Method for providing an adaptation notification to an ego road user to be notified, method for determining a swarm prediction variable of a movement path, and method for determining global swarm movement data

Abstract

The invention relates to a method for providing, by means of a warning device (12), an adaptation notification, preferably a warning notification, to at least one ego road user (10) to be notified to adapt at least one driving function to at least one surroundings road user (32) located and / or moving in the surrounding region (2) of the road user, said warning device being provided with respective movement paths (PCL, PR) which are predicted for the ego road user and for the at least one surroundings road user (32) and along which the movements of the road users (10, 32) are expected. The warning device (12) determines, on the basis of the predicted movement paths (PCL, PR), at least one notification variable, which is characteristic of a provision of the adaptation notification and which is preferably characteristic of whether an adaptation notification (I1, A32) is to be provided and / or at what time the adaptation notification (I1, A32) is to be provided. According to the invention, the warning device (12) determines the notification variable on the basis of at least one swarm prediction variable, in particular at least one statistical swarm prediction variable, for determining the prediction quality of at least one of the predicted movement paths (PCL, PR), wherein the at least one swarm prediction variable is determined on the basis of global swarm movement data of a plurality of road users, said movement data being generated in a plurality of different traffic regions (2).

Description

[0001] Description Method for providing an adaptation notification to an ego road user to be notified, method for determining a swarm prediction variable of a movement path and method for determining global swarm movement data The present invention relates to a method for providing an adaptation notification to at least one ego road user to be notified, method for determining at least one swarm prediction variable of a predicted movement path and method for determining global swarm movement data.The present invention further relates to a warning device for an ego road user for providing an adaptation notification, a swarm data acquisition device for a road user, a map generation device, and a machine-readable and, in particular, computer-implementable movement model for predicting a movement path of a road user. Vehicles with communication devices that can communicate locally with other vehicles in the vehicle's surroundings are known from the prior art. A technology preferably used for this purpose is known as Car2Car communication (also abbreviated as C2C communication, i.e., in German, "car-to-car" communication) or V2V communication (abbreviation for the term Vehicle-to-Vehicle Communication, which is often used in the Anglo-Saxon world).V2V communication is a special case of V2X communication (Vehicle-to-X communication, i.e. "Vehicle-to-everything" communication), also known under the term Car2x communication (for "Car to x"). This generally refers to the communication between a vehicle and its environment, such as neighboring vehicles or infrastructure facilities. Communication between vehicles is currently used, for example, to warn each other of dangerous situations in which the driver would be unable to react at all or would only react very late without a warning. For example, this can be used to warn drivers approaching a traffic jam. This communication technology is also advantageously used to warn drivers of a special emergency vehicle, such as an ambulance, approaching (usually from behind at high speed).In addition to traffic jams, intersections also pose an increased risk of collision with another road user compared to intersection-free lanes. The goal is to reduce the risk of accidents in such traffic situations by warning the driver as early as possible. At the same time, however, the driver of an ego vehicle should be prevented from receiving false warnings, for example, regarding a collision when this dangerous situation would not occur even without a warning.If, for example, another road user approaching the intersection reaches the intersection point where the movement paths of the other road user and the ego vehicle intersect, at a sufficient time distance from the ego vehicle so that both road users can pass the intersection without interference from each other, the driver of the ego vehicle should ideally not be unnecessarily warned of a potential collision. This represents a complex problem in that the warning depends on the movement behavior of several road users, which must be predicted for this purpose. Because the warning function must determine with sufficient lead time whether the situation is becoming critical, a prediction of the road users involved over a corresponding period of time is necessary. However, the further into the future the objects are predicted, the lower the confidence of the prediction.One goal in the development of such a function is to specify a probability for the actual encounter or the determined distance when both road users pass the intersection. The warning should only be displayed when the probability of a critical encounter at the intersection exceeds a specified value. The more reliable and less error-prone the prediction is, the earlier the warning can be generated with the same false alarm rate (“FP”). However, the exact target probability to which the function is parameterized depends on the design of the function and must take customer requirements into account. The uncertainty of the prediction itself arises from the movement behavior of the road users at the intersection (particularly due to deceleration and acceleration processes) and is dependent on various factors, such as:The layout of an intersection, the presence of other road users on site, or even cultural differences. CN 116129650 B discloses a traffic early warning system in which an initial traffic flow of a road section, a vehicle number change condition at an intersection section, and a vehicle change rate at the corresponding intersection section are calculated. Based on the vehicle change rate at the corresponding intersection section and the road capacity of each road section, a congested road section and the congestion time are determined for each road section. Furthermore, an early warning is issued to a target vehicle based on the traffic congestion information and the travel time required by the target vehicle to pass a target traffic light.While this traffic early warning system can reduce the risk of an accident by offering a driver warned of a congested road section by the traffic early warning system the opportunity to avoid the congested road section, it does not offer the possibility of reducing the risk of collision, for example, with vehicles whose movement paths converge at an intersection. WO 2022 / 110611 A1 discloses a method for predicting the crossing behavior of a pedestrian at an intersection. For this purpose, the development of a prediction model based on a neural network is proposed, which is trained using deep reinforcement learning. This predicts discrete behavioral states of the pedestrian when crossing a street, such as walking, running, and stopping, and provides an early hazard warning for crossing pedestrians and passing vehicles.The reward used is a modified time to collision (MTTC), which takes into account the relative position, relative speed, and relative acceleration between the vehicle and a pedestrian. This takes into account that when a vehicle encounters a crossing pedestrian at an intersection, the vehicle decelerates to slow down or accelerates to pass. The disadvantage of this method is that it is only applicable to this one specific pedestrian intersection, where the corresponding sensors are installed on the side of the road and for which the training procedure was carried out. If, for example, a prediction of a pedestrian's crossing behavior at other pedestrian intersection areas is to be made, a correspondingly complex setup and a correspondingly complex, new training procedure are required.The present invention is based on the object of overcoming the disadvantages known from the prior art and of providing methods and devices by means of which a road user can be warned as early as possible and, at the same time, can be prevented as far as possible from being warned in the event of an essentially non-critical encounter at an intersection point. The present invention is further based on the object of providing a method and devices by means of which, by predicting the movement pattern of two road users whose movement behavior potentially influences one another and whose current movement behavior could, for example, cause a potentially critical traffic situation, the most precise and valid statement possible can be made with regard to a possible adjustment of the movement behavior of at least one of the road users involved.In particular, this information should be provided as early as possible. According to the invention, this object is achieved by the subject matter of the independent claims. Advantageous embodiments and further developments of the invention are the subject matter of the dependent claims. In an inventive (preferably computer-implemented) method for providing an adjustment notification (in particular for output and / or transmission and / or provision) to (orfor) at least one ego road user to be notified (in particular to an ego vehicle) for adapting at least one driving function and / or vehicle function to at least one surrounding road user located and / or moving in an area surrounding the ego road user, by means of a warning device, predicted movement paths are and / or are provided to the warning device for the ego road user and preferably for the at least one surrounding road user (in each case), along which movement of the road user is preferably to be expected. The method can be a computer-implemented method. However, it is also conceivable that the method partially comprises computer-implemented method steps and furthermore also, for example, an output of the adaptation notification by means of an output device to a person (such as a user / driver).In other words, a predicted movement path for the ego road user is and / or will be provided to the warning device, along which movement of the ego road user is to be expected. Preferably, a predicted movement path for the surrounding road user is and / or will be provided to the warning device, along which movement of the surrounding road user is to be expected. The warning device provides the adaptation notification, which is to be transmitted and / or output and / or made available to at least the ego road user. A driving function and / or a vehicle function can be understood, for example, as a V2X awareness function and / or a (V2X) safety function. “Adaptation of at least one driving function” can also be understood – for example in the case of pedestrians – as an adaptation of a movement (of the pedestrian), such as a change in the speed of travel.A "driving function" is understood in particular to mean a function for influencing the longitudinal and / or lateral movement and preferably a longitudinal and / or lateral guidance (of a road user such as a vehicle), such as accelerating, decelerating, swerving, and / or changing a route and / or aborting and / or accelerating an overtaking maneuver. A "driving function" can also be understood in particular to mean a vehicle function that is executed and / or performed during a movement of the road user, such as an (acoustic and / or visual) warning function, such as activating a horn and / or flashing headlights, of a road user in the surrounding area, or a lighting function of the road user, such as (road) illumination (adapting the lighting or illumination to the surrounding road user, for example, to prevent glare).The driving function is preferably a (preferably V2X-based) intersection assistance function that warns of VRUs, in particular cyclists, on a collision course. The driving function is preferably a driver assistance function or mobility assistance function and / or a safety function of a vehicle and / or a mobility aid. The adaptation of the at least one driving function (and / or vehicle function) preferably takes place in the ego road user and / or in the surrounding road user. It is also conceivable that a relative movement behavior of the ego road user to the surrounding road user is changed and / or adapted by adapting the at least one driving function and / or vehicle function. Preferably, only the driving function of the ego road user (to be notified) is adapted.The driving function and / or vehicle function can be a driving function or vehicle function that can be carried out at least partially automatically and / or completely automatically, and which is carried out in particular without any (manual) intervention by the driver or user of the (ego / surrounding) road user. However, it is also conceivable that the adaptation of the driving function and / or the vehicle function is or must be carried out and / or triggered by a user or a driver (manually) by intervening in the vehicle control. The term “provision of an adaptation notification to at least one ego road user to be notified” can be understood to mean that the adaptation notification (preferably by the warning device) is sent to a control device of the road user orof the vehicle is provided and / or transmitted in order to at least partially automatically, preferably completely automatically, trigger the adaptation of the driving function and / or vehicle function by the control device. Furthermore, this can preferably also be understood to mean that the adaptation notification (preferably by the warning device) is provided for transmission to the control device of the road user. "Providing an adaptation notification to at least one ego road user to be notified" can additionally or alternatively be understood to mean providing the adaptation notification (preferably by the warning device) for output to a user or a driver of the ego road user.Such an output can be graphic (for example, via a display device, such as a display and / or a head-up display and / or an augmented reality output device of the vehicle and / or via a (mobile) terminal device of the user or the ego road user), and / or acoustically and / or haptically (in particular via an output device of the ego road user and / or the vehicle and / or via a (mobile) terminal device and / or a mobility aid of the user or the ego road user). The adaptation notification can be characteristic of an instruction for a user / driver of the ego road user or the vehicle, from which the user / driver can determine which interventions they must or should undertake to adapt the at least one driving function and / or vehicle function.Preferably, the adaptation of the at least one driving function and / or vehicle function to the at least one surrounding road user serves to prevent and / or mitigate a potentially emerging critical traffic situation (particularly with regard to road safety), in which the ego road user and / or the surrounding road user are / are involved. For example, the critical traffic situation could be a collision between the ego road user and the surrounding road user. However, it would also be possible for only the ego road user and / or the surrounding road user to be involved in the potentially emerging critical traffic situation.For example, in an overtaking maneuver in which the ego road user overtakes another road user despite oncoming surrounding road users in the lane used by the ego road user as an overtaking lane, the other road user could be endangered if the ego road user prematurely merges into the intended lane (and the oncoming surrounding road user could not be involved in the critical situation). Preferably, the adaptation notification is characteristic of a (potentially) critical traffic situation and / or a cause that is causing and / or potentially causing the (potentially) critical traffic situation, such as a cyclist (e.g., poorly visible or obscured by the driver) who is likely to cross the vehicle's path of travel.“Adaptation [of at least one driving function and / or vehicle function]” can be understood as a change in an operating parameter of the driving function, a switching on and / or off of the driving function and / or vehicle function. “Road user” is understood to mean, in particular, a vehicle, for example a motor vehicle and preferably a road vehicle, agricultural machinery, public buses, trams, rescue vehicles, emergency vehicles, electrified micromobility such as pedelecs, and / or a two-wheeler such as a bicycle and / or a motorized two-wheeler such as a motorcycle and / or scooter, and / or e-scooter, and / or road users such as pedestrians, and / or users of other (prescribed) lanes (general movement lanes), such as cycle paths and / or sidewalks. “Road user” is understood to mean, in particular, motorized and / or non-motorized road users.The ego road user could, for example, be a (motor) vehicle, such as a passenger vehicle (e.g., a car). The surrounding road user could, for example, be a so-called "VRU" (abbreviation for "vulnerable road user"). VRUs are, in particular, road users who are not surrounded by a (protective) (driver's) cabin, for example, pedestrians, cyclists, and motorcyclists. The present invention is particularly advantageously applicable to a combination in which either the ego road user or the surrounding road user is a VRU (and the other road user is not a VRU, for example, a (motor) vehicle with a passenger cabin).It would also be conceivable for the road user to be a person with a mobility aid, such as a walker or a wheelchair, or a person, for example a person doing sports such as a jogger, with a mobile device, such as a smartwatch and / or a smartphone and / or a fitness device. Joggers can also be particularly at risk as road users due to their increased speed and the intention to reach a personal running goal. The mobile device can preferably have the warning device. The adaptation notification is preferably a warning notification. This warning notification should preferably be provided to the at least one ego road user to be notified, who is to be warned or is warned preferentially by the provision of the warning notification.Preferably, the at least one surrounding road user is a potentially colliding (with the ego road user) surrounding road user (in the surrounding area of ​​the ego road user). Preferably, the warning notification is provided to the at least one ego road user to be warned in order to avoid a potential collision with the at least one (potentially colliding) surrounding road user in a collision area of ​​the surrounding area, in particular a traffic area, preferably road traffic. In particular, at least one driving function and / or vehicle function is adapted to the at least one surrounding road user located and / or moving in a surrounding area of ​​the road user in order to avoid a potential collision with the at least one surrounding road user.The adjustment of the driving function and / or vehicle function could, for example, involve braking, accelerating, swerving, and / or changing the route, which should be or will be caused and / or triggered by the provision of the warning notification (e.g., automatically or after a warning notification is issued to a driver or user of the vehicle through manual intervention by the driver / user in the vehicle's control or movement of the road user). A collision zone is preferably an area in which movement lanes (e.g., traffic lanes, cycle lanes, safety lanes) or movement tracks of road users, preferably of the ego road user and the surrounding road user, intersect.A collision area is preferably an area (such as an intersection) in which a movement path of an ego road user, provided and / or predetermined by traffic control, can intersect with a movement path of a surrounding road user, provided and / or predetermined by traffic control (such as a crossing of a bicycle lane and a movement path provided by a car turning left when turning left). It is conceivable that the collision area is a (traffic) area (in particular a collision point) in which the predicted movement paths of two road users intersect. In addition to stationary collision areas or collision points, which are predetermined, for example, by the structure of an intersection and / or traffic routing at this intersection, the collision areas can also be non-stationary collision areas.Such a non-stationary collision zone could, for example, also be collision areas or collision points that arise during an overtaking maneuver carried out by an overtaking road user in combination with an oncoming road user in the same lane as the overtaking road user. Here, too, as with a turning maneuver, it is crucial that a sufficient safety distance is maintained between the overtaking road user and the oncoming road user. The warning device is, in particular, a processor-based warning device. The warning device is preferably located in the surrounding area. The warning device is preferably a warning device for the ego road user (to be notified) and / or the surrounding road user (e.g., potentially colliding).The warning device is preferably (fixed and / or unambiguously) assigned to the ego road user or the surrounding road user. The warning device is preferably vehicle-bound, i.e. a component of a vehicle that cannot be removed non-destructively. It would also be conceivable (additionally or alternatively) for the warning device to be a road user-bound warning device, which is, for example, a (particularly fixed or non-destructively removable) component of a mobile device (e.g., smartwatch, smartphone, fitness device) and / or a mobility aid of the (ego and / or surrounding) road user. The warning device preferably moves together with the (ego / surrounding) road user. It would also be conceivable (additionally or alternatively) for the warning device to be a stationary warning device, for example as (part of) a traffic infrastructure facility.Such warning devices can advantageously be used, for example, in particularly critical traffic areas in which accidents or (serious) critical traffic situations occur frequently and / or are to be feared. A “predicted movement path” of the (ego or surrounding) road user is to be understood in particular as a (determined and / or retrieved) future movement path of the (ego or surrounding) road user, in particular (preferably at least in sections) in the surrounding area. The “predicted movement path” of the (ego or surrounding) road user is to be understood in particular as a (future) movement path of the (ego or surrounding) road user along which a movement or movement course of the (ego or surrounding) road user is to be expected. The warning device cansurrounding road user from a, preferably internal, storage device and / or receive it (for example from a particularly stationary traffic infrastructure facility). It is also conceivable, however, that the warning device determines or calculates the predicted movement path (in real time). It is also conceivable that an ego-road user-linked warning device retrieves navigation data and / or data for route guidance of the ego-road user in order to determine the predicted movement path. Preferably, the warning device determines the predicted movement path of the ego-road user and / or the surrounding road user based on at least one, in particular current, (detected and / or transmitted) vehicle position of the respective road user. In the case of an ego-road user-linked warning device, the vehicle position of the ego-road user can be determined on the basis of GNSS data (GNSS)."Global Navigation Satellite System", abbreviation GNSS, a collective term for the use of existing and future global satellite systems for position determination). In the case of an ego-road user-linked warning device, the vehicle position of the surrounding road user can be determined based on environmental data recorded by an environment detection device of the ego road user for detecting the road user's environment or the vehicle's environment. Preferably, the environment detection device for detecting the vehicle environment (of an (ego) vehicle) is selected from a group of sensors comprising a (color) camera, a front camera, a rear camera, an infrared camera, a LIDAR sensor (abbreviation for light detection and ranging or light imaging, detection and ranging), a radar sensor, ultrasonic sensors and the like, as well as combinations thereof.The environment detection device preferably generates spatially resolved (in particular 2D and / or 3D) environment data (from a vehicle environment of the respective (ego) vehicle) for detecting a vehicle environment. However, it is also conceivable that a (in particular stationary) traffic infrastructure facility detects environment data relating to a (in particular current) position of an environment road user (and / or ego road user) and / or position data (e.g. via V2X communication) and transmits it to the warning device (e.g. of the ego road user) (to determine the respective predicted movement path). It is also conceivable that the traffic infrastructure facility determines a predicted movement path of at least the ego road user and / or the environment road user and transmits it to the warning device.Alternatively or additionally, V2X communication data and / or V2V communication data are used to determine the (current) position of the ego road user and / or the surrounding road user and / or to determine the respective predicted movement path (emitted by the ego road user or the surrounding road user). These preferably include a (current) position and / or movement data and / or information about a genus / type and / or information about a geometric extent (length / width) and / or information about a role of the respective road user.According to the invention, the warning device determines (in particular in a computer-implemented method step) on the basis of the at least one predicted movement path and preferably on the basis of the predicted movement paths (on the basis of the predicted movement path for the ego road user and on the basis of the predicted movement path for the surrounding road user) at least one notification variable (preferably warning variable) characteristic for providing (and / or outputting) the adaptation notification (preferably the warning notification).The notification variable (preferably a warning variable) is preferably characteristic of whether an adaptation notification (preferably a warning notification) is to be provided and / or at what point in time (warning time) the adaptation notification (preferably a warning notification) is to be provided—for example, for output to a user and / or to trigger the adaptation of the at least one driving function and / or vehicle function. According to the invention, the warning device determines the (at least one) notification variable (preferably a warning variable) as a function of at least one, in particular statistical, swarm prediction variable for determining a prediction quality of the predicted movement path and preferably for determining a prediction quality of at least one of the predicted movement paths (in particular in a computer-implemented method step).Preferably, the at least one swarm prediction variable is determined on the basis of global swarm movement data of a large number of road users, which data is generated in a large number of different traffic areas. “Global swarm movement data” is to be understood in particular as meaning that it was not only recorded and / or recorded locally, i.e. in relation to exactly one spatially delimited, contiguous traffic area, but in relation to several, preferably non-contiguous and / or non-overlapping traffic areas. The traffic areas are preferably distributed across at least one country, preferably across several countries, and particularly preferably across at least two continents. It is conceivable that the traffic areas are evenly distributed. Preferably, each traffic area has at least one collision area and preferably a large number of collision areas.The swarm prediction variable preferably relates to at least (preferably exactly) one driving maneuver of the ego road user and / or the surrounding road user. Thus, the swarm prediction variable, with respect to the predicted movement path of the ego road user, could, for example, relate to a (for example, left-turn) maneuver of the ego road user at an intersection. For example, the swarm prediction variable, with respect to the predicted movement path of the surrounding road user, could relate to a (for example, straight-line) crossing of the intersection by the surrounding road user. Preferably, the plurality of traffic zones each comprises at least one collision zone, which results from the at least one driving maneuver of the ego road user and from the at least one driving maneuver of the surrounding road user.Preferably, the at least one swarm prediction variable for determining a prediction quality of a movement path predicted for a road user is characteristic of global swarm movement data of a plurality of road users (preferably of the same type as the road user for whom the movement path was predicted), which are generated in a plurality of different traffic zones. A plurality of different traffic zones is understood to mean at least five, preferably at least 10, preferably at least 50, preferably at least 100, preferably at least 1000 (non-overlapping) traffic zones (preferably each with at least one collision zone).The traffic zones preferably have a geometric extension of more than 20 m, preferably more than 50 m, preferably more than 100 m, preferably more than 200 m, preferably more than 500 m, and particularly preferably more than 800 m in at least one direction, preferably in at least two (preferably mutually perpendicular) directions. The plurality of mutually different traffic zones preferably each comprise at least one and preferably a plurality of collision zones and / or intersection zones (of lanes). Each traffic zone preferably comprises at least one collision zone, preferably several collision zones, and particularly preferably at least five collision zones (where the collision zones preferably refer to collision zones of the involved types of road users, i.e., the type or type of the ego road user and the type or type of the surrounding road user).A plurality of road users (which are in particular of the same type or category as the road user for whom the movement path was predicted) is understood to mean in particular at least 50 (in particular different), preferably at least 100, preferably at least 500, preferably at least 1000, preferably at least 2000, and particularly preferably at least 10,000 different road users. The warning device preferably determines the notification variable as a function of a (in particular statistical) swarm prediction variable for determining a prediction quality of the movement path predicted for the ego road user and as a function of a (in particular statistical) swarm prediction variable for determining a prediction quality of the movement path predicted for the surrounding road user.The proposed method specifically concerns the use and, in particular, the analysis of big data for a particularly optimized application of warning timing in V2X awareness functions. Unlike conventional ADAS (Advanced Driver Assistance Systems), innovative V2X functions make it possible to inform the driver of a potential hazard significantly earlier. Awareness). One example of this is a V2X-based intersection assistance system that warns of cyclists on a collision course. A longer warning time of > 5 s allows the driver to react to the situation without stress and defuse it in time, for example, by braking early. The proposed method offers the advantage of adapting the function to the respective circumstances using a sufficient number of data sets. Data sets representing the behavior of cyclists and cars at an intersection are recorded, for example, by camera drones. Preferably, at least one statistical and / or probabilistic movement path variable and / or at least one, in particular, statistical,A swarm prediction variable is provided for determining a prediction quality of at least one predicted movement path. Preferably, the (in particular each) movement path variable is characteristic of a movement path determined from global swarm movement data of the respective species or type of road users. Preferably, the (respective) swarm prediction quality is based on global swarm movement data of a plurality of road users, which are generated in a plurality of different traffic areas.Preferably, the movement path size and / or the prediction size (for determining a prediction size) of the ego road user of a given type or species is determined exclusively (or essentially exclusively) on the basis of global swarm movement data from a large number of road users of the same type or species. The global swarm movement data can be based on movement data from road users determined in traffic situations in which they performed their intended driving maneuver unhindered by other road users (e.g., because no other road user was present). For example, the global swarm movement data can be based on movement data from a large number of road users,who perform the intended left-turn maneuver without interaction (e.g., without fear of collision with a cyclist). The same preferably applies to the surrounding road user. Preferably, the movement path predicted for the ego road user refers to a driving maneuver to be performed and / or intended by the ego road user (in the future and / or expected). Preferably, the global swarm movement data of the plurality of road users exclusively (or essentially exclusively) takes into account movement data from a plurality of road users.which are characteristic of a driving maneuver of the same type as the (future and / or expected) driving maneuver to be performed and / or intended by the ego road user. The same preferably applies to the surrounding road user. The type or genus of a road user is preferably selected from a group of genuses or types comprising a vehicle, a motor vehicle, a passenger car, a public means of transport, a private means of transport, a role of the means of transport, a road vehicle, an agricultural machine, a purpose of the means of transport, (public) buses, trams, rescue and emergency vehicles, electrified micromobility such as pedelecs, a two-wheeler, a bicycle and / or a motorized two-wheeler such as a motorcycle and / or scooter, and / or an e-scooter and / or road users, such as pedestrians, cyclists, motorcyclists, motorized and / or non-motorized road users,VRUs and the like, as well as combinations thereof. In a preferred method, the surrounding area and the plurality of traffic areas are pairwise different from one another. In other words, the swarm prediction quality determined on the basis of global swarm movement data generated from a plurality of traffic areas is applicable to (and meaningful for) an surrounding area different from this plurality of traffic areas. In a further preferred method, the warning device determines, depending on a predetermined and / or predeterminable minimum prediction probability value based on the swarm prediction value, whether the adaptation notification should be provided. This offers the advantage that the adaptation notification is only provided once a minimum prediction probability is reached. This ensures that the critical situation, such as a potential collision,occurs sufficiently safely and a road user is not unnecessarily distracted and / or warned. In a further preferred method, a safety value characteristic of a minimum safety distance, preferably to be maintained during their movement, between the ego road user and the surrounding road user is predefined and / or can be predefined for the warning device. The safety value can depend on a type of intended driving maneuver and / or a movement size of the road user and / or on a species and / or a geometric dimension (e.g. length / width) of the road user. However, it is also conceivable that the safety value can be predefined and / or predefined depending on the person, whereby a personal sense of security of the driver, for example, can be taken into account. Preferably, when at least temporarily below a (or the), in particular predefined and / or predeterminable,minimum safety distance of the predicted movement paths of at least two road users (in particular the ego road user and the surrounding road user), a warning notification to warn at least one road user (preferably the ego road user and / or the surrounding road user) of a potentially dangerous situation resulting from a potential collision with another road user can be provided / issued and / or is provided and / or issued (e.g. to a driver). This offers the advantage that the movement of the road user can be changed,to counteract an actual undershoot of the minimum safety distance. Preferably, the warning device (for determining the notification variable) is determined based on the safety variable as a function of the predicted movement path of the ego road user and at least one swarm prediction variable for determining a prediction quality of the predicted movement path of the ego road user, and as a function of the predicted movement path of the surrounding road user and at least one swarm prediction variable for determining a prediction quality of the predicted movement path,how high the probability is for the ego road user and the surrounding road user to approach each other closer than the specified and / or predeterminable safety distance. In a further preferred method, the (at least one) swarm prediction variable is characteristic of a genus and / or a type and / or kind of the ego road user and / or the surrounding road user and is characteristic of a topological variable characteristic of a topology of at least one movement lane of a traffic area or (the) surrounding area, preferably comprising a collision area, in particular a structure of an intersection, and is characteristic of a,for a geometric extension of an intersection area of ​​predetermined movement paths of the road users, characteristic extension variables. Preferably, the at least one swarm prediction variable is dependent on a genus and / or a type (and / or a) species of the ego road user and / or the surrounding road user and dependent on a topological variable characteristic of a topology of at least one movement path of a traffic area or (the) surrounding area, preferably comprising a collision area, in particular a structure of an intersection, and dependent on a,for a geometric extension of an intersection area of ​​predetermined movement paths of the road users characteristic extension variables. In a further preferred method, an environmental situation determination device of the ego road user and / or the surrounding road user (and / or of a particularly stationary traffic infrastructure facility) collects and / or determines and / or receives (for example as a result of a data retrieval) environmental situation data which are characteristic of at least one influencing variable and preferably of a plurality of influencing variables (preferably within the scope of a computer-implemented method step). In this case, the at least one influencing variable and preferably the plurality of influencing variables is selected from a group of influencing variables which are characteristic of a topology of at least one movement path of a collision area, in particular of a structure of an intersection.characteristic topological variable (X-intersection; T-intersection; angle of the road / movement lane / lane merging into another carriageway or movement lane or lane; curvature of a carriageway and / or lane and / or movement lane; permitted directions of movement along a lane, such as cycle paths which may be used in both directions), a variable characteristic of the presence of at least one other road user, in particular in the collision area, a traffic density, a speed variable characteristic of at least one (involved) road user in the collision area, a traffic regulation variable characteristic of a given traffic regulation, in particular right-of-way regulation,an extent variable characteristic of a geometric extent of the collision area and / or the intersection area of ​​predetermined movement paths of road users, a road user variable characteristic of a class and / or group of a road user involved, a cultural variable characteristic of a geographical area in particular, an environmental variable characteristic of at least one environmental condition that can be influenced in particular by a visibility ratio, a time variable characteristic of a season and / or time of day, a weather variable characteristic of a current weather condition, and the like, as well as combinations thereof. Preferably, the swarm prediction variable is characteristic of a prediction quality of a, in particular statistical, swarm movement behavior of a plurality of road users in a plurality of environmental situations,which are characteristic of the at least one influencing variable (and preferably of the plurality of influencing variables). Preferably, at least a portion of the plurality of environmental situations relates to mutually different traffic areas. Preferably, the at least one (and preferably each) swarm prediction variable is determined as a function of the at least one influencing variable and preferably as a function of the plurality of influencing variables. Preferably, the predicted movement path of the ego road user and / or the surrounding road user is determined on the basis of the at least one influencing variable and preferably as a function of the plurality of influencing variables. Preferably, the plurality of influencing variables comprises at least two, preferably at least three, preferably at least four, preferably at least five,preferably at least seven influencing variables (selected from the above group). In a further preferred method, an environmental situation determination device of the ego road user and / or the environmental road user (and / or a particularly stationary traffic infrastructure facility) collects environmental situation data that are characteristic of at least two, preferably at least three, preferably at least four, preferably at least five, preferably at least eight, preferably at least 10, preferably at least 15 influencing variables selected from the above-mentioned group of influencing variables. The environmental situation determination device can be an environmental detection device (such as that described above). It is also conceivable that the environmental situation determination device can access the environmental data generated by the environmental detection device or can retrieve this or data derived therefrom. However, it is also conceivablethat the environmental situation determination device retrieves and / or receives and / or determines map data and / or navigation data and / or temperature data and the like, so that (for example, by the warning device) at least one value for the respective influencing variables can be determined. Preferably, the at least one swarm prediction variable is determined as a function of the aforementioned variables,i.e., depending on the species and / or type (and / or species) of the ego road user and / or the surrounding road user and depending on a topological variable of the surrounding area (determined, for example, by the warning device and / or the road user and / or by the surrounding situation determination device), and depending on a characteristic expansion variable, for example, of the intersection area (preferably determined by the surrounding situation determination device). A topological variable characteristic of a topology of at least one movement lane of a collision area, in particular a structure of an intersection, can be, for example, the number of arms of carriageways and / or lanes and / or movement lanes terminating at an intersection or at an intersection, a number of lanes, an intersection geometry (a length of widenings, radii), a shape of an intersection, such as an X-intersection, a T-intersection,an angle of a road / movement lane / lane merging into another roadway or movement lane or lane, a curvature of a roadway and / or a lane and / or a movement lane, a connection between at least two and preferably several lanes merging into a junction or intersection, i.e. from which lane of an incoming arm one can drive onto which lane of the outgoing arm, permitted directions of movement along a lane, such as cycle paths that may be used in both directions, and the like, as well as combinations thereof. Preferably, a driving maneuver is predetermined and / or provided to the warning device,which is intended and / or (future and / or expected) to be performed by the ego road user or by the surrounding road user. Preferably, the predicted movement path is characteristic of the intended and / or (future and / or expected) driving maneuver. The driving maneuver can, for example, be selected from a group of driving maneuvers, which include a left-turn maneuver, a right-turn maneuver, a (straight-line) crossing of an intersection,an overtaking maneuver and the like, as well as combinations thereof. Preferably, at least one predetermined influencing variable and preferably a predetermined plurality of influencing variables are assigned to one and particularly preferably to each driving maneuver (from the group of driving maneuvers). Preferably, the environmental situation determination device retrieves the influencing variables assigned to this driving maneuver depending on the intended and / or the (future or expected) driving maneuver and collects and / or determines data on these influencing variables. Preferably, the warning device determines the at least one swarm prediction variable and / or data characteristic of the predicted movement data depending on the influencing variables. It is conceivable that the swarm movement data are at least partially vehicle-to-X communication data (V2X communication data),preferably vehicle-to-vehicle communication data (V2V communication data). Preferably, the swarm movement data are generated and / or determined at least partially (preferably exclusively) on the basis of vehicle-to-X communication data (V2X communication data), preferably on the basis of vehicle-to-vehicle communication data (V2V communication data). In a further preferred method, the warning device determines the predicted movement path of the at least one (in particular potentially colliding) surrounding road user on the basis of data detected by a communication device,Vehicle-to-X communication data (V2X communication data) and preferably vehicle-to-vehicle communication data (V2V communication data) transmitted by the at least one (in particular potentially colliding) surrounding road user. Preferably, the predicted movement path of the at least one (in particular potentially colliding) surrounding road user is determined on the basis of swarm data of a plurality of road users generated in a plurality of different traffic areas (collision areas). In a further preferred method, the warning device is configured for at least one species and / or at least one type of (surrounding) road user, preferably for a plurality of species or types of road users.(in particular, in each case) a movement model with a plurality of movement parameters is provided. Preferably, the movement model or data characteristic thereof and / or the plurality of movement parameters are stored on a storage device internal to the ego road user and / or the warning device. Preferably, the (respective) predicted movement path is determined based on the movement model. This offers the advantage that the road user can determine a respective predicted movement path of (each) road user in real time without any loss of time based on the movement model and particularly preferably based on the position of the surrounding road user (without, for example, requiring communication with a backend server). It is conceivable thatthat a movement with constant speed or a movement with constant acceleration is assumed as the movement model. The movement model can depend on the driving maneuver intended and / or (future or expected) to be performed by the respective road user. For example, for a cyclist who wants to cross an intersection in a straight line, a movement model with constant speed (by the warning device) could be selected, while for a cyclist turning right at the intersection, a movement model with constant acceleration can be selected. Preferably, the movement model depends on at least one influencing variable and preferably on a plurality of influencing variables. Preferably, the movement model dependswhich is assigned to a driving maneuver intended and / or (future or expected) to be performed by the respective road user and / or is selected as a function of such a maneuver, depends on at least one of the influencing variables assigned to the driving maneuver (intended and / or future or expected to be performed by the respective road user). It is conceivable thatthat at least one and preferably more data characteristic of the movement parameters are recorded (such as speed and / or position data of the road user), and based on any changed movement parameters and / or position data, the movement path predicted on the basis of the movement model is updated (repeatedly and / or at regular intervals). Preferably, the swarm prediction variable indicates a confidence of the prediction (of the predicted movement path) for determining a prediction quality. In a further preferred method, at least one time point predicted for the ego road user and / or for the surrounding road user for reaching a collision area, in particular a collision point, is determined. Preferably, a probability is determined as a function of a predetermined time period ("gap"),that both the ego road user and the surrounding road user reach the collision zone (and / or are located in this collision zone) within the specified period around the predicted time. Preferably, a (potential or predictive) collision point and / or collision zone is determined depending on the predicted movement path of the ego road user and / or the surrounding road user. In particular, its position is determined. However, it is also conceivable that, for example, based on map data and / or navigation data, (at least) one potential collision point and / or collision zone (in particular its position) is already specified depending on the position of the ego road user and / or the position of the surrounding road user. For example, for a given intersection toward which the ego road user is moving,where potential collision points with other surrounding road users are located. Preferably, based on the position of the collision point or collision area and the predicted movement path of the ego road user, a predictive time until the collision point or collision area is reached is determined (referred to in the figures as the so-called "time to path intersection point", abbreviated to TTPIP, here TTPIP of the ego road user). Preferably, based on the position of the collision point or collision area and the predicted movement path of the surrounding road user, a predictive time until the collision point or collision area is reached is determined (referred to in the figures as the so-called "predictive "time to path intersection point", abbreviated to TTPIP, pred referred to, here TTPIP predof the surrounding road user). Preferably, the predicted movement path (in each case) as well as the collision point are based on a one-dimensional approach (so that a movement of the road users is modeled by a linear or path-like movement and so that there is no two-dimensional collision area, but rather a point-like collision point). This offers the advantage of a less computationally intensive determination (for example of the notification size). Preferably, at least one swarm prediction quality is provided and / or is (in particular ascertainable by the warning device), which specifies a probability density function for a prediction error distribution (in particular for the respective type or genus of the respective road user and / or preferably the driving maneuver or the predicted movement path) or is characteristic thereof. The prediction error distribution can be for each different TTPIPpred be provided or determined and be characteristic of the difference between the predicted value TTPIP pred and the current time duration TTPIP true for reaching the collision point (“path intersection point”, abbreviation PIP), where the current time period is preferably determined from the global swarm movement data. For example, a statistical correlation of a probability P depending on a given (preferably a plurality of) TTPIP pred be given (and / or determined on the basis of the swarm prediction quality) with the difference of the TTPIP pr–d - TTPIP true (see also explanations to Fig.12). Preferably, on the basis of the probability density function for a prediction error distribution of the ego road user (in particular for a large number of TTPIP predof the ego-road user) and based on the probability density function for a prediction error distribution of the ego-road user and preferably depending on a given time period (in the figures as “gap [T gap,1 , T gap,2 ]“ or “T gap"), within which both the ego road user and the surrounding road user must reach the collision point (PIP) in order to trigger the provision of an adaptation notification (or warning notification), a measure of the probability that both the ego road user and the surrounding road user will reach the collision point (PIP) within the specified time period can be determined. This measure of probability can be determined based on an integration of the probability density function for a prediction error distribution of the ego road user and the probability density function for a prediction error distribution of the surrounding road user. For example, the measure of probability can be calculated according to the formula for P given in the figure description for Fig. 13 X (car, VRU, TTPIP pred , T gap,1 , T gap,2), where the example of a car (car) is given as the ego road user and a VRU as the surrounding road user. In general, the probability density function for the ego road user should be used instead of the probability density function for a prediction error distribution of the car (“car”), and the probability density function for the surrounding road user should be used instead of the probability density function for a prediction error distribution of the VRU. The notification size can be determined depending on the determined probability measure and the or a specified minimum prediction probability size. For example, the determined probability measure can be compared with a minimum probability. Only when the determined probability measure exceeds the specified minimum prediction probability can the adaptation notification be provided and / or issued. In the case of a furtherIn a preferred method, the warning device determines at least one local collision variable as a function of determined (recorded) environmental data and / or map data relating to the environmental area (in particular the collision area), as a function of which the predicted movement path of the at least one (in particular potentially colliding) surrounding road user is determined. In a further preferred method, the warning device determines at least one local collision area variable as a function of determined (recorded) environmental data and / or (navigation) map data stored on an internal storage device of the road user and / or the warning device in relation to the collision area, as a function of which the predicted movement path of the at least one potentially colliding surrounding road user is determined. In a further preferred method, the warning device is provided with a plurality ofinfluencing variables determined from global swarm movement data are provided, which can be updated, in particular independently of one another, by an external server, in particular a backend server (for example as part of an update process, in particular via a communication connection with an external communication partner). This offers the advantage that when the global swarm movement data is expanded, the influencing variables used to determine the swarm prediction variable can be updated. The present invention is further directed to a method for determining, in particular at least local and preferably (at least section-wise) global swarm movement data for generating an at least section-wise, global map of swarm movement data (preferably predicted movement paths) which are characteristic of a plurality of movement paths of a plurality of different road users, by means of (at least) oneSwarm data determination device of at least one road user determining swarm data, in particular a vehicle. At least partially global map of swarm movement data is to be understood in particular as meaning that the global map has swarm movement data for a plurality of different traffic areas. The plurality of different traffic areas can preferably be selected as described in the context of the method for providing an adaptation notification. According to the invention, the swarm data determination device determines local swarm movement data on the basis of vehicle-to-vehicle communication data received by a communication device of the at least one swarm data determining road user and transmitted by at least one road user sending vehicle-to-vehicle communication data, and the local swarm movement data is transmitted to a communication device, in particular theSwarm data-determining road user, for transmission to an external storage device, preferably to an external backend server. Preferably, the communication device transmits the local swarm movement data to the external storage device. Preferably, a plurality of swarm data determination devices of a swarm data-determining road user is provided, each of which determines local swarm movement data and provides it for transmission to the external storage device. The respective method steps are described below only with reference to one swarm data determination device of a swarm data-determining road user, but are analogously transferable to the plurality of swarm data determination devices. Preferably, the plurality of swarm data determination devices moves and / or is located in different traffic areas. The local swarm movement data can bethe vehicle-to-vehicle communication data of a large number of (different) road users, which comprise position data of the respective road user as well as a point in time at which the position data of the respective road user was determined and / or recorded, or characteristic and / or derived variables for this. Preferably, the local swarm movement data comprise a movement data for a type of road user sending the vehicle-to-vehicle communication data. The local swarm movement data preferably relates to movement data of a large number of road users, preferably vehicles, determined (in particular within a predetermined period of time, for example within one minute and / or within 10 minutes and / or within one hour), which were recorded and / or generated within a traffic area having a geometric extent of at least 25 m, preferably 50 m, preferably 100 m,preferably 200 m, preferably 500 m and particularly preferably at least 800 m. Preferably, (all) vehicle-to-vehicle communication data of all road users sending vehicle-to-vehicle communication data, which are received by the road user determining the swarm data (within a predetermined and / or predeterminable period of time), are taken into account when determining the local swarm movement data. Preferably, the at least one road user sending vehicle-to-vehicle communication data is located in an environment (in particular in a communication area of ​​the communication device of the road user determining the swarm data, in which a data exchange, in particular via a transmission and / or reception of vehicle-to-vehicle communication data is possible) of the road user determining the swarm data. For example, data records recorded by camera drones that record the behavior of cyclists and cars at aThere are very few such systems worldwide that represent an intersection, which makes it difficult to parameterise a predicted movement path (which is characteristic of the movement behaviour of a large number of road users of the same species or type), which depends on the previously mentioned influencing variables and other factors, and only an attempt can be made to create a generally valid parameterisation of the function using individual samples. In order to take into account the large number of possible factors for determining the warning time, it is proposed to collect a corresponding large number of, as heterogeneous as possible, data using swarm data from road users. For this purpose, anonymised Car2X data is preferably recorded by fleet vehicles (VE test vehicles, ideally customer vehicles) when approaching an intersection, for example, whereby the trajectories of cyclists and vehicles, for example inintersection area. Thanks to Car2X communication, the recording vehicles receive high-quality data as soon as they come within radio range. The recording ego-vehicle or the road user collecting the swarm data does not necessarily have to be the vehicle that is in the actual intersection situation with the cyclist – it is sufficient to simply pass by or be within reception range. On a backend, this data (local swarm movement data) is then preferentially enriched with additional data (such as map data), and factors influencing the warning time are preferably determined. First, the recorded trajectories of cars and cyclists can be clustered according to direction of movement and maneuver performed at the intersection. Subsequently, it can be investigated, for example, whether the direction of travel of the cyclists (along / against the regular direction of travel on the cycle path) has an influence.The influence of the presence of cars at the intersection on movement behavior and thus also on prediction can be measured. The influencing factors calculated using this "Big Data" approach are then fed back to the vehicles as parameter sets. For example, a vehicle at an intersection A in a rural area, where bicycles are constantly traveling at high speed, can display the warning 7 seconds before crossing the intersection, but at an intersection B in a residential area, only 5 seconds before. Both intersection-specific parameters and formalized parameters (e.g., for intersection type / size / ...) are conceivable for the parameterization of the vehicles. The vehicle-to-vehicle communication data can, for example, be communication data transmitted, in particular, in accordance with the ETSI ITS G5 communication technology. Such vehicle-to-vehicle communication data can, for example, be sensor data (or variables derived therefrom) whichrecorded and / or collected and / or generated by an (on-board sensor) environment detection device of the road user sending the vehicle-to-vehicle communication data. For example, so-called "Collective Perception Messages" (CPM) can be used as such vehicle-to-vehicle communication data. In particular, such CPM messages can contain information about the kinematic dynamics and position as well as other attributes of objects (e.g., vehicles, pedestrians, animals, and others) that are detected by (an) environment detection device(s) (in particular of the road user sending the vehicle-to-vehicle communication data), such as radar, lidar, and camera(s), to which the transmitting road user has access. The Collective Perception Message (CPM) enables, in particular, the interoperable exchange of basic information about the (necessary for the interpretation of the transmitted data) basic information about theThe transmitting road user, their sensory capabilities, the perceived objects, and / or road-related perception areas. CPMs are generated quasi-periodically, as determined by the CPM generation event. This offers the advantage that the transmitted data volume is kept small by transmitting the CPM triggered by the perception of a new object and / or a new attribute. Such vehicle-to-vehicle communication data offers the advantage that information (particularly regarding movement behavior or position and / or time) relating to road users can be captured and thus taken into account in the local swarm data (and thus global swarm movement data), which themselves do not have the ability to transmit (or receive) vehicle-to-vehicle communication data. The vehicle-to-vehicle communication data are preferably so-called "Cooperative Awareness Messages."(CAMs). These vehicle-to-vehicle communication data are used to inform surrounding communication partners (such as other road users) of their presence (by sending the CAM message) cyclically or periodically or at regular intervals. The use of CAM messages offers the advantage that a new CAM message is sent if the distance to the transmission position of the previous CAM message exceeds a predetermined distance (e.g. 4 m). This offers the advantage of enabling very dense data collection. Furthermore, the use of CAM messages offers the advantage that a new CAM is sent if the road user changes course, i.e., changes the orientation in the World Geodetic System of 1984 (WGS84) by more than 4° compared to the last transmitted CAM. This offers very precise recording of the movement path of aRoad user. Furthermore, the use of CAM messages offers the advantage that if the current speed of the road user exceeds the speed value sent in the last CAM by more than 0.5 m / s, a new CAM is sent, so that even large acceleration values ​​can be well represented in the swarm data. Preferably, the transmitted vehicle-to-vehicle communication data comprise information relating to the time of generation of the vehicle-to-vehicle communication data and a position of the road user. Preferably, the transmitted vehicle-to-vehicle communication data additionally or alternatively comprise information relating to at least one geometric variable and / or movement variable of the road user, which is selected from a group of variables, comprising a direction of travel, an orientation of the road user, an orientation of the road user, a speed, a length, a widthof the road user, a longitudinal acceleration, a curvature, a yaw rate, an acceleration control, a lane position, a lateral acceleration, a steering wheel angle, a vertical acceleration, a power class, a toll area, and the like, as well as combinations thereof. Preferably, the transmitted vehicle-to-vehicle communication data additionally or alternatively comprise information relating to a role of the road user (such as a special vehicle, a regular passenger vehicle, and the like). This offers the advantage that data relating solely to the behavior of special vehicles or other road users in the vicinity of special vehicles can be collected at once. Preferably, the transmitted vehicle-to-vehicle communication data additionally or alternatively comprise information relating to a path history. This offers the advantage that, for example, the position of the road user can be determined even more precisely andfurthermore has the advantage that even swarm data is collected which dates back to a time at which the road user sending the vehicle-to-vehicle communication was not yet within (communication) range of the road user collecting the swarm data. In a preferred method, the vehicle-to-vehicle communication data which are sent at a transmission time by the at least one road user sending the vehicle-to-vehicle communication data within a common data packet comprise, with respect to the transmission time, current movement data of the road user sending the communication data and historical movement data of the road user sending the communication data at a plurality of different recording times preceding the transmission time. In a further preferred method, the swarm data determination device determines within apredetermined and / or predeterminable period of time, local swarm data relating to a plurality of different road users sending vehicle-to-vehicle communication data and makes these available, in particular in the form of a common data packet of the communication device (in particular simultaneously) for transmission to the external storage device. In a further preferred method, the swarm data determination device analyses a plurality of vehicle-to-vehicle communication data received by it, in particular transmitted by exactly one road user sending communication data, or data derived therefrom, for redundant movement data of the road user sending communication data at essentially simultaneous recording times and removes these redundant movement data to determine the local swarm data. In other words, the swarm data determination device canClean up vehicle-to-vehicle communication data. This offers the advantage that the data volume to be transmitted to the external storage device is kept as small as possible. In a further preferred method, the swarm data determination device determines local swarm data at regular intervals and makes this available to the communication device for transmission to the external storage device. Preferably, the communication device (in particular of the road user determining the swarm data) transmits the local swarm data to the external storage device. Preferably, the local swarm data are aggregated in the external storage device to form global swarm movement data. Preferably, a plurality of movement paths of the individual road users are determined on the basis of the swarm movement data. Preferably, for each type or genus of road user and preferably with respect to a performedDriving maneuvers or movement maneuvers (such as turning left) generate an averaged or statistical movement path, which is preferably used as a predicted movement path. As already mentioned, the recording of the swarm data offers the advantage of being able to obtain data from almost every conceivable situation when a fleet vehicle drives along there. One could assume that this is also possible without Car2X equipment, i.e., by recording the trajectories of fleet vehicles, e.g., at intersections. This applies to cars. As a car manufacturer that proceeds in this way, however, one does not receive any additional data from other road users, e.g., not from cyclists (which only the bicycle manufacturer has). Furthermore, interactions between road users can only be observed to a very limited extent, namely only when the other road users are recorded by the sensors throughout the entire maneuver. Based on the large number oftransmitted data sets of local swarm movement data, a map of global swarm movement data is generated, in which the corresponding movement data is uniquely assigned to each detected road user (in particular anonymized) and is preferably stored depending on the corresponding position data of the road user. In particular, this enables a map display which indicates movement data (and / or movement paths derived therefrom) assigned to a plurality of road users depending on a location or position. Preferably, for each detected road user for whom movement data was transmitted to the external storage device, a movement path of the road user can be determined (based on the movement data uniquely assigned to the road user). The present invention is further directed to a (in particular computer-implemented) method for determiningat least one, in particular statistical, swarm prediction variable of a predicted movement path of a road user of a given type located in a given traffic area, based on a plurality of determined (measured and in particular driven) movement paths, which are recorded on the basis of a plurality of swarm movement data of a plurality of road users. According to the invention, the plurality of swarm movement data is a plurality of global swarm movement data, which is generated and / or collected at a plurality of mutually different traffic areas. In other words, the at least one, in particular statistical, swarm prediction variable is determined on the basis of a plurality of determined movement paths, which are generated on the basis of a plurality of global swarm movement data, which is generated and / or collected at a plurality of mutually different traffic areas.are collected, a large number of road users are recorded. Preferably, the large number of determined (measured and in particular driven) movement paths can be determined by means of the method described above for determining global swarm movement data. The large number of determined (measured) movement paths of a large number of different traffic areas can be recorded from a view from above (bird's eye view) of a respective traffic area, for example by means of drone recordings. From the drone recordings, in particular a sequence of drone recordings, the object trajectories of the objects (road users) moving in the sequence are preferably extracted or determined. Preferably, drone recordings or sequences of drone recordings of a large number of different traffic areas are used. Preferably, the large number of different traffic areas has those preferred features and / or properties individually or in combinationwhich were described above in the context of the method for providing an adaptation notification. Preferably, the plurality of determined (measured) movement paths of the plurality of different traffic areas are generated or determined by means of the above-described method for determining local and / or global swarm movement data. In this case, the local swarm movement data is preferably generated and / or determined (as described above) by means of vehicle-to-vehicle communication data. Preferably, the at least one, in particular statistical, swarm prediction variable is determined on the basis of the plurality of (measured) movement paths which were determined and / or generated by means of a plurality of different road users of the specified type. The proposed method offers the advantage of adapting the function to the respective circumstances using a sufficient number of data sets. Data sets that describe the behaviorof cyclists and cars at an intersection are recorded, for example, by camera drones. On a backend, this data (local swarm movement data) is then preferably enriched with further data (such as map data) and influencing factors for a prediction variable and / or a warning time are preferably determined. When determining the at least one prediction variable, (measured) movement paths of road users who are moving below a predetermined and / or predeterminable minimum speed (for example, 2.0 m / s) are preferably not taken into account and / or removed from the map of movement data. The aim here is to disregard vehicles that stop while turning. Alternatively, these can also be clustered and, in particular, combined into a separate cluster (instead of excluding them completely from the data). Furthermore, (measured) determined movement paths, which have aIf the specified and / or definable minimum distance is exceeded (approximately 5 m), it will not be taken into account when determining the prediction variable or will be removed from the data set or map of (local) movement data. First, the recorded trajectories of cars and cyclists can be clustered according to direction of movement and maneuver performed at the intersection. Subsequently, it can be investigated, for example, whether the direction of travel of cyclists (along / against the regular direction of travel on the cycle path) and the presence of cars at the intersection point can influence movement behavior and thus also the prediction. The influencing factors calculated using this method or this "Big Data" approach are preferably fed back to the vehicles as parameter sets. For example, a vehicle at an intersection A in a rural area, where bicycles are constantly traveling at high speed, candisplay the warning, but at an intersection B in the stationary area only 5 s beforehand. For the parameterization of the vehicles, both intersection-specific parameters and formalized parameters (e.g. for intersection type / size / ...) are conceivable. The at least one swarm prediction variable can be characteristic of a predicted movement path (in particular of a given type of road user and / or a given traffic area). In a preferred method, the at least one swarm prediction variable is characteristic of a statistical prediction quality of a given predicted movement path. The swarm prediction variable can be a swarm prediction variable for determining a prediction quality as described above in the method for providing an adaptation notification. In particular, it can be a probability density function of a prediction error (preferably for aspecific species or type of road user and / or a driving maneuver and / or a particularly predicted path) (as described in more detail above or in the figures). A major disadvantage of the methodology for creating prediction models from drone data or other recorded data sets is that the current data sets only show a small section (e.g.) of an intersection. The data therefore all have a local bias, and influencing factors can probably only be determined with effort and luck using other, different data sets. In addition, most of the data sets are quite small or contain comparatively few trajectories, so that statistically valid statements could be difficult. The disadvantage of a data set derived from drone data is that a high level of effort is required (operation of the drone or measurement location, labeling of the data). In a further preferred method, the swarm prediction variable is determinedthe multitude of determined movement paths are clustered into driving maneuvers of various types. Such clustering can be carried out depending on deceleration and / or acceleration variables and / or steering angle variables, structure and topology of a traffic area traveled (e.g. determined by map data) and / or a criticality of a traffic situation perceived by the road user (e.g. during an overtaking maneuver). This offers the advantage that similar driving maneuvers can be examined for further influencing factors in order to be able to determine the most precisely predicted movement paths (e.g. depending on the type of intended driving maneuver). Furthermore, in the application case, an intended driving maneuver can be derived relatively easily from navigation data or map data or a fixed and / or predetermined route (guidance). In a further preferred method, the multitudedetermined movement paths of the plurality of road users are statistically evaluated as a function of at least one and preferably a plurality of influencing variables. In this case, the influencing variable is or are selected from a group of influencing variables, which include a topological variable characteristic of a topology of at least one movement lane of a collision area, in particular a structure of an intersection, a variable characteristic of a presence of at least one other road user, in particular in the collision area, a traffic density, a speed variable characteristic of at least one (involved) road user in the collision area, a traffic control variable characteristic of a predetermined traffic control, in particular right-of-way control, a geometric extension of the collision area and / or the intersection area of ​​predeterminedMovement traces of road users, characteristic extent variables, a road user variable characteristic of a class and / or group of a road user involved, a cultural variable particularly characteristic of a geographical area, an environmental variable that can be influenced in particular by a visibility ratio and is characteristic of at least one environmental condition, a time variable characteristic of a season and / or time of day, a weather variable characteristic of a current weather condition and the like, as well as combinations thereof. Preferably, on the basis of the plurality of (measured) movement paths, a statistical and / or averaged and / or average movement path is determined, in particular for a given driving maneuver of a road user. It is also conceivable that a predetermined parameterization of a movement path is predetermined and the parameters of the parameterized movement path are determined on the basis of theA large number of (measured) movement paths are determined (e.g. using optimization algorithms). It is conceivable that when assessing whether a given influencing factor influences the statistical movement behavior of a road user, a standard deviation (or variance) of the measured movement paths in relation to a statistical movement path is determined, and the size of the standard deviation is used to assess whether the influencing factor has an influence. Preference is given to using (e.g. AI-based) pattern recognition methods which recognize a pattern of the presence of certain influencing factors (e.g. the above group of influencing factors) and a specific (determined and / or predetermined) course of a statistical and / or averaged movement path. Preference is given to using regression analysis to describe and analyze relationships between the influencing factors and the (global) swarm movement data and / or the large number of movement paths.Preferably, a parameterization of a predicted movement path (for example for a predetermined driving maneuver) of a predetermined road user is carried out depending on the determined influencing variables. Preferably, on the basis of the parameterization of the predicted movement path, a movement model for predicting a movement path of a road user depending on the influencing variables is generated. The present invention is further directed to a machine-readable and in particular computer-implementable movement model for predicting a movement path of a road user (of a predetermined type, preferably determined and / or to be determined by the road user), preferably detected by an environment detection device of a road user, depending on a position of the road user, wherein the movement model is provided with an input variable for the positioncharacteristic variable (such as position data) and at least one variable characteristic of a (or the) type of road user are supplied, wherein the movement model, on the basis of a plurality of model parameters, maps the input variables to output variables characteristic of the movement path to be predicted. Preferably, the plurality of model parameters is determined on the basis of global swarm movement data of a plurality of road users, which are generated in a plurality of different traffic areas. Preferably, the plurality of different traffic areas has those preferred features and / or properties, individually or in combination, which were described above in the context of the method for providing an adaptation notification. Preferably, at least one and preferably the plurality of model parameters are characteristic of a plurality of statistical movement paths of aA plurality of different traffic areas. The movement model can preferably be generated by the method described above for determining at least one swarm prediction variable of a predicted movement path according to a preferred embodiment. At least one intended and / or predicted driving maneuver can preferably be fed to the movement model as an input variable. At least one influencing variable and preferably a plurality of influencing variables can preferably be fed to the movement model as an input variable. Depending on the intended and / or predicted driving maneuver, the plurality of influencing variables assigned to the driving maneuver (as described above) can preferably be fed to the movement variable. The model parameters can be, for example, the influencing variables and / or movement data characteristic of a (current) movement state of the road user for whom the movement path is to be predicted.Preferably, the movement model can be supplied with environmental data acquired as input variables by the environmental detection device of a road user and / or environmental situation data generated by an environmental situation determination device of a road user (as raw data or data derived therefrom), from which the movement model preferably determines values ​​characteristic of the influencing variables. The present invention is further directed to a warning device for an ego road user, in particular for a vehicle, for providing an adaptation notification, preferably a warning notification, to at least one ego road user to be notified, for adapting at least one driving function to at least one environmental road user located and / or moving in an environmental area of ​​the ego road user. In this case, the warning device for the ego road user and for the at least oneThe warning device is suitable and intended to determine, on the basis of the predicted movement paths, at least one notification variable which is characteristic for the provision of the adaptation notification, which is preferably characteristic of whether an adaptation notification is to be provided and / or at what point in time the adaptation notification is to be provided. According to the invention, the warning device is suitable and intended to determine the notification variable as a function of at least one, in particular statistical, swarm prediction variable for determining a prediction quality of at least one of the predicted movement paths. The at least one swarm prediction variable is preferably based on global swarm movement data of a plurality ofRoad users, which are generated in a plurality of different traffic areas, are determined. Preferably, the warning device is configured, suitable and / or intended to carry out the above-described method for providing an adaptation notification as well as all method steps already described above in connection with the method, individually or in combination with one another. Conversely, the method can be equipped with all features described in the context of the warning device, individually or in combination with one another. The present invention is further directed to a vehicle, in particular a motor vehicle, comprising a warning device for a vehicle as described above, according to an embodiment. The vehicle can in particular be a (motorized) road vehicle. A vehicle can be a motor vehicle, which in particular has a vehicle controlled by the driver himself.Motor vehicle (“Driver only”), a semi-autonomous, autonomous (for example, autonomy level 3 or 4 or 5 (of the SAE J3016 standard)) or self-driving motor vehicle. The autonomy level Level 5 refers to fully automated vehicles. The vehicle can also be a driverless transport system. The vehicle can be controlled by a driver or drive autonomously. Furthermore, in addition to a road vehicle, the vehicle can also be an air taxi, an aircraft and another means of transport or another type of vehicle, for example an aircraft, watercraft or rail vehicle. The present invention is further directed to a swarm data determination device for a road user, in particular for a vehicle, in particular a motor vehicle, for determining, in particular at least local and preferably (at least sectionally) global swarm movement data for generating a,at least section-wise, global map of swarm movement data and / or predicted movement paths, which are characteristic for a plurality of movement paths of a plurality of different road users. According to the invention, the swarm data determination device is suitable and intended to determine local swarm movement data on the basis of vehicle-to-vehicle communication data received by a communication device of the swarm data determination device and transmitted by at least one road user sending vehicle-to-vehicle communication data, and to provide the local swarm movement data to a communication device of the swarm data determination device for transmission to an external storage device, preferably to an external backend server. Preferably, the swarm data determination device is configured, suitable and / or intended to carry out the above-described method for determininglocal and preferably global swarm movement data as well as all method steps already described above in connection with the method, individually or in combination with one another. Conversely, the method can be equipped with all features described in the context of the swarm data determination device, individually or in combination with one another. The present invention is further directed to a map generation device (and / or a movement path size determination device and / or a prediction size determination device) for determining at least one, in particular statistical, swarm prediction size of a predicted movement path of a road user of a given type located in a given traffic area on the basis of a plurality of determined movement paths of a plurality of road users. According to the invention, the plurality of determined movement paths is aA plurality of global swarm movement data, which are generated and / or collected in a plurality of different traffic areas. Preferably, the map generation device (and / or a movement path size determination device and / or a prediction value determination device) determines at least one, in particular statistical, swarm prediction value based on the plurality of movement paths, which were determined and / or generated by a plurality of different road users of the specified type. Preferably, a plurality of swarm prediction values ​​(such as predicted movement paths and / or swarm prediction values) are determined for a plurality of traffic areas and assigned to their position data in the form of a map representation. Preferably, the map generation device (and / or a movement path size determination device and / or a prediction value determination device)configured, suitable and / or intended to carry out the above-described method for determining at least one swarm prediction variable of a predicted movement path as well as all method steps already described above in connection with the method, individually or in combination with one another. Conversely, the method can be equipped with all features described in the context of the map generation device (and / or a movement path variable determination device and / or a prediction variable determination device), individually or in combination with one another. The present invention is further directed to a computer program or computer program product, comprising program means, in particular a program code, which represents or encodes at least some and preferably all method steps of one of the methods according to the invention and preferably of one of the described preferred embodiments and is designed to be executed by aProcessor device is formed. The present invention is further directed to a data memory on which at least one embodiment of the computer program according to the invention or a preferred embodiment of the computer program is stored. Further advantages and embodiments emerge from the attached drawings: Therein: Fig. 1 shows a representation of a collision area from road traffic to illustrate an application area of ​​a preferred embodiment of the present invention; Fig. 2 shows a further example of a traffic situation with a collision area in road traffic to illustrate an advantageous use of the warning devices for road users according to preferred embodiments in particular; Fig. 3 shows an illustration of a process for generating a warning notification for output to a driver of a vehicle; Fig. 4 shows three exemplary display contents for outputtingWarning notifications to a driver to increase the driver's awareness or attention with regard to a crossing cyclist; Fig. 5 shows another traffic situation with a vehicle equipped with a warning device according to the invention in accordance with a preferred embodiment; Figs. 6a - 6c show schematic representations to illustrate prediction probabilities; Figs. 6a - 6c show the effect of selecting a specific prediction model for the predicted movement paths of the road users in comparison to statistically collected data of the respective road users. Figs. 8, 9 show exemplary collision areas in road traffic in which the movement paths of various road users, here motor vehicles and cyclists, can cross; Fig. 10 shows a representation to illustrate or visualize a predicted movement path of a left-turn scenario; Fig. 11 shows an exemplary histogram with regard to the temporalDifference between a predicted time duration for a (motor) vehicle and the actual time duration required to reach the (specified) collision point; Fig. 12 shows a further exemplary histogram or a kernel density estimation with a Gaussian kernel (“Gaussian kernel density estimation”) for a difference between the predicted or predicted TTPIP for the (motor) vehicles (cars) on the one hand and the actual TTPIP; Fig. 13 shows a visualization of an exemplary determination of a probability that a vehicle crosses the movement path of a VRU within a specified time period; Fig. 14 shows a boxplot diagram to illustrate the prediction error in determining the predicted or predicted time duration of a (motor) vehicle until reaching a specified collision point in a comparison of two different movement models for the movement of the vehicle; Fig. 15 shows the resulting probability that the (motor) vehicle andthe cyclist reaches the PIP within a defined time period (“gap”) in the event that their TTPIP pred arrival at the same time (i.e. TTPIP pred,car = TTPIP pred,cyc ) is indicated; Fig.16 a heatmap representation of the probability that the (motor) vehicle will follow the cyclist's movement path under a time period (“gap”) T gap of [-4 s, 2 s] for different combinations of TTPIP pred,carand TTPIP; Fig. 17 shows an illustration of a method according to the invention according to a preferred embodiment for determining swarm movement data to generate a global map of swarm movement data and / or movement paths of a large number of road users; and Fig. 18 shows a schematic structure of a vehicle collecting swarm movement data. Fig. 1 shows a schematic representation of a collision area 2, here an intersection area, in road traffic, in which various road users, here a car as a vehicle identified by reference number 10, which is approaching an intersection on lane 3 in order to turn right into lane 5, as well as other road users 22, 24, 30, encounter each other at the intersection. In the case shown in Fig.The intersection shown in Fig. 1 is a four-way intersection, where two straight lanes 3, 5 (each with two lanes traveling in opposite directions) meet. One of the two intersecting lanes, namely lane 5, provides a zebra crossing for pedestrians 24 on both sides of the intersection to cross lane 5 for traffic control. Reference numeral 7 designates the zebra crossing provided on lane 5 into which vehicle 10 turns after turning right at the intersection. In the current traffic situation shown in Fig. 1, a cyclist 22 is currently crossing zebra crossing 7, while a pedestrian 24 is still on the sidewalk in front of zebra crossing 7 on the opposite side of lane 5 from vehicle 10.Somewhat further away from the zebra crossing 7, but coming from the same side as the pedestrian, a cyclist 30 is approaching the zebra crossing 7. A traffic light system 20 is provided here for traffic control, which is designed here as a networked infrastructure facility. This can, for example, exchange V2X messages with the networked vehicle 10 and transmit a warning 13 to a communication device 14 of the vehicle 10 and / or provide ADAS data 13 (ADAS stands for "Advanced Driver Assistance System"), which an assistance system of the vehicle can process. For example, the traffic light system 20 can have an environmental detection system for the zebra crossing area 7, which is surrounded by a substantially rectangular line and which, in Fig. 1, includes the zebra crossing as well as adjacent areas of the sidewalk and, for example, an adjoining cycle path.The traffic light system can evaluate the recorded environmental data using object recognition and report detected objects, in this case, for example, cyclist 22 and pedestrian 24, preferably together with their position, to V2X-capable road users, such as vehicle 10 (as well as cyclist 30). The fact that infrastructure facility 20 has detected cyclist 22 and pedestrian 24 in the zebra crossing area is indicated by the bounding boxes drawn around these two road users in Fig. 1. However, cyclist 30, who is still outside the detection range of infrastructure facility 20, cannot (yet) be detected (and recognized) by infrastructure facility 20.A warning device of the vehicle 10, in particular a processor-based one, can use the object data received from the communication device 14 of the vehicle 10, for example, to supplement the environmental data obtained from the on-board sensors of the vehicle 1 and perform a driver assistance function based on the environmental data. The driver assistance function can, for example, be an "awareness" function, in which the attention of the driver or user of the vehicle is increased, for example in the event of a (potential) dangerous situation predicted by the driver assistance function, and is preferentially directed to the (potential) danger location. The "awareness" function shown in Fig. 1 is, in particular, a VRU awareness function, where "VRU" is used as an abbreviation for "Vulnerable Road User.""VRUs" are, in particular, road users who are exposed to a particularly high risk. "VRUs" include, for example, pedestrians, cyclists, and users of small electric vehicles such as e-scooters. VRUs are not surrounded by a protective driver's cabin like those in cars or vans and therefore generally represent a vulnerable road user. A VRU awareness function (performed, for example, by the warning device 12) determines, based on the received object data, whether the user or driver of the vehicle 10 should be warned by issuing a warning notification 16 (here graphically represented as a symbolic image) or whether they should be stopped to exercise particular caution because VRUs are (or may be) located along a predicted path of the vehicle.When determining whether and / or when a warning notification 16 should be issued to the user or driver of the vehicle, the warning device 12 according to the invention, in a preferred embodiment, takes into account the likelihood of a collision between the vehicle and one of the other road users or the likelihood of a dangerous situation resulting from an insufficient safety distance between the vehicle and another road user. For this purpose, the warning device 12 can determine, based on predicted movement paths determined for the ego vehicle 10 and for the other road user, whether the two predicted movement paths intersect and / or whether the two predicted movement paths are at a sufficient (safety) distance from each other, so that safe and / or unimpaired movement (along the originally planned movement path) is possible for both road users.In the traffic situation illustrated in Fig. 1, the cyclist 20 (detected or recorded by the infrastructure device 20) is already on the zebra crossing 7. At normal driving speed, the cyclist 20 will have long since left the zebra crossing on the opposite side by the time the vehicle 10 turns right into lane 5. Therefore, if the warning device 12 of the vehicle 10 compares the movement path predicted for the (ego) vehicle 10 with the movement path predicted for the cyclist 22, it can be seen that these two predicted movement paths do not (will not) intersect and that a (specified) sufficient safety distance between the two road users 10, 20 is maintained.In this case, a warning variable determined by the warning device 12 could therefore be characteristic of the assessment result determined by the warning device that no warning notification needs to be issued or that there is no indication to provide a warning notification (for issuance to the user). The situation may be different for the pedestrian 24 (also detected or recognized by the infrastructure device 20), who is moving in the opposite direction of movement with respect to the vehicle 10, is still on the sidewalk on the opposite side of the roadway with respect to the vehicle 10, and is about to enter the zebra crossing 7 to cross the roadway 5. Although the vehicle 10 is still further away from the pedestrian crossing 7 than the pedestrian 24, it is generally moving faster than the pedestrian and could therefore meet at the zebra crossing at approximately the same time.The warning device 12 could therefore receive a warning value as an assessment result when comparing the two respective predicted movement paths, which is characteristic that the driver or user of the vehicle 10 needs to be warned by issuing a warning notification. The same could apply to the cyclist 30. Although the cyclist is further away from a potential intersection point than the pedestrian 24 in the snapshot of a traffic situation shown in Fig. 1, since cyclists 30 usually move at a higher speed than pedestrians, the evaluation of the predicted movement paths of the vehicle and the cyclist could also result in an assessment result that a collision is to be expected (without further intervention). Therefore, the warning device 12 of the vehicle 10 could also determine a warning value here, which is characteristic that the user orthe driver of vehicle 10 is to be warned. In the example shown in Fig. 1, cyclist 30 is not detected by infrastructure facility 20. If the driver's view and / or the on-board sensors or the vehicle-bound environment detection device of vehicle 10 is restricted with respect to road users 24 and / or 30, infrastructure facility 20 can transmit the detected object data relating to pedestrian 24 to the vehicle for ADAS data provision, but not (yet) to cyclist 30. Only with the environment data relating to VRUs in zebra crossing area 7 provided by infrastructure facility 20 could therefore pose a risk that vehicle 10 does not receive environment data relating to cyclist 30 early enough and therefore, for example, the driver or user of vehicle 10 cannot be warned early enough by issuing a warning notification.In a preferred embodiment, the communication device 14 of the vehicle 10 can also receive the V2X communication data, preferably V2V communication data, transmitted by a networked road user, as in the traffic situation illustrated here, in which the cyclist 30 is a networked cyclist who transmits in particular V2X communication data and particularly preferably V2V communication data to his or her surroundings. In the preferred embodiment, the warning device 12 is suitable and intended to evaluate the V2X communication data (preferably V2V communication data) received by the communication device 14 and to take it into account when determining a warning variable and / or when determining and / or retrieving a predicted movement path of the road user sending V2X communication data and / or V2V communication data.The best possible benefit for a driver or user of vehicle 10 who potentially needs to be warned is achieved if the driver is warned in advance of an excessive proximity of the two road users, i.e., in this case, the vehicle and, for example, the pedestrian 24 or the cyclist 30, to each other (in particular, a potential collision event) only by a warning notification issued to them (without further intervention) with a (sufficiently high) predetermined probability. The warning device according to the invention, according to a preferred embodiment, is suitable and intended to take into account a statement derived from swarm data (from a vehicle fleet) regarding the prediction quality of a (respective) predicted movement path when determining the warning variable, in particular the warning time.Only when there is a sufficiently high probability that the road users involved will come too close to one another at an intersection or meet at the intersection is a warning notification provided or issued to the driver. When determining this probability, the respective prediction quality or a probability of location of the predicted movement paths of the road users involved is taken into account. The warning time is preferably determined based on the probability of the paths crossing (in particular by the warning device). Fig. 2 shows a further example of a traffic situation with a collision zone in road traffic to illustrate an advantageous use of the warning devices 12 according to the invention (not shown in Fig. 2) for road users 10, in particular according to preferred embodiments.2 shows that the direct line of sight of a user of a vehicle 10 to a cyclist 30 traveling on a bicycle lane that crosses the roadway of the vehicle 10 and moving toward the intersection of the vehicle's roadway and the bicycle lane is obscured by a building 4. In other words, from the vehicle 10, the area in which the cyclist 30 is currently located is not visible, in particular not even by the vehicle-mounted environment detection device of the vehicle 10. Unlike in the example shown in Fig. 1, the cyclist 30 is not a networked road user. Since such spatial configurations often lead to an accident blackspot, an infrastructure facility 20 can be provided which is arranged such that the lane of a VRU, here the bicycle lane, which is crossed by a roadway, can be detected.The reference numerals 20a and 20b here indicate a limitation of the spatial detection range of the infrastructure facility 20. The cyclist 30, who is (not yet) perceptible by the vehicle 10, is detected by the infrastructure facility 20, and with regard to the cyclist 30, the infrastructure facility 20 sends V2X communication data, which is received by the vehicle 10, in particular its communication device, and (or data derived therefrom) is provided to the warning device 12 of the vehicle for determining a warning variable and / or warning notification. The wireless V2X communication between the infrastructure facility 20 and the vehicle 10 is identified in Fig. 2 by the reference numeral S. Fig. 3 shows an illustration of a process for generating a warning notification S4 for output to a driver of a vehicle. In step S1, relevant Car2X objects orV2X objects, i.e. objects (or road users) I1, I2 detected via V2X communication, are selected. Reference symbol I3 denotes an icon which illustrates the use of ego data, for example a speed of the vehicle bus. This results in "bicycle awareness" or increased attention of the driver with regard to a cyclist, which can be taken into account in further vehicle guidance (S4), marked in Fig. 3 with reference symbol S3. S4 denotes an output (here display) of a graphical awareness notification or warning notification in the central display, which shows a live 3D environment, and via smartlight in a method step S2. Fig. 4 shows three example display contents for outputting warning notifications to a driver to increase the driver's awareness or attention with regard to a crossing cyclist.In the upper segment of the figure, reference numeral 18 indicates a roadway currently being traveled by a user's vehicle, which the driver sees, for example, through the vehicle's windshield. In a display, such as a center display, or a (front) display in the driver's field of vision and / or a head-up display, a symbolically represented cyclist can be displayed together with a warning triangle IW as a graphic warning symbol I1. Preferably, a further graphic symbol R indicates a direction from which the cyclist is expected from the perspective of the warning driver. The direction symbol R can, for example, be displayed in the form of an arrow. Preferably, additionally or alternatively, a text notification IT can be displayed to the driver for further explanation, such as "Attention, cyclist." The lower left part of Fig. 4 shows a 3D representation of the surroundings, which is preferably displayed to the driver.Reference numeral 17 denotes a display in which the 3D representation of the surroundings in front of the vehicle in the direction of travel is output. Reference numeral A8 denotes the roadway currently traveled by the driver's vehicle. The driver's (ego) vehicle itself is also represented symbolically in the 3D representation of the surroundings, so that the driver can quickly orientate himself. Also shown in the 3D representation of the surroundings is the bicycle lane A6, which crosses the currently traveled roadway A8, with a cyclist A32 on it. Cyclist A32 is, in particular, a road user who has a potential for collision with the (ego) vehicle.Preferably, the 3D environment representation only displays objects that are necessary to communicate a potential collision area and the relative position of the (ego) vehicle to the current vehicle position, as well as the relative position of the potentially colliding road users, to the driver. In this way, the driver can acquire all necessary information as quickly as possible. The lower right sub-image shows another preferred option for outputting warning notifications R to a driver / user of a (ego) vehicle regarding a potential collision with a cyclist, here using an AR HUD (abbreviation for augmented reality head-up display). Reference numeral 8 shows the roadway currently being traveled, visible to the driver through the windshield.The reference symbol AVL denotes a traffic sign visible through the windshield of the vehicle to the driver / user of the vehicle, warning of cyclists. The partial image also shows that no cyclist is yet visible to the driver. The AR-HUD displays three arrowheads R pointing to the right as a navigation instruction for the driver. Preferably, a color of the navigation instruction is changed as a warning notification for the driver (for example, a change from the blue arrowheads to an orange (or red) arrowhead) and / or a form of the navigation instruction is changed (for example, a change of a (last) dash to a vertical bar) to signal the danger. It is conceivable, for example, that the warning notification(s) are displayed above a (current) speed display 19 of the vehicle and / or a navigation indication 19. Fig.5 shows a further traffic situation in which a vehicle 10 equipped with a warning device 12 according to the invention according to a preferred embodiment can be warned of a dangerous situation resulting from a possible collision with cyclists 32 crossing the roadway of the vehicle 10 or from insufficient safety distance by issuing a warning notification. For this purpose, the warning device 12, which receives the position information of all five cyclists 32 located in the collision zone 2, identifies those cyclists 32 whose predicted movement paths intersect with the movement path predicted for the (ego) vehicle 10. In the traffic situation shown in Fig. 5, these are the cyclists 32 marked with the reference symbol M. Figs. 6a - 6c show schematic representations to illustrate prediction probabilities. Fig.Figure 6a shows a traffic situation from road traffic with a collision zone, which in this case is a T-junction. The reference symbol M8 denotes the road side lines delimiting the roadway for motor vehicles. The reference symbol M6 marks the boundary line of the bicycle lane. The reference symbol M7 marks the central reservation between two opposing lanes of the roadway. The reference symbol 10, in turn, denotes a vehicle (here a car) as a road user. This vehicle has a warning device 12, which is attached to the vehicle in this case. The vehicle 10 is moving along a roadway in the direction of the arrow FR towards the T-junction. The roadway on which the vehicle 10 is located in the traffic situation shown in Figure 6a merges into a roadway that runs in a straight line here.On the side of the roadway facing vehicle 10, into which vehicle 10 merges, there is a bicycle lane, bordered by bicycle lane markings M6 on the one hand and roadway markings M8 on the other. A cyclist is located in the bicycle lane, schematically represented here as a circle 32. This cyclist 32 is moving in the direction of the arrow, i.e., in a direction from right to left in the plane of the figure. Thus, cyclist 32 is moving toward the intersection or junction area of ​​the roadway on which vehicle 10 is currently located. Reference symbol P10 denotes a predicted path of movement of vehicle 10 during an (intended) left turn from roadway MF, which merges into the (straight-line) roadway GF, into the left-hand section of the (straight-line) roadway GF, as seen from vehicle 10.The (dashed) ellipses designated by reference numerals 34 and 36 indicate prediction probabilities for a (future) location of the cyclist 32 at two different (future) points in time. Both ellipses 34, 36 intersect the predicted movement path P10 of the vehicle 10 and illustrate that, given the corresponding movement behavior of the vehicle 10 and the cyclist 32, a collision between the vehicle 10 and the cyclist 32 is or could be expected. As the ellipses 34, 36 already illustrate in Fig. 6a with respect to the cyclist 32, Fig. 6b again shows that a prediction of a future position of both the vehicle 10 and the cyclist 32 in Fig. 6a is associated with a certain prediction inaccuracy. Instead of a probability of being in 2D space (as illustrated by the ellipses in Fig.6a), a one-dimensional approach is now used.For example, a location probability along a predicted line, the movement path of a road user, is used, i.e., a determination in 1D space. This offers the advantage that determining a one-dimensional location probability is less computationally intensive than a two-dimensional location probability, yet still provides a sufficiently precise model for predicting the relative position of two road users. With respect to cyclist 32, the cyclist has a determinable position (indicated by the circle) at a given initial time (see arrow). At a later time, the cyclist's probable location can be approximated using a Gaussian function GRO.At an even later point in time, approximately time t, a probable location of the cyclist can again be described by a location probability function GRT, which can also essentially have a Gaussian curve, which has a greater variance than the location probability function GRO at the previous point in time. G10 characterizes the location probability function for the vehicle 10 at time t. Here, this has a greater variance or dispersion than the location probability function GRT of the cyclist at time t, also due to the higher vehicle speed. Using the two location probability functions G10 and GRT, it is possible, for example, to calculate the probability that the vehicle 10 and the cyclist 32 will collide at time t (in the collision area). This is illustrated in Fig. 6c.Here, an example probability curve is shown, which indicates the probability of a collision between vehicle 10 and cyclist 32 as a function of a time t measured from a starting point. The assumption here is that both location probability functions, i.e., of vehicle 10 and cyclist 32, reach a maximum at time t. K at the intersection point of their two movement path lines. A movement path line can be understood as the line along which a movement path of a road user runs. In other words, the probability for both road users here, vehicle 10 and the cyclist, is highest that they are at time t Kboth reach the intersection point (simultaneously). Furthermore, the question of whether a warning notification should be issued to one of the road users should take into account not only an exact meeting at the (geometric) intersection point of their (geometric) movement path lines, but also reaching the intersection point within a specified very short time interval or having already reached the intersection point less than a specified time period (e.g. between -t s and + t S ). This can, for example, ensure that the two road users maintain a sufficient safety distance from each other when passing the collision area (particularly by taking into account the probability curve within the time interval T G). Fig.7 a - c show the effect of selecting a specific prediction model for the predicted movement paths of the road users in comparison to statistically collected data of the respective road users. In the boxplot diagrams shown in Fig.7a - 7c, the predicted time duration TTPIP based on the movement path prediction model selected for each road user (averaged over a bin) is shown along the abscissa axis (horizontal coordinate axis). pred until the (specified) crossing point is reached. The abscissa shows the time period (averaged over a bin) TTPIP pred(in seconds), which is determined using the selected path prediction model for the period within which the road user will reach the (prescribed) intersection point (“PIP”) (after calculation by the path prediction model). Along the ordinate axis (vertical coordinate axis), the difference between the predicted time and the actually measured time until reaching the PIP (in seconds), i.e., TTPIP pred – TTPIP true , for the respective predicted time period (on the abscissa). The center line t M in the box indicates the median of the data. Half of the data lies above this value, the other half below. The upper tail of the distribution t oP or lower end of distribution t uPThe box indicates the 75th and 25th quantiles, respectively. The lines extending from the box are called "whiskers" (also "antennae") and represent the expected variation of the data. The whiskers extend beyond the box by 1.5 times the interquartile range, i.e., 1.5 times the upper and lower distribution ends of the box. If the data do not extend to the end of the whiskers, the whiskers extend to the minimum and maximum data values. The points P A In Figs. 7a – 7c, so-called outliers are indicated, which fall above or below the end of the whiskers. Fig. 7a shows such an evaluation for turning cars, assuming that the speed remains constant, i.e.: where: : time to intersection point : distance to path intersection point (distance to the PIP “Path Intersection Point”) : initial velocity (initial velocity) In comparison, Fig. 7b shows the same representation for the turning cars assuming that the acceleration remains constant: Comparing Fig.7a and Fig.7b, it can be clearly seen that the deceleration of the turning vehicle when calculated with the constant speed model leads to the vehicle actually being in front of the intersection point at the predicted time (predicted TTPIP pred compared to measured TTPIP true negative, so E TTPIP = TTPIP pred – TTPIP true ). When using the model shown in Fig.7b with assumed constant acceleration (in this case, of course, negative), the variance is similarly large, the median t M or the percentile toP , t uPare, however, close to the zero line. However, due to the fairly constant speed on the straight stretch, cyclists can be well predicted with the corresponding model in the driving maneuver considered here by cyclists (not turning, but crossing the intersection in a straight line, during which cyclists usually do not brake), as shown in Fig. 7c. Fig. 8 and Fig. 9 illustrate example collision zones in road traffic in which the movement paths of various road users, in this case motor vehicles and cyclists, can cross. The two collision zones are each shown from a bird's eye view, i.e., from above, from which a drone, for example, preferentially takes (camera) images to capture the collision zone and / or a surrounding ground area.Such (camera) recordings are preferably used to record and / or (especially statistically) determine the movement paths of various road users. From the recorded or measured movement paths of various road users, a (especially statistical) uncertainty value with respect to a forecast accuracy can be determined. In Fig. 8, the reference symbol P denotes this. R a (geometric) movement path line of a cyclist, along which a (measured, statistical and / or averaged) movement path extends, i.e., along which the cyclist moves. The reference symbol P CLdenotes a (geometric) movement path line of a left-turning (motor) vehicle (e.g., a car), along which a (measured, statistical, and / or averaged) movement path extends, i.e., along which the left-turning vehicle moves before and / or during and / or after the left turn. The reference symbol P C indicates the (expected and / or expected) collision point, which here is the intersection point of the motion path line P CL of the left-turning vehicle and the movement path line P R of the cyclist crossing the intersection. In Fig.9, the reference symbols P R , P r1 , P r2 , P r3 , P r4 , P r5 the (geometric) movement path lines of cyclists who cross the intersection in a straight line (starting from different lanes and sides of the road) (P r5 flowing into P r2 ; P r3 ; Pr1 flowing into P R ) or turn left at the intersection (P r4 flowing into P r2 ). Again, the respective movement path lines show a (at least partial) course in the collision area (or a surrounding area) along which (statistically and / or averagely and / or on average) the movement paths of the cyclists extend or along which the cyclists move. The reference symbols P CR1 , P CL2 , P C3 , P C6 mark the (geometric) movement path lines of (motor) vehicles (such as cars) (moving on a roadway), along which the movement paths of these vehicles are located during their various maneuvers, for example when driving straight ahead (P C3 ; P C6 ), when turning right (P CL2 ) and when turning left (P CL2), and lanes extend. In the traffic routing shown here, which is predetermined by the intersection, there are a multitude of collision points between (motor) vehicles and cyclists (or between cyclists and cyclists or between (motor) vehicles and (motor) vehicles). Fig. 10 shows a representation to illustrate or visualize a predicted and / or a predetermined movement path P P a left-turn scenario based on a large number of mapped movement paths P G , here, for example, from (motor) vehicles (such as cars or cars) using the left-turn lane, whose measured movement trajectories were mapped or matched to a (parameterized) movement path. With the parameterization (selected here), the driven (motor) vehicle movement paths P G at most 5 m from the predicted movement path P PFig.11 shows an example histogram showing the time difference TTPIP pred - TTPIP true a predicted or forecasted time duration for a (motor) vehicle (or car) and the actual time required to reach the (predetermined) collision point (given a predetermined predicted time duration to reach the collision point). The H axis indicates the frequency of the prediction time errors, and the horizontal axis the time difference TTPIP pred - TTPIP true (measured in seconds). Fig. 12 shows another example histogram or kernel density estimation (solid line marked with reference symbol y) with a Gaussian kernel for a difference TTPIP pred - TTPIP truebetween the time predicted for the (motor) vehicles (cars) on the one hand and the actual TTPIP (TTPIP abbreviation for “Time to Path Intersection Point”) 3.0 s before the predicted arrival at the collision point (“path intersection point”). A probability density P is plotted on the ordinate. The histogram visualizes the prediction error. The Gaussian kernel density estimation y preferably makes it possible to determine and / or estimate fpred , the probability density function, abbreviated as “PDF” for “Probability Density Function” of the prediction error distribution y for this road user type and path. Fig. 13 shows a visualization of an example determination of a probability that a vehicle will cross the movement path of a VRU within a time period (“gap”) of [-0.5 s, 0.5 s].The probability density P is plotted on the left ordinate. The quantity “E” is plotted on the abscissa (horizontal axis). pred + TTPIP pred ” where E pred = TTPIP pred - TTPIP true and TTPIP pred = TTPIP pred,car - TTPIP pred,cyc . TTPIP pred therefore indicates the difference between the predicted time of the (motor) vehicle (e.g., car) and the predicted time of the bicycle (or VRUs) until the PIP is reached. These two predicted time periods TTPIP pred,car and TTPIP pred,cyc exactly the same size, so at TTPIP pred =0, this means that both road users arrive at the PIP at the same time. "y1: Car" denotes the probability density of the (motor) vehicle or car. "y1: VRU ( TTPIPpred = 0 s)" and "y1: VRU ( TTPIPpred = 3 s)" denote the probability density distribution P Xfor a VRU, here a cyclist. Based on a probability density function f pred,car (“Probability Density Function”, abbreviation “PDF”) of the prediction error distribution for the (motor) vehicle, illustrated in Fig.13 by the curve “y1: Car”, and a probability density function f pred,VRU The prediction error distribution for a cyclist, illustrated in Fig. 13 by the curve “y1:VRU (TTPIPpred = 0 s)”, can be used to determine a measure of the probability that the vehicle crosses the movement path of the VRU within the time period (“gap”) of [-0.5 s, 0.5 s] by the relationship: On the right ordinate is P X The hatched area dt shows the differential for the vehicle or car and the area marked with “t±T gap, 1,2 " marked area is the integration area of ​​the inner integral with the integration limits t+T gap,1 and T+T gap,2the probability density for the VRU (cyclist). This marked area represents a factor of the function to be integrated in the outer integral with differential dt. Curve y2 shows the integration result for TTPIP pred = 0 s at t. To determine the probability for the other cases in which TTPIP pred is not zero, i.e., if the two road users do not arrive at the PIP at the same time according to their predicted time until reaching the PIP, the PDF for the cyclist, for example, must be shifted by the difference in the predicted arrival times at the PIP, as illustrated in the thinner line “y1: VRU (TTPIPpred = 3 s)” in Fig. 13. For each shifted PDF, a measure of the probability can be calculated as in the above-mentioned determination for Px (car, VRU, TTPIP pred , T gap,1 ,T gap,2 ) can be calculated, resulting in a set of probabilities for different combinations of TTPIP predThe resulting probability allows for estimating the positive predictive value (PPV) based on the number of true positive (nTP) and false positive (nFP) awareness notifications for this specific scenario between the (motor) vehicle (e.g., car) and the VRU, including the prediction algorithm used, the specific time period, and the desired "gap" (time period) that must be exceeded to trigger the awareness notification: This makes it possible to determine whether the awareness cue (attention stimulus) is considered to be relevant for the driver at a given time TTPIP pred,car and a given "gap" (time period) TTPIP pred relevant is expected. Furthermore, the calculation of an average value of P X about the values ​​of TTPIP predfor a specific TTPIP pred,car , the positive predictive value for the application parameterization for T gap to be estimated as follows: Fig.14 shows a boxplot diagram to illustrate the prediction error in determining the TTPIP pred(TTPIP abbreviation for "Time to Path Intersection Point") of the predicted or predicted time of a (left-) turning (motor vehicle) until reaching a given collision point in comparison of two different motion models for the movement of the vehicle. The reference symbol "CV" denotes the boxplots for a selected motion model for determining the predicted or precise motion path in which the (motor vehicle) moves at a constant speed. The reference symbol "CA" denotes the boxplots (shown in bold dashed lines) for a selected motion model for determining the predicted or precise motion path in which the (motor vehicle) moves with constant acceleration. Along the horizontal axis is the predicted time TTPIP pred, caruntil the PIP is reached. Along the vertical axis, TTPIP err, car = TTPIP pred, car – TTPIP true, car (in seconds) plots the difference between the predicted or forecasted TTPIP and the actual (measured) TTPIP. The measured TTPIP was determined based on movement paths obtained from drone footage of the collision area (here, the intersection). Positive values ​​of TTPIP err, car correspond to cases in which the predicted time TTPIP pred, car until reaching the PIP (Path Intersection Point) was higher than the measured TTPIP true, carTime to reach the PIP. The vehicle (e.g., car) therefore reaches the PIP earlier than predicted. Conversely, with negative values, the vehicle (e.g., car) arrived at the PIP later than predicted. The graph shows that the uncertainty of the forecast / prediction, visualized by the boxes representing the 0.25 and 0.75 percentiles and those with larger TTPIP pred, car increase. Furthermore, the graph shows that the absolute value of the median prediction error increases with the prediction duration for the constant-speed motion model CV up to approximately 1.8 s for a TTPIP pred, car of 5 s. 50% of the vehicles in the drone dataset reached the intersection point 0.56 s to 2.66 s later than predicted. In contrast, the absolute median prediction error for the constant acceleration motion model (CA) is significantly lower at -0.46 s for a TTPIP. pred,carof 5 s. Since drivers decelerate when turning, which is ignored in the constant-speed-acceleration model, the constant-acceleration model is more accurate for predicting this scenario. For a cyclist traveling in a straight line across the collision zone, no discernible difference was found between the two models (not shown here). For the two subsequent figures, only the constant-acceleration model was used for both the left-turning car driver and the cyclist crossing the intersection in a straight line (cf. the traffic situation and collision zone shown in Fig. 8). Fig. 15 shows a measure of the resulting probability that the (motor) vehicle (or car) and the cyclist reach the PIP (path intersection point), i.e. the collision point, within the defined time period (“gap”) in the event that their TTPIP pred arrival at the same time (i.e. TTPIPpred,car = TTPIP pred,cyc ) is indicated. A measure of the probability could, for example, be determined by the formula, where f pred,car and f pred,VRU is the respective probability density function (“Probability Density Function”, abbreviation “PDF”) of the prediction error distribution for the road user and the path (see also Fig. 12). P X or the y-axis indicates the probability that the (motor) vehicle (or the car) follows the cyclist's path within a certain time period T gap Along the abscissa (horizontal axis) is the predicted time period TTPIP pred, car until the PIP is reached. The graph shows two things: that the resulting probability increases with increasing TTPIP preddecreases and that the probability for a given TTPIP is greater, the greater the critical time gap for the attention stimulus or for the awareness cue T gap For example, for a time gap of -2 s to 2 s, the resulting probability that both road users, predicting the same arrival at the PIP (collision point), will cross the PIP within the defined gap or time gap is 0.76 for a TTPIP predof 5 s. For an asymmetric time span of -4 s to 2 s, which takes into account the acceptable time span of 4 s, but allows a passing of 2 s behind the cyclist, the resulting probability is 0.93. Fig. 16 similarly visualizes the resulting probability for the time span (“gap”) [-4 s, 2 s], but also considers various combinations of TTPIP for the (motor) vehicle (or car) and the cyclist. Fig. 16 shows a heatmap representation of the probability that the (motor) vehicle (or car) will follow the cyclist's movement path within the defined time span (“gap”) for various combinations of TTPIP. pred,car and TTPIP. The grayscale axis shows the probability that the vehicle will cross the cyclist's path within a time gap T gap of [-4 s, 2 s]. At low TTPIP predFor the vehicle, the probability drops sharply at the limits of the defined time period (“gap”). At higher TTPIP pred The probability decreases more slowly. The result allows the conclusion that the resulting probability that the (motor) vehicle and the cyclist will cross within the defined time period (“gap”) from the predicted time of the (motor) vehicle to the PIP (TTPIP pred,car ) and the difference between the predicted arrival time of the cyclist. For example, the resulting probability for a TTPIP pred,car = 5 s and a TTPIP pred of 1 s (it is expected that the cyclist reaches the PIP before the (motor) vehicle or car) 0.84 and at a TTPIP predof -1 s (the cyclist is expected to reach the PIP after the (motor) vehicle or car) is 0.94. For 2 s and -2 s the probabilities are 0.68 and 0.81. If the cyclist reaches the PIP (i.e. the collision point) 4 s later than the (motor) vehicle (or car), the probability is only 0.32. Averaging the probabilities for a time span from -4 s to 2 s results in a positive predictive value (PPV) for this parameterization of 0.77. Of course, the resulting probability is lower for the case in which the (motor) vehicle and the cyclist are expected to arrive at the PIP at the same time.With a = 0 at the time the attention notification is issued, the cyclist could, for example, still arrive up to 4 s later than predicted and would still be within the defined time period for a successful awareness cue. If an offset of, for example, 3 s is used, a false positive result is obtained if the cyclist arrives 1 s later than predicted. This results in the lower overall probability. Fig. 17 shows an illustration of a method according to the invention according to a preferred embodiment for determining swarm movement data to generate a global map of swarm movement data and / or movement paths that is characteristic of the movement of a large number of (different) road users. Fig.17 shows a section of a road network of a traffic area with a large number of streets or carriageways that are connected to one another here, exemplarily identified by the reference symbols S1, S2, S3, S4. The reference symbol 1 designates a road user who is collecting (swarm movement) data and who is moving in a straight line, i.e. in particular without turning, along the road S1, in the traffic situation shown here from the road section of the road S1 at the bottom left in Fig. 17 in the direction of the top right of the road S1 in Fig. 17. The reference symbol 110 designates another road user, here a motor vehicle, which, starting from the road S4, is moving along it in the direction of the road S1, turns left at the junction of the road S4 and the road S1 and then drives along it.Reference numeral 112 designates another road user, here a motor vehicle, which, starting from road S3, turns into road S2, then turns right onto road S4, and moves toward road S1. During their journey, the two road users 110, 112 transmit V2V communication data, which vehicle 1 can receive via a communication device 114. Figure 18 shows a schematic structure of a vehicle 1 with the communication device 114, by means of which V2V communication data or V2X communication data can be received.The communication connection created by the transmission and reception of V2V communication data between the road users 110 and 112 sending the V2V communication data, on the one hand, and the vehicle 1 receiving the V2V communication data, on the other hand, is illustrated by the lines labeled "V2V-110" and "V2V-112." The different road users are represented by different symbols. Identical symbols indicate the position of the respective road users at different times.The symbols indicate the respective positions of vehicle 1 and the other road users 110 and 112, as they can be determined by determining their own position data (vehicle 1) and by means of the V2V communication data (in particular received by vehicle 1), which is particularly characteristic of the current position of the road users transmitting the V2V communication data. The V2V communication data is transmitted at specific times. For each of these times, one or the vehicle 1 therefore receives, in particular, information on the position, speed, acceleration, heading, and the like of these road users 110, 112 transmitting V2V communication data, as long as they are within the reception range of the respective V2V communication.The arrow marked “V2V-110” illustrates that the vehicle 1 itself can no longer receive V2V communication data from the vehicle 110 at a time when, due to the large distance and the different direction of travel of the vehicles 1 and 110, the vehicle 1 can no longer detect the vehicle itself using on-board sensors, such as cameras and / or LIDAR sensors and / or radar sensors. The situation is similar with vehicle 112 transmitting V2V communication data. The communication connection “V2V-112” shown as an example in Fig. 17 illustrates that vehicle 1 is already receiving V2V communication data from road user 112 at a time when the road user is outside the field of vision of the driver of vehicle 1, outside the detection range of the vehicle’s own environment detection devices (i.e. the on-board sensors) and is not even on a road S4 that leads into road S1.Between vehicle 1 and vehicle 112, several buildings, such as houses, could also be located along the direct connection to the communication link designated by the communication link "V2V-112." The vehicle 1 further comprises, as shown in the schematic structure in Fig. 18, a swarm data determination device 116, which determines local swarm (movement) data based on vehicle-to-vehicle communication data (V2V communication data) received from the communication device 114 and provides the local swarm (movement) data, in particular to a communication device 115 of the vehicle 1 (schematically shown in Fig. 18), for transmission (designated by reference numeral 118) to an external (in particular cloud-based) storage device 120, preferably to an external backend server and / or a cloud-based server.A map generation device 130 can then retrieve the swarm data or swarm movement data stored on the storage device 120 and use it to generate a global map of movement data or movement paths, in particular depending on a type of road user. Reference numeral 100 denotes a system for generating a global map of swarm movement data and / or swarm movement paths, and preferably of predicted movement paths. The system 100 comprises, in particular, at least one vehicle 1 collecting movement data and / or receiving vehicle-to-vehicle communication data (V2V communication data), and preferably a plurality of such vehicles. These preferably each have a swarm data determination device, which, based on the received V2V communication data, sends swarm movement data (by means of a communication device) to a storage device 120, in particular an external one.The system preferably also comprises the external storage device 120 and particularly preferably a map generation device. The applicant reserves the right to claim all features disclosed in the application documents as essential to the invention, provided that they are novel, individually or in combination, over the prior art. It is further noted that the individual figures also describe features which may be advantageous in themselves. The person skilled in the art will immediately recognize that a specific feature described in a figure may also be advantageous without adopting further features from that figure. Furthermore, the person skilled in the art will recognize that advantages may also arise from a combination of several features shown in individual or different figures.

[0002] List of reference symbols 1 Vehicle 2 Collision area 3, 5 Roadway 4 Building 7 Zebra crossing area with zebra crossing 8 Currently used roadway 10 Vehicle 12 Warning device 13 Warning message / ADAS data provision 14 Communication device 16 Warning notification, here VRU Awareness 17 Display 18 Currently used roadway 19 Speed ​​and / or navigation information 20 (networked) infrastructure facility, e.g.Traffic light system 22 Detected cyclist 24 Detected pedestrian 30 Cyclist 32 Cyclist 4, 36 Representation of the prediction probability of a future location zone S V2X communication – S4 steps R Direction symbol I1, IW Warning symbol IT Text notification IR Graphic direction symbol A10 Displayed ego-vehicle A8 Displayed traveled roadway A6 Crossing cycle lane A32 Cyclist AVL Displayed traffic sign M Marking M8 Road shoulder M7 Central reservation M6 Bicycle lane boundary line P10 Predicted movement path of the vehicle 10 GF Straight-line roadway MF Roadway merging into the straight-line roadway GF G10, GRT, GRO Location probability distribution t. oP , t uP Percentile t M Median t oA , t uA Whisker P A Outlier P R Movement path line of straight-ahead cyclists P CL Movement path line of left-turning cars P CCollision area, especially collision point P r1 , P r2 , P r3 , P r4 , P r5 Movement path line of cyclists P CR1 , P CL2 , P C3 , P C6 Movement path line of (motor) vehicles P P Visualization of a predicted movement path P G Visualization of traveled movement paths H Frequency y Prediction error distribution 110, 112 V2V communication data sending road users 114 Communication device 115 Communication device 116 Swarm data acquisition device S1 – S4 roads 130 Map generation device 120 External storage device

Claims

1. Method for providing an adaptation notification, preferably a warning notification, to at least one ego-road user (10) to be notified, for adapting at least one driving function to at least one surrounding road user (32) located and / or moving in a surrounding area (2) of the ego-road user, by a warning device (12) which has predicted movement paths (P CL , P R ) are and / or are provided, along which a movement of the road users (10, 32) is to be expected, wherein the warning device (12) is based on the predicted movement paths (P CL , P R) determines at least one notification variable which is characteristic for providing the adaptation notification and which is preferably characteristic of whether an adaptation notification (I1, A32) is to be provided and / or at what point in time the adaptation notification (I1, A32) is to be provided, characterized in that the warning device (12) determines the notification variable as a function of at least one, in particular statistical, swarm prediction variable for determining a prediction quality of at least one of the predicted movement paths (P CL , P R), wherein the at least one swarm prediction variable is determined on the basis of global swarm movement data of a plurality of road users, which are generated in a plurality of different traffic areas (2).

2. Method according to claim 1, characterized in that the warning device (12) determines, depending on a predetermined and / or predeterminable minimum prediction probability variable, on the basis of the swarm prediction variable, whether the adaptation notification is to be provided.

3. Method according to at least one preceding claim, characterized in that the warning device (12) is provided with a a characteristic safety value is predetermined and / or can be predetermined, and on the basis of the safety value, depending on the predicted movement path of the ego road user and at least one swarm prediction value for determining a prediction quality of the predicted movement path of the ego road user, and depending on the predicted movement path of the surrounding road user and at least one swarm prediction value for determining a prediction quality of the predicted movement path, the warning device (12) determines how high the probability is for the ego road user and the surrounding road user to approach each other closer than the predetermined and / or predeterminable distance.

4. Method according to at least one preceding claim, characterized in that at least one characteristic safety value for the ego road user and / or for the surrounding road user to reach a collision area,in particular a collision point, a predicted time is determined and, depending on a predetermined period of time, a probability is determined that both the ego road user and the surrounding road user will reach the collision area within the predetermined period of time around the predicted time.

5. Method according to at least one preceding claim, characterized in that an environmental situation determination device of the ego road user and / or the surrounding road user collects and / or determines environmental situation data which are characteristic of at least one influencing variable which is selected from a group of influencing variables which include a topological variable characteristic of a topology of at least one movement lane of a collision area, in particular a structure of an intersection, a topological variable characteristic of a presence of at least one further road user,in particular in the collision area, a characteristic quantity, a traffic density, a speed quantity characteristic of at least one speed regulated for at least one (involved) road user in the collision area, a traffic control quantity characteristic of a predetermined traffic regulation, in particular right-of-way regulation, an expansion quantity characteristic of a geometric extent of the collision area and / or the intersection area of ​​predetermined movement lanes of road users, a road user variable characteristic of a class and / or group of a road user involved, a cultural variable characteristic of a geographical area in particular, an environmental variable characteristic of at least one environmental condition which can be influenced in particular by a visibility ratio, a time variable characteristic of a season and / or time of day, a weather variable characteristic of a current weather condition and the like, as well as combinations thereof, wherein the at least one swarm prediction variable is determined as a function of the at least one influencing variable.

6. Method according to at least one preceding claim, characterized in that the warning device (12) determines the predicted movement path of the at least one surrounding road user (32) on the basis of data detected by a communication device (14),vehicle-to-X communication data (V2X communication data) and preferably vehicle-to-vehicle communication data (V2V communication data) transmitted by the at least one surrounding road user (32) are determined.

7. Method for determining at least local and preferably at least section-wise global swarm movement data for generating a global map of swarm movement data and preferably of predicted movement paths, which are characteristic of a plurality of movement paths of a plurality of different road users, by means of at least one swarm data determination device (116) of at least one swarm data-determining road user, in particular a vehicle (1), characterized in that the swarm data determination device (116) is based on data received from a communication device (114) of the at least one swarm data-determining road user (1),determines local swarm movement data from vehicle-to-vehicle communication data (V2V-110, V2V-112) sent by at least one vehicle-to-vehicle communication data-sending road user (110, 112) and provides the local swarm movement data to a communication device (115), in particular of the swarm data-determining road user (1), for transmission to an external storage device (120), preferably to an external backend server.

8. Method according to the preceding claim, characterized in that the vehicle-to-vehicle communication data, which are transmitted at a transmission time by the at least one road user sending vehicle-to-vehicle communication data within a common data packet, comprise, with respect to the transmission time, current movement data of the road user sending the communication data and historical movement data of the road user sending the communication data at a plurality of different detection times preceding the transmission time. 9.Method according to at least one of the two preceding claims, characterized in that the swarm data determination device determines local swarm data relating to a plurality of different road users sending vehicle-to-vehicle communication data within a predetermined and / or predeterminable period of time and provides it, in particular in the form of a common data packet, to the communication device for transmission to the external storage device. 10.Method according to at least one of the three preceding claims, characterized in that the swarm data determination device analyzes a plurality of vehicle-to-vehicle communication data received by it, in particular transmitted by exactly one road user transmitting communication data, or data derived therefrom, for redundant movement data of the road user transmitting communication data at essentially simultaneous acquisition times and removes this redundant movement data to determine the local swarm data.

11. Method according to at least one of the four preceding claims, characterized in that the swarm data determination device determines local swarm data at regular time intervals and provides it to the communication device for transmission to the external storage device. 12.Method for determining at least one, in particular statistical, swarm prediction variable of a predicted movement path of a road user of a predetermined type located in a predetermined traffic area (2) on the basis of a plurality of determined movement paths which are based on a. A plurality of swarm movement data of a plurality of road users are recorded, characterized in that the plurality of swarm movement data is a plurality of global swarm movement data which is generated and / or collected in a plurality of mutually different traffic areas (2).

13. Method according to the preceding claim, characterized in that the swarm prediction variable is characteristic of a prediction quality of a predetermined predicted movement path.

14. Method according to the preceding claim, characterized in that, to determine the swarm prediction variable, the plurality of determined movement paths are clustered into driving maneuvers of different types.

15. Method according to one of the three preceding claims, characterized inthat to determine the swarm prediction variable, the plurality of determined movement paths of the plurality of road users are statistically evaluated as a function of at least one and preferably a plurality of influencing variables, wherein the influencing variable or the influencing variables are selected from a group of influencing variables which include a topological variable characteristic of a topology of at least one movement lane of a collision zone, in particular a structure of an intersection, a variable characteristic of a presence of at least one other road user, in particular in the collision zone, a traffic density, a speed variable characteristic of at least one (involved) road user in the collision zone, a traffic control variable characteristic of a predetermined traffic control, in particular a right-of-way control,an extent characteristic of a geometric extent of the collision area and / or the intersection area of ​​predetermined movement paths of road users, a road user size characteristic of a class and / or group of a road user involved, a cultural size characteristic of a geographical area in particular, an environmental size characteristic of at least one environmental condition in particular which can be influenced by a visibility ratio, a time size characteristic of a season and / or time of day, a weather variable characteristic of the current weather and the like, as well as combinations thereof.

16. A machine-readable and, in particular, computer-implementable movement model for predicting a movement path of a road user of a predetermined type, preferably determined and / or to be determined by the road user, preferably detected by an environment detection device of a road user, as a function of a position of the road user, wherein a variable characteristic of the position and at least one variable characteristic of a type of road user are fed to the movement model as input variables, wherein the movement model, based on a plurality of model parameters, maps the input variables to output variables characteristic of the movement path to be predicted, characterized in thatthat the plurality of model parameters is determined on the basis of global swarm movement data of a plurality of road users, which are generated in a plurality of different traffic areas.

17. A warning device (12) for an ego road user (10), in particular for a vehicle, for providing an adaptation notification, preferably a warning notification, to at least one ego road user (10) to be notified, for adapting at least one driving function to at least one surrounding road user (32) located and / or moving in an environmental area (2) of the ego road user, wherein the warning device (12) is provided with respectively predicted movement paths (P, CL , P R) are and / or are provided, along which a movement of the road users (10, 32) is to be expected, wherein the warning device (12) is suitable and intended to, on the basis of the predicted movement paths (P CL , P R ) to determine at least one notification variable which is characteristic for providing the adaptation notification and which is preferably characteristic of whether an adaptation notification (I1, A32) is to be provided and / or at what point in time the adaptation notification (I1, A32) is to be provided, characterized in that the warning device (12) is suitable and intended to determine the notification variable as a function of at least one, in particular statistical, swarm prediction variable for determining a prediction quality of at least one of the predicted movement paths (P CL , P R), wherein the at least one swarm prediction variable is determined on the basis of global swarm movement data of a plurality of road users, which are generated in a plurality of mutually different traffic areas (2).

18. Vehicle (10), in particular a motor vehicle, comprising a warning device (12) according to the preceding claim.

19. Swarm data determination device (116) for a road user, in particular for a vehicle, in particular a motor vehicle, for determining, in particular at least (section-wise) local and preferably global swarm movement data for generating an at least section-wise global map of swarm movement data and / or predicted movement paths, which are characteristic of a plurality of movement paths of a plurality of different road users, characterized in thatthat the swarm data determination device (116) is suitable and intended to determine local swarm movement data based on vehicle-to-vehicle communication data (V2V-110, V2V-112) received by a communication device (114) of the swarm data determination device (116) and transmitted by at least one road user (110, 112) sending vehicle-to-vehicle communication data, and to provide the local swarm movement data to a communication device (115) of the swarm data determination device (116) for transmission to an external storage device (120), preferably to an external backend server.

20. Map generation device (130) for determining at least one, in particular statistical, swarm prediction variable of a predicted movement path of a road user of a given type located in a given traffic area (2) based on a plurality of determined movement paths,which are recorded on the basis of a plurality of swarm movement data of a plurality of road users, characterized in that the plurality of swarm movement data are global swarm movement data which are generated and / or collected in a plurality of mutually different traffic areas (2).

Images