According to a further aspect, the invention provides a computer program product comprising instructions which, when executed by a computer, such as the processing unit of a control unit, cause the computer to carry out the method, which can proceed as described above. The same advantages described above in connection with the method according to the invention can be achieved with the computer program product. These advantages are fully referenced herein.

[0030] A corresponding control unit provides a further aspect of the invention. A computer program in the form of code can be stored in a memory unit of the control unit. When the code is executed by a processing unit of the control unit, this program performs a procedure that can proceed as described above. The same advantages described above in connection with the method according to the invention can be achieved with the help of the control unit. These advantages are fully referenced herein.

[0031] Furthermore, it may be provided that an artificial neural network is stored in the storage unit, which is designed to evaluate the driving behavior of at least one other vehicle and / or to determine the degree of automation in the operation of at least one other vehicle.

[0032] In principle, it is conceivable that the artificial neural network could be designed as a neural network with forward feedback, a convolutional neural network, or a recurrent neural network.

[0033] A corresponding vehicle, comprising a corresponding control unit, also represents an aspect of the invention. The same advantages described above in connection with the method according to the invention can be achieved with the vehicle. These advantages are fully referenced herein.

[0034] Further advantages of the invention will become apparent from the following description, in which exemplary embodiments of the invention are described in detail with reference to the drawing.

[0035] It shows schematically: Fig. 1. An exemplary sequence of a proposed procedure.

[0036] As it is Fig.Figure 1 illustrates a procedure that was developed for operating a vehicle F1.

[0037] Vehicle F1 can be configured as an automated or autonomous vehicle. Vehicle F1 can also be referred to as an EGO vehicle.

[0038] As it is Fig. As illustrated in Figure 1, the procedure comprises the following steps: 110 Detection of at least one other vehicle F2 (which can also be referred to as a foreign vehicle) in the vicinity of vehicle F1, e.g. in front of, behind and / or next to vehicle F1, 120 Evaluating the driving behavior of at least one other vehicle F2 (especially depending on the detection), 130 Determining a degree of automation L2, L3, L4, L5 (or, in short, degree) in the operation of at least one other vehicle F2 (especially depending on the evaluation), 140 Providing feedback on the level of automation L2, L3, L4, L5 in the operation of at least one other vehicle F2 to a user B (the user can be, for example, a driver) of vehicle F1 (especially depending on the determination).

[0039] The invention recognizes that modern vehicles F2 increasingly employ methods for machine learning and / or machine vision (so-called computer vision) to enable automated and / or autonomous driving functions in the operation of the vehicles F2.

[0040] If other vehicles F2 in the vicinity of the own vehicle F1 or the EGO vehicle use machine learning and / or machine vision techniques during operation, this has an impact on the safety of the own vehicle F1.

[0041] The user B of vehicle F1 can use the procedure to detect whether other vehicles F2 are using machine learning and / or machine vision procedures during operation or not.

[0042] User B of vehicle F1 is thus given the opportunity to check whether their own safety is endangered by the use of machine learning and / or machine vision by other vehicles F2.

[0043] The procedure helps to raise awareness among user B that other vehicles F2 may use machine learning and / or machine vision, so that user B can be warned and take better care.

[0044] The procedure can therefore contribute to increasing road safety not only in the operation of one's own vehicle F1, but also of other vehicles F2.

[0045] To detect at least one other vehicle F2 and / or to evaluate the driving behavior of at least one other vehicle F2, at least one of the following vehicle sensors can be used: - optical sensor, - Camera, - Radar, - Lidar and / or - acoustic sensor.

[0046] Furthermore, it is conceivable that sensor data from different sensors can be checked and / or weighed against each other in order to detect at least one other vehicle F2, to evaluate its driving behavior and / or to determine the level of automation used L2, L3, L4, L5.

[0047] Advantageously, at least one machine learning method and / or at least one machine vision method and / or at least one artificial neural network (ANN) can be used when carrying out the procedure. In this way, the procedure can utilize modern machine learning and / or vision techniques to process sensor data in order to detect at least one other vehicle F2, to evaluate its driving behavior, and / or to determine the level of automation used (L2, L3, L4, L5).

[0048] The following automation levels L2, L3, L4, L5 can be determined using the procedure and reported to user B of vehicle F1: Grade 2: Partially automated driving, in particular level 2.5: which can provide adaptive distance and speed control (ACC), Grade 3: Highly automated driving, Grade 4: Fully automated driving, Grade 5: Autonomous driving.

[0049] Firstly, the feedback to user B can include a warning that at least one other vehicle F2 is driving automatically and / or autonomously. This may be sufficient to attract user B's attention so that user B can pay closer attention while controlling vehicle F1.

[0050] Secondly, the feedback to user B can include a specific level of automation L2, L3, L4, L5.

[0051] In principle, it is conceivable that feedback to user B can be provided acoustically, e.g. through voice input, and / or visually, e.g. through a display.

[0052] A corresponding computer program product, a corresponding control unit ECU and a corresponding vehicle F1 also represent aspects of the invention.

[0053] The storage unit of the control unit ECU may contain an artificial neural network KNN, which is designed to evaluate the driving behavior of at least one other vehicle F2 and / or to determine the level of automation L2, L3, L4, L5 in the operation of at least one other vehicle F2.

[0054] The artificial neural network ANN can be designed as a forward-feedback neural network, a convolutional neural network, or a recurrent neural network.

[0055] The preceding explanation of the embodiments describes the present invention solely by way of examples. Naturally, individual features of the embodiments can be freely combined with one another, provided this is technically feasible, without departing from the scope of the present invention.

Claims

[1] Method for operating a vehicle (F1), in particular an automated or autonomous vehicle, comprising: - Detection of at least one other vehicle (F2) in the vicinity of the vehicle (F1), in particular in front of, behind and / or beside the vehicle (F1), - Evaluating the driving behavior of at least one other vehicle (F2), - Determining a level of automation (L2, L3, L4, L5) in the operation of at least one other vehicle (F2), - Providing feedback on the level of automation (L2, L3, L4, L5) in the operation of at least one other vehicle (F2) to a user (B) of the vehicle (F1). [2] Method according to claim 1, where the purpose is to detect at least one other vehicle (F2) and / or to assess the driving behavior of at least one other vehicle (F2) at least one of the following vehicle sensors is used: - optical sensor, - Camera, - Radar, - Lidar and / or - acoustic sensor. [3] Method according to one of the preceding claims, wherein when detecting at least one other vehicle (F2) in a vicinity of the vehicle (F1), an area in front of the vehicle (F1), an area behind the vehicle (F1) and / or an area next to the vehicle (F1) is monitored. [4] Method according to any one of the preceding claims, where the driving behavior of at least one other vehicle (F2) and / or to determine the level of automation (L2, L3, L4, L5) in the operation of at least one other vehicle (F2) at least one machine learning method and / or at least one machine image processing method and / or at least one artificial neural network (ANN) is used. [5] Method according to any of the preceding claims, wherein at least one of the following levels of automation (L2, L3, L4, L5) can be determined: Grade 2: Partially automated driving, Grade 2.5: an adaptive distance and speed control (ACC), Grade 3: Highly automated driving, Grade 4: Fully automated driving, Grade 5: Autonomous driving. [6] Method according to any one of the preceding claims, where the feedback to the user (B) includes a warning that at least one other vehicle (F2) is driving automatically and / or autonomously, and / or where the feedback to the user (B) includes a specific level of automation (L2, L3, L4, L5). [7] Method according to one of the preceding claims, wherein the feedback to the user (B) is provided acoustically, in particular by means of a voice input, and / or visually, e.g. by means of a display. [8] Computer program product comprising instructions which, when executed by a computer, cause the computer to perform a method according to any of the preceding claims. [9] Electronic control unit (ECU) comprising a computing unit and a storage unit in which a code is stored which, when at least partially executed by the computing unit, performs a method according to any one of the preceding claims 1 to 7, wherein in particular an artificial neural network (ANN) is stored in the storage unit, which is designed to evaluate the driving behavior of at least one other vehicle (F2) and / or to determine the level of automation (L2, L3, L4, L5) in the operation of at least one other vehicle (F2), wherein the artificial neural network (ANN) is preferably designed as a forward-feedback neural network, a convolutional neural network or a recurrent neural network. [10] Vehicle (F1) comprising a control unit (ECU) according to the preceding claim.

Images