DE102017201728B4 - Control device for generating control data for actuators for configuring an interior situation of a motor vehicle and motor vehicle and method for operating a control device - Google Patents

Abstract

Control device (12) for generating control data (15) for actuators (16) in order to configure an interior situation of a motor vehicle (10), wherein a hardware interface (21) is set up to consist of sensors (14) of the motor vehicle (10), respectively Receiving sensor data (22) by means of a respective sensor-specific driver (24), generating input data (23) from the received sensor data (22) and providing the input data (23) for retrieval by sensor-unspecific retrieval functions, - a model level (25) set up for this purpose is to operate a plurality of models (28), each of the models (28) describing a respective interior-related object by means of attribute values (29) and being set up to retrieve at least one of the attribute values (29) by means of one of the retrieval functions on the Set the basis of at least part of the input data (23), - an interpretation level (26) is set up to use a number of interpretation functions (30) to be carried out, each interpretation function (30) being set up in each case to determine a property (31) of at least one of the modeled objects on the basis of at least one set attribute value (29) of at least one of the models (28), - a function level (27) for this is set up to operate control functions (32), each control function (32) being set up in each case to process some of the control data (15) as a function of at least one set attribute value (29) of at least one of the models (28) and/or as a function of to generate at least one determined property (31), and - a network interface (19) is set up to transmit the control data (15) via a data network (17) of the motor vehicle (10) to the actuators (16), characterized in that at least one of the interpretation functions (30) is set up, depending on the determined property (31), at least one further attribute value (29) in at least est one of the models (28) and thus at least one further attribute value (29) is derived by interpreting the situation in the interior simulated by the models (28), for which no sensor is then required to detect the respective attribute value (29).

Description

The invention relates to a control device for generating control data for actuators, with which an interior situation of a motor vehicle can be configured or influenced. The invention also includes a motor vehicle with the control device. The invention also includes a method for operating the control device in the motor vehicle.

For detecting the situation in an interior of a motor vehicle, i.e. the passenger compartment, a separate control unit with its own specific sensor system is usually used for each individual application, which is then highly optimized for the application, but in the case of two applications, who need similar sensors, the double sensor system is also available accordingly. Examples of specialized sensor technology are seat occupancy mats for airbag control, ultrasonic sensors in the anti-theft alarm system, capacitive sensors in the steering wheel or a single optical sensor for driver observation during piloted driving. Because the control units are provided separately from one another, these sensors cannot also be used for other control functions or synergies can be achieved by combining the sensor data.

However, the use of several sensors has advantages. From the DE 10 2014 219 326 A1 a sensor fusion with a smartphone in a motor vehicle is known, for example.

From the DE 10 2013 224 917 A1 a method for detecting an attention variable describing the attentiveness of a driver is known. A control device receives measurement data as well as a steering angle from the steering wheel and an accelerator pedal angle from the accelerator pedal. The controller then monitors the driver. A further function is not intended for this control device.

From the DE 103 46 846 A1 a data processing unit for an assistance device of a motor vehicle is known. This unit is able to merge road network data with information from environmental information sources.

From the U.S. 2002/0091473 A1 a method is known for automatically checking a driver of a motor vehicle to determine whether the driver is receptive to signals from the motor vehicle. Signals from the motor vehicle are then matched to one another as a function of this. Driver alertness is determined using a model based on sensor fusion. This model can then be used to draw conclusions about the driver's receptivity.

From the EP 1 754 621 A1 a driver assistance system is known, which monitors the surroundings of the motor vehicle and provides a description of the surroundings by means of sensor fusion and surroundings model. An interpretation of the situation in the environment can be made based on the model. The scene interpretation can influence the control of the model.

From the U.S. 2016/027276 A1 a method is known for assessing an insurance risk that can result from the behavior of a driver in a motor vehicle. A 3D model of the driver can be formed here, whose attribute values are set by sensor data.

The object of the invention is to use resources of the motor vehicle multiple times for the control of actuators in an interior of a motor vehicle.

The object is solved by the subject matter of the independent patent claims. Advantageous developments of the invention are disclosed by the dependent patent claims, the following description and the figure.

• ein Sitzmotor, ein Fensterhebermotor, ein jeweiliger Spiegelmotor für einen Rückspiegel und/oder Seitenspiegel, ein Gebläse einer Klimaanlage, eine Heizeinrichtung und/oder Kühleinrichtung einer Klimaanlage, eine Lenkradstelleinrichtung, ein Medienwiedergabegerät (z.B. ein MP3-Abspielgerät).

The invention provides a control device for generating control data for actuators for configuring an interior situation of a motor vehicle. A respective example of such an actuator is:

• a seat motor, a power window motor, a respective mirror motor for a rearview mirror and/or side mirror, a fan of an air conditioning system, a heating device and/or cooling device of an air conditioning system, a steering wheel adjustment device, a media playback device (e.g. an MP3 player).

The control device is characterized by a modular design. A hardware interface is set up to receive sensor data from sensors in the motor vehicle using a respective sensor-specific driver and to generate input data from the received sensor data and to provide the input data for retrieval by sensor-unspecific retrieval functions. Using a respective retrieval function, at least part of the input data can be queried or read out at the hardware interface without the sensors having to be controlled in order to access the sensor data on which the respective input data are based. The hardware interface takes care of this. The retrieval function thus provides a uniform protocol for retrieval of input data, which remains unchanged even when sensors are exchanged. A sensor-specific signal processing of the sensor data is carried out by the drivers, which can be adapted to the replaced sensors. The hardware interface can also have the electrical connections for the sensors.

A model plane is configured to operate multiple models, each of the models describing a respective interior-related object by attribute values. In order to adapt to the interior situation, each object is set up to set at least one of the attribute values using one of the retrieval functions based on at least part of the input data. In other words, each object can be represented by a model. The input data can be used to adapt the model to the object, with the model not requiring any adaptation to the sensor in the manner described, since the input data can be queried sensor-unspecifically or sensor-independently using the retrieval functions. For example, you are independent of a protocol that the manufacturer of the sensor specifies for receiving the sensor data. For example, a temperature value can be retrieved using a retrieval function without having to know how to read the temperature value from the connected temperature sensor.

It does not have to be possible to fully adapt every model based on the input data, i.e. it does not have to be possible to determine every attribute value based solely on the sensor data. An interpretation layer is set up to perform a plurality of interpretation functions, each interpretation function being set up in each case to determine a property of at least one of the modeled objects based on at least one set attribute value (of at least one of the models). Each interpretation function thus interprets the situation in the interior based on the models. A property of the at least one modeled object can be recognized here. If, for example, the position of the driver's upper body is sought while the driver is leaning out of the detection range of a camera (and is therefore no longer recognizable in the camera image), the object "glove compartment" can be based on the sensor data, for example Attribute value "position" changed to "open" (the glove compartment was opened), an interpretation function "driver's position" recognized that the driver must have obviously bent towards the glove compartment.

With the model level and the interpretation level, a description of the situation in the interior is available, which can be created without knowledge of the control of the sensors, i.e. sensor-agnostic, since the control of the sensors, i.e. the reading of the sensor data, is carried out by the hardware interface. Likewise, the situation in the interior is determined regardless of what the situation is to be observed for. So there is no provision for a dependency on a control function. The situation in the interior is thus also provided in a function-agnostic manner.

Any control function can now use this description of the interior. For this purpose, a function level is set up to operate control functions, with each control function being set up in each case to process some of the control data for at least one of the actuators of the motor vehicle as a function of at least one set attribute value of at least one of the models and/or as a function of at least one determined property of the at least one modeled object. The control functions therefore access the models and/or the properties determined by the interpretation level in order to control the actuators for configuring the interior situation.

Finally, a network interface is set up to send the control data to the actuators via a data network of the motor vehicle. Such a data network can, for example, be based on an Ethernet network and/or a CAN bus (CAN—Controller Area Network) and/or a LIN bus (LIN—Local Interconnect Network) and/or a FlexRay bus and/or or a MOST bus (MOST - Media Oriented System Transport).

The advantage of the invention is that a central architecture for sensing the interior function is provided, which is highly modular and capable of providing a uniform, global model of the interior in a sensor- and function-agnostic manner and making this available to a large number of comfort and safety functions to deliver.

The models do not need to be able to adjust all of their attribute values based on the input data (processed sensor data). It is therefore provided that at least one of the interpretation functions is set up to set at least one further attribute value in at least one of the models as a function of the determined property. Thus, by interpreting the situation in the interior simulated by the models, at least one additional attribute value is derived or ascertained, for which no sensor is then required to detect the respective attribute value.

The invention also includes developments whose features result in additional advantages.

The hardware interface is preferably set up to generate said input data by aggregating and/or merging and/or plausibility checking of the received sensor data. An aggregation results in the advantage that sensor data is combined, so that sensor data from different sensors and/or recorded at different times can already be combined through the hardware interface and therefore no corresponding administration effort is necessary in the models. A merger has the advantage that sensor data from different sensors are combined in order to thereby link information sources in the form of the multiple sensors, as a result of which the input data represent a more precise and/or more detailed measurement than the sensor data in each case taken individually. A plausibility check results in the advantage that measurement errors are already recognized in the hardware interface and can be compensated for or eliminated there in order to thereby avoid an influence of a measurement error on one of the models.

At least one of the following objects is represented in the model level by one of the models: a vehicle occupant, the motor vehicle, an object arranged in the interior. A model for a vehicle occupant has the advantage that the effect of the actuator control on the vehicle occupant is represented or recognizable by the occupant model. In addition, non-measurable properties of the vehicle occupant can be derived using the model. The vehicle status can be represented by the model of the motor vehicle in order to represent the boundary conditions that currently exist in the interior for the control functions, e.g. a viewing angle from which you can see an outside mirror. An object model of an object arranged in the interior has the advantage that the position of the object in the interior and/or the influence of the object on the interior situation is represented. For example, a backpack placed in front of an airbag can be modeled as an object and the airbag can be switched off by a control function depending on the size and/or the weight and/or the position of the backpack.

In order to describe an object advantageously for the control functions by means of a model, the following is particularly preferably provided. At least one of the following attributes is described by a respective attribute value of the model of the vehicle occupant: a 2D/3D avatar representation (two-dimensional or three-dimensional model of the vehicle occupant), a position, an orientation, a mass (body weight), a pulse rate, a Respiratory rate, a song beat rate, a lip movement (e.g. movement rate), a temperature, a skin moisture. A control function for seat occupancy, for example, can be operated using the attribute value of the position of the vehicle occupant. An attribute value of a heart rate can be used, for example, for an interpretation function to identify the stress level. An attribute value of a respiratory rate can be used, for example, for an interpretation function for detecting a state of health (shortness of breath). An attribute value of a mass of the vehicle occupant can be used for a control function of a belt tensioner, for example. An attribute value of a song beat frequency can be used by an interpretation function for a vigilance measurement of the vehicle occupant. An attribute value of a temperature, for example a skin temperature, can be used for a control function of an air conditioning system. An avatar representation can include several attribute values and offers the advantage that the respective spatial position of at least one limb can be recorded or determined using the modeled limbs even if these limbs are outside the detection range of the motor vehicle's sensors.

With regard to the model of the motor vehicle (vehicle status model), at least one of the following attributes is described, for example, by a respective attribute value of the model of the motor vehicle: a mirror position, a seat position, a belt status, a driving speed, a terminal status of a voltage supply, a steering angle, a door status, a lid state (e.g., a trunk lid). An attribute value of a mirror position can be used to predict a movement of a driver's head when the driver has to observe the traffic behind. An attribute value of a seat position can be used to decide which controls in the interior can be reached by hand by a given vehicle occupant and which, on the other hand, must be able to be operated, for example, by voice control or line of sight recognition. An attribute value of a driving speed can be used, for example, to predict a lateral force when cornering. A terminal status of a power supply can be used to predict an available motor force of at least one actuator. An attribute value of a steering angle has the advantage that an internal state of the driver can be inferred based on a time profile of the changing attribute value. An attribute value of a door status has the advantage that depending on an open position and/or a closed position of the door, at least one actuator can become more expensive. For example, an attribute value of a door state may be used to control a "trunk open" indication.

With regard to an object model, at least one of the following attributes is described by a respective attribute value of the model of the object: a position, a class, an orientation. An attribute value of a position can be used, for example, to control an airbag in the manner already described. An attribute value of the class can indicate whether the object is a purse or a digital camera or a smartphone, for example. An attribute becomes an orientation can be used, for example, to recognize an orientation of a detection area of a camera of a smartphone, which means that the camera can be used as a sensor external to the vehicle.

Provision is preferably made for at least one of the following properties to be determined in the interpretation level by one of the interpretation functions: a segmented 2D/3D avatar representation, an NCAP classification (NCAP—New Car Assessment Program), a hand movement, a seat occupancy, a body pose, a degree of distraction, a degree of fatigue, an emotion, a stress level, a geometric identification, a gaze direction, a head orientation, an age, a gender, a clothing class (e.g. "warmly dressed"). Each of these properties can now be determined based on the total available attribute values of the models, without the control of the sensors having to be implemented in each of the interpretation functions. Instead, the situation in the interior simulated by the models can be used as a basis for recognizing the respective property.

Provision is preferably made for a context evaluation to be carried out in the interpretation level by at least one of the interpretation functions, which uses a description of the surroundings of the motor vehicle to determine how the surroundings of the motor vehicle and/or what of the surroundings can be seen by a vehicle occupant from the interior. For example, a respective position of at least one road user in the vicinity of the vehicle can be described, so that a respective reaction and/or attitude of the driver and/or another vehicle occupant can be placed in the context of the situation (e.g. "driver looks in the side mirror to see another motor vehicle to be observed").

Said environment description can be received from an environment model, which can be formed based on environment sensors such as radar, lidar, ultrasound, to name just a few. Such an environment model can be provided for piloted driving, for example. Supplemented by navigation data and/or a digital road map, the immediate environment, i.e. other vehicles, other road users such as pedestrians or cyclists, traffic signs, traffic lights, prominent points, can be described by the environment model.

There is a direct connection to various attributes of the interior situation, such as the driver's line of sight, a brake intervention or an unexpected heart rate increase as a result.

The environment model can be located outside the concept outlined here, i.e. also outside the control device, and only needs to be queried to evaluate the situation context of the interior situation, but not filled or synchronized or updated like the own, internal model level.

At the function level, at least one of the following controls is provided by one of the control functions: gesture control, occupant notification, airbag control, driver-dependent seat and/or cockpit adjustment, lighting control, driver authorization check, climate control, 3D sound focusing, HRV biofeedback control (HRV - heart rhythm variability). The gesture control can advantageously use the avatar described, for example, so that not every limb actually has to be captured by a camera. Using a model of an object, it can even be additionally recognized in gesture control whether a vehicle occupant is actually performing an operating gesture or is thinking about the object. An occupant notification can, for example, control a display device and/or voice output device and/or a haptic actuator of the motor vehicle in order to output a message to a vehicle occupant. In the manner described, the airbag controller can, for example, deactivate an airbag if the assigned or associated vehicle seat is occupied by an object. A driver-dependent seat and/or cockpit adjustment can adjust the seat position and/or a design of graphical display elements depending on an identity of a person and/or physiological characteristics of the person, for example their height and/or eyesight. A lighting control can, for example control an ambient light (indirect lighting) in the motor vehicle and/or align a light spot for a reading vehicle occupant. Here, for example, the occupant model can be used to determine the viewing direction and/or head inclination of the vehicle occupant to recognize reading activity and a model of a book that the vehicle occupant is looking at can be used to align the light spot. Depending on an orientation of a field of vision (direction of view and/or head inclination), floor lighting for illuminating a footwell can also be activated, for example. A driver authentication check may utilize geometric features from an occupant model. An air conditioning controller can, for example, control seat air conditioning and/or zone air conditioning (different temperatures and/or ventilation intensities for different room areas). In the motor vehicle, the feedback control can, for example, adapt air conditioning and/or lighting to an internal condition of a vehicle occupant that is signaled by the heart rhythm.

In order to operate the model level, the interpretation level and the functional level, a computing device is preferably provided which is set up for this purpose. A respective model, a respective interpretation function and a respective control function can each be designed as a program module that can be executed by the computing device. In this case, the computing device is preferably additionally set up to subsequently install at least one new model and/or at least one new interpretation function and/or at least one new control function and to supply it with input data from the hardware interface and/or attribute values from other models. A new model and/or a new interpretation function and/or a new control function can thus be retrofitted in the control device. No adaptation to the existing sensors is necessary here, since a sensor-independent retrieval function can be used to retrieve the described input data from the hardware interface.

So far, only sensors have been described as external data sources. Provision is preferably made for the network interface, which leads to the data network, to be set up to receive data from the data network from at least one data source external to the vehicle, for example an Internet server and/or a mobile end device connected to the motor vehicle (for example smartphone and/or Smartwatch and/or tablet PC) and to make them available at the model level and/or at the interpretation level and/or at the functional level. In this way, for example, a database and/or a sensor of a mobile terminal device can also be used to determine or provide at least one attribute value and/or at least one property and/or to generate control data.

The invention also includes a motor vehicle with sensors and actuators that are coupled via an embodiment of the control device according to the invention. As a result, different control functions for controlling the actuators in the motor vehicle can be provided in a flexible and interchangeable and retrofittable manner on the basis of sensor data. The motor vehicle according to the invention is preferably designed as a motor vehicle, in particular as a passenger car or truck.

The operation of the motor vehicle results in an embodiment of the method according to the invention for operating a control device in a motor vehicle. The hardware interface receives sensor data from sensors in the motor vehicle using a respective sensor-specific driver and generates input data from the received sensor data and provides the input data for retrieval by sensor-unspecific retrieval functions. A model layer operates multiple models, each of the models describing a respective object related to an interior of the motor vehicle through attribute values and setting at least one of the attribute values using a respective retrieval function based on the input data. By "related" is meant that the object is in the interior space or bounds the interior space or is viewable from the interior space. An interpretation level executes at least one interpretation function, each interpretation function determining a property of at least one of the modeled objects on the basis of at least one set attribute value of at least one of the models. A function level operates at least one control function, with each control function generating control data depending on at least one set attribute value of at least one of the models and/or depending on at least one determined property. A network interface sends the control data to at least one actuator of the motor vehicle.

The invention also includes developments of the method according to the invention, which have features as have already been described in connection with the developments of the motor vehicle according to the invention. For this reason, the corresponding developments of the method according to the invention are not described again here.

Exemplary embodiments of the invention are described below. For this purpose, the single figure (FIG.) shows a schematic representation of an embodiment of the motor vehicle according to the invention.

The exemplary embodiments explained below are preferred embodiments of the invention. In the exemplary embodiments, the described components of the embodiments each represent individual features of the invention to be considered independently of one another, which also develop the invention independently of one another and are therefore also to be regarded as part of the invention individually or in a combination other than the one shown. Furthermore, the described embodiments can also be supplemented by further features of the invention that have already been described.

Elements with the same function are each provided with the same reference symbols in the figure.

The figure shows a motor vehicle 10, which can be, for example, a motor vehicle, in particular a passenger car or truck. A control device 12 can be provided in order to be able to influence the situation in a passenger compartment or interior 11 of motor vehicle 10, for example the air conditioning and/or arrangement of seats/mirrors and/or the presentation of media content. The control device 12 can be designed as a single control device or as a combination of control devices. In particular, the control device 11 can have a processor device 13, which can have one or more microprocessors and/or one or more microcontrollers. Sensors 14 of motor vehicle 10 can be connected to control device 12, for example a 3D camera and/or a 2D camera and/or at least one temperature sensor and/or at least one microphone and/or at least one to be worn on the body of the accessory for measuring body signals. In order to influence the situation in the interior 11, the control device 12 can generate control data 15 for controlling actuators 16 of the motor vehicle 10. Examples of an actuator 16 are: a seat motor and/or a mirror motor and/or a power window motor and/or an infotainment system (information and entertainment system), an air conditioning system, a display device in an instrument cluster of motor vehicle 10. Control device 12 can be connected to the actuators 16 be coupled via a data network 17. The data network 17 can include, for example, an Ethernet network 18 and/or a bus system 18', for example a CAN bus. The control device 12 can be connected to the data network 17 via a network interface 19, e.g. an NIC (Network Connect Interface). Different network technologies can be coupled to one another via a gateway 20 .

Different program modules can be executed by the processor device 13, which can be installed or implemented individually or modularly in the control device 12. It is provided here that the program modules do not have to be adapted to the existing sensors 14 . Rather, a hardware interface 21 is provided, which can receive sensor data 22 from the sensors 14 and generate input data 23 therefrom, which can be retrieved within the control device 12 using a sensor-independent retrieval routine. For controlling or sensor-specific retrieval of the sensor data 22, the hardware interface 21 can have a suitable driver 24, i.e. driver software, for each sensor type. The sensor data received by the driver 24 can be converted into the input data 23 by aggregation and/or fusion and/or plausibility checking.

An architecture consisting of three levels, namely a model level 25 , an interpretation level 26 and a function level 27 , can be provided in the computing device 13 within the control device 12 . Each level 25, 26, 27 is formed by combining similar program modules in it.

In the model level 25, models 28 can be provided as program modules. Each model 28 can represent an object in the interior 11 or on the motor vehicle 10 . There may be an occupant model, a vehicle state model, and/or an object model. Each model 28 can describe the object it describes or represents by means of attribute values 29, only some of which are provided with a reference number in the figure for the sake of clarity. Each model 28 can set at least some of the attribute values 29 by retrieving at least part of the input data 23 from the hardware interface 21 . For this purpose, the model 28 does not have to know how to control or read out the associated sensor 14 . Instead, a sensor-independent retrieval function can be used, which requests or reads out the required input data 29 at the hardware interface 21 . “Sensor-independent retrieval function” means in particular a retrieval function that can be part of an API (Application Programming Interface) of the hardware interface 21 and/or provides a predetermined, sensor-independent protocol for reading out the input data 23 . The retrieval function does not change even if at least one of the sensors 14 is replaced. For example, a retrieval function to retrieve of temperature values can be provided, through which input data with temperature values can be retrieved, regardless of which sensors 14 or sensor type the temperature values originate from.

Interpretation functions 30 can be provided in the interpretation level. An interpretation function 30 can read out at least one respective attribute value 29 from one or more models and determine a property 31 of an object of the motor vehicle 10 therefrom. This property 31 can then be described again as an attribute value or by several attribute values. These new attribute values can be stored in the appropriate model 28 again.

The model-based description of the interior 11 provided in this way can be used by different control functions 32 of the function level 27 . Based on at least one respective attribute value 29 of at least one model 28 and/or based on at least one property 31 determined by a respective interpretation function 30, each of the control functions 32 can generate control data 15 with which it can control at least one actuator 16. For example, for control functions 32, such as gesture control, airbag control, seat adaptation and/or cockpit adaptation, lighting control and/or other control functions 32 on the basis of different sensors 14, the situation in the interior 11 are described so that the control functions 32 can adapt the control data 15 to the current situation in the interior 11 . In this case, the control functions 32 can use the sensors 14 jointly via the model layer 25 and the interpretation layer 26 . A new control function can also be added, which can then use the existing sensors 14.

Through synergy in hardware and software, this allows the costs per control function 32 to be reduced and thus also the implementation of control functions 32 that would otherwise not be economical due to their own system costs, such as would be necessary through the separate provision of sensors. Furthermore, an increase in robustness is possible, since multiple sensors can be used for each control function 32 as a result of the fusion plausibility check. Depending on the application and equipment feature, the control device 12 allows the complete substitution of other control units.

By increasing the detection potential of the control functions 32 and the fusion of the sensor data 22, there is also redundancy, as is provided for automated driving from level 3 to level 5 (level - automation level) according to SAE J 3016, for example.

The realization of the control device 12 shown in the figure has an architecture with a model level 25 , an interpretation level 26 and a functional level 27 .

On the model level 25, occupants and objects are classified and the vehicle condition is taken into account in a vehicle model. The models 28 are assigned attribute values 29 by means of fusion and aggregation of the sensor parameters. Sensor-specific drivers 24 define the hardware interface 21 and the respective data types of the attribute values that are used for modelling.

At the interpretation level, the attribute values are processed into information describing the characteristics of the occupants and/or objects and/or the motor vehicle, which in turn can be written back into the models 28 or synchronized as attribute values.

Control functions 32 run at the function level, which use the sensors 14 of the models 28 independently of the sensor technology. As an output of the control functions 32, the control data 15 is routed to the individual control units of the respective actuators 16 via the data network 10, for example via the gateway 20.

Overall, the examples show how the invention can provide a fusion of different sensors in the motor vehicle to form a central situation model of the vehicle interior.

Claims (11)

Control device (12) for generating control data (15) for actuators (16) in order to configure an interior situation of a motor vehicle (10), wherein - a hardware interface (21) is set up for this purpose from sensors (14) of the motor vehicle (10), in each case Receiving sensor data (22) by means of a respective sensor-specific driver (24), generating input data (23) from the received sensor data (22) and providing the input data (23) for retrieval by sensor-unspecific retrieval functions - a model level (25) set up for this is to operate a plurality of models (28), each of the models (28) describing a respective object related to the interior through attribute values (29) and being set up to retrieve at least one of the attribute values (29) by means of one of the retrieval functions on the basis of at least part of the input data (23), - an interpretation level (26) is set up to execute a number of interpretation functions (30), each interpretation function (30) being set up in each case to specify a property (31) of at least one of the modeled objects on the basis of at least one set attribute value (29) to determine at least one of the models (28), - a function level (27) is set up to operate control functions (32), each control function (32) being set up in each case to to generate some of the control data (15) depending on at least one set attribute value (29) of at least one of the models (28) and/or depending on at least one determined property (31), and - a network interface (19) is set up for this purpose, sending the control data (15) to the actuators (16) via a data network (17) of the motor vehicle (10), characterized in that that at least one of the interpretation functions (30) is set up to set at least one further attribute value (29) in at least one of the models (28) depending on the determined property (31) and thus by interpreting the model (28) adjusted Situation in the interior at least one further attribute value (29) is derived for which no sensor is then required to detect the respective attribute value (29). Control device (12) after claim 1 , The hardware interface (21) being set up to generate the input data (23) by an aggregation and/or fusion and/or plausibility check of the received sensor data (22). Control device (12) according to one of the preceding claims, wherein in the model level (25) at least one of the following objects is represented by one of the models (28): a vehicle occupant, the motor vehicle (10), an object arranged in the interior. Control device (12) after claim 3 , At least one of the following attributes being described by a respective attribute value (29) of the model (28) of the vehicle occupant: a 2D/3D avatar representation, a position, an orientation, a mass, a pulse rate, a respiratory rate, a song beat rate , a lip movement, a temperature, a skin moisture; and/or at least one of the following attributes is described by a respective attribute value (29) of the model (28) of the motor vehicle (10): a mirror position, a seat position, a belt status, a driving speed, a terminal status of a voltage supply, a steering angle, a door status , a valve state; and/or at least one of the following attributes is described by a respective attribute value (29) of the model (28) of the object: a position, a class, an orientation. Control device (12) according to one of the preceding claims, wherein in the interpretation level (26) at least one of the following properties (31) is determined by one of the interpretation functions (30): a segmented 2D/3D avatar representation, an NCAP classification, a hand movement, a seat occupancy, a body pose, a degree of distraction, a degree of fatigue, an emotion, a stress level, a biometric identification, a gaze direction, a head orientation, an age, a gender, a clothing class. Control device (12) according to one of the preceding claims, wherein in the interpretation level (26) at least one of the interpretation functions (30) carries out a context evaluation which uses a description of the surroundings of the motor vehicle to determine how the surroundings of the motor vehicle and/or what of the surroundings is recognizable from the interior of a vehicle occupant. Control device (12) according to one of the preceding claims, wherein in the function level (27) at least one of the following controls is provided by one of the control functions (32): gesture control, occupant notification, airbag control, driver-dependent seat and/or cockpit adjustment, lighting control, driver authorization check, climate control, 3D sound focusing, heart rhythm variability biofeedback control. Control device (12) according to one of the preceding claims, wherein a computing device (13) is set up to operate the model level (25), the interpretation level (26) and the function level (27) and in this case at least one new model and/or at least one subsequently installing a new interpretation function and/or at least one new control function and supplying it with input data (23) from the hardware interface (21) and/or attribute values (29) from the models (28). Control device (12) according to one of the preceding claims, wherein the network interface (19) is set up to receive data from the data network (17) at least one vehicle-external data source and in the model level (25) and / or in the Interpreta tion level (26) and/or in the function level (27). Motor vehicle (10) with sensors (14) and actuators (16) and a control device (12) according to one of the preceding claims. Method for operating a control device (12) in a motor vehicle (10), wherein - a hardware interface (21) from sensors (14) of the motor vehicle (10) receives sensor data (22) by means of a respective sensor-specific driver (24), from the received sensor data (22) generates input data (23) and provides the input data (23) for retrieval by non-sensor-specific retrieval functions, - a model level (25) operates a number of models (28), each of the models (28) having a respective interior (11 ) of the motor vehicle (10) describes the object related to it by attribute values (29) and sets at least one of the attribute values (29) by means of a respective retrieval function on the basis of the input data (23), - an interpretation level (26) at least one interpretation function (30) executes, each interpretation function (30) each having a property (31) of at least one of the modeled objects on the basis of at least one - a function level (27) operates at least one control function (32), each control function (32) generating control data (15) depending on at least one set attribute value (29) at least generates one of the models (28) and/or as a function of at least one determined property (31), and - a network interface (19) which transmits control data (15) to at least one actuator (16) for configuring an interior situation of a motor vehicle (10). , characterized in that at least one of the interpretation functions (30) sets at least one further attribute value (29) in at least one of the models (28) as a function of the determined property (31) and thus by interpreting the situation simulated by the models (28). at least one further attribute value (29) is derived in the interior, for which there is then no sensor for detecting the respective A attribute value (29) is required.

Images