Method for generating a data set

Abstract

The invention relates to a method for generating a data set, comprising the following steps: receiving motor-vehicle-generated environment data representing an environment of the motor vehicle from the perspective of the motor vehicle; receiving infrastructure-generated environment data representing the environment of the motor vehicle from the perspective of an infrastructure; preparing the motor-vehicle-generated environment data based on the infrastructure-generated environment data in order to determine prepared, motor-vehicle-generated environment data; and generating a data set comprising the prepared, motor-vehicle-generated environment data. The invention also relates to a method for testing at least one component of a process chain for a partially automated driving function of a motor vehicle, a method for generating at least one part of an algorithm for an at least partially automated driving function of a motor vehicle, a device, a computer program and a machine-readable storage medium.

Description

[0001] Description

[0002] title

[0003] Method for generating a data set

[0004] The invention relates to a method for generating a data set, a method for testing at least one component of an operational chain for an at least partially automated driving function of a motor vehicle, a method for generating at least part of an algorithm for an at least partially automated driving function of a motor vehicle, a device, a computer program and a machine-readable storage medium.

[0005] State of the art

[0006] The published patent application DE 10 2013 001 326 A1 discloses a motor vehicle which is designed to exchange operating data with a traffic object located in the vicinity of the motor vehicle and thereby to coordinate a driving maneuver of the motor vehicle with the traffic object.

[0007] Vehicle-generated environmental data is used to test an algorithm for executing an at least partially automated driving function of a motor vehicle. This means that a motor vehicle uses environmental sensors to detect its surroundings while driving and provides environmental data corresponding to the detection. This environmental data can be used to test an algorithm for executing an at least partially automated driving function of a motor vehicle.

[0008] Disclosure of the invention The object underlying the invention is to provide a method for generating a data set.

[0009] The object underlying the invention is also to provide a method for testing at least one component of an operational chain for an at least partially automated driving function of a motor vehicle.

[0010] The object underlying the invention is also to provide a method for generating at least part of an algorithm for an at least partially automated driving function of a motor vehicle.

[0011] The object underlying the invention is also to provide a device.

[0012] The object underlying the invention is also to provide a computer program.

[0013] The object underlying the invention is also to provide a machine-readable storage medium.

[0014] These objects are each achieved by the corresponding subject matter of the independent claims. Advantageous embodiments of the invention are the subject matter of the respective dependent subclaims.

[0015] According to a first aspect, a method for generating a data set is provided, comprising the following steps:

[0016] Receiving motor vehicle-generated environment data representing an environment of the motor vehicle from the perspective of the motor vehicle, receiving infrastructure-generated environment data representing the environment of the motor vehicle from the perspective of an infrastructure, processing the motor vehicle-generated environment data based on the infrastructure-generated environment data in order to determine processed, motor vehicle-generated environment data, and generating a data set comprising the processed, motor vehicle-generated environment data.

[0017] According to a second aspect, a method for testing at least one component of an operational chain for an at least partially automated driving function of a motor vehicle is provided, comprising the following steps:

[0018] Receiving a data set comprising vehicle-generated environmental data prepared using infrastructure-generated environmental data representing an environment of the motor vehicle from the perspective of an infrastructure, and testing the at least one component of the chain of effects based on the data set.

[0019] According to a third aspect, a method is provided for generating at least part of an algorithm for an at least partially automated driving function of a motor vehicle, comprising the following steps:

[0020] Receiving a data set comprising vehicle-generated environmental data, prepared using infrastructure-generated environmental data, representing an environment of the motor vehicle from the perspective of an infrastructure, and generating at least part of an algorithm for an at least partially automated driving function of a motor vehicle based on the data set.

[0021] According to a fourth aspect, a device is provided which is configured to carry out all steps of the method according to the first aspect and / or according to the second aspect and / or according to the third aspect. According to a fifth aspect, a computer program is provided which comprises instructions which, when the computer program is executed by a computer, for example by the device according to the fourth aspect, cause the computer to carry out a method according to the first aspect and / or according to the second aspect and / or according to the third aspect.

[0022] According to a sixth aspect, a machine-readable storage medium is provided on which the computer program according to the fifth aspect is stored.

[0023] The invention is based on and includes the discovery that a data set is generated that comprises processed, motor vehicle-generated environmental data, wherein this environmental data describes or represents the environment of a motor vehicle from the perspective of the motor vehicle. Thus, motor vehicle-generated environmental data is processed, with the processing being carried out based on infrastructure-generated environmental data that represents the environment of the motor vehicle from the perspective of an infrastructure.

[0024] This means that information is available for processing the vehicle-generated environment data that the vehicle itself does not have. The vehicle's environment sensors, which were used to detect the vehicle's environment to generate the vehicle-generated environment data, have a limited range for detecting the environment. Objects in the vehicle's environment can also hide other objects in the vehicle's environment from the environment sensors, i.e., shield them, so that the vehicle's environment sensors cannot detect such other objects.

[0025] In contrast, an infrastructure environment sensor can detect such other objects better. An infrastructure environment sensor can also have a greater range than the vehicle's environment sensor, meaning that more of the vehicle's surroundings can be detected by the infrastructure environment sensor than by the vehicle's environment sensor itself.

[0026] Thus, the vehicle-generated environmental data can be efficiently processed so that processed, vehicle-generated environmental data can be made available in the form of the data set.

[0027] Based on such a data set, at least one component of an action chain for an at least partially automated driving function of a motor vehicle can then be tested in an advantageous and efficient manner and / or at least part of an algorithm for an at least partially automated driving function of a motor vehicle can be generated.

[0028] The approach described here contrasts with the state of the art described above, according to which only vehicle-generated environmental data as such, i.e., unprocessed environmental data, was used to test an algorithm for an at least partially automated driving function of a motor vehicle. According to the state of the art, only the environment as it appears from the perspective of the motor vehicle was used to test such an algorithm, not from the perspective of an infrastructure.

[0029] For example, such an algorithm or a component of a chain of effects for an at least partially automated driving function of a motor vehicle would be very time-consuming and very expensive to develop, based solely on unprocessed, vehicle-generated environmental data, even though a correspondingly smaller hardware change to a motor vehicle's environmental sensor, for example, 1° more field of view or 20 m more range for the environmental sensor, would be a better solution for the future. Since, according to the concept described here, the vehicle-generated environmental data is processed based on the infrastructure-generated environmental data, questions such as whether, for example, a correspondingly smaller hardware change might not be the better solution, can be efficiently answered.

[0030] The concept described here therefore involves creating a dataset that contains processed, vehicle-generated environmental data. This means that the vehicle-generated environmental data is processed. This is particularly important in contrast to a fusion of vehicle-generated environmental data and infrastructure-generated environmental data. A data fusion creates a completely new dataset: the fused dataset.

[0031] The wording "processed," however, specifically clarifies that the original data set, i.e., the vehicle-generated environmental data, is still recognizable or present. This means, in particular, that the original vehicle-generated environmental data is still recognizable or present in the processed, vehicle-generated environmental data.

[0032] This can be seen analogously to treating water or treating a piece of furniture. After treatment, the original piece of furniture is still recognizable or is still present. The corners may have been rounded or the piece of furniture polished during the treatment. However, a completely new piece of furniture is not created.

[0033] The same applies to treated water. Here, for example, substances are filtered out so that the original water is still present and recognizable, and no new water is produced.

[0034] It is therefore clear that data fusion is not the processing of vehicle-generated environmental data. Thus, processing, as described above, is distinct from data fusion, i.e., the merging of data.

[0035] Processed, motor vehicle-generated environmental data are therefore in particular not fused environmental data, and therefore do not correspond to fused environmental data.

[0036] In one embodiment of the method according to the first aspect, it is provided that the processing comprises validating the motor vehicle-generated environmental data in order to determine a validation result, so that the processed, motor vehicle-generated environmental data comprise the validation result.

[0037] This provides the technical advantage, for example, that vehicle-generated environmental data can be efficiently processed. Validation can, for example, involve checking the dimensions of an object detected in the vehicle's surroundings based on the vehicle-generated environmental data. The vehicle moves, at times, very dynamically (speed) in an unknown world. The infrastructure is stationary and is part of a partially known world (all stationary objects, their properties, positions, etc.) are known. These objects are therefore defined as correct, i.e., as a reference. Using the "correct" data, i.e., the reference, of the infrastructure ("ground truth"), the recorded and determined vehicle data can be checked for "correctness," i.e., validated.

[0038] In one embodiment of the method according to the first aspect, it is provided that the processing comprises a plausibility check of the motor vehicle-generated environmental data in order to determine a plausibility check result, so that the processed, motor vehicle-generated environmental data include the plausibility check result. This results in, for example, the technical advantage that the motor vehicle-generated environmental data can be processed efficiently. Validation can, for example, comprise a plausibility test. Validation can, for example, comprise a step of checking whether information determined from the motor vehicle-generated environmental data exactly matches information determined from the infrastructure-generated environmental data.Plausibility checking may, for example, include a step of checking whether information determined from the vehicle-generated environmental data agrees, within a tolerance range, with information determined from the infrastructure-generated environmental data.

[0039] In one embodiment of the method according to the first aspect, it is provided that the processing comprises cleaning the motor vehicle-generated environmental data of erroneous data, so that the processed, motor vehicle-generated environmental data comprise cleaned motor vehicle-generated environmental data.

[0040] This provides the technical advantage, for example, that the vehicle-generated environmental data can be processed efficiently. An error or problem could be, for example, that a vehicle's environmental sensor is dirty, resulting in incorrect environmental sensor data in the vehicle-generated environmental data. These data errors can be found and removed by comparing them with the infrastructure's environmental data and, for example, replaced with the correct infrastructure's environmental data. One advantage, for example, is that the vehicle-generated environmental data may still be usable.

[0041] The cleaning thus comprises, in particular, removing the erroneous data and, in particular, replacing the erroneous data with correct environmental data of the infrastructure (infrastructure-generated environmental data). In one embodiment of the method according to the first aspect, it is provided that the processing comprises analyzing the infrastructure-generated environmental data in order to obtain environmental information about the environment of the motor vehicle, wherein the processing further comprises enriching the motor vehicle-generated environmental data with the environmental information, so that the processed, motor vehicle-generated environmental data comprise the environmental information.

[0042] This provides the technical advantage, for example, that the vehicle-generated environmental data can be processed efficiently. The infrastructure's environmental data also includes data from areas in the environment that the vehicle cannot detect using its environmental sensors, for example, because the area(s) are not within the field of view of the environmental sensors and / or there are obscurations in the scene, for example, because a pedestrian is standing behind a cyclist. The area(s) not detected can be added to the vehicle-generated environmental data as environmental information to enhance the data.

[0043] In one embodiment of the method according to the first aspect, it is provided that the infrastructure-generated environmental data comprise context data which represent a context of the environment, so that the motor vehicle-generated environmental data are prepared based on the context data.

[0044] This provides the technical advantage, for example, that the vehicle-generated environmental data can be processed efficiently.

[0045] Based solely on the vehicle-generated environmental data before it is processed, the context in which the vehicle was while driving is usually not clear. For example, a certain behavior of traffic in the vicinity of the vehicle only makes sense if the context is known. If, for example, all vehicles in the vicinity of the vehicle stop or move to the left or right side of the road, this is rather unusual behavior without knowledge of the context. However, if the context is known that emergency services were on the move with their blue lights flashing at the time, such behavior makes sense, because vehicles must make way for emergency services with their blue lights flashing.

[0046] In one embodiment of the method according to the first aspect, it is provided that the context comprises one or more of the following contexts: structural context, traffic context, event context, weather context, road condition context.

[0047] This provides the technical advantage, for example, that the vehicle-generated environmental data can be processed efficiently.

[0048] A structural context indicates, for example, whether there was a construction site in the vicinity of the motor vehicle.

[0049] A traffic context indicates, for example, how a traffic light system switched, i.e. which signal pattern was present, and / or whether there was a traffic diversion, for example regulated by a traffic sign, and / or whether there was a change in a speed limit.

[0050] An event context indicates, for example, whether there is an emergency vehicle in the vicinity of the motor vehicle that is on an operation.

[0051] A weather context, for example, indicates the weather in the surroundings of the motor vehicle.

[0052] A road condition context indicates, for example, whether there are one or more holes in the road, whether the road is icy, and / or whether the road is wet. In one embodiment of the method, the vehicle-generated environmental data is generated based on a detection of the environment by means of an environmental sensor system of the motor vehicle, wherein at least one technical parameter, in particular range and / or field of view, of the environmental sensor system of the motor vehicle is received, wherein the data set is generated such that the data set comprises the at least one technical parameter of the environmental sensor system of the motor vehicle.

[0053] This results in the technical advantage, for example, that the data set contains additional information about the processed, motor vehicle-generated environmental data, which information can be used, for example, advantageously for testing at least one component of an action chain for an at least partially automated driving function of a motor vehicle and / or for generating at least part of an algorithm for an at least partially automated driving function of a motor vehicle.

[0054] In one embodiment of the method according to the first aspect, it is provided that the infrastructure-generated environment data were generated based on a detection of the environment by means of an environment sensor system of the infrastructure, wherein infrastructure environment sensor data is received which indicates at least one technical parameter of the environment sensor system of the infrastructure, wherein the data set is generated such that the data set comprises the at least one technical parameter of the environment sensor system of the infrastructure.

[0055] This results in the technical advantage, for example, that the data set containing the processed, vehicle-generated environmental data contains additional information, which can be used, for example, advantageously for testing at least one component of an operational chain for an at least partially automated driving function of a motor vehicle and / or for generating at least part of an algorithm for an at least partially automated driving function of a motor vehicle. In one embodiment of the method according to the first aspect, it is provided that the vehicle-generated environmental data represent a respective environment of several motor vehicles from the perspective of the respective motor vehicle during a respective journey through a respective infrastructure area of ​​the infrastructure, wherein the infrastructure-generated environmental data represent the respective environment of the motor vehicles from the perspective of the infrastructure.

[0056] This provides, for example, the technical advantage that vehicle-generated environmental data from multiple vehicles is available for generating the data set. Thus, a data set can be generated that is advantageously particularly suitable for use in a method according to the second aspect and / or the third aspect.

[0057] In one embodiment of the method according to the first aspect, it is provided that the respective infrastructure area is in each case an element selected from the following group of infrastructure areas: road junction, construction site, tunnel, motorway entrance, motorway exit, motorway, stop, entrance, exit, narrowing, road narrowing, lane narrowing, underpass, place with adjacent public traffic, in particular football stadium or train station.

[0058] This results in the technical advantage, for example, that particularly important infrastructure areas are taken into account with regard to testing at least one component of an operational chain for an at least partially automated driving function of a motor vehicle and / or for generating at least part of an algorithm for an at least partially automated driving function of a motor vehicle.

[0059] A road junction is, for example, an intersection, a junction, a T-junction, a roundabout, or a motorway interchange. In one embodiment of the method according to the first aspect, it is provided that at least some of the respective infrastructure areas are identical to one another, where identical means that they are of the same infrastructure area type and the same infrastructure area, and / or where at least some of the infrastructure areas are different from one another, where different means that they may be of the same infrastructure area type but different infrastructure areas, or that they may be different infrastructure area types.

[0060] This provides, for example, the technical advantage that vehicle-generated environmental data can be used by an identical infrastructure area and / or by different infrastructure areas.

[0061] In one embodiment of the method according to the first aspect, it is provided that the data set is generated in such a way that it comprises the motor vehicle-generated environmental data and / or the infrastructure-generated environmental data.

[0062] This results in, for example, the technical advantage that the data set is generated efficiently. According to this embodiment, the data set includes not only the processed, vehicle-generated environment data, but also the unprocessed, vehicle-generated environment data and / or the infrastructure-generated environment data used to process the vehicle-generated environment data.

[0063] Thus, a data set is advantageously available which is particularly suitable for use for the method according to the second aspect and / or the method according to the third aspect.

[0064] In one embodiment of the method according to the second aspect, it is provided that the at least one component comprises an algorithm for executing the at least partially automated driving function, wherein the testing comprises simulating a motor vehicle which drives at least partially automatically based on the at least partially automated driving function while executing the algorithm.

[0065] This provides, for example, the technical advantage that an algorithm for executing the at least partially automated driving function can be efficiently tested.

[0066] In one embodiment of the method according to the second aspect, it is provided that the motor vehicle-generated environmental data were generated based on a detection of the environment by means of an environmental sensor system of the motor vehicle, wherein the data set comprises at least one technical parameter, in particular range and / or field of view, of the environmental sensor system of the motor vehicle, wherein the at least one component comprises an environmental sensor system of the motor vehicle, wherein the testing comprises changing the at least one technical parameter based on the data set in order to obtain at least one changed technical parameter, and wherein the testing comprises simulating a motor vehicle having a simulated environmental sensor system with the at least one changed technical parameter, which motor vehicle drives at least partially automatically based on the at least partially automated driving function using the simulated environmental sensor system.

[0067] This provides the technical advantage, for example, of efficiently testing the vehicle's environmental sensors. In particular, it is advantageous to test whether the modified technical parameter results in an improvement. For example, a slightly larger field of vision and / or greater range is sufficient for the vehicle to drive more efficiently, at least partially automated, using its environmental sensors. For example, an algorithm for an at least partially automated driving function does not require lengthy and expensive development if such a small technical change to the vehicle's environmental sensors achieves a similar effect.An application case in which the vehicle did not drive sufficiently well, at least partially automated, so that an algorithm would need to be improved accordingly. However, in this specific application case, the algorithm is not to blame, but rather inadequate environmental sensors. For example, a pedestrian enters the field of view of the vehicle's environmental sensors late due to insufficient range and / or field of view, causing the vehicle to brake late. Increasing the range and / or field of view can defuse such a situation without the need for lengthy algorithm development.

[0068] In one embodiment of the method according to the third aspect, it is provided that at least one part of the algorithm is generated using an artificial neural network.

[0069] This results in the technical advantage that at least part of the algorithm can be generated efficiently.

[0070] According to an embodiment of the method according to the second aspect and / or according to the third aspect, it is provided that the data set comprises the motor vehicle-generated environmental data and / or the infrastructure-generated environmental data.

[0071] This provides the technical advantage, for example, that a particularly suitable data set can be used.

[0072] In one embodiment of the method according to the second aspect and / or according to the third aspect, it is provided that the data set was generated according to the method according to the first aspect.

[0073] This results in the technical advantage, for example, of using a particularly suitable data set. Statements made in connection with the method according to the first aspect and / or the second aspect apply analogously to the method according to the third aspect, and vice versa.

[0074] Technical functionalities and features of the method according to the first aspect and / or according to the second aspect result analogously from corresponding technical functionalities and features from the method according to the third aspect and vice versa.

[0075] The embodiments described here can be combined with each other in any way, even if this is not explicitly described.

[0076] This means in particular that features from embodiments of the method according to the first aspect can be used in embodiments of the method according to the second aspect and / or according to the third aspect and vice versa.

[0077] Furthermore, device features result from corresponding process features and vice versa.

[0078] In one embodiment of the method according to the first aspect and / or according to the second aspect and / or according to the third aspect, the respective method is a computer-implemented method.

[0079] An at least partially automated driving function results in at least partially automated driving of the motor vehicle.

[0080] The term "at least partially automated driving" encompasses one or more of the following cases: assisted driving, partially automated driving, highly automated driving, or fully automated driving. The term "at least partially automated" therefore encompasses one or more of the following cases: assisted, partially automated, highly automated, or fully automated. Assisted driving means that a driver of the motor vehicle continuously performs either the lateral or longitudinal control of the motor vehicle. The other driving task (i.e., controlling the longitudinal or lateral control of the motor vehicle) is performed automatically. This means that with assisted driving of the motor vehicle, either the lateral or longitudinal control is controlled automatically.

[0081] Partially automated guidance means that the longitudinal and lateral guidance of the vehicle is automatically controlled in a specific situation (for example, driving on a highway, driving within a parking lot, overtaking an object, driving within a lane defined by lane markings) and / or for a specific period of time. The driver of the vehicle does not have to manually control the longitudinal and lateral guidance of the vehicle. However, the driver must continuously monitor the automatic control of the longitudinal and lateral guidance in order to intervene manually if necessary. The driver must be ready to fully assume control of the vehicle at any time.

[0082] Highly automated guidance means that for a certain period of time in a specific situation (for example: driving on a motorway, driving in a parking lot, overtaking an object, driving in a lane defined by lane markings), the longitudinal and lateral guidance of the motor vehicle are automatically controlled. The driver of the motor vehicle does not have to manually control the longitudinal and lateral guidance of the vehicle. The driver does not have to constantly monitor the automatic control of the longitudinal and lateral guidance in order to be able to intervene manually if necessary. If necessary, a takeover request is automatically issued to the driver to take over control of the longitudinal and lateral guidance, in particular with a sufficient time reserve. The driver must therefore potentially be able to take over control of the longitudinal and lateral guidance.Limits of the automatic control of lateral and longitudinal guidance are automatically detected. With highly automated guidance, it is not possible to automatically achieve a minimal-risk state in every initial situation.

[0083] Fully automated guidance means that in a specific situation (e.g. driving on a motorway, driving within a parking lot, overtaking an object, driving within a lane defined by lane markings), the longitudinal and lateral guidance of the vehicle are automatically controlled. The driver of the vehicle does not have to manually control the longitudinal and lateral guidance of the vehicle. The driver does not have to monitor the automatic control of the longitudinal and lateral guidance in order to intervene manually if necessary. Before the automatic control of the lateral and longitudinal guidance is terminated, the driver is automatically prompted to take over the driving task (controlling the lateral and longitudinal guidance of the vehicle), particularly with a sufficient time reserve. If the driver does not take over the driving task, the vehicle is automatically returned to a state with minimal risk.Limits of the automatic control of lateral and longitudinal guidance are automatically detected. In all situations, it is possible to automatically return to a system state with minimal risk.

[0084] The phrase “at least one” includes the phrase “one or more”.

[0085] An environmental sensor system within the meaning of the description, i.e. in particular the environmental sensor system of the motor vehicle and / or the environmental sensor system of the infrastructure, comprises, for example, one or more environmental sensors.

[0086] For example, the environmental sensors of the infrastructure's environmental sensor system are spatially distributed within the infrastructure.

[0087] The environmental sensors of the motor vehicle's environmental sensor system are, for example, spatially distributed on the motor vehicle and / or installed in the motor vehicle, i.e., are comprised of the motor vehicle. An environmental sensor within the meaning of this description is, for example, one of the following environmental sensors: radar sensor, lidar sensor, image sensor, in particular image sensor of a video camera, magnetic field sensor, infrared sensor, and ultrasonic sensor.

[0088] The environmental sensors of the infrastructure can, for example, be referred to as infrastructure environmental sensors.

[0089] The environmental sensor system of the motor vehicle can, for example, be referred to as a motor vehicle environmental sensor system.

[0090] Motor vehicle-generated environmental data is therefore created or generated in particular in the motor vehicle.

[0091] Infrastructure-generated environmental data is therefore created or generated particularly within the infrastructure.

[0092] The terms "create" and "generate" can be used synonymously.

[0093] A data set within the meaning of the description can, for example, also be referred to as a test data set, insofar as the data set can be used in particular for testing at least one component of an action chain for an at least partially automated driving function of a motor vehicle.

[0094] Such a test data set can be used in particular for the method according to the third aspect.

[0095] Embodiments of the invention are illustrated in the drawings and explained in more detail in the following description. They show:

[0096] Fig. 1 is a flowchart of a method according to the first aspect, Fig. 2 is a flowchart of a method according to the second aspect,

[0097] Fig. 3 is a flowchart of a method according to the third aspect,

[0098] Fig. 4 a motor vehicle driving through an infrastructure,

[0099] Fig. 5 shows a device according to the fourth aspect and

[0100] Fig. 6 shows a computer-readable storage medium according to the sixth aspect.

[0101] Fig. 1 shows a flowchart of a method for generating a data set, comprising the following steps:

[0102] Receiving 101 motor vehicle-generated environment data representing an environment of the motor vehicle from the perspective of the motor vehicle, receiving 103 infrastructure-generated environment data representing the environment of the motor vehicle from the perspective of an infrastructure, processing 105 the motor vehicle-generated environment data based on the infrastructure-generated environment data in order to determine processed, motor vehicle-generated environment data, and generating 107 a data set comprising the processed, motor vehicle-generated environment data.

[0103] According to one embodiment of the method according to the first aspect, the generated data set is output. According to one embodiment, the output data set is stored on a computer-readable storage medium. Outputting comprises, according to one embodiment, sending the generated data set over a communications network.

[0104] Fig. 2 shows a flow diagram of a method for testing at least one component of an action chain for an at least partially automated driving function of a motor vehicle, comprising the following steps: receiving 201 a data set which comprises motor vehicle-generated environment data which represent an environment of the motor vehicle from the perspective of an infrastructure and which has been prepared using infrastructure-generated environment data which represent an environment of the motor vehicle from the perspective of the motor vehicle, testing 203 the at least one component of the action chain based on the data set.

[0105] Fig. 3 shows a flowchart of a method for generating at least part of an algorithm for an at least partially automated driving function of a motor vehicle, comprising the following steps:

[0106] Receiving 301 a data set comprising motor vehicle-generated environmental data, which represent an environment of the motor vehicle from the perspective of an infrastructure, prepared using infrastructure-generated environmental data, which represent an environment of the motor vehicle from the perspective of the motor vehicle, generating 303 at least part of an algorithm for an at least partially automated driving function of a motor vehicle based on the data set.

[0107] Fig. 4 shows a motor vehicle 401 during a journey, in particular an at least partially automated journey, on a road 403 of an infrastructure 405.

[0108] The infrastructure may, for example, comprise one or more of the following infrastructure areas, which are represented by pictograms with corresponding reference symbols: intersection, 407, construction site 409, tunnel 411 .

[0109] This means, in particular, that the road 403 can lead through one or more of the above-mentioned infrastructure areas 407, 409, 411. Within the infrastructure 405, an environmental sensor system 413 is arranged, which is configured, for example, to monitor the road 403 and, for example, an infrastructure area, for example, one of the infrastructure areas 407, 409, 411, on the infrastructure side. The environmental sensor system 413 comprises, for example, a video camera 415 comprising an image sensor (not shown), a microphone 417, a temperature sensor 419, and a gas sensor 421, symbolically identified by a nose symbol. Furthermore, three points are identified by the reference numeral 423, which is intended to symbolize that the environmental sensor system 413 can comprise further environmental sensors.

[0110] The video camera 415, the microphone 417, the temperature sensor 419, and the gas sensor 421 are spatially distributed within the infrastructure 405. These environmental sensors detect their respective environments and output environmental sensor data corresponding to the detection. The environmental sensor data is infrastructure-generated environmental data.

[0111] The environment sensor data can be processed, for example, by a first computer system 425 to detect objects in the respective environment of the environment sensors. Upon detection of an object, information about the detected object, such as a position, speed, and direction of movement, can be sent to the motor vehicle 401, for example, via a first wireless communication interface 427, in order to provide infrastructure-supported assistance to the motor vehicle during its journey through the infrastructure 405.

[0112] The motor vehicle 401 includes an environmental sensor system 429, which has a further video camera 431 with an image sensor (not shown) and a radar sensor 433. These environmental sensors detect the environment of the motor vehicle 403 and output environmental sensor data corresponding to the respective detection. The environmental sensor data is motor vehicle-generated environmental data. The motor vehicle-generated environmental data describes the environment of the motor vehicle 401 from the perspective of the motor vehicle 401.

[0113] The infrastructure-generated environmental data of the one or more environmental sensors of the infrastructure 405 that detect the environment of the motor vehicle 401 describe the environment of the motor vehicle 401 from the perspective of the infrastructure 405.

[0114] The motor vehicle 401 sends the vehicle-generated environmental data via a wireless communication network, for example, WLAN and / or mobile radio, to a data center 435, which receives this environmental data via a second wireless communication interface 437. Alternatively or in addition to the wireless transmission of the vehicle-generated environmental data from the motor vehicle 401 to the data center 435, a wired or cable-based communication can be provided after the journey of the motor vehicle 401, i.e., when it has arrived at a destination and is parked there, and / or this environmental data can be stored on a machine-readable storage medium, for example, a hard disk, in particular an SSD, wherein the machine-readable storage medium is sent to the data center 435.

[0115] A second computer system 439 is located within the data center 435, which processes the vehicle-generated environmental data. This processing includes preparing the vehicle-generated environmental data. For this preparation, the second computer system 439 uses the infrastructure-generated environmental data, which is sent from the first computer system 425 to the data center via the first wireless communication interface 427.

[0116] The processing is characterized by a function block 441. As a result, a data set 443 is generated, which includes the processed, vehicle-generated environmental data.

[0117] Data set 443 can be used, for example, for a simulation 445, according to which another motor vehicle 447 travels through infrastructure 405 in an at least partially automated manner. For this at least partially automated journey, an at least partially automated driving function is used, which is executed by an algorithm. The algorithm can be generated for the first time or can be further developed, so that the correspondingly generated or further developed algorithm can be tested within the framework of the simulation.

[0118] Furthermore, for example, the infrastructure-side environmental sensor system 413 and / or the motor vehicle-side environmental sensor system 429 can be tested and / or further developed on the basis of the data set 443. The infrastructure-side environmental sensor system 413 and the motor vehicle-side environmental sensor system 429 as well as the algorithm are exemplary components of an action chain for an at least partially automated driving function of a motor vehicle.

[0119] Fig. 5 shows a device 501 which is configured to carry out all steps of the method according to the first aspect and / or according to the second aspect and / or according to the third aspect.

[0120] Fig. 6 shows a machine-readable storage medium 601 on which a computer program 603 is stored. The computer program 603 comprises instructions which, when executed by a computer, cause the computer program 603 to execute a method according to the first aspect and / or according to the second aspect and / or according to the third aspect.

[0121] In the known state of the art, only information from the perspective of the motor vehicle is used for such testing or development of an algorithm.

[0122] What happened outside the area detectable by the vehicle's environmental sensors is therefore not available according to the state of the art. The concept described here proposes using, in addition to the vehicle's view, the infrastructure's view of the respective traffic situation to process the vehicle-generated environmental data. A correspondingly generated data set can be used, for example, for the further development of a vehicle's environmental sensor system.

[0123] For example, expensive and lengthy further development of an algorithm or an environmental sensor can be avoided by simply making a minor hardware change, such as more visibility, for example 1 degree more visibility, and / or more range, for example 20 m more range, which could only be determined through testing based on the data set.

[0124] By using the infrastructure-generated environmental data, the vehicle-generated environmental data can be checked, for example, validated, during the processing process.

[0125] The vehicle-generated environmental data generally does not provide any context for the current traffic situation.

[0126] Such a context includes, for example:

[0127] Traffic control data (for example, a traffic light changes its status (signal pattern), lanes are closed, speed limits are changed, dynamic traffic signs change their display, ...) and / or an event influencing the environment, for example, an ambulance is approaching from a distance and road users behave accordingly and make way, which would be strange behavior without knowledge of the approaching ambulance, and / or

[0128] Road condition data and / or construction site data and / or weather data. Environmental data within the meaning of the description includes, for example, raw data, for example images, and / or evaluated data, for example object data from evaluated images).

[0129] In this way, a data set can be generated efficiently which is particularly suitable for generating and developing an algorithm for an at least partially automated driving function of a motor vehicle.

[0130] The vehicle-generated environmental data can be automatically validated. This means that the environmental data can be verified to be accurate.

[0131] This allows additional data (processed, vehicle-generated environmental data) to be made available to the algorithm. This means that the algorithm can be tested and / or validated against this data.

[0132] This also allows a triggering algorithm for recording vehicle-generated environmental data to be verified / further developed.

[0133] This allows a component of a chain of effects for an at least partially automated driving function of a motor vehicle to be optimally and efficiently developed and improved. For example, the optimal range and / or field of view of an environment sensor can be determined.

Claims

Claims 1. A procedure for generating a data set, comprising the following steps: Receiving (101) vehicle-generated environment data, which represents the vehicle's environment from the vehicle's perspective; receiving (103) infrastructure-generated environment data, which represents the vehicle's environment from the perspective of an infrastructure; processing (105) the vehicle-generated environment data based on the infrastructure-generated environment data to determine processed vehicle-generated environment data; and Generating (107) a data set which includes the processed, motor vehicle-generated environment data.

2. Method according to claim 1, wherein the processing comprises validating the vehicle-generated environmental data to determine a validation result, such that the processed, vehicle-generated environmental data includes the validation result.

3. Method according to claim 1 or 2, wherein the processing comprises plausibility checking of the vehicle-generated environmental data in order to determine a plausibility check result, such that the processed vehicle-generated environmental data includes the plausibility check result.

4. Method according to one of the preceding claims, wherein the processing comprises cleaning the vehicle-generated environment data of erroneous data, such that the processed vehicle-generated environment data comprises cleaned vehicle-generated environment data.

5. Method according to one of the preceding claims, wherein the processing comprises analyzing the infrastructure-generated environment data to obtain environment information about the environment of the motor vehicle, wherein the processing further comprises enriching the motor vehicle-generated environment data with the environment information, such that the processed, motor vehicle-generated environment data includes the environment information.

6. Method according to any of the preceding claims, wherein the infrastructure-generated environment data includes context data which represents a context of the environment, such that the vehicle-generated environment data is processed based on the context data.

7. The method of claim 6, wherein the context comprises one or more of the following contexts: structural context, traffic context, event context, weather context, road condition context.

8. Method according to one of the preceding claims, wherein the vehicle-generated environmental data were generated based on a detection of the environment by means of environmental sensors of the vehicle, wherein at least one technical parameter, in particular range and / or field of view, of the environmental sensors of the vehicle is received, wherein the data set is generated in such a way that the data set includes the at least one technical parameter of the environmental sensors of the vehicle.

9. Method according to one of the preceding claims, wherein the infrastructure-generated environmental data were generated based on a recording of the environment by means of environmental sensors of the infrastructure, wherein infrastructure environmental sensor data are received which specify at least one technical parameter of the environmental sensors of the infrastructure, wherein the data set is generated in such a way that the data set includes the at least one technical parameter of the environmental sensors of the infrastructure.

10. Method according to one of the preceding claims, wherein the vehicle-generated environment data represent a respective environment of several motor vehicles from the perspective of the respective motor vehicle during a respective journey through a respective infrastructure area of ​​the infrastructure, wherein the infrastructure-generated environment data represent the respective environment of the motor vehicles from the perspective of the infrastructure.

11. Method according to claim 10, wherein the respective infrastructure area is an element selected from the following group of infrastructure areas: road junction, construction site, tunnel, motorway entrance, motorway exit, motorway, stop, entrance, exit, narrowing, road narrowing, carriageway narrowing, underpass, place with adjacent public traffic, in particular football stadium or railway station.

12. Method according to claim 10 or 11, wherein at least some of the respective infrastructure areas are identical to each other, where identical means that they are of the same infrastructure area type and the same infrastructure area, and / or wherein at least some of the infrastructure areas are different to each other, where different means that they may be of the same infrastructure area type but different infrastructure areas, or that they may be different infrastructure area types.

13. Method according to one of the preceding claims, wherein the data set is generated such that it includes the vehicle-generated environment data and / or the infrastructure-generated environment data.

14. Method for testing at least one component of an action chain for at least a partially automated driving function of a motor vehicle, comprising the following steps: Receiving (201) a data set comprising processed, vehicle-generated environmental data representing an environment of the vehicle from the perspective of the vehicle, using infrastructure-generated environmental data representing an environment of the vehicle from the perspective of the vehicle, and testing (203) at least one component of the chain of effects based on the data set.

15. Method according to claim 14, wherein the at least one component comprises an algorithm for executing the at least partially automated driving function, wherein the testing comprises simulating a motor vehicle which drives at least partially automatically based on the at least partially automated driving function by executing the algorithm.

16. Method according to claim 14 or 15, wherein the vehicle-generated environmental data were generated based on the detection of the environment by means of environmental sensors of the vehicle, wherein the data set comprises at least one technical parameter, in particular range and / or field of view, of the environmental sensors of the vehicle, wherein the at least one component comprises environmental sensors of the vehicle, wherein the testing comprises changing the at least one technical parameter based on the data set in order to obtain at least one changed technical parameter, and wherein the testing comprises simulating a vehicle comprising simulated environmental sensors with the at least one changed technical parameter, which drives at least partially automatically based on the at least partially automated driving function using the simulated environmental sensors.

17. Method for generating at least part of an algorithm for at least a partially automated driving function of a motor vehicle, comprising the following steps: Receiving (301) a data set comprising processed, vehicle-generated environment data representing an environment of the vehicle from the perspective of the vehicle, using infrastructure-generated environment data representing an environment of the vehicle from the perspective of the vehicle, generating (303) at least part of an algorithm for at least a partially automated driving function of a vehicle based on the data set.

18. The method of claim 17, wherein at least part of the algorithm is generated using an artificial neural network.

19. Method according to any one of claims 14 to 18, wherein the data set comprises the vehicle-generated environment data and / or the infrastructure-generated environment data.

20. Method according to any one of claims 14 to 18, wherein the data set was generated according to the method according to any one of claims 1 to 13.

21. Device (501) configured to perform all steps of the method according to any of the preceding claims.

22. Computer program (603) comprising instructions which, when the computer program (603) is executed by a computer, cause it to execute a method according to any one of claims 1 to 20.

23. Machine-readable storage medium (601) on which the The computer program (603) according to claim 22 is stored.

Images