DE102022213126A1 - Method, computer program and device for influencing a state of a user of a machine - Google Patents

Abstract

The present invention relates to a method, a computer program with instructions and a device for influencing a state of a user of a machine. The invention also relates to a machine that uses such a device or such a method. In a first step, states and behavior of the user can be recorded as user-related data (S1). In addition, states and an operating environment of the machine can be recorded as machine-related data (S2). An actual state of the user is then determined on the basis of the user-related data (S3). A target state of the user is also determined on the basis of the machine-related data (S4). Based on the actual state and the target state, a user-specific interaction strategy is generated (S5). Finally, at least one interaction system of the machine is controlled according to the user-specific interaction strategy (S6).

Description

The present invention relates to a method, a computer program with instructions and a device for influencing a state of a user of a machine. The invention also relates to a machine that uses such a device or such a method.

• Level 0: „Nur Fahrer“, der Fahrer fährt selbst, lenkt, beschleunigt, bremst usw.
• Level 1: Bestimmte Assistenzsysteme helfen bei der Fahrzeugbedienung.
• Level 2: Teilautomatisierung. Dabei werden automatisches Einparken, Spurführung, allgemeine Längsführung, Beschleunigung, Verzögerung usw. von den Assistenzsystemen übernommen, einschließlich der Kollisionsvermeidung.
• Level 3: Bedingte Automatisierung. Der Fahrer muss das System nicht ständig überwachen. Das Fahrzeug führt selbständig Funktionen wie das Auslösen des Blinkers, den Spurwechsel und die Spurführung aus. Der Fahrer kann sich anderen Dingen zuwenden, muss aber auf Anforderung innerhalb einer Vorwarnzeit die Kontrolle übernehmen.
• Level 4: Vollständige Automatisierung. Die Führung des Fahrzeugs wird permanent vom System übernommen. Ist das System nicht mehr in der Lage, die Aufgaben zu bewältigen, kann der Fahrer aufgefordert werden, die Steuerung zu übernehmen.
• Level 5: Kein Fahrer erforderlich. Abgesehen vom Setzen des Ziels und Starten des Systems ist kein menschliches Eingreifen erforderlich.

Automated driving, also known as automatic driving or controlled driving, is the movement of vehicles, mobile robots and driverless transport systems that are largely autonomous. There are different degrees of automated driving. In Europe, various transport ministries, e.g. in Germany the Federal Highway Research Institute, have defined the following levels of automation:

• Level 0: “Driver only”, the driver drives, steers, accelerates, brakes, etc.
• Level 1: Certain assistance systems help with vehicle operation.
• Level 2: Partial automation. Automatic parking, lane guidance, general longitudinal guidance, acceleration, deceleration, etc. are taken over by the assistance systems, including collision avoidance.
• Level 3: Conditional automation. The driver does not need to constantly monitor the system. The vehicle performs functions such as activating the indicator, changing lanes and lane guidance independently. The driver can attend to other things, but must take control on request within a pre-warning period.
• Level 4: Complete automation. The system is in permanent control of the vehicle. If the system is no longer able to handle the tasks, the driver can be asked to take over control.
• Level 5: No driver required. No human intervention is required other than setting the destination and starting the system.

In vehicles with partially or fully automated functionality, the rights and obligations of the driver at different levels of automation are often non-transparent and difficult to understand. This applies in particular to the automation levels "partial automation" (L2) and "conditional automation" (L3) as well as to frequent switching between different automation levels and different driving contexts, such as when driving through towns while driving on country roads. Here, the driver must continuously monitor the vehicle's behavior and be able to take over control of the vehicle as quickly as possible in safety-critical situations.

This conflicts with the human ability to maintain monotonous tasks over a longer period of time. At the same time, the driver bears a great deal of responsibility, as intervention is necessary, particularly in complex situations in which the automation system can no longer control the vehicle independently.

The problem is that active intervention by the driver may be necessary during long and monotonous phases of a partially automated journey, and there is only a very limited preparation time available for this. In addition, longer, incident-free use of a partially automated vehicle can lead to subjective trust in the system rising to unjustified levels and the assessment of the need for monitoring by the driver not corresponding to the actual need. This phenomenon is known as "over-trust".

It therefore makes sense to continuously adjust the driver's behavior and attention to the requirements of the current level of automation and the respective driving situation and to support the driver even in the supposedly monotonous phases in order to counteract the problem mentioned.

This problem is less acute in higher levels of automation (L4/L5), as the preparation times for a transition to lower levels of automation or manual driving are significantly longer, or in the case of L5, are eliminated altogether. The vehicle is able to master even complex situations without the help of the driver or user. The aim here is more to strengthen trust in such a system, especially among inexperienced users of an automated vehicle, and to make the rare transitions to lower levels of automation or manual driving as pleasant as possible.

In the area of vehicle occupant monitoring, there are different approaches to detecting the current state of an occupant, e.g. a driver, and to ensure that he or she behaves appropriately. Different characteristics are considered. For example, it can be determined whether a driver has his or her hand on the steering wheel or not. The driver's behavior can also be monitored, for example by looking at the direction of gaze, head position and head movement. In addition, the emotional or affective state can be determined by analyzing facial expressions and language, for example. Physiological parameters such as heart rate, heart rate variability or breathing rate can be used as additional characteristics to obtain a comprehensive picture of the driver's state.

Against this background, EP 1 914 106 A2 an attention management system. The driver of a motor vehicle is continuously monitored and his current level of attention is recorded. If the attention for the driving task is too low, measures are taken to redirect attention back to the driving task. Alternatively, assistance functions can also be switched off.

EP 3 730 331 A1 describes a method for ensuring driver attention during autonomous driving. The method comprises determining a state of a driver of the vehicle by means of a driver monitoring system, wherein the state of the driver comprises at least one attention parameter, and comparing the determined state of the driver with a predefined attention model. The method further comprises controlling the driver assistance feature based on the comparison between the determined state of the driver and the predefined attention model.

WO 2016/169585 A1 describes a device for controlling a change in the situational awareness of a user of a vehicle. An input module receives a user detection signal comprising information that can be used to determine a situational awareness level of the user. A processing module determines a currently required situational awareness level. To do so, it selects a predefined situational awareness level from a plurality of predefined situational awareness levels based on driving information. The processing module also determines a current situational awareness level of the user based on the information contained in the user detection signal. Finally, the processing module generates a control signal for a module for changing the situational awareness based on the current and the currently required situational awareness level.

A similar problem to that encountered when operating a motor vehicle also occurs when driving aircraft or rail vehicles or when operating machinery.

Current systems lack suitable interaction strategies for effective interaction between machine and user, which in combination can ensure user behavior that conforms to the level of automation.

It is an object of the present invention to provide improved solutions for influencing a state of a user of a machine.

This object is achieved by a method having the features of claim 1, by a computer program with instructions having the features of claim 8, by a device having the features of claim 9 and by a machine having the features of claim 10. Preferred embodiments of the invention are the subject of the dependent claims.

• - Bestimmen eines Istzustands des Nutzers;
• - Bestimmen eines Sollzustands des Nutzers;
• - Generieren einer nutzerspezifischen Interaktionsstrategie; und
• - Ansteuern zumindest eines Interaktionssystems der Maschine entsprechend der nutzerspezifischen Interaktionsstrategie.

According to a first aspect of the invention, a method for influencing a state of a user of a machine comprises the steps:

• - Determining the current status of the user;
• - Determining a target state of the user;
• - Generating a user-specific interaction strategy; and
• - Controlling at least one interaction system of the machine according to the user-specific interaction strategy.

• - Bestimmen eines Istzustands des Nutzers;
• - Bestimmen eines Sollzustands des Nutzers;
• - Generieren einer nutzerspezifischen Interaktionsstrategie; und
• - Ansteuern zumindest eines Interaktionssystems der Maschine entsprechend der nutzerspezifischen Interaktionsstrategie.

According to a further aspect of the invention, a computer program comprises instructions which, when executed by a computer, cause the computer to perform the following steps to influence a state of a user of a machine:

• - Determining the current status of the user;
• - Determining a target state of the user;
• - Generating a user-specific interaction strategy; and
• - Controlling at least one interaction system of the machine according to the user-specific interaction strategy.

The term computer is to be understood broadly. In particular, it also includes control units, embedded systems and other processor-based data processing devices.

The computer program can, for example, be made available for electronic retrieval or stored on a computer-readable storage medium.

• - ein Zustandsmodul zum Bestimmen eines Istzustands des Nutzers und zum Bestimmen eines Sollzustands des Nutzers;
• - ein Strategiemodul zum Generieren einer nutzerspezifischen Interaktionsstrategie; und
• - ein Steuermodul zum Ansteuern zumindest eines Interaktionssystems der Maschine entsprechend der nutzerspezifischen Interaktionsstrategie.

According to a further aspect of the invention, a device for influencing a state of a user of a machine comprises:

• - a state module for determining an actual state of the user and for determining a target state of the user;
• - a strategy module for generating a user-specific interaction strategy; and
• - a control module for controlling at least one interaction system of the machine according to the user-specific interaction strategy.

In the solution according to the invention, a personalized, i.e. user-specific interaction takes place between the machine and the user in order to bring the user into a desired target state, if this is necessary. On the basis of the user's current state and a desired target state, a personalized interaction strategy is determined to optimally support the user in the given context. The use of a personalized interaction strategy makes it possible to take into account the interpersonal variance of human behavior, abilities, perception, etc. To implement the interaction strategy, a number of interaction systems with different modalities and parameterization options are preferably available, by means of which interactions optimized for the individual user are generated. The personalized interaction between machine and user increases the security of automated systems by increasing system transparency and effectively communicating the requirements placed on the user. In this way, a reduction in misuse and an increase in acceptance can be achieved.

According to one aspect of the invention, the user's states and behavior are recorded as a basis for determining the user's current state. Based on data from a corresponding sensor system or by fusing different sensors, an actual state can be determined using algorithms for state recognition, for example. Examples of states and behavior are tiredness, attention, actions such as making phone calls, stress, workload, body posture such as hands on the steering wheel, or situational awareness. A comprehensive view of the actual state is thus taken by fusing different sensors and different partial aspects of the user's state.

According to one aspect of the invention, states and an operating environment of the machine are recorded as a basis for determining the target state of the user. By recognizing the operating environment of the machine, a basis is created for evaluating the current actual state of the user in relation to the current state of the machine and the current operating context. For example, in a vehicle, the current automation level can be determined based on the current environment, e.g. the traffic density or the type of road, and the machine state, e.g. the currently active automated driving functions, and the target state of the user can be derived from this.

According to one aspect of the invention, the generation of the user-specific interaction strategy is based on an interaction model. Such interaction models can in particular comprise a number of target states for the various automation levels, a number of reference states and transition rules for transitions between the states. The transition rules define which interaction strategy can be used to achieve a desired target state or a reference state. The use of interaction models has the advantage that possible interaction strategies can also be derived in a simple manner for new states not contained in the interaction model, e.g. by calculating the distance to the nearby reference states.

According to one aspect of the invention, a prediction is made on the basis of the interaction model as to when a user leaves a target state. The interaction models can be implemented in the form of Markov models, for example. With such a model, a prediction can be made on the basis of probabilities as to when the user no longer behaves in accordance with the current automation level or leaves a target state. This prediction makes it possible to take preventive countermeasures using suitable interaction strategies.

According to one aspect of the invention, the user-specific interaction strategy is generated by a trained artificial intelligence (AI) system. The generation and monitoring of a user-specific interaction strategy is preferably realized as an artificial intelligence-based selection of parameterized and combined elements from the available interaction strategies. This creates optimized and, above all, personalized human-machine interaction models. The AI system takes into account the user's reaction to past interactions in order to adapt future interactions to the individual needs and preferences of the respective user. For this purpose, a reinforcement learning method is advantageously used, which takes the user's reaction and the change in his behavior to previously applied interactions are analyzed in terms of conformity to the respective automation level, compared with the desired target behavior and, on the basis of this evaluation, iteratively approaches an optimal interaction strategy. A reward value used in the reinforcement learning process can also take into account the length of time that passes until the user reacts to an action or until he reaches the desired target state by applying an interaction strategy.

According to one aspect of the invention, the interaction system uses at least one human-machine interface or at least one machine function. Interaction with the user can take place through conventional human-machine interfaces, e.g. through visual outputs on a display or using light elements, acoustic outputs via a loudspeaker or haptic outputs using haptic actuators. All of these human-machine interfaces are well suited to influencing the state of a user in a desired manner. In addition, machine functions can also be used to carry out actions. For example, slight steering movements or braking processes in a vehicle can be used to bring the user into a target state.

Preferably, a solution according to the invention is used in a machine, e.g. in a means of transport or an industrial machine. The means of transport can be, for example, a motor vehicle, an aircraft, a rail vehicle or a watercraft.

Further features of the present invention will become apparent from the following description and the appended claims taken in conjunction with the figures.

Figure overview

• 1 shows schematically a method for influencing a state of a user of a machine;
• 2 shows schematically a first embodiment of a device for influencing a state of a user of a machine;
• 3 shows schematically a second embodiment of a device for influencing a state of a user of a machine;
• 4 shows schematically a machine using a solution according to the invention;
• 5 shows schematically further details of the generation of interaction strategies;
• 6 shows a schematic diagram of a reinforcement learning procedure;
• 7 shows schematically the contents of a rule base;
• 8th shows schematically the learning of a new state;
• 9 shows a schematic model with the learned new state; and
• 10 shows a schematic model with several target states.

Character description

In order to better understand the principles of the present invention, embodiments of the invention are explained in more detail below with reference to the figures. The same reference numerals are used in the figures for the same or equivalent elements and are not necessarily described again for each figure. It is understood that the invention is not limited to the embodiments shown and that the described features can also be combined or modified without departing from the scope of the invention as defined in the appended claims.

1 shows a schematic of a method for influencing a state of a user of a machine. In a first step, states and behavior of the user can be recorded as user-related data S1. In addition, states and an operating environment of the machine can be recorded as machine-related data S2. On the basis of the user-related data, an actual state of the user is then determined S3. On the basis of the machine-related data, a target state of the user is also determined S4. Based on the actual state and the target state, a user-specific interaction strategy is generated S5, e.g. by a trained artificial intelligence system. The generation S5 of the user-specific interaction strategy can, for example, be based on an interaction model. Finally, at least one interaction system of the machine is controlled S6 in accordance with the user-specific interaction strategy. The interaction system can, for example, use at least one human-machine interface or at least one machine function. If an interaction model is used, a prediction can also be made on the basis of the interaction model as to when a user will leave a target state.

2 shows a simplified schematic representation of a first embodiment of a device 10 for influencing a state of a User of a machine. The device 10 has an input 11 via which user-related data ND, e.g. data from an interior sensor system 33, and machine-related data MD, e.g. machine status data MZD from a status sensor system 32 or environmental data UD from an environmental sensor system 34, can be received. A status module 12 is set up to determine an actual status Z _I of a user on the basis of the received user-related data ND. The status module 12 is also set up to determine a target status Z _S of the user on the basis of the received machine-related data MD. A strategy module 13 is set up to generate a user-specific interaction strategy IS based on the actual status and the target status, e.g. using a trained artificial intelligence system AI. The user-specific interaction strategy can be generated, for example, on the basis of an interaction model. A control module 14 is set up to control at least one interaction system 31 ₁ -31 _n of the machine in accordance with the user-specific interaction strategy IS. For this purpose, the control module 13 can output corresponding control commands SB to the at least one interaction system 31 ₁ -31 _n via an output 17 of the device 10. The interaction system 31 ₁ -31 _n can, for example, use at least one human-machine interface or at least one machine function. If an interaction model is used, a prediction can also be made on the basis of the interaction model as to when a user leaves a target state.

The state module 12, the strategy module 13 and the control module 14 can be controlled by a control module 15. Settings of the state module 12, the strategy module 13, the control module 14 or the control module 15 can be changed via a user interface 18. The data generated in the device 10 can be stored in a memory 16 of the device 10 if necessary, for example for later evaluation or for use by the components of the device 10. The state module 12, the strategy module 13, the control module 14 and the control module 15 can be implemented as dedicated hardware, for example as integrated circuits. Of course, they can also be partially or completely combined or implemented as software that runs on a suitable processor, for example on a GPU or a CPU. The input 11 and the output 17 can be implemented as separate interfaces or as a combined interface.

3 shows a simplified schematic representation of a second embodiment of a device 20 for influencing a state of a user of a machine. The device 20 has a processor 22 and a memory 21. For example, the device 20 is a control unit or an embedded system. Instructions are stored in the memory 21 which, when executed by the processor 22, cause the device 20 to carry out the steps according to one of the described methods. The instructions stored in the memory 21 thus embody a program which can be executed by the processor 22 and which implements the method according to the invention. The device 20 has an input 23 for receiving information. Data generated by the processor 22 are made available via an output 24. In addition, they can be stored in the memory 21. The input 23 and the output 24 can be combined to form a bidirectional interface.

The processor 22 may include one or more processor units, such as microprocessors, digital signal processors, or combinations thereof.

The memories 16, 21 of the described devices can have both volatile and non-volatile memory areas and can comprise a wide variety of storage devices and storage media, for example hard disks, optical storage media or semiconductor memories.

4 shows a schematic of a machine 30 in which a solution according to the invention is implemented. In this example, the machine 30 is a motor vehicle. A device 10 according to the invention for influencing a state of a user N ₁ , N ₂ , e.g. a central processing unit, is supplied with information about the vehicle environment by an environment sensor system 34 and with information about the vehicle interior and in particular the vehicle occupants N ₁ , N ₂ by an interior sensor system 33. A state sensor system 32 also supplies information about the vehicle state. On the basis of the data from the interior sensor system 33, the current actual state, in particular of the driver, is derived with the help of algorithms for occupant state recognition and compared with the target state derived from the data from the environment sensor system 34 and the current automation level. If a discrepancy between these states is detected, an interaction is generated and carried out with the help of a network 37 and a number of interaction systems 31 ₁ -31 _n . By means of a data transmission unit 35, a connection can be established to a backend, for example to obtain updated software for components of the motor vehicle, or to other motor vehicles or infrastructure facilities. ments in the vicinity of the motor vehicle. A memory 36 is provided for storing data.

In the following, further details of a solution according to the invention are given by means of 5 until 10 to be discribed.

5 shows schematically further details of the generation of interaction strategies. The basic sequence of the method is shown together with the data flow. Here, "start" and "end" represent the start and end of the journey or the activation and shutdown of the system. Basically, the system is executed continuously as a monitoring program, which is also illustrated by the loop back to the "start" node. In function block 200, the states and behavior of the individual users are recorded, e.g. their respective level of attention or tiredness. These are estimated by analyzing sensor data, e.g. using machine learning methods. The analysis of the states and behavior patterns can be carried out separately or partially combined. In function block 210, the independent estimates of the states and behavior patterns are merged individually for each user. Dependencies between the estimated variables can be taken into account. In function block 220, the actual state for each user is determined on the basis of the merged estimates together with global measurements, e.g. the interior temperature. Based on the current environment, e.g. the traffic density or the type of road, and the machine status, e.g. the currently active automated driving functions, the current automation level is determined in function block 230. From this automation level, it is derived which target behavior the users must have in order to ensure a safe state or a state that conforms to the automation level. Critical or atypical operating states can bring about a change in the automation level. In accordance with the actual and target states determined in the previous processing steps, a comparison is carried out in function block 240 and a decision is made in function block 250 as to whether interaction is required. If interaction is required, an interaction strategy is generated in function block 260 on the basis of a user-specific interaction model M ₁ , M ₂ taken from a memory 280. The behavior is continuously monitored. If it turns out when generating the interaction strategy that a desired target state cannot be achieved, the degree of automation can be adjusted if possible, for example by deactivating selected active automated driving functions. Finally, the system can be terminated in function block 270 by an abort criterion, e.g. when the system is switched off or at the end of the journey.

The generation and monitoring of a user-specific interaction strategy is preferably realized as an artificial intelligence-based selection of parameterized and combined elements from the available interaction strategies. This creates optimized and, above all, personalized human-machine interaction models. The AI system takes into account the user's reaction to past interactions in order to adapt future interactions to the individual needs and preferences of the respective user.

For this purpose, a reinforcement learning method is advantageously used, which analyzes the user's reaction and the change in his behavior to previously applied interactions in terms of conformity to the respective automation level, compares it with the desired target behavior and, on the basis of this evaluation, iteratively approaches an optimal interaction strategy. 6 shows a schematic of a sequence of such a reinforcement learning process. In the process, an environment 310, in this case the users and the user environment, is observed by sensors and the sensor data SD is analyzed using AI-supported perception 320. The result of the perception 320 is the current actual states Z _I of the users, such as attention, direction of gaze or activity. In step 330, a comparison is now carried out for each user individually based on the current automation level and its requirements for user behavior with the current actual state Z _I of the user. As a result, the target state Z _S , the current actual state Z _I and a reward value B from the comparison are made available to an agent 340. If the user to be observed is not yet known to the system, an interaction model M is generated from the rule base 300, which is stored in a memory 280 and continuously updated. The rule base 300, which describes which interaction strategy is used to get from one state to a target state, can be constructed, for example, from empirical experiments and expert knowledge. The agent 340 derives an interaction strategy from the current actual state Z _I and the target state Z _S , which is implemented in actions A, e.g. in the form of activating various functions of a human-machine interface 350. The human-machine interface 350 carries out the actions, which in turn affect the environment 310. This closes the cycle and the agent 340 can determine over time through the reward value B whether an interaction strategy was successful or not. Over the time the machine is in use, a collection of interaction models M is created in the memory 280, each consisting of a rule basis and observations of the agent 340. These can be transferred to another machine if necessary. In addition, it is possible to clean up the interaction models M, for example by clustering states.

In the following, the user-specific interaction models and the rule base 300, which forms the basis of the agent decision, are discussed in more detail. 7 shows a schematic of the contents of a rule base 300. A target state 400 is shown, which describes, for example, the stage-compliant behavior for the 3rd automation level in the form of rules. These rules describe the permitted state ranges of the user, e.g. "eyes off the road for less than 1 second". In addition to the target state 400, there are also reference states 410, which describe states in which a user can be, and transition rules 420. These define the interaction strategy with which the target state 400 can be reached. Such an interaction strategy can consist, for example, of showing a message on a display. Transition rules 420 can also exist between reference states 410. In addition, several possible transition rules 420 can exist between two states 400, 410. The transition rules 420 can be determined in particular from prior knowledge or empirical experiments. Other examples of states are fatigue, attention, actions such as talking on the phone, stress, workload, body posture such as hands on the steering wheel, or situation awareness.

The rule base 300 is used to create an interaction model for new users of the system, which is personalized by the agent over time. 8th and 9 show how a new state can be learned. 8th schematically the learning of a new state 440. 9 shows schematically the model M with the learned new state 460. In 8th a fictitious current state 440 of a user is marked, based on the current user states. By calculating the distance 450 to the nearby reference states 410, a possible interaction strategy is derived and applied. The reward value is used to determine the success. In 9 the current state was added to the model as a learned new state 460 with a new transition rule 470. In addition, existing transition rules 420 can also be adapted and thus the initial rule base 300 can be changed and customized for the user with adapted transition rules 430.

10 shows a schematic of an interaction model M with several target states 400, which represent the different automation levels, and a number of reference states 410. The different target states 400 can also be linked to transition rules 420, which are bidirectional. This is particularly advantageous when it is necessary to switch from one target state 400 to another because the automation level has changed. Furthermore, the agent learns over time which interaction strategies 420 are the most successful. The interaction strategies 420 are then assigned success probabilities, which are calculated in 10 are visualized by different line thicknesses. A possible implementation of the interaction models M can be in the form of a Markov model, for example.

List of reference symbols

10

contraption

11

Entrance

12

State module

13

Strategy module

14

Control module

15

Control module

16

Storage

17

Exit

18

User interface

20

contraption

21

Storage

22

processor

23

Entrance

24

Exit

30

machine

311-31n

Interaction system

32

Condition sensors

33

Interior sensors

34

Environmental sensors

35

Data transmission unit

36

Storage

37

Network

200-270

Function blocks

280

Storage

300

Rule base

310

Vicinity

320

perception

330

Condition comparison

340

agent

350

Human-machine interface

400

Target state

410

Reference state

420

Transition rule

430

Adapted transition rule

440

New condition

450

distance

460

Learned new state

470

New transition rule

A

action

B

Reward value

IS

Interaction strategy

AI

Artificial intelligence system

M, M1, M2

Interaction model

MD

Machine-related data

MZD

Machine status data

N1, N2

Users

ND

User-related data

SB

Control command

SD

Sensor data

UD

Environmental data

ZI

It's on

ZS

Target state

S1

Recording the states and behavior of a user

S2

Recording the machine’s conditions and operating environment

S3

Determining the current state of the user

S4

Determining a target state of the user

S5

Generating a user-specific interaction strategy

S6

Controlling an interaction system

QUOTES INCLUDED IN THE DESCRIPTION

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

Cited patent literature

• EP1914106A2 [0009]
• EP 3730331 A1 [0010]
• WO 2016/169585 A1 [0011]

Claims (10)

Method for influencing a state of a user (N ₁ , N ₂ ) of a machine (30), comprising the steps of: - determining (S3) an actual state (Z _I ) of the user (N ₁ , N ₂ ); - determining (S4) a target state (Z _S ) of the user (N ₁ , N ₂ ); - generating (S5) a user-specific interaction strategy (IS); and - controlling (S6) at least one interaction system (31 ₁ -31 _n ) of the machine (30) in accordance with the user-specific interaction strategy (IS). Procedure according to Claim 1 , whereby states and behavior of the user (N ₁ , N ₂ ) are recorded (S1) as a basis for determining (S3) the actual state (Z _I ) of the user (N ₁ , N ₂ ). Procedure according to Claim 1 or 2 , wherein states and an operating environment of the machine (30) are recorded (S2) as a basis for determining (S4) the target state (Z _S ) of the user (N ₁ , N ₂ ). Method according to one of the preceding claims, wherein the generation (S5) of the user-specific interaction strategy (IS) is based on an interaction model (M ₁ , M ₂ ). Procedure according to Claim 4 , whereby on the basis of the interaction model (M ₁ , M ₂ ) a prediction is made as to when a user (N ₁ , N ₂ ) leaves a target state (Z _S ). Method according to one of the preceding claims, wherein the user-specific interaction strategy (IS) is generated by a trained artificial intelligence system (AI) (S5). Method according to one of the preceding claims, wherein the interaction system (31 ₁ -31 _n ) uses at least one human-machine interface or at least one machine function. A computer program comprising instructions which, when executed by a computer, cause the computer to carry out the steps of a method according to any of the Claims 1 until 7 to influence a state of a user (N ₁ , N ₂ ) of a machine (30). Device (10) for influencing a state of a user (N ₁ , N ₂ ) of a machine (30), with: - a state module (12) for determining (S3) an actual state (Z _I ) of the user (N ₁ , N ₂ ) and for determining (S4) a target state (Z _S ) of the user (N ₁ , N ₂ ); - a strategy module (13) for generating (S5) a user-specific interaction strategy (IS); and - a control module (14) for controlling (S6) at least one interaction system (31 ₁ -31 _n ) of the machine (30) according to the user-specific interaction strategy (IS). Machine (30), wherein the machine (30) comprises a device (10) according to Claim 9 has or is arranged to implement a method according to one of the Claims 1 until 7 to influence a state of a user (N ₁ , N ₂ ) of the machine (30).

Images