Automatic repositioning for vehicles for area monitoring

Abstract

The invention relates to a method for monitoring a geographical area by a plurality of vehicles (1), comprising the steps of: defining (S1) an area in the geographical area for which continuous monitoring is desired; determining (S2) desired positions or position ranges for the vehicles (1) in order to enable continuous monitoring; automatically reparking (S3) at least a subset of the vehicles (1) to their desired positions or position ranges if a respective current location of the vehicles (1) does not allow continuous monitoring; carrying out (S4) monitoring of a respective sub-area of ​​the area of ​​the geographical area by a respective one of the plurality of vehicles (1); and combining (S5) data about the monitored sub-areas to form an area data set.

Description

[0001] The invention relates to a computer-implemented method for monitoring a geographical area by a plurality of vehicles.

[0002] It is known in the prior art to create an overall image by combining images taken with cameras of a vehicle by determining overlapping areas of the individual images from the cameras.

[0003] In this context, US 2022 / 0207756 A1 relates to a method comprising: determining an overlap region corresponding to an overlap between a first and a second image; determining differences between the first image and the second image in the overlap region; transforming at least the first image at least partially based on the differences.

[0004] In particular, the teaching of US 2022 / 0207756 A1 is limited to the perspective of a vehicle's sensors—the sensor fusion based on it is not supported by external data. For example, if the vehicle is parallel to a house wall, events on a street within the house wall or behind a corner at one end of the house wall cannot be observed. However, cases are conceivable in which a larger overall picture of a geographical area needs to be monitored.

[0005] DE 10 2024 001 755 A1 describes a method for application-specific area observation of an area, wherein sensor data of at least one selected sensor are provided by a central computing device to at least one motor vehicle in the area and / or a receiving unit depending on the application.

[0006] The object of the present invention is to monitor a geographical area as seamlessly as possible using a large number of vehicles.

[0007] The invention is based on the features of the independent claims. Advantageous developments and refinements are the subject of the dependent claims.

[0008] A first aspect of the invention relates to a computer-implemented method for monitoring a geographical area by a plurality of vehicles, comprising the steps: - Defining an area in the geographical area for which continuous monitoring is desired; - Determining desired positions or positioning ranges for the vehicles to enable continuous monitoring; - automatic re-parking of at least a subset of the vehicles to their desired positions or position ranges if the current location of the vehicles does not allow for continuous monitoring; - carrying out surveillance of a respective sub-area of ​​the area of ​​the geographical area by a respective one of the plurality of vehicles; and - Combining data about the monitored sub-areas into an area data set.

[0009] The current location of each of the vehicles is characterized by the fact that the respective vehicle is stationary, i.e., parked, at that location. Preferably, only a parked vehicle is used to monitor the area.

[0010] Depending on the embodiment of the computer-implemented method, the steps are carried out in the order mentioned or not. The automatic re-parking of at least one of the vehicles, if continuous monitoring is not possible in the current position of the vehicles, can take place before the monitoring of the sub-areas of the geographical area by one of the vehicles begins. In this case, the vehicles are positioned in advance in such a way that when the sub-areas are subsequently monitored by the individual vehicles, continuous monitoring is possible by combining the data on the monitored sub-areas. This is done in particular with the help of a digital map, based on this data, a computing unit can automatically optimize where the individual vehicles should be positioned in order to be able to cover the area of ​​interest in the geographical area without gaps.

[0011] In an alternative embodiment, the monitoring of the sub-areas by the individual vehicles is initiated in an initial step, and a reconstruction determines whether the area of ​​interest within the geographical area is already being monitored without interruption. This embodiment has the advantage that the exact orientation and position of each of the vehicles does not need to be known; instead, only the result of monitoring a respective sub-area is analyzed to determine whether it can contribute to complete, i.e., seamless, monitoring of the area of ​​interest.

[0012] The area of ​​interest of the geographical area includes, in particular, streets within the geographical area if traffic on those streets is to be monitored. For example, house surfaces, roofs, and other areas irrelevant to road traffic can be omitted and not counted towards the definition of continuous monitoring. If the purpose of continuous monitoring is to monitor the speed of other road users in the area of ​​interest within the geographical area, the area of ​​interest is preferably limited to streets with a uniform speed limit, such as a contiguous speed zone.

[0013] The data from the monitored sub-areas can be compiled into an overall image, the area dataset. This area dataset can be presented in a graphically understandable way, especially if the individual vehicles are monitored by cameras. The data can be compiled centrally or decentrally. In an "autonomous vehicle speed sensing array," the vehicles connect independently based on their position, orientation, and the availability of their vehicle sensors.

[0014] The defined area should be monitored in such a way that it can be carried out seamlessly. Seamless monitoring is achieved when the monitoring objective can be achieved without the object of monitoring being obscured at certain locations within the defined area. For example, if speed monitoring is carried out in the defined area by other road users, the defined area is preferably a road network on which the other road users can move. Seamless monitoring is ensured when a bicycle path through the road network can be clearly identified, particularly using the area data set.

[0015] The means by which seamless monitoring is achieved is by positioning the vehicles at desired positions or desired positioning ranges if their current location does not allow for seamless monitoring. If this is the case, at least one of the vehicles is automatically reparked so that the vehicles assume their desired positions or positioning ranges and can seamlessly monitor the defined area with the help of their own sensors. An algorithm can check which of the vehicles are actually suitable for reparking. In particular, a distinction can be made according to SAE level, and in particular only vehicles with an automation level of SAE Level 3 or higher, or 4 or higher can be commanded to automatically repark. It can also prove advantageous if only vehicles with SAE Level 5 are automatically reparked.

[0016] In this case, a respective vehicle can also check whether it can fully monitor its assigned sub-area. If this is not the case, the vehicle can transmit this information to a central coordination point, in particular a central processing unit, whereupon the central processing unit directs the vehicle to repark or removes it from monitoring entirely. In the first and / or second case, a recalculation is advantageously performed by a further execution of the optimization algorithm in order to determine an optimal distribution of the vehicles, taking into account this limited monitoring at this position or this position range.

[0017] The vehicle sensors are preferably those already provided in vehicles, as is standard in modern vehicles. Every modern vehicle has a camera unit configured for environmental monitoring. This camera can be used to observe the surroundings and other road users within this environment, and by determining overlapping areas of individual camera images from the respective vehicles, a consistent area dataset can be generated. In other words, the data recorded by the respective vehicles across their monitored sub-areas is consistently and coherently combined to create an area dataset as an overall image.

[0018] As mentioned above, automatic reparking occurs after determining a desired position or a desired position range for a respective vehicle. For this purpose, an optimization algorithm is advantageously executed automatically, which determines the desired positions or position range for the vehicles. Multi-objective optimization can be performed here to consider several objectives in a cost function, which is minimized. Restrictions can also be defined, for example, a maximum travel distance when reparking a vehicle from its current location or from the home of its owner / user. Furthermore, parking regulations and, if available, other locally applicable standards are preferably taken into account in the optimization algorithm, particularly as respective restrictions.

[0019] On the other hand, in contrast to restrictions, incentives can also be used, so that, for example, a re-parked vehicle receives a charging slot at an electric charging station, especially an inductive charging station. Further incentives can be provided, such as electric charging at a cheaper rate.

[0020] In a further embodiment, restrictions can be defined by owners or users of the vehicles. For example, it can be predetermined that a vehicle may not be reparked. It can also be provided that an owner / user can specify a maximum distance that must be observed when reparking.

[0021] According to an advantageous embodiment, the desired positions or position ranges for the vehicles are determined using a digital map in order to enable seamless monitoring.

[0022] The digital map already contains relevant geometric features, allowing the optimization of vehicle positioning to be carried out using this data. Shadows and obscurations, particularly from buildings, can also be taken into account. If the digital map also contains current information about other parked road users, these can also be taken into account.

[0023] According to a further advantageous embodiment, the desired positions or position ranges for the vehicles are determined by applying an optimization algorithm to determine an optimal vehicle distribution.

[0024] In particular, nonlinear optimization techniques are used, such as gradient-based optimization methods, quadratic optimization methods, or evolutionary algorithms. Pre-trained artificial neural networks can also be used for optimization purposes.

[0025] According to a further advantageous embodiment, an optimization goal is that a respective vehicle from the subset is positioned a minimum distance away from the current location and / or from the home of the owner or user of the respective vehicle when reparking.

[0026] On the one hand, minimizing this distance can be used as the optimization objective in a multi-objective optimization; on the other hand, a maximum distance can also be specified as a constraint. The usual degrees of freedom and techniques in nonlinear optimization can be applied for this purpose.

[0027] According to a further advantageous embodiment, the monitoring of a respective sub-area of ​​the geographical area by a respective one of the vehicles is initially carried out at already existing locations of the vehicles, wherein the data on the monitored sub-areas are used to check whether there is continuous monitoring, and if it is determined that there is no continuous monitoring, the automatic re-parking is initiated.

[0028] While optimal positions or positioning ranges for the vehicles are determined a priori using the digital map, and automatic re-parking is performed if necessary based on this determination, empirically based optimization can also be achieved by checking the current surveillance of the defined area within the geographical region and automatically re-parking accordingly. Both methods can also be combined, in particular by first determining the optimal positions using the digital map and then verifying continuous surveillance by monitoring the sub-areas with individual vehicles. If this check is negative, further re-parking can be performed.

[0029] According to a further advantageous embodiment, the monitoring of the sub-areas by the vehicles is carried out using vehicle cameras. Lidar systems can also be used appropriately under certain circumstances.

[0030] According to a further advantageous embodiment, traffic monitoring is carried out using the area data set.

[0031] Traffic monitoring can be used to locate another road user, particularly through license plate recognition. Traffic monitoring can also be carried out anonymously initially, for example, to detect a road user traveling at a speed exceeding the speed limit. The vehicle sensor data is then used to detect and document speeding by other road users within the defined area, e.g., using license plate recognition. This can be done permanently or for a temporary measurement.

[0032] According to a further advantageous embodiment, it is classified as continuous monitoring if a route of another road user within the defined area can be traced completely by means of the area data record.

[0033] According to the invention, a transport network is defined as an area within the geographical region and mathematically modeled as a graph with a multitude of possible routes.

[0034] Various algorithms from the field of graph theory can be used to search a road network for junctions. A road network can be modeled as a graph in which intersections are represented as nodes (also called "vertices") and roads as edges (also called "edges"). There are various state-of-the-art algorithms for this. This graph must be constructed once. An optimal target configuration preferably consists of a monitoring vehicle at each edge (i.e. road) in each direction of travel. In other words, in the best case scenario, there are two vehicles between two nodes (road intersection / junction), each covering one direction. Alternatives are possible, for example if a vehicle has appropriate sensors (e.g. lidar or radar) in both directions.Preferably, the immediate vicinity (i.e., adjacent edges) is always checked first for overcrowding, so that vehicles have to travel the shortest possible distance when reparking. If several vehicles are suitable, the vehicle selected for reparking is the one that keeps the distance to the driver, owner, or keeper as small as possible, or even reduces it if necessary. If none of the vehicles is suitable, the process continues with another edge of the graph.

[0035] According to a further advantageous embodiment, if there is no parking space at a desired position or in a position area, i) the vehicle is driven back to its original location and, after a predetermined period of time, is re-parked at the desired position or position area, or ii): the vehicle reads out the registration numbers of other parked road users at the desired position or in the position area without parking space and transmits them to a central processing unit, and if the central processing unit detects that another road user has left the occupied desired position or position area, the re-parking to the desired position or position area is attempted again.

[0036] Further advantages, features and details emerge from the following description, in which - if necessary with reference to the drawing - at least one embodiment is described in detail.

[0037] They show: Fig. 1: An exemplary situation for the application of the procedure of Fig. 2. Fig. 2: A method for monitoring a geographical area according to an embodiment of the invention.

[0038] The representations in the figures are schematic and not to scale.

[0039] Fig. Figure 1 shows a bird's-eye view of an exemplary situation of a geographical area whose traffic network is to be monitored for speeding road users. In this exemplary situation, a method for monitoring by a plurality of vehicles 1 is applied, the method steps of which are schematically shown in Figure 1. Fig. 2. Therefore, the following explanations can also be based on the sketches of the Fig.2 can be followed, which may be helpful for further understanding. In a first step of the method, a relevant area S1 is defined from the geographical area that includes an urban area with streets and houses. This area includes the streets but neglects the houses as irrelevant. The reason for this is that other road users are only expected on the streets and therefore only the streets need to be monitored. The defined area is also characterized by the fact that there is a uniform speed limit here. In order to monitor compliance with the speed limit by other road users, continuous observation of the streets in the desired area is desirable. A large number of vehicles 1 are located in the geographical area, each of which has a sensor unit with a camera.The vehicles 1 are in particular passenger cars; for clarity, one of the schematically sketched vehicles 1 is marked as such and shown in detail enlarged, although there are actually several vehicles 1 in the area shown. A digital map of the geographical area is available, in which the position profiles of the roads in the defined area are plotted. These roads can be modeled as a mathematical graph, with intersections being nodes and roads between the intersections being edges. This mathematical modeling makes it easier to apply an optimization algorithm to determine S2 the desired positions or position ranges for the vehicles 1 in order to enable seamless monitoring. If seamless monitoring of the defined area is not possible with the current location distribution of the vehicles 1, at least a subset of the vehicles 1 is automatically reparked S3, i.e.i.e., at least one of the vehicles 1 to its desired position or within its desired position range, so that the roads within the geographical area with a uniform speed limit can be fully monitored. This means that there is no uncertainty about the route taken by another road user within the geographical area. Each of the vehicles 1 carries out a respective monitoring S4 of a respective sub-area visible to it. By combining S5 the data generated by the individual vehicles 1, an area data set 1 can be generated.

[0040] Although the invention has been illustrated and explained in detail by preferred embodiments, the invention is not limited by the disclosed examples, and other variations may be derived therefrom by those skilled in the art without departing from the scope of the invention. It is therefore clear that a multitude of variations exist. It is also clear that exemplary embodiments are truly only examples and should not be construed as limiting the scope, possible applications, or configuration of the invention in any way.Rather, the preceding description and the description of the figures enable the person skilled in the art to implement the exemplary embodiments in concrete terms, whereby the person skilled in the art, with knowledge of the disclosed inventive concept, can make various changes, for example with regard to the function or the arrangement of individual elements mentioned in an exemplary embodiment, without departing from the scope of protection defined by the claims and their legal equivalents, such as further explanations in the description.

Claims

[1] Computer-implemented method for monitoring a geographical area by a plurality of vehicles (1), comprising the steps: - defining (S1) an area in the geographical area for which continuous monitoring is desired, wherein a traffic network within the geographical area is defined as the area and is mathematically modelled as a graph with a plurality of possible routes; - Determining (S2) desired positions or position ranges for the vehicles (1) in order to enable continuous monitoring; - automatic re-parking (S3) of at least a subset of the vehicles (1) to their desired positions or position ranges if a current location of the vehicles (1) does not allow continuous monitoring; - carrying out (S4) a monitoring of a respective sub-area of ​​the area of ​​the geographical region by a respective one of the plurality of vehicles (1); and - Combining (S5) data on the monitored sub-areas to form an area data set. [2] Method according to claim 1, wherein the desired positions or position ranges for the vehicles (1) are determined using a digital map in order to enable continuous monitoring. [3] Method according to claim 2, wherein the desired positions or position ranges for the vehicles (1) are determined by applying an optimization algorithm to determine an optimal vehicle distribution. [4] Method according to claim 3, wherein an optimization objective is that a respective vehicle (1) from the subset is positioned a minimum distance away from the current location and / or from the home of the owner or user of the respective vehicle (1) when reparking. [5] Method according to claim 1, wherein the monitoring of a respective sub-area of ​​the geographical area by a respective one of the vehicles (1) is initially carried out at already existing locations of the vehicles (1), using the data on the monitored sub-areas it is checked whether there is continuous monitoring, and if it is determined that there is no continuous monitoring, the automatic re-parking is initiated. [6] Method according to one of the preceding claims, wherein the monitoring of the sub-areas by the vehicles (1) is carried out using vehicle cameras. [7] Method according to one of the preceding claims, wherein traffic monitoring is carried out using the area data set. [8] Method according to one of the preceding claims, wherein continuous monitoring is classified if a route of another road user within the defined area can be traced without gaps by means of the area data record. [9] Method according to one of the preceding claims, wherein, in the event of a lack of parking at a desired position or in a position area, i) the vehicle (1) is driven back to its original location and, after a predetermined period of time, is re-parked again at the desired position or position area, or ii) the vehicle (1) reads out the registration numbers of other parked road users at the desired position or in the position area without parking and transmits them to a central processing unit, and when the central processing unit detects that another road user has left the occupied desired position or position area, the re-parking to the desired position or position area is attempted again.

Images