DE102022000169A1 - Method for offering vehicle functions - Google Patents

Abstract

The invention relates to a method for offering vehicle functions (F1 to Fn) to a vehicle occupant (P1 to Pk) in a vehicle (1). According to the invention, settings and behavior of vehicle functions (F1 to Fn), vehicle occupants (P1 to Pk) and sensors (S1 to Sm) are evaluated continuously by means of a control unit (SE), - by means of the control unit (SE) for this Vehicle occupants (P1 to Pk), a vehicle function profile (FP1 to FP3) is created for a predetermined selection of vehicle functions (F1 to Fn), which are determined by the vehicle occupants (P1 to Pk), their activities and/or detected sensor values (SW) can be influenced,- such a vehicle function profile (FP1 to FP3) is automatically determined in a parameter space (R) using machine learning methods,- by means of the control unit (SE) based on the vehicle function profile (FP1 to FP3) determined in each case Vehicle occupants (P1 to Pk) are offered automation for setting this vehicle function (F1 to Fn) as a function of detected sensor values (SW) and/or the automation ing is adjusted as a function of recorded sensor values (SW),- fleet-wide vehicle function profiles (FP1 to FP3) are mapped and compared via a central storage unit (C) and- vehicle occupant-specific preferences recorded using the determined vehicle function profiles (FP1 to FP3) reliably and stored anonymously and shared across vehicles.

Description

The invention relates to a method for offering vehicle functions to a vehicle occupant in a vehicle.

From the DE 10 2020 006 693 A1 a method for operating an anti-theft system for a motor vehicle and an anti-theft system are known.

The method provides that an unauthorized action by a person is detected by means of a detection device of the anti-theft system and, when the impermissible action is detected, a control signal for an alarm system of the anti-theft system and a control signal for a controllable chassis device of the motor vehicle are generated by means of an electronic computing device of the anti-theft system. The control signal for the controllable chassis device is generated in such a way that the chassis device is moved into a completely lowered position when the control signal is received.

The invention is based on the object of specifying a method for offering vehicle functions to an occupant in a vehicle.

The object is achieved according to the invention by a method which has the features specified in claim 1.

Advantageous configurations of the invention are the subject matter of the dependent claims.

• - mittels einer Steuereinheit kontinuierlich Einstellungen sowie ein Verhalten von Fahrzeugfunktionen, des Fahrzeuginsassen und Sensoren ausgewertet werden,
• - mittels der Steuereinheit für diesen Fahrzeuginsassen jeweils ein Fahrzeugfunktions-Profil für eine vorgegebene Auswahl von Fahrzeugfunktionen erstellt wird, welche durch den Fahrzeuginsassen, dessen Aktivitäten und/oder erfasste Sensorwerte beeinflussbar ist,
• - ein solches Fahrzeugfunktions-Profil mittels Verfahren des maschinellen Lernens automatisiert in einem Parameterraum ermittelt wird,
• - mittels der Steuereinheit basierend auf dem jeweils ermittelten Fahrzeugfunktions-Profil dem Fahrzeuginsassen in Abhängigkeit erfasster Sensorwerte eine Automatisierung zur Einstellung dieser Fahrzeugfunktion angeboten wird und/oder die Automatisierung in Abhängigkeit erfasster Sensorwerte angepasst wird,
• - über eine zentrale Speichereinheit flottenübergreifende Fahrzeugfunktions-Profile abgebildet und verglichen werden und
• - anhand der ermittelten Fahrzeugfunktions-Profile erfasste fahrzeuginsassenspezifische Präferenzen sicher und anonymisiert gespeichert und fahrzeugübergreifend geteilt werden.

According to the invention, a method for offering vehicle functions for a vehicle occupant in a vehicle provides that

• - settings and behavior of vehicle functions, the vehicle occupants and sensors are continuously evaluated by means of a control unit,
• - a vehicle function profile for a predetermined selection of vehicle functions is created by means of the control unit for this vehicle occupant, which can be influenced by the vehicle occupants, their activities and/or detected sensor values,
• - such a vehicle function profile is automatically determined in a parameter space using machine learning methods,
• - by means of the control unit, based on the vehicle function profile determined in each case, the vehicle occupant is offered automation for setting this vehicle function as a function of detected sensor values and/or the automation is adapted as a function of detected sensor values,
• - Fleet-wide vehicle function profiles are mapped and compared via a central storage unit and
• - Based on the determined vehicle function profiles, vehicle occupant-specific preferences are stored securely and anonymously and shared across vehicles.

By using the method, a vehicle occupant can be made aware of the presence of certain vehicle functions, or the vehicle occupant is reminded of the presence of certain vehicle functions by means of the method so that they can be used if necessary. The method thus serves to provide transparency to the vehicle occupant.

Using the method, so-called established routines of using vehicle functions that do not offer any variety can be broken up, in particular by means of automated adjustment.

By using the method, vehicle occupants can see a range of vehicle functions present in the vehicle in a transparent manner, with the vehicle occupants experiencing an expanded range of vehicle functions. In addition, the vehicle occupants have the option of making changes to vehicle functions that have been set once.

Exemplary embodiments of the invention are explained in more detail below with reference to drawings.

• 1 schematisch ein Fahrzeug mit mehreren Fahrzeuginsassen und einer Vorrichtung zur Durchführung eines Verfahrens zur Angebotstransparenz von Fahrzeugfunktionen für die Fahrzeuginsassen,
• 2 schematisch die Vorrichtung zur Durchführung des Verfahrens, und
• 3 schematisch einen Parameterraum zur Ermittlung von Fahrzeugfunktions-Profilen.

show:

• 1 schematically a vehicle with several vehicle occupants and a device for carrying out a method for the transparency of vehicle functions for the vehicle occupants,
• 2 schematically the device for carrying out the method, and
• 3 schematically a parameter space for determining vehicle function profiles.

Corresponding parts are provided with the same reference symbols in all figures.

1 shows a vehicle 1 by way of example and in a greatly simplified manner. Several vehicle occupants P1 to Pk are seated in the vehicle 1 and the vehicle 1 has a device for carrying out a method for the transparency of vehicle radio services Functions F1 to Fn for the vehicle occupants P1 to Pk.

In 2 the device for carrying out the method is shown. 3 shows a parameter space R for determining vehicle function profiles FP1 to FP3 based on a function ζ1 to ζk of vehicle functions F1 to Fn, sensors S1 to Sm, activities and preferences of a vehicle occupant P1 to Pk identified therefrom and optionally considering a change over time over the time t To ensure clarity, three vehicle function profiles FP1 to FP3 are shown. However, fewer or more vehicle function profiles FP1 to FP3 in the parameter space R can also be determined.

It is generally known that, in addition to a comparatively high degree of automation of vehicle functions F1 to Fn, preferences of vehicle occupants P1 to Pk with regard to vehicle functions F1 to Fn and their settings in vehicle 1 are also recorded and/or predicted. Alternatively or additionally, combinations of a behavior of vehicle occupants P1 to Pk and vehicle functions F1 to Fn are defined, which are thus set or executed as standard. As a consequence, it can happen that vehicle functions F1 to Fn are only perceived on one side or not at all. In particular, a vehicle 1 has a large number of vehicle functions F1 to Fn, with the respective vehicle function F1 to Fn being able to adopt different settings which can be made by vehicle occupants P1 to Pk, for example.

A method described below can represent a solution to counteract a one-sided use of vehicle functions F1 to Fn, a reduced use of vehicle functions F1 to Fn and an ignorance regarding the presence of available vehicle functions F1 to Fn.

The device for carrying out the method for offering transparency with regard to the vehicle functions F1 to Fn has a plurality of sensors S1 to Sm and a control unit SE.

When the vehicle 1 is in use, the control unit SE continuously evaluates settings of the vehicle functions F1 to Fn present in the vehicle 1, behavior of the vehicle occupants P1 to Pk and detected sensor values SW of the sensors S1 to Sm.

In this way, the control unit SE can influence at least one of the sensors S1 to Sm, for example an interior camera, an interior microphone, etc., for a predefined selection of vehicle functions F1 to Fn, in particular those which are influenced by vehicle occupants P1 to Pk and/or detected sensor values SW are or can be dependent on them, each create a vehicle function profile FP1 to FP3.

For example, the control unit SE can create vehicle function profiles FP1 to FP3 from air conditioning settings, radio station settings, seat positions, air vent settings, sun visor settings, tank level, tire pressure etc. using a respective setting as a function of detected sensor values SW.

These vehicle function profiles FP1 to FP3 are automated in particular using machine learning methods, in particular statistical learning, in the in 3 parameter space R shown in a more detailed and simplified two-dimensional manner. In this case, discrete or discretized settings of vehicle functions F1 to Fn are optionally automatically extracted and features recognized as being relatively important are assigned to the sensors S1 to Sm.

The parameter space R is based on various parameters from the vehicle functions F1 to Fn and the vehicle occupants P1 to Pk, of which, as described above, their settings and/or their behavior are continuously recorded and evaluated using the sensors S1 to Sm. The control unit SE can thus derive the vehicle function profiles FP1 to FP3 for the respective vehicle function F1 to Fn within the parameter space R according to their typical settings or changes in settings as a function of detected sensor values SW.

The parameter space R and its change over time can be recorded using statistical methods, including using clustering, distance measurement methods, deep neural networks, and thus individually define the respective vehicle function profiles FP1 to FP3. In particular, preferences of vehicle occupants P1 to Pk with regard to settings of vehicle functions F1 to Fn can thus be derived. In addition, vehicle occupants P1 to Pk can also implicitly control and coordinate a behavior of the parameter space R. In particular, it is possible to use vehicle function profiles FP1 to FP3, as in 3 is shown, which are specific for a respective vehicle occupant P1 to Pk, each visualized as a person, and for specific vehicles 1. For example, sun visor settings can only be taken into account in vehicles 1 in which a sun visor is present.

A specific exemplary embodiment of a sun blind as a functional unit of a vehicle function F1 to Fn is described below. In the vehicle 1 there is a vehicle occupant P1, a vehicle function F1 relates to the sun blind, a light incidence measuring device is designated as sensor S1 and an interior camera as sensor S2.

Properties and settings of these components F1, S1, S2 listed above are continuously transmitted to the control unit SE over a period of time and a respective value, e.g. Sensor values SW, mapped in the parameter space R. The control unit SE determines the following relationship over time: If the sensor values SW recorded by the sensor S1 increase above a predetermined threshold value, the vehicle occupant P1 activates the vehicle function F1, ie the sun blind. If the recorded sensor values SW of the sensor S1 in the form of the incident light measuring device decrease, then the activation of the sun blind is reversed by the vehicle occupant P1. Based on such a recorded course, it is possible for the control unit SE to offer this vehicle occupant P1 to automatically activate the sun blind as a vehicle function F1 when it is detected that the predefined threshold value for the sensor values SW of the sensor S1 is exceeded. A behavior dependent on the detected sensor values SW and observed by the control unit SE can therefore be automated on the vehicle side for this vehicle occupant P1.

Furthermore, by means of the control unit SE z. For example, it can be recognized that the threshold value assumes a different range of values if it is determined on the basis of sensor values SW of the interior camera as sensor S2 that the vehicle occupant P1 is wearing sunglasses. As a result, the control unit SE can use the created vehicle function profiles FP1 to FP3 for the various vehicle functions F1 to Fn to adapt, for example refine, a desired automation.

The vehicle function profiles FP1 to FP3 determined in this way, their properties and their behavior as well as active or passive influences of the vehicle occupants P1 to Pk can also be mapped and compared across vehicles and/or fleets via a central storage unit C, as in 2 is shown in more detail. Using so-called similarity-based approaches, distance measurement methods, clustering methods, pattern recognition and matching methods and/or other methods of statistical learning, vehicle function profiles FP1 to FP3 can thus be recognized and related to vehicle occupant preferences P1 to Pk.

Similarities of vehicle function profiles FP1 to FP3 and preferences of vehicle occupants P1 to Pk can also be mapped using a distance measure in one embodiment of the method. If it is determined that the distance measure is less than a predefined first threshold value, a similarity is identified. Analogous to this, it is determined on the basis of a distance measure that exceeds a predetermined second threshold value that there is no similarity between vehicle function profiles FP1 to FP3 and preferences. These threshold values cannot only be described linearly in the sense of larger or smaller in relation to a measure. Rather, such a measure can also be determined depending on how recorded sensor values SW of the sensors S1 to Sm, settings of the vehicle functions F1 to Fn and behavior of the vehicle occupants P1 to Pk, z. B. over the course of time or depending on a change in other parameters, behave symmetrically to each other. Optimal threshold values can thus be detected and optimized via vehicle and/or fleet-wide data acquisition.

This makes it possible to compare similar preferences of vehicle occupants P1 to Pk and to offer vehicle occupants P1 to Pk suggestions for settings of vehicle functions F1 to Fn and for automated behavior of vehicle functions F1 to Fn based on the preferences determined in this way. Such a procedure can be compared with online ordering platforms that state, for example, that customers who bought this product also bought such a product. In this analogy, however, only settings of the vehicle functions F1 to Fn are compared as a function of detected sensor values SW of the sensors S1 to Sm. Patterns based on the vehicle occupants P1 to Pk in the vehicle 1 are not detected.

In particular, this specific vehicle 1 is taken into account when the method is used in the control unit SE. It is therefore taken into account whether there is a sun blind at all in the vehicle 1 and whether a recommended alternative preference is thus offered to the corresponding vehicle occupant P1 to Pk.

The method carried out in the control unit SE takes into account the respective vehicle function profile FP1 to FP3 in relation to the specific vehicle occupant P1 to Pk based on recorded sensor values SW of the sensors S1 to Sm, based on recorded parameters in relation to the vehicle occupant P1 to Pk and the settings of the Vehicle functions F1 to Fn, in particular based on temporal and contextual similarities with other vehicle function profiles FP1 to FP3. For example, this determines how the setting Change the length of the sun blind depending on whether you are wearing sunglasses or exposure to the sun, taking into account other vehicle occupants P1 to Pk.

Each vehicle function profile FP1 to FP3 contains found patterns of values from input values of the vehicle functions F1 to Fn, the sensor values SW of the sensors S1 to Sm and the activities of the respective vehicle occupants P1 to Pk determined from the sensor values SW. In this case, patterns define a set of these values, which behave similarly in the parameter space R depending on a contextual and, if appropriate, time profile.

A vehicle function profile FP1 to FP3 in a parameter space R always relates to a specific vehicle 1 and a specific vehicle occupant P1 to Pk. In this way, vehicle function profiles FP1 to FP3 can also be compared with one another across parameter spaces via vehicle fleet data. In particular, characteristics of dissimilar vehicle function profiles FP1 to FP3 can also be recognized and described in this way.

A function ζ1 to ζk in a vehicle function profile FP1 to FP3 for a respective vehicle 1 and a vehicle occupant P1 to Pk forms a relationship between various input signals from vehicle functions F1 to Fn, sensor values SW and, if necessary, considering different points in time. Parameter properties, in particular clustered points in the parameter space R of the vehicle functions F1 to Fn in relation to sensor values SW, form a vehicle function profile FP1 to FP3.

The behavior of the patterns themselves, their change over time or context-related aspects, a density and correlations of the point clouds and their trajectories in the parameter space R are determined using statistical methods. Vehicle function profiles FP1 to FP3 are compared with one another using this information.

For example, the vehicle function profile FP1 to FP3 of the vehicle occupant P1 in the vehicle 1 behaves in relation to the sun blind as a vehicle function F1 depending on a location, just like a behavior of a vehicle occupant P2 in the vehicle 1 in relation to a sunroof as a vehicle function F2 depending on a location , the control unit SE can propose the vehicle function F2 sunroof to the vehicle occupant P1 as an alternative offer from the location.

In a further specific exemplary embodiment, the vehicle function profile FP1 to FP3 of the vehicle occupant P1 stores the fact that the sun blind is activated when the sensor values SW of the sensor S1 in the form of the incident light measuring device increase and a certain threshold value is exceeded. In a vehicle function profile FP1 to FP3 of vehicle occupant P2, it is stored that when the specific threshold value is exceeded by detected sensor values SW of sensor S1, the sun blind is activated as vehicle function F1 and the sunroof is opened as vehicle function F2. It can be suggested to the vehicle occupant P1 that the sliding roof can be opened when the threshold value for activating the sun blind is exceeded.

In this way, u. a. Detect vehicle function profiles FP1 to FP3 that have not yet occurred either in the respective vehicle occupants P1 to Pk or in the respective vehicle 1. Likewise, variants of vehicle function profiles FP1 to FP3 can be detected across fleets. While the latter suggests that this is vehicle behavior that is undesirable for vehicle occupants P1 to Pk, the previously mentioned vehicle function profiles FP1 to FP3 can offer an opportunity to suggest new preferences to vehicle occupants P1 to Pk and thus serve as a digital convenience function.

In one possible embodiment of the method, it is provided that the comparisons of the vehicle function profiles FP1 to FP3 are also carried out in the vehicle 1 itself instead of using the central storage unit C.

Vehicle-to-vehicle communication enables information to be passed on in relation to the parameter space R and a calculation can be carried out in a decentralized manner in the vehicles 1 . In this case, a geoposition of the respective vehicle 1 is to be taken into account in a restricted manner for these parameter spaces R.

In the event that vehicle occupant-specific characteristics with regard to personal preferences are stored and shared securely and anonymously, in particular in the sense of a fingerprint, a comparison can also be carried out directly on these fingerprints in a further development. This makes it possible for alternative settings of vehicle functions F1 to Fn to be suggested to the vehicle occupants P1 to Pk, in particular using a negative of the fingerprint, i. H. a personal fingerprint.

Reference List

1

vehicle

C

central storage unit

FP1 to FP3

Vehicle Function Profile

F1 to Fn

vehicle function

P1 to Pk

vehicle occupant

R

parameter space

SE

control unit

S1 to Sm

sensor

SW

sensor value

t

Time

ζ1 to ζk

function

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 102020006693 A1 [0002]

Claims (5)

Method for offering vehicle functions (F1 to Fn) for a vehicle occupant (P1 to Pk) in a vehicle (1), characterized in that - by means of a control unit (SE) continuously settings and behavior of vehicle functions (F1 to Fn), vehicle occupants (P1 to Pk) and sensors (S1 to Sm) are evaluated, - by means of the control unit (SE) for this vehicle occupant (P1 to Pk) a vehicle function profile (FP1 to FP3) for a predetermined selection of vehicle functions (F1 to Fn ) is created, which can be influenced by the vehicle occupants (P1 to Pk), their activities and/or recorded sensor values (SW), - such a vehicle function profile (FP1 to FP3) is automated using machine learning methods in a parameter space (R) is determined, - by means of the control unit (SE) based on the respectively determined vehicle function profile (FP1 to FP3) of the vehicle occupant (P1 to Pk) as a function of detected sensor values (SW) automation for setting this vehicle function (F1 to Fn) is offered and/or the automation is adapted depending on detected sensor values (SW), - fleet-wide vehicle function profiles (FP1 to FP3) are mapped via a central storage unit (C) and are compared and - based on the determined vehicle function profiles (FP1 to FP3), vehicle occupant-specific preferences are stored securely and anonymously and shared across vehicles. procedure after claim 1 , characterized in that similarities of vehicle function profiles (FP1 to FP3) and preferences of vehicle occupants (P1 to Pk) are mapped using a distance measure, wherein if the distance measure falls below at least a predetermined threshold value, a similarity is recognized. procedure after claim 2 , characterized in that vehicle function profiles (FP1 to FP3) are compared with one another across parameter spaces via vehicle fleet data. Method according to one of the preceding claims, characterized in that the selection of vehicle functions (F1 to Fn) made is marked in terms of time. Method according to one of the preceding claims, characterized in that the settings and the behavior of vehicle functions (F1 to Fn), vehicle occupants (P1 to Pk) and sensors (S1 to Sm) are evaluated taking different points in time into account.

Images