DE102023000709A1 - Method for improving driving safety in a motor vehicle - Google Patents

Abstract

The invention relates to a method for improving driving safety in a motor vehicle (10), according to which feature values (MW) of features (M) are detected by means of a sensor system (11) of the motor vehicle (10) while the motor vehicle (10) is driving characterize a driver physiologically, emotionally and driver-specifically while driving. According to the method, by comparing the recorded characteristic values (MW) with stored reference values (RW) it is checked and/or the characteristic values (MW) are mapped to the probabilities as to whether there is at least one symptom of a health problem that at least limits the driver's ability to drive and, If this is true or a probability threshold is exceeded, an emergency program (NP) is processed.

Description

The invention relates to a method for improving driving safety in a motor vehicle. The invention further relates to a motor vehicle with a control/regulation device which is set up/programmed to carry out the method according to the invention.

The DE 10 2019 204 786 A1 deals with a system and a method for driver condition monitoring in a vehicle, in which the driver's vital functions are determined using sensor data and specified driver and driving data of the vehicle are stored and monitored. These data are processed in combination in a control device in such a way that the driver's physical constitution is determined. In addition, health information is used to determine vital signs. The control device uses this data to prepare a risk assessment, with a warning being issued if the driver's physical constitution no longer corresponds to a predetermined target value. Not only is the driver informed, but also external facilities such as databases or control centers for emergency calls or road monitoring.

The DE 10 2021 003 489 A1 discloses a method for detecting a gradually decreasing driving performance of a vehicle driver, in which individual driving-specific behavior as well as physiological and emotional characteristics of a driver as well as his health data are recorded, which are combined and evaluated to determine driving performance. This makes it possible to counteract the risk of accidents by early detection of vehicle drivers with a gradual loss of driving ability.

It is an object of the present invention to show new ways in the development of methods for improving driving safety when driving a motor vehicle.

This task is solved by the subject matter of the independent patent claims. Preferred embodiments are the subject of the dependent claims.

The basic idea of the invention is therefore to compare feature values of features that describe the driver or another vehicle occupant of a motor vehicle physiologically, emotionally and driver-specifically during the journey with stored reference values and in this way to determine whether a symptom of a health restriction or a health problem of the driver is present that could impair the driver's ability to drive and thus driving safety during the journey. If such a symptom - for example a symptom that indicates an impending heart attack or stroke - is detected during the comparison, the invention proposes to run an emergency program by means of which an obvious health problem is reacted to. In this way, negative consequences associated with such a negative problem - for example the risk of a traffic accident due to the driver's reduced ability to drive associated with the health problem - can be counteracted.

The method according to the invention therefore serves to improve driving safety in a motor vehicle. In the method according to the invention, characteristic values of physiological, emotional and driver-specific features which characterize the driver are recorded by means of a sensor system of the motor vehicle while the motor vehicle is driving. By comparing the recorded characteristic values with stored reference values, it is checked whether there is at least one symptom of a health problem that at least limits the driver's ability to drive, and, if this is the case, an emergency program is processed. The reference values can also be a stored probability distribution which the recorded feature values can be mapped, the probability distribution determining a probability of the driver's limited ability to drive due to at least one limiting health problem depending on the recorded feature values.

Said reference values or images can be taken from an external database. This database can be constructed based on characteristic values that are recorded using a fleet of motor vehicles and transmitted to the database by the respective motor vehicle.

For this purpose, in the event that an electronic emergency call (“eCall”) is triggered automatically or manually by a vehicle in the fleet, the characteristic values belonging to the characteristics are transmitted to an external database at the time the emergency call is triggered in a period before the emergency call is triggered. In order to further specify this transmitted information, the data record is expanded to include a specific symptom or a specific marker for the reason for the emergency call, e.g. acute health problems or accident, as well as, if applicable, the type of acute health problem (e.g. heart attack, stroke, epilepsy attack). , panic attack, etc.) expanded. This can in particular via the input and/or automatic reading of medical findings from an electronic patient file after a diagnosis of the health problem has taken place.

To carry out the method according to the invention, a current copy of the database with said reference values and/or learned images or heuristic functions can be stored in a control/regulation device present in the motor vehicle according to the invention, which carries out the method according to the invention. For example, certain reference values and/or a time course of one or more characteristic values can be assigned to the symptom or marker “heart attack”, “stroke” or “panic attack”.

Said characteristic values of physiological, emotional and driver-specific characteristics of the driver of a motor vehicle can be recorded via the sensors installed in the vehicle and processed in a control/regulation device provided in the motor vehicle. Physiological characteristics such as heartbeat frequency, breathing frequency, skin conductivity, etc. can be measured using optical or electronic sensors in the sensor system. Driver-specific features can also be obtained from acceleration and speed sensors in the motor vehicle's sensor system as well as with the help of other sensors in the sensor system, for example to determine the wheel speed or engine speed. Driver-specific features can be extracted from driver assistance systems (“ADAS”), which form part of the sensor system. For example, the number of vehicle assistance interventions such as emergency braking and steering wheel interventions to stay in lane, as well as information about the lateral positioning of the motor vehicle in a lane and also information regarding the distances selected by the driver from vehicles in front can be used.

At least one interior camera installed in the motor vehicle and at least one interior microphone of the sensor system present in the vehicle interior of the motor vehicle also allow data about the driver of the motor vehicle to be recorded in a time-resolved manner and processed in the control/regulation device mentioned. In this case, computer-aided/machine vision methods make it possible to recognize and localize the position of the driver in the vehicle interior as well as his body movements, for example gestures.

Optionally, the driver can also be clearly identified using said sensors. The identification can be carried out using a variety of methods, in particular through image/sound recordings, in particular from the interior camera of the motor vehicle, and through machine vision methods, so that the recorded characteristic values can be clearly assigned to a specific person.

By comparing the recorded feature values with stored reference values, the method according to the invention checks whether there is at least one symptom of a health problem that at least limits the driver's ability to drive. For this purpose, the stored reference values can be assigned to at least one specific symptom - hereinafter also referred to as a “marker”.

In a preferred embodiment, the stored reference values are obtained from a vehicle fleet by a machine learning model based on manual or automatic emergency calls from the fleet, with the type of emergency, the reference values and the feature values before and during the emergency call as input data for the model be used.

In a preferred embodiment, alternatively or in addition to the comparison with the stored reference values, a mapping can be trained using machine learning methods and/or a heuristic function can be defined which assigns the feature values of features to a symptom of a health problem that at least limits the driver's ability to drive.

For this purpose, a variety of machine learning methods can be used, for example ensembles of decision trees, deep neural networks, which learn a mapping from the feature values - before or at the time the electronic emergency call is made - to the marker itself.

In the course of the method according to the invention, it is also conceivable to generate a heuristic function based on a statistical analysis of the overall data from the characteristic values of the same marker. For example, it can be exploited that a heart attack is preceded by an increase in the heartbeat frequency (“heart palpitations”) and/or the conductivity of the skin (“anxiety” and “sweating”) that deviates greatly from normal behavior, which is determined by a threshold adjustment from a heuristic function of the parameters Heartbeat frequency and skin conductivity can be detected with a previously defined threshold value.

If the recorded characteristic values correspond to certain stored reference values, In the method according to the invention, it is determined that the symptom associated with these reference values or the marker for a health problem associated with these reference values is present, and an emergency program is executed in response to the recognized symptom. Many possibilities are conceivable for determining, in the course of the comparison essential to the invention, which feature values must be present in order to either affirm or deny the presence of a symptom for a health problem when compared with the stored reference values. The relevant expert will select from such variants depending on the application. In particular, it can be provided that individual or all of the recorded feature values deviate from the respective reference value by at most a common threshold value or a threshold value individually defined for each recorded feature.

The emergency program to be executed can be executed depending on the symptom or marker detected. Different detected symptoms or markers can therefore lead to the execution of different emergency programs. In this way, an individual emergency program, ideally adapted to the type of symptom or marker detected, can be carried out for each detected symptom or marker.

In a preferred embodiment, the emergency program can include an automatic transmission of an emergency call to an emergency control center. The emergency control center informed in this way can in turn arrange for medical assistance to be provided to the driver.

According to an advantageous development of the method according to the invention, the emergency program can also include the provision of at least one support measure for safely stopping and parking the motor vehicle. Since a health problem is typically accompanied by an impairment of the driver's ability to drive, the risk of the driver causing an accident can be significantly reduced in this way.

Particularly preferably, the at least one support measure can include automatically reducing the speed of the motor vehicle.

Likewise preferably, the at least one support measure can include automatic preconditioning of a braking system of the motor vehicle for braking the motor vehicle. This makes it possible to prepare for any braking process that may still need to be triggered by the driver, whereby the motor vehicle can be brought to a standstill particularly quickly if the driver presses the brake pedal.

A further preferred variant has proven to be particularly advantageous, in which the at least one support measure includes the calculation of navigation data from the current position of the motor vehicle to a suitable stopping position. This makes it easier for the driver to maneuver the motor vehicle in a relatively safe position - for example a shoulder or a traffic bay - and park it there, especially in the event of a health problem. The risk of a collision between the motor vehicle that has been parked due to an emergency and another vehicle, particularly if this vehicle has been brought to a stop in a comparatively confusing position due to the emergency situation that has arisen, can be reduced to a significant extent in this way.

According to a further advantageous development, the at least one support measure can include a display of path information based on the calculated navigation data about the calculated stopping position. This makes it easier for the driver to navigate the motor vehicle to the desired, safe stopping position, especially if the driver's ability to control the motor vehicle is reduced due to the health problem that has occurred.

In another preferred embodiment of the method according to the invention, the external database containing the reference values will be updated by transmitting the recorded feature values to this database. In this way, the quality of the feature values stored in the database will be continuously improved and made available to other motor vehicles, so that when the method according to the invention is carried out, an improved comparison of the recorded feature values with the continuously improved reference values stored in the database is achieved.

The invention further relates to a motor vehicle with a brake system for braking the motor vehicle and with a navigation system for calculating and displaying navigation data. Furthermore, the motor vehicle according to the invention comprises a sensor system for detecting physiological, emotional and/or driver-specific features that characterize a driver of the motor vehicle, in particular while the motor vehicle is driving. Furthermore, the motor vehicle according to the invention comprises a control/regulation device for evaluating the features detected by the sensor system. For this purpose, the one with the Sensor system interacting control/regulation device is connected to the sensor system in a data-transmitting manner, and the control/regulation device is set up/programmed to carry out the method according to the invention presented above. The advantages of the method according to the invention explained above are therefore transferred to the motor vehicle according to the invention.

In a preferred embodiment of the method according to the invention, the motor vehicle comprises a communication device which is connected to the control/regulation device in a data-transmitting manner for transmitting the recorded features to an external database. As a result, the quality of the data stored in the database will be continuously improved and made available to other motor vehicles, so that when carrying out the method according to the invention, an improved comparison of the recorded feature values with the continuously improved reference values stored in the database is achieved.

Further important features and advantages of the invention emerge from the subclaims, from the drawings and from the associated description of the figures based on the drawings.

It is understood that the features mentioned above and those to be explained below can be used not only in the combination specified in each case, but also in other combinations or alone, without departing from the scope of the present invention.

Preferred exemplary embodiments of the invention are shown in the drawings and are explained in more detail in the following description, with the same reference numbers referring to the same or similar or functionally the same components.

• 1 ein Beispiel eines erfindungsgemäßen Kraftfahrzeugs,
• 2 ein das erfindungsgemäße Verfahren beispielhaft illustrierendes Ablaufdiagramm.
• 3 eine das Anlernen von Merkmalswerten erläuternde Darstellung.

Show schematically:

• 1 an example of a motor vehicle according to the invention,
• 2 a flow chart illustrating the method according to the invention by way of example.
• 3 a representation explaining the learning of feature values.

The 1 schematically illustrates an example of a motor vehicle 10 according to the invention, which 2 a flowchart of the method according to the invention. According to 1 the motor vehicle 10 according to the invention has an in 1 vehicle interior 21, not shown in detail. In addition, the motor vehicle 10 includes a sensor system 11 for detecting physiological, emotional and driver-specific features M, whose feature values MW (cf. 2 ) a driver (in 1 not shown, cf. 2 ) of the motor vehicle 10 while the motor vehicle 10 is traveling. Furthermore, the motor vehicle 10 includes a control/regulation device 12 which interacts with the sensor system 11 for evaluating the features M detected by the sensor system 11. For this purpose, the control/regulation device 12 and the sensor system 11 are connected to one another in a data-transmitting manner. The sensor system 11 can be a device 11a for determining a current heart rate of the driver - for example in the form of a so-called "smartwatch" with an integrated function for measuring the current heart rate of its wearer - as well as an interior microphone 11b and an interior camera 11c for acoustic or optical monitoring of the vehicle interior 21 and thus also the driver located in the vehicle interior 21.

The motor vehicle 10 further includes a brake system 16 for braking the motor vehicle 10 and a navigation system 17 for calculating and displaying navigation data. Both the brake system 16 and the navigation system 17 can be controlled by the control/regulation device 12. Likewise, the motor vehicle 10 includes an emergency call unit 15 that can be controlled by the control/regulation device 12 and by means of which an emergency call can be made to an external rescue control center 14 that is not part of the motor vehicle 10.

As in 1 Also illustrated, the motor vehicle 10 comprises a communication device 13 which is connected to the control/regulation device 12 in a data-transmitting manner for wirelessly transmitting the feature values MW (cf. 2 ) of the recorded features M to an external database 20 and - in the opposite direction (see double arrow 19 in 1 ) - for transferring reference values RW of the features M from the database 20 into a storage unit 12a present in the control/regulation device 12.

The control/regulation device 12 is set up or programmed to carry out the method according to the invention, which is described below using the flow chart of 2 should be explained as an example.

The method according to the invention serves to improve driving safety in the motor vehicle 10. Therefore, in a measure preceding the actual method, it can be checked whether the motor vehicle 10 is being used or whether no user is sitting in the vehicle. This can be done, for example, using interior cameras in the vehicle (not shown) or weight sensors on the seats (not shown) can be checked. Since in the latter case the occurrence of a health problem in the driver cannot be recognized, the method according to the invention can be terminated prematurely (see field “END” in 2 ). If, on the other hand, a driver is recognized, during the journey according to a first measure (a) of the sensor system 11 of the motor vehicle 10, feature values MW of features M are included, which characterize the driver physiologically, emotionally and driver-specifically.

In a second measure (b) of the method according to the invention, it is checked whether at least one symptom of a health problem that at least limits the driver's ability to drive is present. For this purpose, the characteristic values MW recorded in measure a) are compared with reference values RW. The reference values RW stored in a storage unit 12a of the control/regulation device 12 can be assigned to at least one specific symptom. The reference values RW may have been transferred from the external database 20 into the storage unit 12a. If, in the course of the comparison, the determined feature values MW can be assigned to certain stored reference values RW, a symptom assigned to these reference values RW is determined to have occurred in the driver.

If, in turn, the presence of a specific symptom or a specific marker was recognized through said comparison, an emergency program NP is carried out in a third measure (c) of the method according to the invention. The emergency program NP can be executed depending on the symptom or marker detected. Different detected symptoms or markers can therefore lead to the execution of different emergency programs NP (not shown in the figures).

The emergency program NP can include a first program measure PM1, according to which the motor vehicle 10 makes an emergency call to a rescue control center 14. The rescue control center 14 can initiate suitable help measures in response to the emergency call. The emergency call can be transmitted as follows 1 shown can be done wirelessly using an emergency call unit 15, which can carry out a so-called “eCall” function.

Alternatively or in addition to the first program measure PM1 of the emergency program NP, this can include a second program measure PM2, which includes the implementation of one or more support measures UM1, UM2, UM3 for safely stopping and parking the motor vehicle 10.

A first such support measure UM1 can be an automatic reduction of the speed of the motor vehicle 10. This reduces the risk of a collision of the motor vehicle 10 at high vehicle speed. In an extreme case, it is also conceivable that the motor vehicle 10 is automatically brought to a complete standstill as part of the support measure UM1.

A second support measure UM2 can include automatic preconditioning of a brake system 16 of the motor vehicle 10 for braking the motor vehicle 10. This makes it possible to prepare for any braking process to be triggered by the driver, whereby the motor vehicle 10 can be brought to a standstill particularly quickly if the brake pedal is actuated by the driver.

According to a third support measure UM3, navigation data can be calculated from the current position of the motor vehicle 10 to a suitable stopping position, which makes it easier for the driver to stop and park the motor vehicle 10 at a relatively safe position - for example a shoulder or a traffic bay.

Said navigation data can expediently be displayed to the driver in a display device (not shown in the figures) present in the motor vehicle 10, for example in a “head-up display”. This makes it easier for the driver to navigate the motor vehicle 10 to the desired stopping position.

Regardless of said program measures PM1 and PM2, the database 20 containing the reference values RW can be updated in the course of the emergency program NP in a further program measure PM3 using the feature values MW recorded in the course of the method according to the invention.

The 3 shows a flowchart that illustrates the learning of feature values. First, a measure (x0) checks whether the motor vehicle has stopped driving. If this is not the case, different feature values are recorded in a measure (x1) while the motor vehicle 10 is traveling using the sensor system 11. If an emergency call is triggered at a certain point in time, the characteristic values recorded up to this point in time are stored in the control/regulation device 12 and, alternatively or additionally, in a measure (x2) transmitted to the external database 20 via the communication connection 19. If it is recognized that a health problem exists, the corresponding marker of the identified health problem is saved in a measure (x3).

In one measure (x4), training data of a machine learning model can be obtained from a vehicle fleet based on manual or automatic emergency calls from the fleet, with the type of emergency, the marker and the feature values before and during the emergency call as input data for the model be used. For this purpose, a variety of machine learning methods can be used, for example ensembles of decision trees, deep neural networks, which learn a mapping from the feature values MW - in time before or at the time of making the electronic emergency call - to the marker itself. Based on the totality of data sets obtained from the fleet, reference values RW are extracted and/or a measure is created to map the characteristic values MW, which were recorded before the electronic emergency call was made, onto the respective marker. For example, all data sets with the marker “heart attack”, i.e. the MW characteristic values that precede an automatically or manually triggered electronic emergency call and are subsequently assigned to the diagnosis “heart attack”, can be automatically mapped to the “heart attack” class, whereas all data sets with The “panic attack” marker can be mapped to the “panic attack” class. A variety of machine learning methods can be used, e.g. ensembles of decision trees, deep neural networks, etc., which learn a mapping of the feature values to the marker itself before or when the electronic emergency call is made. In general, a probability of the occurrence of a respective marker is at least approximated based on the characteristic values, whereby the probability of a non-safety-relevant or a so-called “normal state marker” can also be added.

Furthermore, it is possible to generate a heuristic function based on a statistical analysis of the overall data from the data sets with the characteristic values of the same marker. For example, it is conceivable that a heart attack is preceded by an increase in the heart rate ("rapid heartbeat") and/or the conductivity of the skin ("anxiety", "sweating") that deviates greatly from normal behavior, which is determined by a threshold adjustment based on a heuristic function of the parameters “Heartbeat frequency” and “skin conductivity” can be detected with a previously defined threshold value. A data record in the database can therefore contain information which includes reference values and/or learned images or heuristic functions.

In a further measure (x5), the training data learned in measure (x4) can be transferred to all motor vehicles in the vehicle fleet, where they are available for carrying out the method according to the invention.

QUOTES INCLUDED IN THE DESCRIPTION

This list of documents listed 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.

Cited patent literature

• DE 102019204786 A1 [0002]
• DE 102021003489 A1 [0003]

Claims (12)

Method for improving driving safety in a motor vehicle (10), - according to which feature values (MW) of features (M) which characterize a driver of the motor vehicle (10) physiologically, emotionally and/or driver-specifically are recorded by means of a sensor system of the motor vehicle (10) while the motor vehicle is driving, - according to which, by comparing the recorded characteristic values (MW) with reference values (RW) stored, in particular in an external database (20), it is checked whether at least one symptom of a health problem that at least limits the driver's ability to drive is present and, if this is the case , an emergency program (NP) is processed. Procedure according to Claim 1 , characterized in that, alternatively or in addition to the comparison with stored reference values (RW), an image is trained using machine learning methods and/or a heuristic function is defined, which determines the feature values (MW) of features (M) of a symptom for a Assign the driver's fitness to drive to at least a limiting health problem. Procedure according to Claim 1 or 2 , characterized in that the reference materials are obtained using a machine learning model from a vehicle fleet of motor vehicles based on manual or automatic emergency calls from the vehicle fleet, with the type of emergency being the marker and the feature values before and during the emergency call as input parameters of the model be used. Procedure according to one of the Claims 1 until 3 , characterized in that the emergency program (NP) includes automatic transmission of an emergency call to a rescue control center (14). Method according to one of the preceding claims, characterized in that the emergency program (NP) comprises the provision of at least one support measure (UM1, UM2, UM3) for stopping and parking the motor vehicle (10). Procedure according to Claim 5 , characterized in that the at least one support measure (UM1) includes automatically reducing the speed of the motor vehicle (10). Procedure according to Claim 5 or 6 , characterized in that the at least one support measure (UM2) comprises automatic preconditioning of a brake system (16) of the motor vehicle (10) for braking the same. Procedure according to one of the Claims 5 until 7 , characterized in that the at least one support measure (UM3) includes the calculation of navigation data from the current position of the motor vehicle (10) to a suitable stopping position. Procedure according to Claim 8 , characterized in that during the course of the support measure (UM3), path information from the current position of the motor vehicle (10) to the calculated stopping position is displayed. Method according to one of the preceding claims, characterized in that the reference values (RW) contained in the database (12) are updated by means of the feature values (MW) recorded and transmitted to the external database (12). Motor vehicle (10), - with a brake system (16) for braking the motor vehicle (10), - with a navigation system (17) for calculating and displaying navigation data, - with a sensor system (11) for detecting characteristic values (MW) of physiological, emotional and/or driver-specific features (M) characterizing a driver of the motor vehicle (10) while the motor vehicle (10) is driving, - with a control/regulation device (12) which interacts with the sensor system (11) and is connected to transmit data and which is set up/programmed to carry out the method according to one of the preceding claims. Motor vehicle according to Claim 9 , characterized in that the motor vehicle (10) comprises a communication device (13) which is connected to the control/regulation device (12) for transmitting data and for transmitting the detected features to an external database (20).

Images