DE102018211831A1 - A method for determining a seat adjustment of a seat of a motor vehicle, computer program product, server device, communication terminal and motor vehicle - Google Patents

Abstract

The invention relates to a method for determining a seat adjustment (23) of a seat (3) of a motor vehicle (1), in which the following steps are carried out: - providing a body dimension (12) of a user (6) of the motor vehicle (1); a visual characteristic (13) of the seat (3) of the motor vehicle (1); - providing an artificial neural network (14) trained with at least one reference subject feature (20), at least one reference body dimension (21) and at least one reference seat adjustment (22); and determining the seat adjustment (23) by entering the body dimension (12) and the subject property (13) into the artificial neural network (14).

Description

The invention relates to a method for determining a seat adjustment of a seat of a motor vehicle. Furthermore, the invention relates to a method for generating an artificial neural network. The invention also relates to a computer program product and a seat adjuster. Furthermore, the invention relates to a server device, a communication terminal and a motor vehicle, each with a corresponding seat adjuster.

The optimal seating position in a vehicle is an important aspect for comfort and safety of the driver or passenger. Finding the optimum seating position is often not easy and requires some effort, especially in an unknown vehicle, such as a rental car. For motor vehicle users who are taller or smaller than the average, it is particularly difficult to find the optimum seating position. There are many seat adjustment options, such as seat height, distance to the steering wheel and seat tilt. This makes it even more difficult to find the optimum seating position.

Usually, the seat adjustment is adjusted manually by the driver or the passenger of the motor vehicle, at least during the first use of the motor vehicle. The manually adjusted seat adjustment is then usually stored in the motor vehicle, so that the driver or passenger can resort to the stored seat adjustment when reusing the motor vehicle and does not have to manually adjust the seat each time the motor vehicle begins to operate.

Usually, this is particularly relevant when the motor vehicle is used by different people and the seat adjustment is changed regularly, as is the case for example with rental vehicles.

It is an object of the invention to provide a method, a computer program product, a seat adjuster, a server device, a communication terminal and a motor vehicle, in which or with which a seat adjustment of a seat can be determined more easily.

This object is achieved by a method, a computer program product, a seat adjuster, a server device, a communication terminal and a motor vehicle according to the independent claims.

• - Bereitstellen einer Körperabmessung eines Nutzers des Kraftfahrzeugs;
• - Bereitstellen einer Sitzeigenschaft des Sitzes des Kraftfahrzeugs;
• - Bereitstellen eines mit zumindest einer Referenzeigenschaft, zumindest einer Referenzkörperabmessung und zumindest einer Referenzsitzeinstellung trainierten künstlichen neuronalen Netzes; und
• - Bestimmen der Sitzeinstellung durch Eingabe der Körperabmessung und der Sitzeigenschaft in das künstliche neuronale Netz.

In a method according to the invention, a seat adjustment of a seat of a motor vehicle is determined. The following steps are carried out:

• Providing a body dimension of a user of the motor vehicle;
• - Providing a Sitzeigenschaft the seat of the motor vehicle;
• Providing an artificial neural network trained with at least one reference characteristic, at least one reference body dimension and at least one reference seat setting; and
• Determine the seat adjustment by entering the body dimension and the sit characteristic into the artificial neural network.

The invention is based on the finding that the seat adjustment by the artificial neural network can be determined more easily.

Furthermore, the invention is based on the recognition that the body size and the seat dimension are particularly suitable for serving as input for the artificial neural network. By entering the body dimension and the sit characteristic into the artificial neural network, the seat adjustment can be determined more accurately than is the case with other input parameters.

Artificial neural networks are networks of artificial neurons. This forms an abstraction or modeling for information processing. Artificial neural networks are usually based on the networking of many McCulloch-Pitts neurons. Other artificial neurons can also be used, for example the high-order neuron. In particular, following the construction of the neural network, the training phase in which the network learns follows. During learning, new connections are developed, existing connections are deleted, weightings are changed, thresholds of the respective neurons are adapted, if these have threshold values, neurons are deleted or added and / or modification of activation functions, propagation functions or output functions.

The seating feature comprises in particular at least one seat dimension, in particular a respective dimension of seat parts, such as a dimension of a seat line or a dimension of the seat surface of the seat. However, the seating property may also include a curvature or a degree of hardness of a seat surface or a leaning surface of the seat. Also, the seating feature may include the size and / or position of a headrest of the seat. Also, for example, a position of an armrest of the seat is included in the seating property.

The artificial neural network is thus in particular trained with the reference subject property and the reference body dimension as input training data and trained with the reference seat setting as training target data. The reference seating feature and / or the reference body dimension is preferably identified by the reference seat setting or has a label. As a result, the training data is suitable for a supervised learning procedure, such as for the artificial neural network. The reference seat adjustment thus specifies, in particular, how the seat is adjusted with the reference body characteristic assigned and the reference body dimension.

It may also be provided that the seat of the motor vehicle is adjusted by the particular seat adjustment (and not only the seat adjustment is determined). When adjusting the seat, the seat adjustment is used to adjust the seat according to the output of the neural network. Thus, the determination of the seat adjustment does not necessarily mean that the seat will actually be set accordingly. Thus, it is also possible to output a seat adjustment proposal to the user before setting the seat, and only when the user agrees to the seat adjustment proposal, which is visually output via a display device in the motor vehicle, the seat of the motor vehicle is finally adjusted based on the particular seat adjustment.

Preferably, it is provided that the seat adjustment is transmitted to the motor vehicle before entering the user in the motor vehicle and the seat is adjusted by the seat adjustment before boarding the user in the vehicle. The transmission of the seat adjustment can take place, for example, wirelessly, for example, the seat adjustment can be transmitted from an external vehicle server or from a mobile communication terminal of the user to the motor vehicle before the user gets into the vehicle. It is advantageous to adjust the seat before boarding the driver, since the user can take the motor vehicle faster into operation or can drive faster and does not require time to adjust or set the seat accordingly.

Furthermore, it is preferably provided that the body dimension of the user is provided by a mobile communication terminal of the user, which can be coupled to the motor vehicle. As the mobile communication terminal, for example, an intelligent mobile phone ("Smartphone"), a tablet computer or a smart watch ("Smartwatch") are available. It is advantageous to provide the body measurement by the mobile communication terminal, since the mobile communication terminal is often carried by the user with him and the user thus always has his body dimensions with him and this does not have to capture each time new or retrieve from a remote location. The body dimensions can therefore also be provided if there is a radio hole at the location of the motor vehicle. Furthermore, with the storage of the body dimension in the mobile communication terminal, data protection specifications can also be met. Thus, the user always has his body dimensions with him or always under his own control.

Furthermore, it is preferably provided that at least one image of the user is detected by means of a camera of a mobile communication terminal and the body dimension is determined on the basis of the image. Thus, the user can determine his body size at any time by, for example, a self-photo ("selfie") or determine the body size of an already existing on the mobile communication terminal image of the user. By means of the mobile communication terminal, the user therefore has the opportunity to independently determine his own body dimensions at various times. The mobile communication terminal is then in particular designed to determine the body dimensions of the user on the basis of the at least one image. For this purpose, for example, a plurality of images of the user can be made from different positions by means of the camera integrated in the mobile communication terminal. For example, the principle of stereo perspective can be used to create a three-dimensional image of the user. The dimension of the user can then be derived from the image, for example by means of photogrammetric methods. By way of example, the length and / or width of various limbs of the user can be determined on the basis of the image. Furthermore, based on the image, a posture of the user can be determined. For example, a hollow backbone or a spinal deformity can be taken into account when determining the body dimension.

Furthermore, it is preferably provided that the body dimension of the user is transmitted to a rental car company providing the motor vehicle and the seat adjustment is generated by entering the body dimension and the viewing characteristic into the artificial neural network of the rental car company. Thus, the rental car company can determine the seat adjustment before the presence of the user and set up the vehicle accordingly. The rental car company can then, for example, but also determine whether the motor vehicle can be adjusted to the user's requirements for the seat adjustment at all. That's the way it works For example, the user may have a body dimension that deviates significantly from an average body dimension, and the vehicle of the rental car company may not be adjusted with the seat adjustment output from the neural network. The rental car company can then, for example, inform the user in advance, for example, before the arrival of the user, or to provide another rental vehicle. The user can then, for example, also turn to another rental car company to book there another rental car.

Furthermore, it is preferably provided that a body height of the user and / or a leg length of the user and / or an upper body length of the user and / or an arm length of the user is provided as the body dimension. The various dimensions of limbs or sections of the user and the resulting detailed body size, the seat adjustment can be determined more precisely.

Furthermore, it is preferably provided that the user is given a possibility to manually correct the seat adjustment predetermined by the artificial neural network. This manual correction can then in turn be incorporated into the artificial neural network, so that the artificial neural network in the future, the manual change or correction of the user takes into account the seat adjustment.

It is preferably also provided that an additional setting criterion, in particular a comfort criterion and / or a safety criterion, is specified for the seat adjustment. For example, if the user puts more emphasis on comfort, the seat can be made more comfortable with the seat adjustment, but not with the utmost security, for example, or the seat can be adjusted with utmost safety, perhaps with sacrificing comfort.

• - Bereitstellen zumindest einer Referenzsitzeigenschaft;
• - Bereitstellen zumindest einer Referenzkörperabmessung;
• - Bereitstellen zumindest einer Referenzsitzeinstellung; und
• - Erzeugen des künstlich neuronalen Netzes mit der Referenzsitzeigenschaft und der Referenzkörperabmessung als Referenzeingangsgrößen und der Referenzsitzeinstellung als Referenzausgangsgröße durch Lösen eines Optimierungssystem von den Referenzeingangsgrößen und der Referenzausgangsgröße.

The invention also relates to a method for generating an artificial neural network. In the method according to the invention the following steps are carried out:

• - providing at least one reference title property;
• Providing at least one reference body dimension;
• - providing at least one reference seat setting; and
• Generating the artificial neural network having the reference subject property and the reference body dimension as reference input quantities and the reference seat adjustment as a reference output quantity by solving an optimization system from the reference input quantities and the reference output variable.

The invention is based on the knowledge that the seat adjustment can be more accurately determined by the use of the artificial neural network.

Furthermore, the artificial neural network can continue to be trained even after the initial training or can be adapted with further reference data, in particular further reference reference properties and / or further reference body dimensions.

Furthermore, the invention relates to a computer program product. The computer program product according to the invention has at least one computer-readable storage medium. A computer-executable program code instruction is stored on the computer-readable storage medium, which is designed to carry out the method according to the invention for determining the seat adjustment and / or the method according to the invention for generating the artificial neural network.

The computer-readable storage medium can be designed, for example, as a flash memory, for example of a mobile communication terminal.

In particular, the computer program product is designed as an application program ("app") for a mobile operating system.

The invention also relates to a seat adjuster. The seat adjustment device according to the invention has a reception interface which is set up to receive a body dimension of a user of the motor vehicle, a seat characteristic of the seat of the motor vehicle and an artificial neural network which is trained with at least one reference title feature, at least one reference body dimension and at least one reference seat setting. Furthermore, the seat adjustment device has an evaluation unit which is set up to determine a seat adjustment of the seat by inputting the body dimension and the subject property into the artificial neural network.

The receiving interface can be designed, for example, as a programming interface (API - Application Programming Interface). The seat adjuster may receive the body size, the subject property, and the artificial neural network, for example, wirelessly and / or by wire.

It is preferably provided that the seat adjustment device is at least data-transmitting connected to a seat drive device of a seat of the motor vehicle, wherein the seat drive means generates a control current for operating at least one actuator, which is designed to adjust a seat of the motor vehicle. By the Sitzinstelleinrichtung the Sitzansteuerungseinrichtung is notified of the seat adjustment. The seat adjustment is then translated by the seat driver into the at least one control stream. By the control current can then be operated at least one actuator of the seat. By the actuator then, for example, a backrest of the seat and / or a seat of the seat and / or a lumbar support of the seat can be adjusted. The seat adjusting device and the seat driving device may be formed in a common housing or separately from each other. Thus, the seat adjuster can also be arranged, for example, outside the seat at a diverse position in the motor vehicle or even outside the vehicle. The seat adjuster can also be integrated in the seat.

Furthermore, the invention also relates to a server device. The server device according to the invention has the seat adjustment device according to the invention. Furthermore, the server device is formed outside the vehicle. For example, the server device can be designed as a centralized or decentralized or distributed server. This is advantageous because the server device preferably provides the seat adjustment device around the clock. Thus, a third party, for example, with the consent of the user, use the seat adjuster to determine the seat adjustment for the user in advance of a use of a motor vehicle.

The invention also relates to a communication terminal. The communication terminal according to the invention has the seat adjustment device according to the invention. The communication terminal is formed outside the vehicle and mobile. The communication terminal or the mobile communication terminal can be present, for example, as a smartphone, as an intelligent watch, as a fitness bracelet or as a tablet computer. By the communication terminal with the seat adjustment device, the user can determine the seat adjustment himself at various times without having to rely on a server device, for example. The user can carry the communication terminal with him at any time and for example also transport from a motor vehicle to another motor vehicle.

The invention also relates to a motor vehicle. The motor vehicle according to the invention comprises the seat adjustment device according to the invention, a seat drive device and an actuator. The seat drive device is designed to generate a control current for operating the actuator. The actuator is configured to set a seat of the motor vehicle. The arrangement of the seat adjusting device in the motor vehicle is advantageous, since the seat adjusting device can then preferably resort to free computing capacity of the motor vehicle in order to determine the seat setting on the basis of the neural network.

Advantageous embodiments of the inventive method are to be regarded as advantageous embodiments of the computer program product, the seat adjuster, the server device, the communication terminal and the motor vehicle. The subject components of the seat adjuster, the server device, the communication terminal and the motor vehicle are each adapted to perform the respective process steps.

Further features of the invention will become apparent from the claims, the figures and the description of the figures.

Embodiments of the invention are explained in more detail below with reference to schematic drawings.

• 1 eine schematische Seitenansicht eines Ausführungsbeispiels eines Kraftfahrzeugs und einer erfindungsgemäßen Sitzeinstelleinrichtung;
• 2 eine schematische Darstellung eines Ausführungsbeispiels eines erfindungsgemäßen Verfahrens zum Bestimmen einer Sitzeinstellung mittels eines künstlichen neuronalen Netzes; und
• 3 eine schematische Darstellung des künstlichen neuronalen Netzes zum Bestimmen der Sitzeinstellung.

Show it:

• 1 a schematic side view of an embodiment of a motor vehicle and a seat adjustment device according to the invention;
• 2 a schematic representation of an embodiment of a method according to the invention for determining a seat adjustment by means of an artificial neural network; and
• 3 a schematic representation of the artificial neural network for determining the seat adjustment.

In the figures, identical or functionally identical elements are provided with the same reference numerals.

1 shows a motor vehicle 1 with a seat drive device 2 , The seat drive device 2 is in a seat according to the embodiment 3 of the motor vehicle 1 arranged. The seat drive device 2 generates a control current 4 with which an actor 5 of the seat 3 is operated. By the actor 5 becomes the seat 3 then, for example, adjusted with respect to a seat back inclination, a seat surface length, a headrest height, a side support and / or a lumbar support position.

The seat 3 is formed according to the embodiment as a driver's seat but can also be designed as a passenger seat or as a backseat or backseat.

Further shows 1 a user 6 with a communication terminal 7 , According to the embodiment, the communication terminal 7 a seat adjuster 8th on. The communication terminal 7 can be configured for example as a smartphone. The seat adjuster 8th is according to the embodiment data transmitting with the seat drive device 2 connected. The data transmitting connection between the seat adjuster 8th and the seat drive device 2 can be configured for example as a radio link, in particular Bluetooth connection. The data connection between the seat adjuster 8th and the seat drive device 2 can be either directly or indirectly via a server device 9 available.

According to a further, not shown embodiment, the seat adjuster 8th also in the server setup 9 be arranged.

The seat adjuster 8th has a receiving interface according to the embodiment 10 and an evaluation unit 11 on.

The receiving interface 10 is set up, a body dimension 12 of the user 6 , a site property 13 of the seat 3 and an artificial neural network 14 to recieve.

As the body dimension 12 For example, is a body height 15 , one leg length 16 , a torso length 17 and / or an arm's length 18 of the user 6 in front.

As the subject property 13 For example, the dimension of individual seat components, for example, a length and / or width of a seat surface of the seat 3 and / or a length and / or a width and / or a curvature of a seat back of the seat 3 and / or a length and / or a height of an armrest and / or a headrest of the seat 3 ,

As the subject property 13 but can also complement or alternatively a massage function of the seat 3 available. By the massage function can then, for example, at least a portion of the user 6 be massaged or moved, for example, by adaptive air cushion or acted upon with force.

The communication terminal 7 has a camera according to the embodiment 19 on. With the camera 19 becomes an image of the user 6 detected. The image then becomes the body dimension 12 certainly. To determine the body dimension 12 For example, multiple images can also be taken, or the camera can be designed as a so-called TOF camera (TOF - time of flight) and provide an image with depth information. On the basis of several pictures the body dimension can 12 then determined, for example, on the principle of the stereo perspective.

2 shows how the artificial neural network 14 is generated or trained. Preferably, the artificial neural network becomes 14 modeled before training, ie it is for example determined how many hidden layers the artificial neural network 14 having.

In one step S1 becomes a reference subject property 20 provided. Preferably, a plurality of reference keyword properties are provided. In one step S2 becomes a reference body dimension 21 provided. Preferably, a plurality of reference body dimensions are provided. In one step S3 becomes a reference seat adjustment 22 provided. Preferably, a plurality of reference seat settings are provided. The reference subject property 20 , the reference body dimension 21 and the reference seat adjustment 22 are available as so-called training data. In particular, the relationship between the reference subject property 20 , the reference body dimension 21 and the reference seat adjustment 22 known. For example, what reference seat adjustment is known 22 is present when the reference subject property 20 exists in a certain way and also the corresponding reference body dimension 21 exists in a certain way. Thus, for example, a reference user is sitting with his associated reference body dimension 21 in a reference seat with the reference subject property 20 and is this reference seat with the reference seat adjustment 22 set, this information is used as training information for the artificial neural network 14 used.

As training data, a large number of these reference data is used. It is assumed that if this reference seat adjustment, for example, has been performed manually in a plurality of cases such that the reference user has subsequently sat comfortably or as desired in his reference seat, then this reference setting may 22 serve as a template to the artificial neural network 14 to generate or train. In the future, with the trained neural network 14 also unknown users with unknown body dimensions and / or unknown seats with unknown sit properties as input parameters in the trained neural network 14 be entered to a seat adjustment 23 to determine.

In one step S4 becomes an optimization problem of the reference input quantities, ie according to the embodiment of the reference body dimension 21 and the reference seat adjustment 22 , and the reference output, that is, according to the embodiment of the reference seat adjustment 22 , solved.

Solving the optimization problem will be done in one step S5 the artificial neural network 14 in the seat adjustment 23 provided.

3 shows the artificial neural network 14 with input variables, such as the body dimension 12 and the subject property 13 and an output variable, namely the seat adjustment 23 , According to the embodiment, the artificial neural network 14 at least one hidden layer 24 on. The artificial neural network 14 can also have several hidden layers 24 respectively.

A method for determining seat adjustment 23 then, for example, runs as follows. It becomes the body dimension 12 of the user 6 provided. This can be done, for example, in which a picture with the camera 19 from the user 6 is recorded.

The body size 12 but can already be present in a database. So the user can 6 the body dimension 12 for example, enter or enter yourself.

It becomes the site property 13 of the seat 3 provided. This can for example be done by a seat type of the seat 3 or a type of motor vehicle 1 with the seat 3 is determined or known.

Furthermore, the artificial neural network 14 provided. The artificial neural network 14 can for example on the communication terminal 7 be stored or in the vehicle 1 or in the server device 9 ,

By entering the body dimension 12 and the subject property 13 into the artificial neural network 14 eventually becomes the subject property 23 certainly.

Through the seat adjustment 23 Finally, the seat driver 2 be instructed. The seat drive device 2 generated by the seat adjustment 23 the control current 4 with which the actor 5 is controlled to the seat 3 adjust. The seat adjustment 23 So is in particular a parameter set, which is the setting of the seating position of the seat 3 describes.

Preferably, it is provided that the seat adjustment 23 even before the user has boarded 6 in the motor vehicle 1 is transmitted to the motor vehicle and the seat 3 thereby before the entrance of the user 6 in the motor vehicle 1 is set.

According to an embodiment not further described, it may also be provided that the body dimension 12 is transmitted to a rental car company. The rental company can seat 3 of the motor vehicle 1 then early through the artificial neural network 14 for the user 6 to adjust.

The server device 9 is according to the embodiment outside of the motor vehicle 1 arranged and can, for example, by a mobile data transmission method with the motor vehicle 1 and / or with the communication terminal 7 be connected. The seat adjuster may additionally or alternatively be in the server device 9 be arranged. Through the server device 9 can the seat adjuster 9 For example, more computing power and / or the ability to train at night be provided.

The method can be executed, for example, as an app or as a computer program product on the communication terminal. In this case, it may be provided, for example, that the user 6 create a profile that gets some characteristic information about him, such as body height 15 , Leg length 16 , Upper body length 17 , Arm length 18 , In an extended version, the app can determine this version of a picture or photo independently.

With this information, the method determines the safest and / or most comfortable seat adjustment 23 , In vehicles with automatic or motorized seat control and Internet connection, the seat can already be adjusted before the user 6 the vehicle enters.

Furthermore, it is preferably provided that the user 6 a possibility is given by the artificial neural network 14 predetermined seat adjustment 23 manually correct or readjust and this manual correction in turn into the artificial neural network 14 so that the artificial neural network 14 in the future, the manual change or correction of the user 6 at the seat adjustment 23 considered.

It is also preferably provided that an additional setting criterion, for example a comfort criterion and / or a safety criterion by the user 6 can be specified. Sets the user 6 For example, more emphasis on comfort, so can the seat 3 with the seat adjustment 23 comfortable, for example, not be set with the highest security or the seat 3 For example, it can be adjusted with the utmost security, perhaps with a compromise in comfort.

LIST OF REFERENCE NUMBERS

1

motor vehicle

2

Seat driving device

3

Seat

4

control current

5

actuator

6

user

7

communication device

8th

seat adjuster

9

server facility

10

Reception interface

11

evaluation

12

body dimensions

13

seat property

14

artificial neural network

15

body height

16

leg length

17

Waist length

18

arm's length

19

camera

20

Reference headquarters property

21

Reference body size

22

Reference seat adjustment

23

seat adjustment

24

hidden layer

Claims (14)

Method for determining a seat adjustment (23) of a seat (3) of a motor vehicle (1), in which the following steps are carried out: - Providing a body size (12) of a user (6) of the motor vehicle (1); - Providing a Sitzeigenschaft (13) of the seat (3) of the motor vehicle (1); - providing an artificial neural network (14) trained with at least one reference subject feature (20), at least one reference body dimension (21) and at least one reference seating setting (22); and - determining the seat adjustment (23) by entering the body dimension (12) and the subject property (13) into the artificial neural network (14). Method according to Claim 1 , characterized in that the seat (3) of the motor vehicle (1) by the specific seat adjustment (23) is adjusted. Method according to Claim 1 or 2 Characterized in that the seat adjustment (23) is received in the motor vehicle (1) to the motor vehicle (1) prior to boarding of the user (6) and the seat (3) through the seat adjustment (23) prior to boarding of the user ( 6) in the motor vehicle (1) is set. Method according to one of the preceding claims, characterized in that the body dimension (12) of the user (6) by a mobile communication terminal (7) of the user (6), which is coupled to the motor vehicle (1) is provided. Method according to one of the preceding claims, characterized in that at least one image of the user (6) by means of a camera (19) of a mobile communication terminal (7) is detected and the body dimension (12) is determined based on the image. Method according to one of the preceding claims, characterized in that the body dimension (12) of the user (6) is transmitted to a rental car company providing the motor vehicle (1) and the seat adjustment (23) by entering the body dimension (12) and the seat characteristic (13 ) is generated in the artificial neural network (14) of the rental company. Method according to one of the preceding claims, characterized in that a body height (15) of the user (6) and / or a leg length (16) of the user (6) and / or an upper body length (17) of the user (6) and / or an arm length (18) of the user (6) is provided as the body dimension (12). Method for generating an artificial neural network (14), in which the following steps are carried out: - providing at least one reference title property (20); - providing at least one reference body dimension (21); - providing at least one reference seat adjustment (22); and Generating the artificial neural network (14) having the reference subject property (20) and the reference body dimension (21) as reference input quantities and the reference seat adjustment (22) as a reference output by solving an optimization problem of the reference inputs and the reference output. A computer program product comprising at least one computer-readable storage medium having computer-executable program code instructions stored thereon, the computer-executable program code instructions adapted to perform a method according to any one of Claims 1 to 7 or 8th perform. Seat adjuster (8) with: - A receiving interface (10) which is arranged, a body dimension (12) of a user (6) of the motor vehicle (1), a Sitzeigenschaft (13) of the seat (3) of the motor vehicle (1) and with at least one Referenzseitzeigenschaft (20 ) to receive at least one reference body dimension (21) and at least one reference seat setting (22) trained artificial neural network (14); and - An evaluation unit (11) which is adapted to determine a seat adjustment (23) of the seat (3) by entering the body dimension (12) and the Sitzeigenschaft (13) in the artificial neural network (14). Seat adjustment (8) after Claim 10 Characterized in that the seat adjuster (8) with a seat adjustment device (2) of a seat (3) of the motor vehicle (1) is connected at least data-transmitting, wherein the seat drive means (2) a control current (4) for operating at least one actuator (5) , which is designed to set a seat (3) of the motor vehicle (1) generated. Server device (9) with a seat adjuster (8) according to Claim 10 or 11 , wherein the server device (9) is formed outside the vehicle. Communication terminal (7) with a seat adjuster (8) according to Claim 10 or 11 , wherein the communication terminal (7) is formed outside the vehicle and mobile. Motor vehicle (1) with a seat adjuster (8) according to Claim 10 or 11 , with a seat drive device (2) which generates a control current (4) for operating at least one actuator (5) which is designed to set a seat (3) of the motor vehicle (1).

Images