Method for adapting an at least partially automated driving function for a vehicle

Abstract

The invention relates to a method (100) for adapting at least one at least partially automated driving function for a vehicle (12), in particular an at least partially autonomous vehicle, comprising the steps: - Providing (101) and / or detecting (101) environmental data, wherein the environmental data are specific to a, in particular direct and / or indirect, environment of the vehicle (12); - determining (102) a current and / or future situation, in particular traffic situation, of the vehicle (12) based on the environmental data; - determining (103) a safety criticality based on the at least one at least partially automated driving function and the current and / or future situation; - determining (104) at least one safety limitation for the at least one at least partially automated driving function based on the, in particular current, safety criticality; - adapting (105) the at least one at least partially automated driving function based on the at least one safety limitation.

Description

The invention is based on a method for adapting at least one at least partially automated driving function for a vehicle, in particular a vehicle that is at least partially autonomously driving.Prior ArtDE 10 2013 016 488 A1 discloses a driver assistance system which is designed to temporarily switch into a second operating mode when a triggering condition or at least one triggering condition of a plurality of triggering conditions is fulfilled, in which the control of the motor vehicle is carried out autonomously by the driver assistance system without the possibility of intervention by the driver, wherein the triggering condition is designed to evaluate at least the driver property data describing the prediction data and at least one driver property.Disclosure of the InventionAgainst this background, the approach presented here provides a method having the features of claim 1, an apparatus, a system and a computer program having the features according to the subordinate claims.The method for adapting at least one at least partially automated driving function for a vehicle, in particular a vehicle driving at least partially autonomously, has the following steps:providing and / or detecting environment data, wherein the environment data is specific to an, in particular direct and / or indirect, environment of the vehicle;determining a current and / or future situation, in particular traffic situation, of the vehicle based on the environmental data;determining a safety criticalness based on the at least one at least partially automated driving function and the current and / or future situation;determining at least one safety limit for the at least one at least partially automated driving function based on the, in particular current, safety criticalnessadapting the at least one at least partially automated driving function based on the at least one safety limit.The term "method for adjusting" may be understood as a method that aims to adjust one or more at least partially automated driving functions for a vehicle. The method comprises a plurality of steps starting from the provision and / or acquisition of environment data which are specific to the environment of the vehicle. The current and / or future situation of the vehicle is then determined on the basis of these environment data. Based on the at least one at least partially automated driving function and the current and / or future situation, a safety critical is determined. At least one safety limit for the at least one at least partially automated driving function is then determined on the basis of the current safety criticalness. Finally, the at least one at least partially automated driving function is adapted based on the at least one safety limit. The method for adapting can be calculated or generated by means of AI-based algorithms and can also take into account the preference of the driver of the vehicle. The adaptation of the driving function can take place directly via an output of corresponding control commands for an actuator system assigned to the driving function and / or indirectly via a change in a behavior plan assigned to the driving function.The term "at least one at least partially automated driving function" can be understood as a function that allows the vehicle to automatically perform certain driving tasks without the driver having to intervene manually. This function may be either fully automated or partially automated, which means that the driver still has control of the vehicle, but the system performs certain tasks, such as braking or accelerating. The function may also be limited, such as driving on the freeway or parking. The automation can be calculated or generated by AI-based algorithms and is controlled by the method for adapting the driving function based on environmental data, safety criticality and safety limitation.The term "in particular at least partially autonomously driving vehicle" can be understood as a vehicle that is capable of automatically executing specific driving functions without a human driver having to monitor or monitor them. The vehicle may be capable of self-accelerating, braking, and steering, but may still require human monitoring to ensure that it is safely and effectively driving on the road. The vehicle may also be able to detect and respond to certain traffic situations, such as stopping at a red traffic light or avoiding an obstacle. It is important to note, however, that the vehicle is not always fully autonomous and may still require human control, particularly in unpredictable situations or in areas where the technology is not yet mature.The term "providing" can be understood as the action in which the environment data is specifically captured or made available for an, in particular direct and / or indirect, environment of the vehicle. This can be done by various methods, such as sensors, cameras, or other sensing devices mounted in or on the vehicle. Making available can also be understood to mean that the data have already been recorded or even temporarily stored, in particular for a long time, and are subsequently available for the further method steps. The captured data can then be used for determining the current and / or future situation of the vehicle in order to adapt the at least one at least partially automated driving function. Providing the environmental data is thus an important step in the method for adapting the driving functions of the vehicle.The term "sensing" may be understood to mean collecting data or information about the environment of the vehicle. This can be done by various sensors or cameras, which capture specific features of the environment and convert them into digital data. The capturing may also include processing and interpreting these data to create an accurate representation of the environment. The aim of the detection is to obtain a current and accurate representation of the environment of the vehicle in order to enable a safe and effective adaptation of the automated driving functions.The term "environment data" may be understood as information collected about the environment of the vehicle to determine a current and / or future situation of the vehicle. This information may be from various sources, such as sensors, cameras, GPS systems, or other data sources. The environmental data may be specific to an immediate and / or indirect environment of the vehicle and may include various aspects of the environment, such as the location of other vehicles, traffic signs, road conditions, weather conditions, and obstacles on the road. The environmental data is used to enable accurate and comprehensive situation detection, which serves as a basis for the determination of the safety criticality and the safety limitation, in order to adapt the at least one at least partially automated driving function of the vehicle.The term "direct" may be understood to mean a physical vicinity or direct connection between the vehicle and its environment. It relates to objects or events that are in close proximity to the vehicle and thus may have a direct impact on the driving functions. For example, immediate environment data may include information about the condition of the road, obstacles, or other vehicles in the immediate vicinity of the vehicle.The term "indirect" may be understood as a type of environment data that is not directly captured by the vehicle, but is provided indirectly by other sensors or sources. This data can originate, for example, from traffic guidance systems, weather services or other vehicles and provide information about the traffic situation or other relevant factors which are relevant for the adaptation of the automated driving functions of the vehicle. In contrast, "immediate" environmental data refers to information directly captured by the vehicle itself, such as cameras, radar, or lidar sensors. The consideration of indirect environmental data can contribute to the vehicle having a more comprehensive and more accurate idea of its environment and thus being able to react more safely and effectively to changes in the traffic situation.The term "environment" may be understood as the area surrounding and in which the vehicle is moving. This area may include both the immediate environment of the vehicle and the immediate environment related to the traffic situation and other factors that may affect the vehicle. The environmental data provided and / or captured may include information about road conditions, weather, traffic density, position of other vehicles and obstacles, as well as other relevant factors important for adjusting the automated driving functions of the vehicle. The environment may also refer to the physical environment, including the road infrastructure, buildings, and landscape surrounding the vehicle. Collectively, the term "environment" refers to all factors that may affect the vehicle and its driving functions, and includes both the immediate and immediate environment of the vehicle.The term "determining" may be understood as the process of making a decision or result based on existing information or data. In the context of claim 1, the determining refers to the steps necessary to determine a current or future situation of the vehicle, including the traffic situation and the safety criticality. The determination also includes the determination of safety limits for the automated driving function based on the safety criticality and the adaptation of the driving function accordingly. The determination can be made by using algorithms, models or look-up tables and can be made in predetermined steps or continuously. It can also take into account the preference of the driver and adapt the driving function directly or indirectly.The term "current" may be understood as a description of the current state or situation of the vehicle. It refers to the environment and traffic situation surrounding the vehicle at that moment. The "current" situation can be determined instantaneously, preferably within 5 seconds, particularly preferably within 1 second, particularly preferably within 0.1 second and very particularly preferably within 0.01 second. It is important to note that the "current" situation may change continuously and therefore continuous monitoring and adjustment of the driving functions is required to ensure safe driving.The term "future situation" may be understood as a prediction of the environment in which the vehicle will be in the near future. This prediction is based on the acquired environment data and other relevant information such as the current speed of the vehicle. The future situation may include various aspects, such as the position and speed of other vehicles on the road, traffic signs and traffic lights, weather conditions, and road conditions. Predicting the future situation is important to appropriately adapt the at least partially automated driving function of the vehicle and to ensure that the vehicle can navigate safely and effectively.The term "traffic situation" can be understood as the current or future situation in road traffic, which relates to the vehicle and its environment. This may include, for example, the number and speed of other vehicles, the traffic rules and signs, the weather conditions, the road conditions, and other factors that may affect driving safety. The traffic situation may also include the position of the vehicle on the road, the direction in which it is travelling, and other relevant information relevant to the adaptation of the vehicle's automated driving functions. The determination of the traffic situation is based on the captured environmental data and can take place in real time in order to enable a rapid adaptation of the driving functions.The term "safety criticality" may be understood as an assessment of the potential risk emanating from a particular driving function based on the current and / or future situation of the vehicle and optionally the environment. Safety criticality is determined based on various factors, such as vehicle speed, distance to other vehicles or obstacles, weather conditions, and road conditions. The higher the safety criticality of a driving function, the greater the risk for accidents or other dangerous situations. It is therefore important to take account of the safety criticalness in the adaptation of automated driving functions and to define corresponding safety limits in order to minimize the risk.The term "safety limitation" can be understood as a restriction or limitation of the at least one at least partially automated driving function, which is determined based on the safety criticality. The safety limitation serves to minimize the risk of potential dangerous situations and to ensure the safety of the vehicle and its occupants. The safety limitation can take place in predeterminable or predetermined levels or continuously and can also be adapted according to a preference of the driver of the vehicle. The safety limitation can be adjusted directly by outputting corresponding control commands for an actuator system assigned to the driving function and / or indirectly by changing a behavior plan assigned to the driving function.The term "current" may be understood as a description of the current state or situation of the vehicle. It refers to the time at which the safety criticality and safety limitation are determined to adjust the at least one at least partially automated driving function. The term "current" may also mean that the safety limit is determined based on the current situation of the vehicle and the environment in which it is located. It can relate to a rapid and precise determination of the safety criticality and the safety limitation in order to enable a rapid adaptation of the driving function.The term "adjusting" may be understood to mean the change of an at least partially automated driving function based on the safety limit derived from the safety criticality and the current or future situation of the vehicle. The adaptation can take place directly by the output of control commands to the actuator system or indirectly by the change in the behavior plan of the driving function. The adaptation thus takes place in real time and enables a fast reaction to changing traffic situations and safety concerns. The safety limitation can take place in predeterminable or predetermined levels or continuously and can also take into account the preferences of the driver.Advantageously, the method according to the claim enables a situation-dependent limitation of automated driving functions depending on the current and / or future traffic situation and the safety criticality. In contrast to the prior art, in which the limitation of the automated driving functions takes place on the basis of the system state, the method enables a more precise and situation-dependent limitation, which is matched to the specific environmental data of the vehicle. This increases the safety of the vehicle and its occupants and reduces the risk of accidents.Further advantages are evident from the dependent claims.It is furthermore advantageous if the at least one at least partially automated driving function is calculated or generated by means of Kl-based algorithms.The term "AI-based algorithms." can be understood as a method of calculating or generating automated driving functions based on artificial intelligence (AI). Kl-based algorithms use machine learning and other technologies to learn from large amounts of data and recognize patterns. These patterns are then used to make decisions and take actions that are matched to the specific conditions of the environment of the vehicle. AI-based algorithms may also be able to improve and optimize themselves by continually collecting and analyzing new data. With respect to the described method, Kl-based algorithms can be used to adapt and improve the at least partially automated driving functions of the vehicle by reacting to the specific conditions of the environment of the vehicle. Advantageously, the feature of the claim that the at least one partially automated driving function is calculated or generated by means of AI-based algorithms enables a more precise and effective control of the vehicle. Compared to the prior art, where the driving functions are calculated in a conventional manner, the AI-based calculation may allow a faster and more accurate response to changes in the environment of the vehicle. Moreover, the AI-based calculation may also allow better adaptation to individual driving styles and preferences, which may result in an overall safer and more pleasing driving experience.In a further embodiment, it is provided that in the step of determining the current situation, the current situation is determined instantaneously, preferably within 5 seconds, particularly preferably within 1 second, particularly preferably within 0.1 second and very particularly preferably within 0.01 second.The term "determination of the current situation" can be understood as the step in the method for adapting at least one at least partially automated driving function for a vehicle, in which the current situation of the vehicle is determined. In this case, environment data which are specific to the direct and / or indirect environment of the vehicle are provided and / or acquired. The current situation of the vehicle is determined on the basis of these data, wherein the determination is preferably carried out within 5 seconds, particularly preferably within 1 second, particularly preferably within 0.1 second and very particularly preferably within 0.01 second. The current situation may include, for example, information about the position of the vehicle, the speed, the direction, the distance to other vehicles or obstacles, as well as other relevant factors. The determination of the current situation is an important step in the method, since it serves as a basis for the determination of the safety criticality and the safety limitation, which in turn are used for adapting the at least one at least partially automated driving function.The term "instantan" can be understood as a very rapid and immediate determination of the current situation. It means that the determination of the current situation takes place within a very short period of time, preferably within 5 seconds, particularly preferably within 1 second, particularly preferably within 0.1 second and very particularly preferably within 0.01 second. The term "instantan" implies that the determination of the current situation takes place in real time and that no or only a small delay is tolerated. It is important that the determination of the current situation takes place as quickly as possible in order to ensure an accurate and reliable adaptation of the at least partially automated driving functions.Advantageously, the feature of the claim enables a quick and precise response to critical situations. Compared to the prior art, in which the determination of the current situation is carried out on the basis of the system state, the feature enables a more accurate and situation-dependent limitation of AI outputs. This improves the safety of highly automated or autonomous systems (ADS), in particular in situations in which the ego vehicle is at small distances from other objects or road users.In a further embodiment, it is provided that in the step of determining the safety criticality, the safety criticality is determined via a predefinable or predefined model or a look-up table.The term "determination of safety criticality" can be understood as the process in which the potential risk or risk of a particular driving function is evaluated in a particular situation. This is done by using a predetermined model or look-up table that contains or computes the safety criticality for various situations and driving functions. The safety criticality is determined based on factors such as the traffic situation, the speed of the vehicle, and other relevant environmental data. Determining safety criticality is an important step in the method for adapting automated driving functions, as it ensures that the driving functions are activated to 100% only when they are safe and do not pose a risk to the driver or other road users.The term "predefinable model" can be understood as a predefined mathematical or statistical model that was created on the basis of data and empirical values. It may serve as a kind of template or reference to determine the security criticality in a particular context. The predefinable model can be based, for example, on historical data that were obtained from the analysis of accidents or other safety-critical events. It may also be based on simulations or tests performed to evaluate the performance of vehicles or driving functions in various situations. The predefinable model can serve as a basis for developing algorithms or decision rules that are used in adapting driving functions. It can also serve as a reference for evaluating the safety criticality in real time by comparing it with the current environmental data. Overall, the predefinable model enables a precise and reliable determination of the safety criticality, which in turn contributes to increasing the safety of autonomously driving vehicles.The term "predefined model" can be understood as a predefined mathematical or statistical model that is used to determine the safety criticality in the method for adapting at least one at least partially automated driving function for a vehicle. The model may be created based on data from various sources, such as sensors or cameras, and includes a set of parameters used to calculate safety criticality. The model can also contain a look-up table which enables an assignment of specific situations and driving functions to a specific safety criticality value. The predefined model can thus serve as a type of reference or standard for determining the safety criticality and enable a rapid and precise adaptation of the driving functions.The term "look-up table" may be understood as a data structure used to quickly and efficiently access predefined values. A look-up table consists of a list of input values and the associated output values. When a certain input value is searched in the table, the corresponding output value is directly retrieved from the table instead of being obtained by a complicated calculation. This allows for a fast and efficient determination of values, especially in applications where a large number of input values have to be processed. In the method described, the safety criticality is determined based on a predefined model or a look-up table, in order to enable a fast and efficient adaptation of the driving functions.Advantageously, the feature of the claim that the safety criticality is determined via a predetermined model or a look-up table enables a more precise and reliable determination of the safety criticality compared to the prior art. By using a predetermined model or look-up table, the safety criticality can be quickly and effectively determined without requiring complex calculations to be performed. This results in a faster reaction time of the system and thus increases the safety of the system as a whole. Compared to the prior art, in which the safety criticality is determined on the basis of the system state, the feature of the claim enables a more accurate determination of the safety criticality depending on the current situation or the scenario. This allows the system to respond better to unpredictable events and improve the overall safety of the system.In a further embodiment, it is provided that the model or the look-up table contains or calculates a value for the safety criticality for a plurality of situations, in particular traffic situations, and a plurality of at least partially automated driving functions.This step is important to determine an appropriate safety limit for the automated driving function and thus ensure the safety of the vehicle and its occupants.The term "model" may be understood as a mathematical or algorithm representation of a system or situation. In this case, the model refers to a representation of the traffic situations and the at least partially automated driving functions that calculates or contains a value for the safety criticality. The model may be based on various data sources, such as vehicle sensor data, traffic data, or weather data. It may also use various algorithms or methods to calculate safety criticality, such as machine learning or statistical analysis. The model may be continuously updated to take into account new data and situations and to improve the accuracy of the safety criticality.The term "look-up table" may be understood as a data structure used to quickly and efficiently access predefined values. A look-up table consists of a list of input values and the associated output values. When a certain input value is searched in the table, the corresponding output value is directly retrieved from the table instead of being obtained by a complicated calculation. This allows for rapid and efficient processing of data, especially in applications requiring a large number of input values to be processed. With respect to the described method, the look-up table is used to calculate or store the value for the safety criticality based on a plurality of traffic situations and at least partially automated driving functions.The term "plurality" may be understood as a number of at least two elements or objects. With respect to the described method, this means that the model or look-up table contains or calculates a value for the safety criticality for a plurality of situations and driving functions. This means that the model or look-up table is able to cover a wide range of situations and driving functions and is not limited to a limited number of scenarios. The use of a plurality of situations and driving functions in the modeling or look-up table enables a more precise determination of the safety criticality and thus a more precise adaptation of the automated driving functions.The term "situations" may be understood as a combination of factors relating to the environment of the vehicle and its movements. These factors may include, for example, the position and speed of other vehicles, traffic signs, road conditions, weather conditions, and other relevant information. The situations may include both current and future events and may change rapidly. The method for adapting the automated driving functions is based on the accurate determination of the current and future situations in order to determine an appropriate safety limit and to adapt the driving functions accordingly.The term "traffic situations" may be understood to mean the entirety of the circumstances and conditions that may occur in connection with traffic on a road or other traffic path. These include, for example, the number and type of vehicles, the speed and direction of the vehicles, the position and movement of pedestrians and cyclists, the weather conditions, the road conditions and traffic signs, and other traffic rules. A traffic situation can change rapidly and therefore requires continuous monitoring and adjustment of the driving functions of the vehicle in order to ensure safe and effective driving. The model or look-up table as described in claim 5 can contribute to evaluating the safety criticality in different traffic situations and to defining corresponding safety limits for the automated driving functions of the vehicle.The term "value" may be understood as a numerical or quantitative representation of a particular aspect or characteristic. In the context of the claims, the value refers to the safety criticality of an at least partially automated driving function in a particular traffic situation. The model or look-up table contains or calculates a value for the safety criticality, which is based on an assessment of the risks and hazards associated with the use of the driving function in the respective situation. The value can serve as a measure for the grading of the safety limit, which is carried out after the determination of the safety criticality. The term "includes" may be understood as an indication that the model or look-up table includes a plurality of situations and a plurality of at least partially automated driving functions for which a safety criticality value is calculated or present. It means that the model or look-up table contains a comprehensive collection of data and information relevant to a variety of situations and driving functions and which make it possible to determine the safety criticality. The term "includes" also implies that the model or look-up table may be continuously updated and augmented to accommodate new situations and driving functions and improve the accuracy of the safety criticality.The term "calculated." may be understood to mean performing mathematical operations or algorithms to determine a value or quantity. In the context of the claims, the calculation refers to the determination of the safety criticality for a particular traffic situation and a particular automated driving function. The model or look-up table contains or calculates a value for safety criticality based on various factors such as the speed of the vehicle, the distance to other vehicles or obstacles, and other relevant environmental data. The calculation is generally automatic and can be supported by using Kl-based algorithms.Advantageously, the feature of the claim that the model or the look-up table contains or calculates a value for the safety criticality for a plurality of situations, in particular traffic situations, and a plurality of at least partially automated driving functions, enables a more precise limitation of the AI output depending on the current situation. Compared with the prior art, in which the limitation of the AI output is performed on the basis of the system state, this feature enables more accurate and situation-dependent limitation, resulting in higher safety and reliability of the system. Also, by using a model or look-up table for a plurality of situations, the system can better respond to unpredictable situations and make appropriate decisions.In a further embodiment, it is provided that, in the step of determining the at least one safety limit, the safety limit is determined via a predefinable or predefined assignment to the value for the safety criticality, in particular on the basis of the predefinable or predefined model or the look-up table, in particular according to the embodiment described above.The term "determination" may be understood as the process of setting or determining a particular value, property, or state based on existing data or information. In the context of the claims, the determination refers to the steps of the method for adapting automated driving functions, in which the current and future situation of the vehicle as well as the safety critical are determined in order to define a safety limit for the driving functions. The determination is made on the basis of environmental data, a predefined model or a look-up table and can be made in predefinable or predefined steps or continuously. The term "predeterminable assignment" can be understood as a predefined relationship between a specific value or a specific variable and another variable or another value. This relationship is set in advance and may be in the form of a table, a diagram, or a mathematical formula. In this context, the safety limit is determined on the basis of such a predefinable mapping by comparing the value for the safety criticality with the corresponding safety limit from the predefined mapping. This enables a quick and effective adaptation of the automated driving functions of the vehicle in order to ensure safety.The model or the table contains or calculates a value for the safety criticality for a plurality of situations, in particular traffic situations, and a plurality of at least partially automated driving functions. This enables a precise determination of the safety limitation in the step of determining the at least one safety limitation in the method according to the invention. Advantageously, the method according to this embodiment enables a precise determination of the safety limitation for AI-based function modules in highly automated or autonomous systems. By using a predefinable or predefined assignment to the value for the safety criticality, in particular on the basis of the predefinable or predefined model or the look-up table according to this embodiment, the safety limitation can be determined depending on the current situation or the scenario. This enables situation-dependent limitation of AI-based actuator interventions, which leads to increased safety in highly automated or autonomous systems.In a further embodiment, it is provided that the safety limitation takes place in predeterminable or predetermined steps or continuously.The term "predeterminable steps" can be understood as a possibility of adapting the safety limitation in predefined steps or intervals. This means that the safety limitation is not adjusted continuously, but in discrete steps, which can be defined in advance. For example, the safety limit could be adjusted in steps of 10%, which means that the driving function is restricted in 10% steps to ensure safety. Alternatively, the steps could be determined at other intervals depending on the requirements of the specific application. The use of predeterminable steps enables a precise and predictable adaptation of the driving function in order to ensure the safety of the vehicle and of its occupants.The term "predetermined levels" may be understood as a method for determining safety limits for the automated driving functions of a vehicle, in which the limits are adjusted at predefined steps or intervals. This means that the boundaries are not adjusted continuously, but at certain steps or intervals that have been determined in advance. For example, a safety limit for an automated driving function could be adjusted in steps of 5 km / h, so that the function is limited by 5 km / h when each step is reached. This limiting method enables a precise and predictable adaptation of the automated driving functions and can contribute to increasing the safety of the vehicle.The term "continuously" can be understood as a type of limitation in which there are no fixed steps or steps, but the limitation can be adapted continuously and without interruption. In other words, this means that the limitation is not limited to specific values or intervals, but rather that it can be adapted in any possible form. For example, the safety limit for an automated driving function could be continuously adjusted to allow finer control over the driving function and to ensure higher safety.Advantageously, the feature of the embodiment allows the safety limitation to take place in predeterminable or predetermined steps or continuously, a more precise and more flexible control of the AI-based actuator interventions depending on the current situation. This allows for higher system safety and effectiveness, especially in critical situations where fast and precise decisions are required.In a further specific embodiment, it is provided that the safety limitation takes place according to a preference of a driver of the vehicle.The term "preference" may be understood as an individual preference or inclination of a person relating to a particular decision or action. In the context of this embodiment, the preference refers to the safety limit in adapting the automated driving functions of a vehicle. This means that the safety limit is adjusted according to the preferences or wishes of the driver of the vehicle. This may mean, for example, that the driver prefers a higher safety margin to ensure a higher level of safety, or that he prefers a lower safety margin to allow for a faster or more efficient trip. The preference of the driver may be determined by manual adjustment or by automatic adjustment based on previous decisions of the driver.The term "driver" may be understood to mean a person controlling the vehicle and having responsibility for safe and proper operation of the vehicle. The driver may be either a natural person that manually controls the vehicle or an artificial intelligence that autonomously controls the vehicle. In either case, the driver is responsible for ensuring that the vehicle complies with the applicable traffic rules and regulations and that the safety limits are set according to the driver's preferences. The driver may also be responsible for monitoring the automated driving functions and intervene where appropriate to ensure that the vehicle is operated safely and effectively.The term "vehicle" may be understood as a motorized means of travel that may travel on wheels or other means of travel such as rails or tracks. A vehicle can be used for transporting people or goods and can be in various sizes and shapes, from small cars to large trucks or trains. A vehicle may be controlled by a driver or autonomously, and may have various types of drives, such as an internal combustion engine, an electric motor, or a hybrid drive. A vehicle may also be equipped with various sensors and cameras to sense its environment and adjust its driving functions.Advantageously, the feature of the embodiment that the safety limitation is performed according to a preference of the driver of the vehicle enables individual adaptation of the safety limitation to the needs and preferences of the driver. The individual adaptation allows the driving experience to be improved since the driver has more control over the safety limit and can feel more secure. Moreover, the individual adaptation can contribute to the vehicle being better matched to the requirements of the driver and thus being able to be operated more safely and effectively overall.In a further embodiment, it is provided that, in the step of adapting the at least one at least partially automated driving function, the driving function is adapted directly via an output of corresponding control commands for an actuator system assigned to the driving function and / or indirectly via a change in a behavior plan assigned to the driving function. In particular, it can be provided here that the driving function is adapted via a change of input data into the behavior plan assigned to the driving function.The term "automated driving function" may be understood as a function that allows a vehicle to perform certain driving tasks without human intervention. This function may be partially or fully automated and typically includes control of the vehicle's throttle, brake, and steering. Automated driving functions can operate at different levels of automation, from simple assistance systems to fully autonomous driving. Such functions may be computed or generated by AI-based algorithms and often require the collection and processing of environmental data to ensure safe and effective travel. Within the scope of the described method, automated driving functions are adjusted directly via an output of corresponding control commands for an actuator system assigned to the driving function and / or indirectly via a change of a behavior plan assigned to the driving function.The term "driving function" may be understood as a function or behavior performed by a vehicle to control or move it. A driving function may include, for example, accelerating, braking, steering, or shifting the vehicle. It may also be a combination of various functions that cooperate to control the vehicle, such as automatic parking or lane keeping. The driving function can be adapted directly via an output of corresponding control commands for an actuator system assigned to the driving function or indirectly via a change in a behavior plan assigned to the driving function.The term "output" may be understood as the process of sending a signal or instruction to a device or component to perform a particular action. In the context of the described method for adapting an at least partially automated driving function for a vehicle, the output relates to the transmission of control commands to the actuator system which is connected to the corresponding driving function. These control commands can be sent directly to the actuator system in order to enable a direct adaptation of the driving function, or indirectly via a change in the behavior plan assigned to the driving function. The output is thus an essential step in the adaptation process, which makes it possible to adapt the driving function of the vehicle according to the environmental conditions and the safety criticality.The term "control commands" may be understood as instructions that are sent to the actuators of a vehicle in order to adapt a specific driving function. These instructions may relate, for example, to the speed, the steering or the brakes and are generally generated by a computer or a control unit. The control commands can be sent directly to the actuator system in order to enable a direct adaptation of the driving function, or indirectly via a change in the behavior plan, which then controls the actuator system accordingly. The control commands must be precise and reliable in order to ensure a safe and effective adaptation of the driving function.The term "actuator system" can be understood as the system of actuators that are responsible for converting control commands into movements or actions.Actuators may be, for example, electric motors, hydraulic cylinders or pneumatic cylinders, which are controlled by electrical or hydraulic signals. Actuator system is thus an important component of automated systems, since it enables the conversion of control commands into actual movements or actions. In the context of the described method for adapting automated driving functions for vehicles, the actuator system is responsible for converting the corresponding control commands for adapting the driving functions into actual movements of the vehicle.The term "behavior planning" may be understood as a process in which a series of decisions are made to determine the behavior of a vehicle in a particular situation. This process takes into account various factors such as the environment of the vehicle, the current traffic situation, and the goals of the vehicle. The behavior planning can directly or indirectly influence the adaptation of the automated driving functions by changing the control commands for the actuator system of the vehicle or the behavior planning itself. Behavior planning is an important part of the method for adapting the automated driving functions and contributes to the vehicle acting safely and effectively on the road.The feature of the claim advantageously enables the adaptation of the driving function to take place directly via an output of corresponding control commands for an actuator system assigned to the driving function and / or indirectly via a change in a behavior plan assigned to the driving function. This leads to a more precise and faster adaptation of the driving function to the current situation, since the actuator system and behavior planning can be directly influenced. Compared to the prior art, in which the adaptation of the driving function may take place via several steps, this is a distinct advantage in terms of the safety and efficiency of the system.The advantages described above also apply in a corresponding manner to a device, in particular a control unit for adapting at least one at least partially automated driving function for a vehicle, in particular a vehicle which is driving at least partially autonomously, which is configured to carry out the method according to one of the embodiments described above.The term "device" can be understood as a technical unit which is specifically designed to carry out the described method for adapting at least one at least partially automated driving function for a vehicle. The device may be embodied as a control unit or as part of a control unit and may contain various components such as processors, memories, sensors, actuators and communication interfaces. The device may be integrated into the vehicle or may be embodied as a separate device that is connected to the vehicle. The apparatus may also be capable of receiving and processing data from other systems or devices to ensure accurate adjustment of the driving functions.The term "control unit" may be understood as an electronic component capable of executing the method for adapting at least one at least partially automated driving function for a vehicle. The control unit may control and monitor various sensors and actuators of the vehicle to adjust the driving functions according to the environmental data and the current and / or future situation of the vehicle. The control unit can also be equipped with AI-based algorithms to calculate and generate the driving functions. It may be capable of determining the safety criticality and safety limits and of correspondingly adapting the driving functions. The controller may be implemented as part of a larger system or as a stand-alone device in a vehicle.The term "automated driving function" may be understood as a function that allows a vehicle to perform certain driving tasks without human intervention. These functions can range from simple assistance systems such as automatic braking or parking to fully autonomous driving functions that the vehicle is able to navigate independently and make decisions. Automated driving functions may be calculated or generated by Kl-based algorithms and typically require a plurality of sensors and actuators to sense and react to the environment of the vehicle. The adaptation of these functions is based on the evaluation of the current and future situation of the vehicle and the safety criticality in order to ensure safe travel. The device, in particular control unit, which is capable of executing the method for adapting the automated driving functions can be implemented as part of an autonomously driving vehicle or as a retrofitting for an existing vehicle.By using this device, the safety criticality of Kl-based function modules in highly automated or autonomous systems (ADS) can be improved. The device is capable of limiting the outputs of function modules that were calculated by means of AI-based algorithms as a function of the current situation or the associated safety criticality. In contrast to the prior art, in which the limitation of the AI output takes place on the basis of the system state (e.g. speed), the limitation in the invention takes place on the basis of the situation or the scenario. This increases the safety of the vehicle and the occupants.The advantages described above also apply in a corresponding manner to a system for adapting at least one at least partially automated driving function for a vehicle, in particular a vehicle which is driving at least partially autonomously, comprising at least the vehicle and the device according to one of the embodiments described above.The term "system" may be understood as a combination of various components that cooperate to perform a particular function. The system operates to ensure safe and effective driving by collecting the environmental data, determining the current and future situation of the vehicle, assessing safety criticality, and adjusting the automated driving functions accordingly.The advantages described above also apply in a corresponding manner to a computer program comprising instructions which, when the computer program is executed by a computer or by an apparatus according to one of the embodiments described above, cause the computer program / s to execute the method according to one of the embodiments described above.The term "computer program" may be understood as a collection of instructions executable by a computer or apparatus to perform a particular task. These instructions may be written in a programming language and are typically stored in a file. The computer program may perform various functions such as data processing, computations, control of hardware, or execution of algorithms. It may also be referred to as software and may be executed on various platforms such as desktop computers, mobile devices, or embedded systems.The invention likewise relates to a computer-readable storage medium which comprises the computer program. The storage medium is embodied, for example, as a data memory such as a hard disk and / or a nonvolatile memory and / or a memory card. The storage medium may be integrated into the computer or an apparatus according to any of the embodiments described above, for example.BRIEF DESCRIPTION OF THE DRAWINGExemplary embodiments of the invention are schematically illustrated in the drawings and explained in more detail in the following description. The same reference numerals are used for the elements shown in the different figures and acting in a similar manner, wherein a repeated description of the elements is omitted.The following are shown: FIG. 1 shows a schematic illustration of a method, a device, a vehicle, a system and a computer program for adapting at least one at least partially automated driving function for a vehicle according to exemplary embodiments; and FIG. 2 shows a schematic illustration of a method for adapting at least one at least partially automated driving function for a vehicle according to a further exemplary embodiment.As already explained above, the present invention describes a method, a device, a system and a computer program which make it possible to limit or restrict an at least partially automated driving function for a vehicle, in particular for at least partially autonomously driving, in a situation-related manner, in order, for example, to give a driver of the vehicle more room for action in the respective situation and / or for the respectively activated driving function.FIG. 1 illustrates, according to exemplary embodiments of the invention, a method 100, a device 10 and a system 30 for adapting at least one at least partially automated driving function for a vehicle 12 which is in particular at least partially autonomously driving.According to a first method step 101, environment data can be provided and / or acquired. The environment data can be specific to an, in particular direct and / or indirect, environment of the vehicle. Subsequently, according to a second method step 102, a current and / or future situation, in particular traffic situation, of the vehicle 12 can be determined on the basis of the environmental data. According to a third method step 103, a safety criticalness can then be determined based on the at least one at least partially automated driving function and the current and / or future situation. According to a fourth method step 104, at least one safety limit for the at least one at least partially automated driving function can then be determined on the basis of the, in particular instantaneous, safety criticality. Subsequently, according to a fifth method step 105, the at least one at least partially automated driving function can be adapted based on the at least one safety limit.In this case, the method steps 101- 105 can be carried out by a device 10. The device 10 is designed, for example, as a computer and / or as a device 10 for data processing and can comprise means for executing the steps of a method 100 according to exemplary embodiments of the invention. The device may be disposed or integrated in the vehicle 12. Alternatively, the device 10 can be designed as a vehicle-external device which is connected to the vehicle 12 by signal technology. Furthermore, the device 10 can have a communication interface for, in particular, wireless networking with further devices, units, vehicles or the like. Furthermore, the apparatus 10 can have a computer program 20 according to exemplary embodiments of the invention. The computer program 20, when executed by a computer or the device 10, can cause it / s to execute the steps of the method 100 according to embodiments of the invention. Together, the vehicle 12 and the device 10 form a system 30 for adapting at least one at least partially automated driving function for a vehicle 12, in particular a vehicle 12 which is driving at least partially autonomously.FIG. 2 shows a schematic illustration of a method 200 for adapting at least one at least partially automated driving function for the vehicle 12 according to a further exemplary embodiment. A surroundings model 210 of the vehicle 12 is generated via provided surroundings data 202 and / or surroundings data acquired by means of a sensor system. This environment model 210 is used both for the behavior planning 220 of the vehicle 12 or its driving functions, in particular at least partially automated driving functions, and for the determination 240 of a current and / or future situation, in particular traffic situation, of the vehicle 12. For determining 240 the current and / or future situation, in particular traffic situation, of the vehicle 12, V2X data 244 and / or map data 242 from the environment, in particular the immediate and / or indirect environment, of the vehicle 12 can optionally be used. Furthermore, a trajectory planning 230, in particular an AI-based trajectory planning, takes place from the behavior planning 220. Depending on the results of the trajectory planning 230 and of the determination 240 of the current and / or future situation of the vehicle 12, the at least one at least partially automated driving function is limited 250. This can be effected, for example, by means of an output signal (indicated by the arrow 252) to a corresponding unit.This is explained in more detail below by way of example. In order to improve the passage of circular traffic using a (partially) automated driving assistance system, in a first step, environment data can be recorded and evaluated. As a result, entry of the host vehicle into a circular traffic can be detected. In this case, it is possible to use both surroundings perceptions and data from a digital map. Furthermore, it is determined whether and if yes which driving function is activated. In this case, this can be an ACC function (distance-controlled cruise control), wherein the host vehicle is oriented at a lead vehicle traveling ahead. Furthermore, a safety criticalness is established for this combination of "ACC function" and "travel through circular traffic". This may be, for example, a medium criticality. In the next step, an acceleration ramp of the own vehicle is lowered or kept at a constant level according to the determined mean criticality in order to prevent the vehicle from losing the lead vehicle from its (sensor) time sequence and driving too quickly through the circular traffic. Finally, the original acceleration value is restored as soon as the vehicle leaves the circumcircle. In this case, it is possible to use environmental perceptions and map information again in order to keep the changing traffic situation in view and, if appropriate, to react appropriately.Such a system can be used in a wide variety of vehicles with a (partially) automated driving assistance system in order to increase the safety in road traffic and to make the driving functions more effective.The method according to the invention is also described once again in other words. The method relates to the safety criticalness of in situative scenarios in which the ego vehicle has small distances to other objects or road users. The method comprises a sequence for determining the situation, determining the safety limits and imposing these limits. Determination of the situation: The situation or the stay in a situation is determined securely by the ADS (Automated Driving System). This can be done by various methods, such as, for example, internally to the vehicle via localization on an HD map with semantic information about zones and their properties, on the basis of the internal environment model from an exteroceptive sensor system or by a dedicated module for situation recognition on the basis of the environment sensor system and / or localization. The situation can also be determined externally to the vehicle via local infrastructure or the central traffic / fleet management.Description and classification of the situations: The relevant situations to be distinguished and described can be described and classified on the basis of PEGASUS ontology. In this case, different parameters such as the presence of VRUs (Vulcanizable Road Users), the width of roads, the distance to dangerous obstacles, turning situations or scenarios and the following travel with a small distance can be differentiated. The classification may also include the NHTSA pre-crash scenarios.Determination of the safety limits: The current safety limits are derived on the basis of models for deriving potentially dangerous interventions in specific situations. This can relate to primary control variables such as trajectory, steering, braking, acceleration or door opening, but also to metadata which can influence the later decision about interventions, such as, for example, the maximum confidence level of the result of a free space detection. Implementation of the Safety Limits: The Safety Limits are implemented by limiting the outputs of the Kl-based function modules accordingly. This can be done either by a hard limitation or by the safety limits being transferred to the functional module and already being taken into account there in the regulation and result finding. Alternatively, it is also possible to select from different AI paths.References included in the specificationThis list of documents cited by the applicant has been produced in an automated manner and is only included for the better information of the reader. The list is not part of the German patent application or utility model application. The DPMA does not take any adhesion for any faults or omissions.Patent Literature citedDE 10 2013 016 488 A1

[0002]

Claims

Method (100) for adapting at least one at least partially automated driving function for a vehicle (12), in particular for at least partially autonomously driving, having the steps: - providing (101) and / or capturing (101) environmental data, wherein the environmental data are specific to an, in particular direct and / or indirect, environment of the vehicle (12); - determining (102) a current and / or future situation, in particular traffic situation, of the vehicle (12) on the basis of the environmental data; - determining (103) a safety criticalness on the basis of the at least one at least partially automated driving function and the current and / or future situation; - determining (104) at least one safety limit for the at least one at least partially automated driving function on the basis of the, in particular current, safety criticalness; adapting (105) the at least one at least partially automated driving function based on the at least one safety limit.Method (100) according to Claim 1, characterized in that the at least one at least partially automated driving function is calculated or generated by means of Kl-based algorithms.Method (100) according to Claim 1 or 2, characterized in that, in the step of determining (102) the current situation, the current situation is determined instantaneously, preferably within 5 seconds, particularly preferably within 1 second, particularly preferably within 0.1 second and very particularly preferably within 0.01 second.Method (100) according to one of Claims 1 to 3, characterized in that, in the step of determining (103) the safety criticality, the safety criticality is determined via a predefinable or predefined model or a look-up table.Method (100) according to Claim 4, characterized in that the model or the look-up table contains or calculates a value for the safety criticality for a plurality of situations, in particular traffic situations, and a plurality of at least partially automated driving functions.Method (100) according to either of Claims 4 and 5, characterized in that, in the step of determining (104) the at least one safety limit, the safety limit is determined by means of a predeterminable or predetermined assignment to the value for the safety criticality, in particular on the basis of the predeterminable or predetermined model or the look-up table according to Claim 5.Method (100) according to one of the preceding claims, characterized in that the safety limitation takes place in predeterminable or predetermined steps or continuously.Method (100) according to one of the preceding claims, characterized in that the safety limitation takes place according to a preference of a driver of the vehicle (12).Method (100) according to one of the preceding claims, characterized in that, in the step of adapting (105) the at least one at least partially automated driving function, the driving function is adapted directly via an output of corresponding control commands for an actuator system assigned to the driving function and / or indirectly via a change in a behavior plan assigned to the driving function.Method (100) according to Claim 9, characterized in that the driving function is adapted via a change in input data in the behavior plan assigned to the driving function.Device (10), in particular a control unit for adapting at least one at least partially automated driving function for a vehicle (12), in particular driving at least partially autonomously, which is configured to carry out the method (100) according to one of Claims 1 to 10.System (30) for adapting at least one at least partially automated driving function for a vehicle (12), in particular for at least partially autonomously driving, comprising at least the vehicle (12) and the device (10) according to claim 11.A computer program (20) comprising instructions which, when the computer program (20) is executed by a computer or by an apparatus (10) according to claim 11, cause the computer program / s to execute the method (100) according to any one of claims 1 to 10.

Images