DE102021112804A1 - Process, vehicle control unit and vehicle for anticipatory data collection from networked road users - Google Patents

Abstract

Method for anticipatory data collection from networked road users by determining an event risk R of an event, with at least one road user repeatedly recording at least one risk parameter P and the event risk R being repeatedly determined as a function of the risk parameter P, with the risk parameter P being determined by a recording device 17 which is assigned to the and/or another road user is determined, with a trigger signal T being transmitted as a function of the determined event risk R by means of a communication device 14, which is assigned to the and/or another road user, which is transmitted by other road users or technical means that comprise at least one said communication device 14, is receivable, characterized in that the said trigger signal T transmitting and receiving road users depending on the trigger signal T a Datenerhebun g start, with one or more event data being recorded and stored and/or transmitted to an external server 7 by means of communication device 14, with the transmission of the trigger signal T and/or the recording of the event data beginning in time before the event.

Description

technical field

The invention relates to a method for anticipatory data collection from networked vehicles, for the predictive provision of at least one piece of information about an event in the public or private space.

State of the art

From the disclosure documents DE102005026517A1 and DE102014015669A1 Methods are known for automatically collecting information about a traffic accident and for forwarding it to third parties, preferably emergency services, by means of wireless communication. The collection of information is triggered here by a trigger, which takes place via data evaluation by accident detection devices of the vehicles involved in the accident, in particular via sensors for detecting the deformation of the vehicle body. The data transmitted in this way includes location data, deformation data and other vehicle-internal operating parameters.

From the disclosure document DE102016216522A1 a method for providing information about an accident event between at least two vehicles is known, the information being able to include video, audio and image recordings, location, damage and personal data and operating parameters of the vehicles. The information is provided by in-vehicle sensors and by passers-by or eyewitnesses. The corresponding information is recorded and made available by sending out a trigger signal, which is activated by technical means for accident detection of the vehicles directly involved in the accident or by an emergency call. The data collected can be transmitted to external servers and authorities and stored using Car2Car or other Car2X communication.

Taking data protection into account, the disclosure document DE102020004792A1 the procurement of information and the exchange of information about a traffic accident are only released after a successful examination of an accident event and only between those involved in the accident and, if necessary, to authorities and only after consent to a data protection declaration. The acquisition and distribution of information by third parties, bystanders is explicitly excluded. Furthermore, camera recordings should anonymize uninteresting areas by means of image recognition and image processing in order to protect the privacy of people who happen to be in the recordings.

From the disclosure document DE102019203543A1 a system is known which calculates an accident risk of a vehicle in real time as a probability value from predetermined risk parameters and outputs a control signal of the vehicle as a function of this. The risk parameters essentially include route parameters that are intended to map the environmental conditions of the vehicle and driving parameters that take into account the driving dynamics properties of the vehicle, in particular the driving speed.

Furthermore, from the disclosure document DE102016003574A1 a procedure for event-related monitoring of the immediate vicinity of a vehicle and from the disclosure document DE102017201426A1 a method for operating a communication between vehicles and safety devices is known.

Disclosure of Invention

The methods presented in the prior art have in common the fact that all information about an accident that can be recorded by vehicles, other road users or people is only collected after the accident has been recognized or validated. With regard to the question of liability of possible accident perpetrators or criminals, however, important available information is missed, so that afterwards only the statements of those involved and eyewitnesses can be used, which are now based on memories of the respective event. However, the aspect of data protection stands in the way of possible approaches to the event-related collection of available information. In the present invention, a method is therefore proposed to provide all information that can be recorded by vehicles, other road users or other people at the moment of an event in public or private space for subsequent clarification of the course of the event and at the same time to take into account the aspect of data protection .

The object of the present invention is now to provide a method for anticipatory data collection from networked road users, with all information available in the spatial environment of an event being recorded and stored for a period of time around the event. Data collection begins before the actual event, so that the course of the event can already be documented. Farther the event does not refer exclusively to traffic accidents between at least two road users, but can rather be used to record any event in public or private space, but preferably to document traffic accidents, criminal offenses or violations of public order or existing legal framework conditions.

The object is achieved by a method according to Claims 1 to 8, a vehicle control unit according to Claim 9 and a vehicle according to Claim 10 in that detectable risk parameters are evaluated by vehicles in real time and repeatedly or at different intervals in order to identify an event such as a dangerous situation, to anticipate a traffic accident, a criminal offense or other violations of public order and to initiate measures. The measures essentially relate to the triggering of a trigger signal, the transmission of that trigger signal to other networked vehicles or technical means in the spatial vicinity of the vehicle or technical means triggering the trigger signal, and the data collection itself. Another vehicle or technical means that receives the trigger signal is prompted to start data collection and to record all available information and to forward it to a central external server in encrypted and anonymous form. All recorded information from all vehicles or technical means that have received the trigger signal is received on such a server, so that the collected data can be further processed, in particular for initiating further measures in real time and/or for later processing of the data received.

The intention of the present invention is now to use all available technical means in the context of an event in order to record all measurable information about that event, whereby the data collection is already started in advance, so that important information and evidence about the course of the event is provided . In order to protect personal data and the privacy of persons not involved in the event or who happen to be in the spatial vicinity of that event, technical means for natural face anonymization and further image anonymization are used. This results in the following advantages.

The repeated or continuous recording and evaluation of risk parameters from vehicles moving or stationary in public or private space can be integrated into the existing safety systems of the vehicles themselves and supplement them, so that dangerous situations can be recognized at an early stage. By receiving a trigger signal from a networked vehicle or other technical means from a networked system, a dangerous situation can also be detected which cannot yet be detected by the vehicle's internal sensors of the respective vehicle even at the moment the trigger is received. The method according to the invention can therefore be used in an advantageous manner as an early warning system for dangerous situations for the self-protection of the vehicle and its occupants.

The data collection by several networked road users or other technical means offers the advantage that the course of the event from several perspectives and with different types of information such as image, video, audio, GPS, traffic, damage and personal data and other operating parameters involved road users, collectively referred to below as event data, are documented. Through the forward-looking data collection based on the evaluation of risk parameters by several networked road users or other technical means, the data collection as evidence can be justified on a case-by-case basis in order to take the aspect of data protection into account. Thus, the reason for the data collection can be argued and proven in an advantageous manner.

The data transmission from all triggered road users or other technical means to a central server in real time as well as transmission to other third parties, such as rescue or safety facilities, offers the advantage that suitable measures can be initiated in real time. For example, rescue control centers can use video and image recordings of an accident to quickly get an overview of the type and number of vehicles involved, the possible number of injured people and even the injuries to be expected, with the necessary number of rescue workers, vehicles, based on this information or required tool can be determined. Image and video recordings can be made anonymous in real time, since it is not necessary to identify personal data when planning a rescue operation, for example, which means that the data protection of the road users involved is taken into account.

Due to the networking of road users in the vicinity of the event, it is not necessary for every road user, not even the road users directly involved in that event, to use the technical means to early detection of dangerous situations, accident detection, recording and evaluation of risk parameters or for communication with other road users or other technical means. This results in the advantage that the method according to the invention can also be used indirectly for road users who do not have the appropriate technical means for self-protection according to the present invention.

• 1 ein System zur vorausschauenden Datenerhebung von vernetzten Verkehrsteilnehmern,
• 2 den Prozessablauf zur vorausschauenden Datenerhebung,
• 3 ein Fahrzeug zur Ausführung des Verfahrens,
• 4 den schematischen Aufbau der Erfassung und Auswertung von Risikoparametern.

The following detailed description of the invention is supported by the representation of the following figures, here showing

• 1 a system for anticipatory data collection from networked road users,
• 2 the process flow for predictive data collection,
• 3 a vehicle to carry out the procedure,
• 4 the schematic structure of the recording and evaluation of risk parameters.

In 1 shows the schematic structure of a system 1 for anticipatory data collection from networked road users. It should be noted that the system 1 shown is only to be understood as a possible embodiment and the present invention in its entirety is not intended to be limited to the features shown, nor is it intended to be fixed to this one embodiment. Rather, it serves in 1 System 1 shown for the qualitative illustration of potential properties and as support for the basic functioning of the method according to the invention for anticipatory data collection from networked road users.

A system 1 according to the invention for predictive data collection from networked road users, hereinafter referred to as system 1, includes at least one road user, which includes at least technical means that are set up to capture at least one piece of information about an event in the public or private space and this to transmit at least one piece of information via a communication link to an external server 7 and/or other road users. Road users within the meaning of the invention are stationary, autonomously moving or person-guided means of transport, such as motor vehicles of all kinds, but in particular passenger cars, hereinafter referred to as vehicles 2, 3, bicycles or other means of transport that are designed to be movable through public or private space, people 5 in the function of pedestrians, passers-by or in any other way in the corresponding rooms. In essence, therefore, all persons or technical devices that move or stay in public or private space and/or are able to record at least one item of information relating to an event and transmit it to an external server 7 or a third party. According to the present invention, however, at least one road user is preferably a passenger car, in particular a vehicle involved 2 or not involved 3 in an event. This requirement results from the aspect of recording and evaluating risk parameters by a recording device, as described in more detail below. Additional technical means for detecting at least one piece of information about an event can include security devices such as surveillance cameras 4, microphones and/or other optical sensors such as lidar or radar sensors and others.

An event within the meaning of the invention is characterized by an interaction between at least a first road user and another road user and/or the physical environment of the first road user, with at least one road user and/or the physical environment suffering damage or violations of public order or the law framework conditions occur. Events preferably relate to traffic accidents, criminal offences, robberies, violent confrontations or other violations mentioned in public or private space. The public space within the meaning of the invention is any spatial area that is publicly accessible, but preferably roads and paths on which road users can move. The private space within the meaning of the invention is any spatial area, including streets and paths, on which road users can move, which is privately owned.

The system 1 for executing the method according to the invention is now built around a described event and in accordance with 1 shown as an example as a traffic accident between two vehicles 2 involved. Other uninvolved vehicles 3 can be located in the immediate spatial environment, for example on the same street. In addition, the spatial area of the traffic accident can lie within the field of view of a surveillance camera 4 . Furthermore, persons 5 who are not directly involved in the traffic accident can be present in the spatial environment and act as eyewitnesses. Involved 2 and uninvolved vehicles 3, surveillance cameras 4, but also electronic media from third parties 5, technical means for data transmission, hereinafter communication devices 14, included in the networked system 1. The data can be transmitted using all existing technologies such as TCP/IP, other Internet data connections, local or global wireless communication networks, Bluetooth, radio, GPS or other satellite-based communication networks, Car2Car or other Car2X communication and others. The type of communication between the individual road users and other technical means is of secondary importance for the invention, insofar as the individual communication devices 14 are set up to transmit data to other road users, data transmission units 6 or directly to an external server 7 and/or from them receive. With regard to the aspect of data protection, corresponding transferrable data can be sent and received anonymously and encrypted.

Parts of the system 1 can also be uninvolved third parties, such as rescue 8 and/or safety devices 10. These can also include communication devices 14 in order to carry out data transmission to road users, surveillance cameras 4, people 5, data transmission units 6 or the external server 7 . These can also include peripherals to support the event or the traffic accident, such as emergency vehicles 9 to ensure help and care for the injured and/or security vehicles 11 to ensure the safety of road users and/or to possibly escape perpetrators 12 pursue.

According to 2 the method according to the invention for anticipatory data collection from networked road users within the system 1 can be described by a higher-level process flow. The first step S1 relates to detecting an event at the time it takes place or immediately thereafter. In one aspect of the invention, in which the event comprises a traffic accident between at least a first road user and another road user and/or their physical surroundings and where at least one road user involved is a vehicle 2 involved, the detection of the event or the traffic accident by a or several detection devices such as in-vehicle sensors, in particular crash detection. Further versions of the vehicle-internal sensor system can be used to detect deformation of the body or other components of the vehicle 2 involved, technical means for optical perception of the environment, such as radar or lidar sensors, distance and collision assistants, video cameras and/or other technical means for optical and acoustic perception of the vehicle environment or include interior. Alternatively, the event can be recognized by involved or uninvolved third parties, such as persons 5, in the form of an emergency call being made. Emergency calls can continue to be made automatically by technical means of image recognition or by security systems such as alarm systems. Alternatively, surveillance cameras 4 can be used, which are linked to appropriate computer programs that can automatically detect traffic accidents and/or violent or criminal offenses on the basis of image recognition, for example using the trajectory of detected objects or the posture and movement of people 5 . According to the configuration of the event detection by the first step S1, it is intended to ensure that an event can be detected with all available technical means and available information if it has already happened. Exact features and embodiments are known here from the prior art, so that the first step S1 can preferably be used to complete the event recognition. The event detection after the first step S1 should be repeated, preferably continuously or at least cyclically, so that the process flow, starting from this step, repeatedly jumps to the second step S2, which leads the process flow back to the first step S1 if (n) an event is not recognized and upon detection (j) of an event leads to a further step S5.

In general, the higher-level process flow shown can be partially or completely executed within a vehicle 2 involved. In 3 For this purpose, the schematic structure of a vehicle 2 is illustrated, which is set up to carry out the method according to the invention. In one embodiment, the vehicle 2 includes a vehicle control unit 15 for this purpose, which is set up to exchange data with other vehicle-internal elements such as a drive unit 16, a detection device 17, a data interface 18 for the user of the vehicle 2 and a communication device 14. The drive unit 16 is to be understood here as representative of the entirety of individual control devices and technical units which can detect, process and output operating parameters of the vehicle 2, such as the driving status, temperatures, quantity and pressures of fluids, gases and components of the vehicle 2. The Detection device 17 is to be understood here as representing all of the vehicle's internal sensors, which, for example, the driving dynamics state of the vehicle 2, information from assistance systems, the position and trajectory of the vehicle 2 within global and local reference systems and relative to the immediate phy sical environment and / or other road users and other information such as all event data mentioned. The data interface 18 is to be seen here as representative of all the technical means that are set up to carry out data transmission between a user of the vehicle 2 and the vehicle control unit 15 . In order to receive, process, store and transmit the corresponding data, the vehicle 2 can include a data memory 19 which can be integrated into the vehicle control unit 15 .

Referring again to 2 In a further step S3 of the process flow, an event is advantageously predicted according to the invention by risk assessment, which can be carried out in parallel and independently of the first step S1. The intention of the risk assessment is to use all information accessible to a road user, preferably an involved 2 or uninvolved vehicle 3, in order to predict a possible risk for an event. Within a vehicle 2, for example, all data that is processed within the vehicle's internal control units 15, operating parameters of the vehicle 2 and data from a detection device 17 and/or assistance systems, including image, video, audio, GPS and traffic data and/or others Event data is used as information. In one embodiment of the method according to the invention, all of the available information mentioned is combined into individual risk parameter groups for improved clarity.

In 4 the schematic structure of the recording and evaluation of risk parameters P is shown for this purpose. The listing and combination of individual risk parameters P into superordinate groups is to be understood here as an exemplary embodiment, with the method according to the invention not being limited in its entirety to the features shown, nor should it be fixed to this one embodiment. Rather, the schematic structure shown serves to provide a qualitative illustration of an exemplary embodiment and as support for understanding the basic functioning of the method according to the invention. A first group of risk parameters is referred to below as vehicle parameters P _F and includes one or more information about the physical properties of the vehicle 2 such as weight, payload, dimensions, wheelbase, track width and driving dynamics properties such as driving speed, power-to-weight ratio, acceleration capacity, relative speed other road users and other parameters describing vehicle 2. The physical properties of a vehicle 2 are typically fixed values and can already be stored in the data memory 19 of the vehicle control unit 15 after the vehicle 2 has been manufactured. If the vehicle properties change, for example due to a load, a trailer load, bodywork conversions or other physical changes, this information can be transmitted to the vehicle control unit 15 via the data interface 18 by the user of the vehicle 2, the manufacturer or a technician during a visit to the workshop. Alternatively, there is also the possibility of transmitting this information via the wireless communication device 14 as part of OTA software updates.

In one embodiment, another group of risk parameters P is hereinafter referred to as route parameters Ps and includes one or more pieces of information about the route conditions, such as the type of road, including expressway, interurban road, urban area, traffic-calmed areas, construction sites, speed limits or other information from traffic signs or traffic control systems as well as road conditions such as road width, road surface, road surface evenness, road damage, curve radii and others as well as road environmental conditions such as traffic density, obstacles or objects located outside the road, number and type of other road users as well as information and statistics on local and global accident or hotspots. Route parameters Ps thus correspond to risk parameters, which are recorded and processed in particular by the detection device 17, in particular by vehicle cameras and other optical and acoustic systems such as lidar or radar sensors, and are essentially based on the perception of the surroundings of the vehicle 2.

In one embodiment, a further group of risk parameters P is referred to below as the friction parameter P _R and includes one or more pieces of information about the type and properties of the tires on the vehicle 2, such as summer or winter tires, static friction values, rolling surface, effective tire radius, wheel speed, slip, state of wear and more. Analogously to the vehicle parameters P _F , constant friction parameters P _R , such as the type of tires, friction values or geometric properties of the tires, can already be entered in the data memory 19 before the start of the journey. In general, the friction parameters P _R listed and summarized here as examples could be regarded as vehicle parameters P _F , so to speak. In this embodiment, however, they become conscious considered separately from each other, since the tires on the one hand make contact between the vehicle 2 and the road and are therefore the most important component in terms of driving safety, and on the other hand individual risk parameters can influence each other, which ultimately influences the entire risk group.

In one embodiment, another group of risk parameters P is referred to below as weather parameters Pw and includes one or more pieces of information about climatic and weather-related environmental conditions such as air temperature, road surface temperature, type and amount of precipitation, intensity of solar radiation, light and visibility conditions, smoothness and Aquaplaning, wind direction and wind speed, frost and others. Analogously to the vehicle parameters P _F and the friction parameters P _R , individual weather parameters Pw can be determined by the detection device 17 and others via data transmission using the communication device 14 .

In one embodiment, another group of risk parameters P is referred to below as occupant parameters P _I and includes one or more items of information about the state of health of the driver or other vehicle occupants, such as tiredness, symptoms of illness, physical impairments, ability to react, anticipatory driving style, general driving profile, ability to concentrate, possible distractions and others. The occupant parameters P _I can also be determined via the detection device 17, in particular by cameras 4 directed into the interior of the vehicle 2, a wide variety of methods for evaluating the state of health of the driver or other occupants being known. Furthermore, the vehicle control unit 15 can be set up to learn the driving behavior of the driver and itself to predict various driving maneuvers or expected reactions of the driver to specific driving situations. Alternatively, the driver's facial expressions can be recognized, with the emotional behavior, tiredness or general alertness being recognized and evaluated.

In an advantageous embodiment, the individual risk parameters P of the risk parameter groups P _F , P _S , P _R , P _W , P _I are recorded repeatedly or at different intervals in order to enable a continuous or at least cyclical risk assessment, with the recording of the individual parameters preferably several times per minute, in particular preferably several times per second. The individual risk parameters P of the various risk parameter groups P _F , P _S , P _R , P _W , P _I now correspond to the detectable information about an event, of which at least one piece of information should be available for risk assessment according to the invention. In one embodiment, the risk group values R _F , Rs, R _R , Rw, R _I are calculated and/or interpreted on the basis of the recorded risk parameters P of the various risk parameter groups P _F , P _S , P _R , Pw, P _I , the risk group values R _F , R _S , R _R , R _W , R _I are configured in an advantageous embodiment as a numerical value within a value range from 0 to 1, preferably as a floating-point number. The risk group values R _F , R _S , R _R , R _W , R _I thus advantageously form a probability value, with the value 0 corresponding in one embodiment to an event probability, or in other words an event risk R of 0%, and to the Value 1 is assigned an event risk R of 100%. Alternatively, any other assignment of numerical values or scaling of the individual numerical values can take place insofar as a qualitative assessment with regard to an event probability can be reproduced by the numerical assessment of the risk group values RF, _RS , _{R R ,} R _W , R _I . The calculation of the individual risk group values RF , R _S , _{R R ,} R _W , R _I is now essentially based on the evaluation of individual risk parameters P.

In an advantageous embodiment, the individual risk parameters P are evaluated by calculating, allocating and interpreting the recorded measured values or information received by means of the communication device 14 . As an example of the driving parameters P _F , the detected driving speed of the vehicle 2 can be determined as a risk parameter P and evaluated in relation to the current recommended speed, with, for example, falling below the recommended speed detected by optical assistance systems by 10%, a risk event of 0% is assigned, any further deviation However, the value of the event risk R increases by 10% of the recommended speed. In an exemplary embodiment, the detection device 17 of a vehicle 2 detects a current driving speed of 64 km/h and a recommended speed of 50 km/h, as a result of which the event risk R with regard to the driving speed is increased to 0.3. The specified numerical values of the exemplary embodiment described are of a purely fictitious nature and can be freely selected in a real application. It is up to the user to assign the percentage steps in which the deviation of the driving speed from the recommended speed is to be applied to the event risk R. The method according to the invention can thus advantageously be modified to suit the application and can be parameterized for any application. Furthermore, different risk parameters P can oppose each other tig or their calculation and interpretation may depend on other risk parameters P. In one aspect, where the driving speed is evaluated as a risk parameter P for the vehicle parameters P _F , the value for the event risk R with regard to the driving speed cannot depend solely on the relationship to the current recommended speed, but also, for example, on route parameters Ps and/or friction parameters P _R . For example, exceeding the recommended speed by a relatively small amount can result in a higher event risk R if a winding and damaged roadway was determined simultaneously by the route parameters Ps and reduced road grip by the friction parameters P _R . The friction parameters P _R can also depend on the weather parameters P _W , with summer or winter tires being determined having the effect of reducing or increasing a possible event risk R depending on the weather conditions. These listed dependencies of the different risk parameters R as well as the way of evaluating and interpreting them are also the responsibility of the user and can therefore be freely selected.

The risk parameters P recorded in this way, which are assigned to risk parameter groups P _F , P _S , P _R , Pw, P _I and converted and/or interpreted into risk group values RF , R _S , _{R R ,} R _W , R _I , are then a higher-level event risk R, whereby a freely selectable calculation method can also be used here. In one embodiment, the arithmetic mean is formed from the individual risk group values R _F , R _S , R _R , R _W , R _I , which correspond to probability values within the numerical values from 0 to 1. In an alternative embodiment, the individual risk group values R _F , R _S , R _R , R _W , R _I can be weighted. Furthermore, as an alternative, any further calculation method can be executed insofar as it determines a value for an event risk R whose numerical value lies within a value range between 0 and 1. Referring to 4 shows the course of the event risk R over time t graphically and for a fictitious sequence. Within the first time steps 0 to 3, the event risk R is in the value range 0.1 to 0.2, which represents a low risk of a possible event. From time steps 4 to 7, the event risk R increases significantly to a value of 0.9 and then falls again over the course of time. According to the invention, this course of the event risk R is to be both calculated and further processed within the vehicle control unit 15, the further processing including the execution of measures or actions as a function of the event risk R.

Referring to 2 according to the present invention, an event is to be predicted after step S3, which is cyclically queried in step S4. The query of the event occurs analogously to step S2 repeatedly, preferably several times per minute, particularly preferably several times per second, and includes the evaluation of the event risk R determined. In an exemplary embodiment, an event is recognized when the event risk R exceeds a first threshold value that triggers the event, a value of 0.7 being assigned to this, for example. Referring to 4 , an event is accordingly recognized at point in time 6, with the event risk having assumed the value 0.75. The determination of the absolute value of the first threshold value is the responsibility of the user and can therefore advantageously be freely selected according to the invention. Through the continuous or at least repeated detection and processing of risk parameters P up to an event risk R that can be evaluated, the entire environment of the respective vehicle 2 is advantageously continuously analyzed for possible events by the method according to the invention, with the corresponding adjustment of the parameter evaluation and the interpretation of the calculated event risk R, an event can be predicted, namely at the moment when that event risk R increases and exceeds a definable first threshold value. According to the repeated evaluation of the event risk R, it is checked in step S4 whether an event was predicted or, in other words, whether the first threshold value was exceeded, with the process flow shown jumping forward to a further step S5 if an event (j) is predicted and without prediction of an event (n) jumps back to step S3.

In step S5, according to the invention, a trigger signal T is now sent, depending on the results of the decisions from steps S2 and S4. Step S2 reveals whether an event such as a traffic accident has been detected, with the event being detected using conventional means and only immediately at the moment of the event or immediately thereafter. Step S4 indicates whether an event was predicted by the process described in step S3. The trigger signal T is sent out immediately when step S2 or step S4 is confirmed, and thus in the case of S4, according to the invention, advantageously in time before a potential event. In one embodiment of the method according to the invention, the transmission of the trigger signal T includes causing the vehicle control unit 15 to send the trigger signal via the communication device 14 Emit T in the networked system 1, which can be received by surrounding vehicles 2, 3 or other technical means, which also have a corresponding configuration of a communication device 14. If other vehicles 2, 3 or technical means receive a signal sent in this way, they are prompted to start data collection. The data collection here includes the collection of event data with all technical means that are set up to record the corresponding data. In one embodiment, in which a surveillance camera 4 receives such a trigger signal T, it is prompted to start video recording for its field of view, even if this surveillance camera 4 is in an idle state, for example. In an alternative embodiment, in which, for example, an electronic medium such as a smartphone of a passer-by who is in the reachable area of the signal receives a corresponding signal, his smartphone is prompted to start video and audio recording. Alternatively, data relating to the position or the digital link to local networks and/or other electronic media could also be provided by a corresponding electronic medium. In one embodiment, the data collection mentioned takes place as a background process on the electronic medium, so that the associated user remains unaffected when using the electronic medium. In an alternative embodiment, the process of data collection is shown to the user of an electronic medium as a visual display or other type of notification and his declaration of consent is requested in order to protect data protection and privacy.

In an alternative embodiment, in which a vehicle 2 involved in the potential event or a vehicle 3 not involved receives a signal as a trigger T, all the technical means of the vehicle 2, 3 in question are prompted to start data collection. The data collection can include video recordings of the interior and surroundings of the vehicle 2, 3 using all cameras belonging to the vehicle 2, 3, audio recordings using microphones, analysis of the surroundings using radar and lidar sensors and/or entire assistance systems, location data, but also the recording of the vehicle operating parameters and the detectable risk parameters P include. In an alternative embodiment, video recordings of the interior are only recorded with the prior consent of the occupants, so that personal rights are protected even during an accident event.

The data collected in this way from all vehicles 2, 3 or other technical means in the vicinity of the potential event can be stored on local data storage devices 19. In an advantageous embodiment, the collected data are transmitted by means of a communication device 14 to an external server 7, on which all collected data on a potential event from all triggered sources are received. With regard to data protection, transferrable data can be encrypted and transferred anonymously. In one embodiment, a mutual data protection declaration, authorization for data collection and processing and/or encryption and decryption precautions can be agreed. In the case of vehicles 2, 3, for example, this can be done by the manufacturer or by the user, contractually when the vehicle 2, 3 is taken over, by an agreement with local authorities, insurance companies or other third parties. In the case of stationary technical means such as surveillance cameras 4 from public or private institutions, data protection agreements can also be made in advance between the respective parties involved in data collection and data processing. In the case of electronic media or other technical means of eyewitnesses, the data can be collected, for example, by confirming the terms and conditions of use of the electronic medium or the software stored on them, and additional digital messages can be requested via data protection queries or declarations of consent. The collection and transmission of data can also be a process that is decoupled from one another, so that data are first recorded and encrypted when a trigger signal T is received, but are only transmitted and/or decrypted if a valid data protection declaration or declaration of consent is present. This is particularly advantageous and also necessary because there can only be a short period of time between the time the trigger signal T is received and the occurrence of the potential event, so that the short-term processing of data protection agreements could negate the advantage of this method of predictive data collection. In an alternative advantageous embodiment of the method according to the invention, vehicles 2, 3 or other technical means, which are prompted to collect data after receiving a trigger T, can include computer programs or other software that anonymizes the collected data. In this way, video recordings can be made using means of natural face anonymization, with essential facial features being concealed, but facial expressions and gestures are essentially retained, as can audio recordings could also be anonymized to make voice recognition unrecognizable.

For the purpose of reducing the data capacity and in order to minimize the collection of superfluous data, the data collection is advantageously carried out according to the invention only in a time frame around a potential event, with the data collection being initiated with the receipt of a trigger signal T and for the duration of the trigger signal T stops. In an advantageous embodiment, before the local storage and/or before the transmission of the detectable data to the external server 7, the detectable data is filtered by the vehicle control unit 15, so that only data useful for initiating suitable measures are detected, stored and/or forwarded . In an alternative embodiment, the forwarded data is evaluated and filtered in a backend by the external server 7 in order to remove duplicate information and to further reduce the amount of data used in terms of data protection. In the event of a recognized event subsequent to step S2, a fixed period of time can be provided within which the data is collected. With regard to the spatial aspect of the event, the data collection from vehicles 2, 3 or mobile technical means can continue to be limited to a fixed radius around the potential event and/or the vehicle 2, 3 emitting the trigger signal T or the corresponding technical means, so that when leaving the specified radius, the data collection is stopped. In the case of a predicted event after step S4, the duration of the data collection can be linked to the duration of the trigger signal transmission. In one embodiment, in which the trigger signal T is triggered by the event risk R exceeding a first threshold value, the trigger signal T remains active as long as the event risk R is above the corresponding threshold value. If the value falls below the specified threshold value, the trigger signal transmission stops and thus also the data collection by surrounding vehicles 2, 3 and other technical means. Referring to 4 stops the trigger signal transmission in an alternative advantageous embodiment only after falling below a second threshold value, which is preferably lower than the first. In this way, a hysteresis is created, which is illustrated in the figure as a hatched area and which advantageously creates a subsequent decay time in the form of debouncing. Such debouncing can prevent the data collection from stopping too early and thus essential information from being lost, and at the same time a possible dynamic in the triggering and stopping of the data collection is reduced, which improves the consistency of the data collected. In the exemplary embodiment according to 4 For example, the second lower threshold value is defined as 0.3, with the trigger signal transmission only stopping in time step 14, when the event risk falls below the value of 0.3. In an alternative embodiment, a debounce time can be used instead of or in addition to the hysteresis and/or the lower threshold value, with the trigger signal transmission being continued for the duration of the corresponding debounce time after the second lower threshold value or the conditions for a detected event have been undershot.

In an advantageous embodiment, additional risk premiums are taken into account, which have priority over other risk group values RF, _RS , _{R R ,} R _W , R _I . Risk premiums can include threshold values that hard set the event risk R to 0 or 1 and thus block or enable the trigger signal transmission. These risk premiums can arise as a function of individual risk parameters P and/or from a large number of different conditions. In an exemplary aspect, a risk premium is caused by the driver's occupant parameters R _I in connection with traffic sign recognition. If, for example, the traffic sign recognition recognizes that the driver of a vehicle ignores 2, 3 traffic signs, a distraction can be applied and a risk premium can be applied to an individual risk parameter P, its parameter group P _I , a risk group R _I or directly to the event risk R. The application of the risk can include setting one of the parameters mentioned to 1, for example. Alternatively, the risk can also be applied in stages or assume a lower value, for example 0.3, so that if the risk is applied three times, an event risk R of 0.9 results from the risk premium alone. For this purpose, however, a risk premium is set for a certain period of time, so that this state cannot be cured in the subsequent time step.

Referring to 2 In a last step S6, the data is processed, including the merging, storing, evaluating and distributing of collected and generated data. The data collected according to the method according to the invention are preferably combined on an external server 7 which is set up to receive, store and evaluate all event data or process them in some other way. The evaluation and further processing of the event data mentioned includes the assignment of various data inputs to an event, the documentation of the chronological sequence of the incoming data, the identification of the event, the number and type of road users involved, the determination of the course of events as well as possible culprits or polluters as well as the transmission of the received and/or evaluated or further processed data to third parties not involved in the event, such as emergency services 8 and/or security services 10 and/or more. In an advantageous embodiment, the measures mentioned comprising process step S6 can be carried out automatically, for example using machine learning and/or neural networks, which can be integrated into the external server 7 in terms of software.

The method according to the invention now offers the further advantage that the actual event itself does not have to be recognized or identified in order to carry out a forward-looking data collection based on the evaluation of risk parameters P in the form of the course of a calculated event risk R. Furthermore, an exact number of detectable risk parameters P is not required, since each individual risk parameter P described allows the determination of an event probability. For example, the vehicle speed, driver drowsiness detection, weather conditions or accident statistics alone could be used to activate the anticipatory data collection, although the accuracy and reliability of the anticipatory detection of an event increases with the number of risk parameters P recorded.

exemplary embodiments

• 5A eine Gefahrensituation eines einzelnen Fahrzeuges auf einer Landstraße,
• 5B den Verlauf einzelner Risikogruppenwerte und des Ereignisrisikos,
• 6A eine Gefahrensituation auf einer Straßenkreuzung, vor einem Ereignis,
• 6B die Auslösung eines Triggers bei einem vorhergesagten Ereignis,
• 6C die Gefahrensituation auf der Straßenkreuzung, unmittelbar nach dem Ereignis.

Exemplary embodiments of the described embodiments are shown below. Further special features and advantages result from the supporting diagrams, which show:

• 5A a dangerous situation of a single vehicle on a country road,
• 5B the course of individual risk group values and the event risk,
• 6A a dangerous situation at a road crossing, before an event,
• 6B the triggering of a predicted event,
• 6C the dangerous situation at the road crossing, immediately after the event.

In 5A the scenario of a single vehicle 2 driving on a winding country road is shown as an exemplary embodiment, the movement of the vehicle 2 being illustrated by different positions, which are identified by the time steps t ₅ , t ₁₀ , t ₁₅ , t ₂₀ . Within this fictitious scenario, vehicle 2 is located on a roadway 13 that has two roadway impairments 21A, 21B, which here include, for example, wet conditions 21A and a pothole 21B. The scenario also includes two traffic signs 20A, 20B, including a speed limit 20A and a curve warning 20B, and three trees 22 at the edge of the road. At time t ₁₀ , the current risk group values R _F , R _S , R _R , R _W , R _I and the event risk R resulting therefrom are also shown as examples. In 5B the courses of the individual risk group values R _F , R _S , R _R , R _W , R _I and the entire event risk R are shown for the time steps t ₀ to t ₂₅ .

At the starting point t ₀ , the weather risk group Rw has a value of 0.7, which remains unchanged over the course of time shown, since the weather does not change within the specified scenario. The value of 0.7 already indicates an increased event risk R due to the weather conditions, which can be caused by precipitation, ice, frost, wind, fog or others. If the weather risk group R _W were the only one to which the value of the event risk R would be based, a trigger signal T could already be sent out with the value 0.7. The driving risk group R _F includes a value of 0.3 at the beginning, which can arise, for example, when the vehicle 2 is moving on an overland route at a speed that is close to the recommended speed or slightly higher, with both the weather conditions R _W and the Static friction conditions R _H can weight the driven speed in their evaluation. The static friction risk group R _H also has a value of 0.3 at the beginning, which can be attributed to adapted tires of the vehicle 2, but depending on the weather, to an increased driving speed. Both the route risk group Rs and the driver occupant risk group R _I have a value of 0.1 at the beginning, which can be traced back to a straight, level route and a safe driving style without possible driving errors in advance. The total event risk R, which is calculated by arithmetic averaging within the vehicle control unit 15 from the current risk group values RF, R _S , _{R H} , R _W , R _I , has a value of 0.3 in time step t ₀ , which is up to Time step t ₄ oscillates around this value under analogous conditions.

At the time of the transition from time step t ₄ to time step t ₅ , environment analysis 23 of vehicle 2 using detection device 17 clearly shows a speed limit 20A, which is below the usual recommended speed for the current route and a roadway impairment 21A, in the form of wetness on the roadway, is detected. Due to the current driving speed, which is increased with regard to the new speed limit, and the small distance to the speed-limiting traffic sign 20A, the value for the driving risk group R _F receives a risk surcharge and jumps to the value 1. Due to the now increased driving speed and the reduced static friction due to the detected wetness of the road 13, the value for the static friction risk group R _H increases to 0.7, as does the value for the driver occupant risk group R _I , since the increased driving speed anticipates the reduced control of the driver over the vehicle 2. The route risk group Rs has a value of 0.2, which can be traced back to the route, which is still flat and largely straight. At this point in time t ₅ the event risk R reaches the value of 0.66, with the first threshold value for the transmission of a trigger signal T being fixed at the limit value 0.7 in this exemplary embodiment, so that no predictive data collection is activated at this point in time. In the further course, the driving speed of the vehicle 2 is reduced by the driver, so that the value for the driving risk group R _F drops, whereas the value for the driver occupant risk group R _I continues to rise, since the current driving speed is still above the recommended speed.

At the time of the transition from time step t ₉ to time step t ₁₀ , the environment analysis 23 of the vehicle 2, using the detection device 17, detects another traffic sign 20B in the form of a warning about a winding roadway and another roadway impairment 21B in the form of a pothole.

Due to the further reduced driving speed and the end of wet roadway 13, the value of the driving risk group R _F drops to 0.7 and the value for the static friction risk group R _H drops to 0.6. Due to the following winding road section, in connection with the still slightly excessive driving speed, the driver occupant risk group R _I has a value of 0.9 and the value for the route risk group Rs has increased to 0.7. At this point in time t ₁₀ , the event risk R is calculated to be 0.72, which thus exceeds the first threshold value of 0.7 for a trigger signal T to be emitted. Accordingly, a trigger signal T is sent from time t ₁₀ , with the vehicle beginning to collect data using acquisition device 17 , to store data in data memory 19 and to transmit data to an external server 7 using communication device 14 . The second threshold value for stopping the trigger signal T is fixed here with the value 0.3, so that the data collection only stops when the event risk R falls below 0.3. In the further course, the calculated event risk R fluctuates between the values 0.7 and 0.8.

At the time of the transition from time step t ₂₀ to time step t ₂₁ , environmental analysis 23 of vehicle 2, using detection device 17, detects that vehicle 2 has left the roadway 13 and a possible collision with a tree 22 on the roadside. At this point in time t ₂₀ , the conventional accident detection of vehicle 2 starts, which, for example, initiates emergency braking. At the same time, the imminent collision is also detected by the risk parameters P, whereby the risk group values R _F , R _S , R _H , R _W , R _I increase and the value for the driving risk group R _F receives a risk surcharge and jumps to the value 1. The process steps S1 and S2 also take effect as a result of the conventional accident detection, with an accident event having been detected and the event risk R hard being set to the value 1.

In the exemplary embodiment listed, anticipatory detection of an imminent accident event advantageously takes place according to the invention, with the anticipatory data collection being started, for example, 10 time steps before conventional accident detection, and corresponding data and information having already been transmitted to an external server 7 . Within the external server 7, a computer program can automatically evaluate the data received and independently contact rescue services 8. There are two further advantages here. On the one hand, an emergency call can be made at a time when the vehicle 2 is still intact, so that successful data transmission is more likely than after a possible traffic accident, in which the vehicle 2 could be damaged in such a way that a possible data transmission failed. On the other hand, valuable time is saved by the method according to the invention, which could save lives in the event of a serious traffic accident.

In the fictitious exemplary embodiment listed, it should be noted that the value for the event risk R, on which the transmission of the trigger signal T and thus the anticipatory data collection are based, fluctuates around the value 0.7 within the times t ₅ and t ₂₀ . However, since the threshold for triggering the trigger T is set to the value 0.7, for example, there is a critical period of time in which a trigger sig nal T is transmitted with an uncertain probability, although the dangerous situation already exists. This circumstance can be remedied by the application-specific parameterization of the calculation of the event risk R, for example by lowering the first threshold value for sending out the trigger signal T. Alternatively, the detection and evaluation of the risk parameters P could be designed more sensitively, a weighting of individual risk group values RF, _RS , _{R H} , R _W , R _I are introduced into the calculation of the event risk R, the effects of risk premiums are adjusted or further conditions for sending out a trigger signal T are added. Thus, the method according to the invention can be parameterized in an advantageous manner in a way that is suitable for the application and versatile.

In 6A Another scenario is shown as an exemplary embodiment, comprising an intersection on which there are several vehicles 2A, 2B, 3, 3A, 3B and cyclists 25, two traffic lights 24A, 24B, a speed-limiting traffic sign 20 and a surveillance camera 4, which insight to the middle of the street crossing. At the time shown, a first traffic light 24A is red and a second traffic light 24B is green, so that the horizontal direction of travel is stopped and the vertical direction of travel is released. In the vertical direction of travel, from bottom to top, a first vehicle 2A involved in a later event is driving on the left-hand lane and two cyclists 25 and a first uninvolved vehicle 3A are on the right-hand lane. The first uninvolved vehicle 3A recognizes by means of an environmental analysis 23 by the detection device 17 the road users located in the vertical direction of travel, the switching of the second traffic light 24B and the current speed limit based on the traffic sign 20. A second vehicle 2B involved in the later event is in the horizontal lane , which is at the stop line of the first traffic light 24A. Some of the road users include communication devices 14 that are set up to receive trigger signals T of the method according to the invention and other data, and to send data. These are marked with a symbol for network transmissions. In this exemplary embodiment, the second vehicle 2B involved and the first uninvolved vehicle 3A include a vehicle control unit 15 that is set up to carry out the method according to the invention by continuously determining an event risk R.

In 6B the same scenario is shown at a second point in time, with the first traffic light 24A switched to green and the second traffic light 24B switched to red, with the first uninvolved vehicle 3A and a cyclist 25 standing in the vertical direction of travel at the stop line to the second traffic light 24B. The second vehicle 2B involved has already entered the intersection. However, the first vehicle 2A involved is still moving and is passing the stop line of the second traffic light 24B, which is red. Both the first uninvolved vehicle 3A and the second involved vehicle 2B recognize an event risk R through the calculations within their vehicle control units 15 by evaluating the recorded risk parameters P, so that a first trigger signal TA comes from the first uninvolved vehicle 3A and from the second involved vehicle 2B second trigger signal TB is sent. Referring to 2 both vehicles 2B, 3A start the automatic data collection according to the process flow of the method according to the invention. The surveillance camera 4 receives the triggered trigger signals T and also starts data collection.

In 6C the same scenario is shown at a third point in time, the two vehicles involved 2A, 2B having already collided in the middle of the intersection. Due to the data collection and data transmission according to the invention to an external server 7, an accident event has already been identified by the computer program integrated in said external server 7 and the type and number of vehicles and people involved have already been forwarded to a rescue service 8, with the accident now being caused by eyewitnesses Emergency calls can be validated. Furthermore, all transmitted data are stored for a short time and transmitted to a security service 10 in order to clarify the course of the accident and to initiate possible criminal prosecution. In this exemplary embodiment, the method according to the invention benefits from the presence of a plurality of vehicles 2B, 3A and technical means which are set up to carry out the method, so that the data is collected from different perspectives, with different types of data and in different quality, in order to provide the most comprehensive information collection possible . Advantageously, data is also available for the first vehicle 2A involved, which does not have a corresponding vehicle control unit 15 and/or a detection device 17 that is set up to carry out the method according to the invention.

In a further development, the method according to the invention can also be used for acts of violence and other crimes. For this purpose, the calculation of the event risk R can include further risk parameters P that can be evaluated, which, for example, record acts of violence using image recognition, with alarm systems or other security devices be linked to security devices or be involved in police radio or other data transmissions. In an alternative embodiment, uninvolved third parties, such as security services 10 or police authorities, can trigger signals T in order, for example, to observe a possible escape route for criminals.

reference list

1

networked communication system

2, 2A, 2B

vehicle involved

3, 3A, 3B

uninvolved vehicle

4

Security Camera

5

person

6

data transfer unit

7

external server

8th

rescue service

9

rescue vehicle

10

security service

11

security vehicle

12

suspect

13

roadway

14

communication facility

15

vehicle control unit

16

drive unit

17

detection device

18

data interface

19

data storage

20, 20A, 20B

traffic sign

21A, 21B

roadway impairments

22

tree

23

environment analysis

24A, 24B

Traffic light

25

cyclist

P

risk parameters

PF

vehicle parameters

PI

occupant parameters

PR

friction parameter

p.s

route parameters

pw

weather parameters

R

event risk

RF

driving risk

RH

risk of stiction

RI

occupant risk

RS

route risk

RW

weather risk

S1

recognize event

S2

event detected?

S3

predict event

S4

event predicted?

S5

release triggers

S6

process data

t

time step

T, TA, TB

triggers

QUOTES INCLUDED IN DESCRIPTION

This list of documents cited by the applicant was generated automatically and is included solely for the better information of the reader. The list is not part of the German patent or utility model application. The DPMA assumes no liability for any errors or omissions.

Patent Literature Cited

• DE 102005026517 A1 [0002]
• DE 102014015669 A1 [0002]
• DE 102016216522 A1 [0003]
• DE 102020004792 A1 [0004]
• DE 102019203543 A1 [0005]
• DE 102016003574 A1 [0006]
• DE 102017201426 A1 [0006]

Claims (10)

Method for anticipatory data collection from networked road users by determining an event risk R of an event involving at least one road user, with at least one risk parameter P being repeatedly recorded by the named and/or at least one other road user and the event risk R repeatedly being determined as a function of the at least one risk parameter P is determined, with the at least one risk parameter P being determined by at least one detection device 17, which is assigned to the named and/or the other road user, with a trigger signal T depending on the determined event risk R by means of a communication device 14, which is assigned to the named and/or the is assigned to other road users, is emitted, which can be received by other road users or technical means, which include at least one communication device 14, characterized in that the road users and/or technical means sending out and receiving said trigger signal T start data collection as a function of trigger signal T, with one or more event data being recorded and stored and/or transmitted to an external server 7 by means of communication device 14, with the sending of trigger signal T and/or begin recording event data prior to the event. procedure after claim 1 , characterized in that the determination of the event risk R includes a conversion of detectable risk parameters P into at least one of the risk group values R _F , R _S , R _R , R _W , R _I , which each represent a probability value. procedure after claim 2 , characterized in that the determination of the event risk R includes the calculation of a mean value from one or more of the risk group values R _F , R _S , R _R , R _W , R _I . procedure after claim 3 , characterized in that the trigger signal T is then transmitted by the at least one communication device 14 when the determined event risk R exceeds a first threshold value. procedure after claim 4 , characterized in that the trigger signal T is transmitted as a function of the detected risk parameters P, the determined risk group values RF, R _S , R _R , R _W , R _I and/or a risk premium until the event risk _R determined falls below a second threshold value or a specified period of time ends. Method according to one of the preceding claims, characterized in that at least one road user determining the risk event R comprises a vehicle 2, 3 which is stationary or moving autonomously or is being driven by a person. Method according to one of the preceding claims, characterized in that the recordable event data includes one or more image, video, audio, GPS, traffic, damage, personal data and/or operating parameters of vehicles 2, 3. Method according to one of the preceding claims, characterized in that event data that can be transmitted to the external server 7 is recorded and forwarded in encrypted and/or anonymized form and that decryption and/or de-anonymization takes place as a function of a declaration of consent under data protection law and that in the case of persons for If there is no declaration of consent mentioned, a natural face anonymization is carried out. Vehicle control unit 15, which is set up to to repeatedly receive at least one risk parameter P from a detection device 17 assigned to a road user, • to repeatedly determine an event risk R as a function of the at least one risk parameter P, • to send out a trigger signal T depending on the event risk R, • and to carry out a method according to any one of the preceding claims. Vehicle 2, 3, comprising a vehicle control unit 15 after claim 9 , At least one detection device 17 and at least one communication device 14, which is set up to a method according to one of Claims 1 until 8th to execute.

Images