Method and system for predicting a change of a static automated vehicle related to a digital map

Abstract

Methods and systems for predicting a change in a static environment of an automated vehicle relative to a digital map. A method (10) for predicting a change in a static environment of an automated vehicle relative to a digital map, the digital map comprising at least information about roadway guidance and static objects in the environment of the automated vehicle, the method comprising the following method steps. Deviations in the surroundings of the automated vehicle from the digital map are detected on the basis of sensor data of at least one sensor and / or at least one database, and the detected deviations are quantified by ascertaining at least one change indicator. At least one probability of a future change in the surroundings of the automated vehicle relative to the digital map is ascertained on the basis of the identified and quantified deviations.

Description

Technical Field

[0001] The invention relates to a method and a system for predicting changes in the static surroundings of an automated vehicle relative to a digital map. Background Art

[0002] The reliable map properties of digital maps for automated driving are specified by a series of safety-related requirements. These requirements include, for example, global and relative accuracy limits that must not be exceeded. If, for example, an accuracy requirement is exceeded and the digital map does not indicate this, a failure of a certain map property occurs. An upper limit on the failure rate (e.g. 1 / 5,000 km) is another important safety requirement for the characterization of digital maps. In order to publish digital maps or map properties, it must be proven that all safety requirements are met.

[0003] The failure rate of the map is also affected by changes over time. As soon as static infrastructure elements change, for example due to an update of the road markings or a change in the speed limit due to the installation of corresponding road signs, wherein the road signs are also included in the digital map, the failure rate of the digital map increases if no countermeasures are taken, because the digital map is out of date.

[0004] Static infrastructure elements have different change characteristics. For example, short-term changes may be made, such as lanes being reduced or narrowed due to accidents, which have no long-term impact on the failure rate of digital maps. Direct long-term changes may also occur, such as changes in speed limit signs, which lead to an immediate increase in the failure rate of maps. In addition, slow and observable changes may occur. For example, it is possible that the geometry of the road changes, which leads to a long-term increase in the failure rate of maps.

[0005] The characteristics or types of map attributes determine whether and what changes are possible. For example, changes in road geometry may require large-scale construction measures. Short-term changes are therefore not possible. In contrast, changes in speed limit signs can be made on a short-term basis. However, changes to traffic signs may also be required due to long-term construction zones.

[0006] It is known from the prior art that changes in the static surroundings relative to a digital map can be predicted. The prediction can concern changes that are not yet recognizable, such as planned construction areas. However, current deviations can also be used to make predictions about future changes. In this case, these changes are not yet effective, for example due to construction work that is still in progress, so that predictions can be made. Changes that are already effective can be detected in order to use them to control the automated vehicle. Summary of the invention

[0007] One object of the present invention is to specify or provide an improved method and an improved system for predicting changes in the static surroundings of an automated vehicle relative to a digital map. This object is achieved by a method and a system according to the present invention for predicting changes in the static surroundings of an automated vehicle relative to a digital map.

[0008] Advantageous developments of the invention are described in the context of advantageous embodiments.

[0009] A method for predicting changes in a static surroundings of an automated vehicle relative to a digital map, wherein the digital map includes at least information about roadway guidance and static objects in the surroundings of the automated vehicle, the method comprising the following method steps: Based on sensor data of at least one sensor and / or at least one database, deviations in the surroundings of the automated vehicle relative to the digital map are detected, and the detected deviations are quantified by ascertaining at least one change indicator. Based on the detected and quantified deviations, at least one probability of a future change in the surroundings of the automated vehicle relative to the digital map is ascertained.

[0010] The automated vehicle is designed as a motor vehicle and can be guided at least partially automatically. The expression “at least partially automatically guided” includes one or more of the following: assisted guiding, partially automated guiding, highly automated guiding, fully automated guiding.

[0011] Assisted guidance means that the driver of the motor vehicle continuously guides the motor vehicle either transversely or longitudinally. The corresponding other driving task (i.e., the control of the longitudinal guidance or transverse guidance of the motor vehicle) is performed automatically. This therefore means that during assisted guidance of the motor vehicle, either the transverse guidance or the longitudinal guidance is automatically controlled.

[0012] Partially automated guidance means that the longitudinal and lateral guidance of the motor vehicle is automatically controlled in certain situations (e.g. driving on the motorway, driving in a parking lot, overtaking an object, driving in a lane defined by lane markings) and / or within certain time periods. The driver of the motor vehicle does not have to manually control the longitudinal and lateral guidance of the motor vehicle himself. However, the driver must constantly monitor the automated control of the longitudinal and lateral guidance in order to be able to intervene manually if necessary. The driver must be ready to take over the full control of the motor vehicle at any time.

[0013] Highly automated guidance means that the longitudinal guidance and lateral guidance of the motor vehicle are automatically controlled in certain situations (e.g., driving on a highway, driving in a parking lot, overtaking an object, driving in a lane determined by lane markings) within a certain period of time. The driver of the motor vehicle does not have to manually control the longitudinal guidance and lateral guidance of the motor vehicle himself. The driver does not have to continuously monitor the automated control of the longitudinal guidance and lateral guidance so that he can intervene manually when necessary. When necessary, a takeover request is automatically output to the driver in order to take over the control of the longitudinal guidance and lateral guidance, especially in a manner with sufficient time margin. That is, the driver must potentially be able to take over the control of the longitudinal guidance and lateral guidance. The limits of the automated control of the lateral guidance and longitudinal guidance are automatically identified. In the case of highly automated guidance, it is not possible to automatically reach a state with minimal risk in any initial situation.

[0014] Fully automated guidance means that in certain situations (e.g. driving on a motorway, driving in a parking lot, overtaking an object, driving in a lane defined by lane markings), the longitudinal and transverse guidance of the motor vehicle is automatically controlled. The driver of the motor vehicle does not have to manually control the longitudinal and transverse guidance of the motor vehicle himself. The driver does not have to monitor the automated control of the longitudinal and transverse guidance so that he can intervene manually if necessary. Before the automated control of the transverse and longitudinal guidance is terminated, a request to the driver to take over the driving task (control of the transverse and longitudinal guidance of the motor vehicle) is automatically implemented, in particular in a manner with sufficient time margin. If the driver does not take over the driving task, the automation returns to the state with the lowest risk. The limits of the automated control of the transverse and longitudinal guidance are automatically identified. In any situation, it is possible for the automation to return to the system state with the lowest risk.

[0015] All information stored in a digital map can generally be referred to as attributes. All changes relative to the digital map related to static objects and roadway guidance should be referred to as deviations relative to the digital map in the static surroundings of the automated vehicle. This may only exist temporarily, so it is not necessarily required to update the digital map when a deviation occurs. For example, it may be that the roadway covering of the roadway has been refurbished, so a short-term deviation from the roadway guidance stored in the digital map can be confirmed. However, this deviation is only temporary and is not permanently valid if the engineering at the roadway is completed. On the contrary, "change" should be understood as such a change: it is expected that these changes will be continuously valid, such as changes in roadway guidance. This change requires matching or updating of the digital map because the digital map has at least one section that incorrectly represents a part of the surroundings in the future.

[0016] The method advantageously enables the prediction of changes in the static surroundings which would otherwise lead to an outdated digital map and thus to an increased digital map failure rate, as would be the case if changes were only detected. This aspect is important in particular in the context of safety-related applications in order to keep below a defined failure rate or error rate which is required for the safe use of digital maps in safety-related functions within the scope of automated vehicle control. The method thus contributes to a safety concept within the scope of which reliable map properties can be provided. Safety in the control of automated vehicles in road traffic can thus be improved.

[0017] By reducing the degradation of digital maps during operation (ie, the increase in the error rate), the error rate budget for creating or updating digital maps can also be advantageously reduced. This can lead to a positive impact on the map production costs.

[0018] For example, the sensor data on which the deviation is identified can be video data of at least one camera of the automated vehicle and / or at least one external camera, i.e. a camera of another automated vehicle and / or a camera arranged in the area of ​​the infrastructure. For the recognition of several static surrounding features, such as several lane marking types, for example, known safety-qualified video-based perception schemes can be used. Alternatively or additionally, the sensor data can be radar data of a radar unit of the automated vehicle or of an external radar unit, data of a lidar unit or of an external lidar unit, odometry data of the automated vehicle and / or of at least one other vehicle, and data of a navigation system. Alternatively or additionally, satellite data and / or aerial photography can also be used as sensor data.

[0019] In addition to the sensor data, information can also be obtained from a database in order to identify deviations in the static surroundings of the automated vehicle. For example, a construction site database can be used that includes information about ongoing and / or planned construction sites. Alternatively or additionally, at least one database that includes information about traffic flow can also be used.

[0020] The method for predicting changes in the static surroundings of an automated vehicle relative to a digital map therefore uses the aforementioned information or data and completes the aforementioned information or data by means of at least one change indicator in order to predict changes in the digital map. This makes it possible to significantly reduce the costs for change prediction information (e.g. traffic flow information) for external suppliers. The use of at least one change indicator offers the advantage that more reliable predictions can be made than predictions without the use of change indicators. The method is particularly suitable for map properties whose changes occur slowly and are observable, for example with respect to changes in road geometry or the construction of toll booths.

[0021] In one specific embodiment, a similarity measure is determined between the information stored in the digital map and the identified deviations. As a change indicator There are a large number of similarity measures available, with the aid of which the similarity between sensor data and / or information from a database and map attributes of a digital map can be determined, for example the Hausdorff metric can be determined as a similarity indicator.

[0022] In one specific embodiment, the at least one similarity indicator is determined based on the cloud or in the vehicle. The cloud-based approach advantageously ensures that the same similarity indicator is evaluated for all available data. In-vehicle approaches can be used, for example, if strict bandwidth restrictions apply within the context of data exchange with automated vehicles and, for example, sensor data cannot be sent to a cloud system outside the vehicle.

[0023] In one specific embodiment, the time evolution of at least one similarity indicator is taken into account when determining at least one probability of future changes in the surroundings of the automated vehicle relative to the digital map. To this end, the similarity indicator must be stored in a memory. The time evolution can be used as a reminder of possible upcoming changes. For example, if a new construction site is built, the similarity indicator may change in a way that has not been observed so far.

[0024] In one embodiment, at least one change indicator includes information about changes in the sensor data. For example, the change indicator may include information about quality degradation and / or disturbances in the sensor data. This advantageously enables sensor degradation to be taken into account when predicting changes.

[0025] In one embodiment, the probability of future changes in the surroundings of the automated vehicle relative to the digital map is determined by means of an algorithm for machine learning or a rule-based algorithm. This means that information about deviations and change indicators is used as input data for the algorithm. Rule-based approaches can also be referred to as classical approaches. Rule-based algorithms are trained with the aid of a set of rules. For example, these rules can be provided in the form of expert knowledge from a database or by another algorithm. Advantageously, the approach of the rule-based algorithm can be particularly simple to implement, thereby enabling the method to be carried out in a resource-efficient manner.

[0026] In one embodiment, an algorithm for machine learning is trained based on information about previous deviations, already determined change indicators, previous changes and already determined probabilities of the previous changes. Known algorithms for machine learning or rule-based algorithms can be used to determine the at least one probability. However, different training data than those mentioned can also be used. For example, information about previous deviations, already determined change indicators, previous changes and already determined probabilities of the previous changes related to other automated vehicles or fleets of automated vehicles can be used for training. This approach advantageously makes it possible to detect complex correlations and dependencies between different prediction sources and take them into account in the prediction.

[0027] In one specific embodiment, the method comprises the following additional method steps. At least one determined probability of a future change in the surroundings of the automated vehicle relative to the digital map is compared with at least one predefinable threshold value. If the determined probability is greater than the threshold value, at least one section of the digital map is deactivated or updated. By deactivating or updating the at least one section of the digital map, the error rate of the digital map can advantageously be reduced.

[0028] In one embodiment, the method includes the following additional method steps. An electronic horizon for the automated vehicle is provided based on a digital map including at least one section that is deactivated or updated. Advantageously, this method makes it possible to provide an error-free electronic horizon, since no segments of the digital map that no longer correctly represent the static surroundings due to changes are used for the generation of the electronic horizon, or updated sections of the digital map are used for the generation of the electronic horizon. In order to provide the electronic horizon, the position and orientation of the automated vehicle relative to the digital map must be known or determined. For this purpose, for example, known, so-called map matching methods can be used.

[0029] A system for predicting changes in a static environment of an automated vehicle relative to a digital map, wherein the digital map includes at least information about roadway guidance and static objects in the environment of the automated vehicle, the system comprising a detection device having at least one sensor for identifying deviations in the environment of the automated vehicle relative to the digital map; an analysis device for quantifying the identified deviations by determining at least one change indicator; and a prediction device for determining a probability of a future change in the environment of the automated vehicle relative to the digital map based on the identified and quantified deviations.

[0030] In one specific embodiment, the system has a memory for storing information about previous deviations, already determined change indicators, previous changes and already determined probabilities of the previous changes. This makes it more advantageous to train algorithms for machine learning.

[0031] In one specific embodiment, the system has a processing device for comparing the ascertained probability of a future change in the surroundings of the automated vehicle relative to the digital map with a predefinable threshold value. As a result, at least one section of the digital map that incorrectly represents the static surroundings of the automated vehicle can be deactivated or updated. It is possible to deactivate at least one section and update at least one other section.

[0032] In one embodiment, the system has a positioning device for determining the position and orientation of the automated vehicle and a horizon generator for generating an electronic horizon for the automated vehicle based on a digital map. The electronic horizon can be provided in particular based on a digital map including at least one section that is deactivated or updated. BRIEF DESCRIPTION OF THE DRAWINGS

[0033] The following is a detailed description of a method and a system for predicting changes in the static surroundings of an automated vehicle relative to a digital map with reference to a schematic drawing. The drawing shows:

[0034] Figure 1 : Method steps of a method for predicting changes of a static surrounding environment of an automated vehicle relative to a digital map; and

[0035] Figure 2 : Elements of a system for predicting changes of the static surroundings of an automated vehicle relative to a digital map. DETAILED DESCRIPTION

[0036] Figure 1Method steps 11 , 12 , 13 , 14 , 15 of a method 10 for predicting changes in the static surroundings of an automated vehicle relative to a digital map are schematically shown.

[0037] Automated vehicles are highly dependent on map-based information in order to support them primarily in terms of route planning, perception and situation understanding. In addition to increased accuracy, digital maps must also be very up-to-date, i.e. digital maps must have high accuracy both temporally and spatially.

[0038] The digital map includes at least information about lane guidance and static objects in the surroundings of the automated vehicle. The digital map may also include additional information. In particular, the digital map may have logical layers with different types of information.

[0039] For example, a digital map may have a planning layer. This planning layer contains information about how the exact lane geometry is constructed, terrain information, and connectivity information. This planning layer is used to plan the maneuvers. Keeping the automated vehicle on the lane. Several static surrounding objects can be taken into account in the safety-compliant planning layer.

[0040] The digital map can also have a positioning layer. The positioning layer contains recognizable objects for different sensor types, for example, the positioning layer can include video data and / or radar data and / or lidar data, which can be stored separately in the positioning layer. The current sensor information can be compared with the information from the positioning layer for the following purpose: in order to estimate the vehicle position.

[0041] The digital map may also have a dynamic layer. The dynamic layer contains dynamically changing information, such as road conditions, traffic flow, weather conditions and / or parking information. The dynamic layer is typically used in particular for comfort functions.

[0042] In a first method step 11 of method 10 , deviations from a digital map are detected in the surroundings of the automated vehicle based on sensor data of at least one sensor and / or at least one database, and the detected deviations are quantified by ascertaining at least one change indicator.

[0043] For example, a similarity measure such as the Hausdorff measure can be determined as a change indicator for the information stored in the digital map and the identified deviation. This can be done either cloud-based or in the vehicle. At least one change indicator can also include information about changes in the sensor data as an alternative or in addition.

[0044] In a second method step 12 , at least one probability of a future change in the surroundings of the automated vehicle relative to the digital map is ascertained based on the identified and quantified deviations. In this case, the temporal evolution of at least one similarity indicator can be taken into account.

[0045] For example, the probability of future changes in the surroundings of the automated vehicle relative to a digital map can be determined using an algorithm for machine learning or a rule-based algorithm, wherein the algorithm for machine learning can be trained prior to method 10 based on information about previous deviations, already determined change indicators, previous changes and already determined probabilities of these previous changes.

[0046] In an optional third method step 13, the at least one determined probability of a future change in the surroundings of the automated vehicle relative to the digital map can be compared with at least one predefinable threshold value. If the determined probability is greater than the threshold value, at least one section of the digital map can be deactivated or updated in an optional fourth method step 14. Deactivating at least one section of the digital map is to be understood as no longer taking into account the relevant section of the digital map for the automated control of the automated vehicle or for the automated control of the automated vehicle.

[0047] In a likewise optional fifth method step 15, an electronic horizon for the automated vehicle is provided based on the digital map comprising the at least one section that is deactivated or updated. The third, fourth and fifth method steps 13, 14, 15 may also be omitted.

[0048] Figure 2 Elements 21 , 22 , 23 , 24 , 25 , 26 , 27 of a system 20 for predicting changes in the static surroundings of an automated vehicle relative to a digital map are schematically shown.

[0049] The system 20 has a detection device 21, which has at least one sensor for detecting deviations from a digital map in the surroundings of an automated vehicle. The detection device 21 of the system can have at least one sensor inside the vehicle and / or at least one sensor outside the vehicle. Therefore, the detection device 21 is not limited to automated vehicles. For example, the sensor outside the vehicle can be designed as an infrastructure sensor, which is arranged in a fixed position in the area of ​​the infrastructure. However, the sensor outside the vehicle can also be a sensor of another vehicle, in particular another automated vehicle.

[0050] The system 20 further comprises an evaluation device 22 for quantifying the identified deviation by determining at least one change indicator. The detection device 21 and the evaluation device 22 are connected to a prediction device 23 of the system 20 and are designed to provide the identified deviation and the determined similarity indicator to the prediction device 23. The prediction device 23 is designed to determine the probability of a future change in the surroundings of the automated vehicle relative to the digital map based on the identified and quantified deviation. This can be done, for example, with the aid of an algorithm for machine learning or a rule-based algorithm.

[0051] The system 20 optionally has a memory 24. The memory 24 is provided for storing information about previous deviations, already determined change indicators, previous changes and already determined probabilities of the previous changes. The memory is connected to the prediction device 23. If an algorithm for machine learning is used for the prediction, the information stored in the memory 24 can be used by the prediction device 23 to train the algorithm for machine learning. Within the scope of this training, parameters are determined, which can be provided to the prediction device 23 in order to take these parameters into account in the prediction.

[0052] Optionally, the system 20 has a processing device 25 for comparing the ascertained probability of future changes in the surroundings of the automated vehicle relative to the digital map with a predefinable threshold value. In this way, reliable predictions about changes in the static surroundings can be made.

[0053] To provide an electronic horizon for an automated vehicle, the system 20 can additionally have a positioning device 26 for determining the position and orientation of the automated vehicle, and a horizon generator 27 for generating an electronic horizon for the automated vehicle based on a digital map including at least one section that is deactivated or updated. To this end, the horizon generator 27 is connected to the positioning device 26 and the processing device 25 so as to be able to receive information about the position and orientation of the automated vehicle relative to the digital map and about the deactivated and / or updated sections of the digital map. However, the positioning device 26 and the horizon generator 27 can be omitted, as can the memory 24.

Claims

1. A method (10) for predicting changes in the static surroundings of an automated vehicle relative to a digital map, wherein: The digital map comprises at least information about roadway guidance and static objects in the surroundings of the automated vehicle, wherein the method comprises the following method steps: - detecting deviations in the surroundings of the automated vehicle from the digital map based on sensor data of at least one sensor and / or at least one database, and quantifying the detected deviations by ascertaining at least one change indicator, Based on the identified and quantified deviations, at least one probability of a future change in the surroundings of the automated vehicle relative to the digital map is ascertained.

2. The method (10) according to claim 1, in, A similarity measure is ascertained as a change indicator between the information stored in the digital map and the detected deviation.

3. The method (10) according to claim 1 or 2, in, The at least one similarity indicator is ascertained in the cloud or within the vehicle.

4. The method (10) according to any one of the preceding claims, in, When ascertaining at least one probability of a future change in the surroundings of the automated vehicle relative to the digital map, the temporal evolution of the at least one similarity indicator is taken into account.

5. The method (10) according to any one of the preceding claims, in, The at least one change indicator includes information regarding a change in the sensor data.

6. The method (10) according to any one of the preceding claims, in, The probability of a future change in the surroundings of the automated vehicle relative to the digital map is ascertained using an algorithm for machine learning or a rule-based algorithm.

7. The method (10) according to claim 6, in, The algorithm for machine learning is trained based on information about previous deviations, already ascertained change indicators, previous changes and already ascertained probabilities of the previous changes.

8. The method (10) according to any one of the preceding claims, comprising the following additional method steps: - comparing at least one determined probability of a future change in the surroundings of the automated vehicle relative to the digital map with at least one predefinable threshold value, If the ascertained probability is greater than the threshold value, at least one section of the digital map is deactivated or updated.

9. The method (10) according to claim 8, comprising the following additional method steps: - providing an electronic horizon for the automated vehicle based on a digital map comprising at least one section which is deactivated or updated.

10. A system (20) for predicting changes in the static surroundings of an automated vehicle relative to a digital map, wherein: The digital map comprises at least information about lane guidance and static objects in the surroundings of the automated vehicle, wherein the system (20) has a detection device (21) having at least one sensor for identifying deviations in the surroundings of the automated vehicle relative to the digital map; an analysis device (22) for quantifying the identified deviations by determining at least one change indicator; and a prediction device (23) for determining the probability of future changes in the surroundings of the automated vehicle relative to the digital map based on the identified and quantified deviations.

11. The system (20) according to claim 10, A memory (24) is provided for storing information about previous deviations, already ascertained change indicators, previous changes and already ascertained probabilities of the previous changes.

12. The system (20) according to claim 10 or 11, A processing device (25) is provided for comparing the ascertained probability of a future change in the surroundings of the automated vehicle relative to the digital map with a predefinable threshold value.

13. The system (20) according to any one of claims 10 to 12, There is a positioning device (26) for determining the position and orientation of the automated vehicle, and a horizon generator (27) for generating an electronic horizon for the automated vehicle based on the digital map.

Images