Method for automatically adapting the behavior of a mobile user device

Abstract

Method (400) for automatically adjusting the behavior of a mobile user device (102) to increase the probability that the mobile user device (102) will successfully complete a task (408), which, for example, involves an interaction with a human (104), wherein successful completion of the task includes, for example, causing the human (104) to engage with a robotic device (102) and / or causing the robotic device (102) to engage with the human (104), wherein the method (400) is executed by a processor (220) in the mobile user device (102), and wherein the method (400) comprises: Receiving sensor data (302) comprising a, in particular first, signal output by a first sensor (404), wherein the first signal indicates that a person (104) is in an environment (101) of the mobile user terminal (102), wherein the method (400) further comprises: Identifying an action to be performed by the mobile robot device (102), wherein the action is identified to increase the probability of successful completion of the task, wherein the identified action is based at least partially on the received sensor data (302) and a model (306) representing successes and failures of past attempts by the mobile robot device (102) to complete the task (408) with other people (104); Transmitting a signal to an actuator (410) of the mobile robot device (102) to cause the mobile robot device (102) to perform the action as an attempt to complete the task (408), wherein a second signal is output by a second sensor (406), the second signal indicating a first state of the mobile user device (102), the first state being subject to control by the mobile user device (102), and the first state being identified as relevant to the probability that the mobile user device (102) will successfully complete the task (408), the method comprising identifying an indication of whether the robot (102) has successfully completed the task (408) based on a human response (104) to the robot (102) performing the action;and providing the sensor data (302), the action performed, and the identified indication of whether the robot (102) has successfully completed the task (408) for the purpose of adapting the model (306).

Description

[0001] The present invention relates to a method for automatically adapting the behavior of a mobile user terminal device, as well as a method for user-dependent operation of at least one data processing system according to the respective preambles of claims 1 and 2.

[0002] Anything disclosed independently for one procedure is disclosed for the other procedure and vice versa.

[0003] The procedure described here for operating at least one data processing system in a user-dependent manner includes, in a first step, the provision of at least one initial computing unit which, based on user data stored in a database, generates and / or retrieves at least one user profile of the user's usage objects that is uniquely assigned to the user.

[0004] In the context of the present invention, a user profile is a configuration of a user account within an operating system, stored in a system administration interface. A user profile can encompass various rights, such as read rights, write rights, connection rights, program installation rights, and deletion rights. These rights are assigned by the respective system administrator as part of user management and can be modified by any other system administrator. The administrators can be one or more users. The administrator, however, only has a privileged user profile. To ensure data protection, the user profiles of others can generally be restricted.

[0005] The user profile preferably also stores the user's own files and settings. This includes program configuration files.

[0006] Particularly sensitive user data can be additionally secured, for example in a root account or protected storage.

[0007] The first processing unit can be a computer, a notebook, or any other type of computer chip.

[0008] In a second step, a second computing unit is provided, which makes at least one usage suggestion to the user for the future use of a usage object, whereby the user can confirm, change and / or reject the usage suggestion, and furthermore, in the event of a change and / or rejection of the usage suggestion by the user, the second computing unit adapts the user profile to the then changed user data.

[0009] Such an adjustment can consist of modifying usage data, which is clearly, preferably uniquely, associated with a corresponding usage object in the user profile in order to reflect a changed usage behavior of the individual usage objects.

[0010] In a third step, a third computing unit is provided, which receives the user data transmitted by the first computing unit, whereby the user data is control and / or regulation data, by means of which at least one usage object is controlled and / or regulated.

[0011] The present invention is therefore divided into at least three steps. A first step is defined by the operation performed by the first computing unit and can be described as a learning phase of the data processing system.

[0012] A second step is defined by the provision of the second computing unit, which can be described as a correction phase of the data processing system, with a third learning phase representing an implementation phase, which is represented by the provision of the third computing unit.

[0013] DE 10 2018 204 740 A1 discloses the provision of an action information learning device, a robot control system, and an action information learning method for facilitating cooperative work between an operator and a robot. An action information learning device includes: a state information acquisition unit that captures the state of a robot, in a case where the robot hands over a workpiece, picked up from a workpiece storage location, to an operator within a transfer area, which is the area in which the workpiece is to be handed over;an action information output unit for outputting an action which is adjustment information for the state; a compensation calculation section for capturing determination information, which is information about a transfer time relating to the transfer of the workpiece, and calculating a compensation value from reinforcement learning based on the determination information thus captured; and a value function update section for updating a value function by performing reinforcement learning based on the compensation value calculated by the compensation calculation section, the state, and the action.

[0014] DE 10 2017 009 223 A1 discloses a control device for a robot for performing a process in cooperation with a person. The control device comprises a machine learning device, which includes a recognition unit for classifying an action of the person and a learning unit for learning the action of the person while the person performs a process in cooperation with the robot; and an action control unit for controlling the action of the robot based on a result of the classification by the recognition unit.

[0015] DE 10 2017 007 729 A1 discloses a machine learning device for learning the movement of a robot involved in a joint task performed by a human and a robot, comprising a state monitoring unit that monitors a state variable indicating the robot's state when the human and robot are collaborating and performing a task, a reward calculation unit that calculates a reward based on control data and the state variable for controlling the robot and a human action, and a value function update unit that updates an action value function for controlling the robot's movement based on the reward and the state variable.

[0016] DE 10 2016 009 113 B4 discloses a machine learning device for a robot that enables a person and the robot to work together, wherein the machine learning device comprises: a state observation unit that observes a state variable which monitors a state of the robot during a period in which the person and the robot work together; a determination data unit that obtains determination data for at least one of a workload level for the person and a work efficiency; and a learning unit that learns a training data set to set an action of the robot based on the state variable and the determination data.

[0017] WO 2017 / 163 251 A2 discloses robotic systems for simultaneous human-robot operations in a collaborative workspace. In some embodiments, the collaborative workspace is defined by a reconfigurable workbench to which robot members can be optionally added and / or removed depending on task requirements. The tasks themselves are optionally defined within a production system, potentially reducing the computational complexity of predicting and / or interpreting human operator actions while maintaining flexibility in the execution of the assembly process itself. In some embodiments, robotic systems include a motion tracking system for monitoring the movements of individual body parts of the human operator.Optionally, the robot system plans and / or adapts robot movements based on movements previously observed during the performance of a current operation.

[0018] US 2016 / 0 260 027 A1 shows that, in order to enable safe work in a space where a robot and a worker coexist without defining an area in a workspace using a monitoring boundary or the like, and thus improve productivity, a robot control device is provided that controls the robot by capturing time-series states of the worker and the robot and includes: a capture unit configured to capture a worker state; a learning information storage unit configured to store learning information obtained by learning the time-series states of the robot and the worker;and a control unit configured to control the operation of the robot based on the worker's state output by the sensing unit and the learning information output by the learning information storage unit.

[0019] According to at least one embodiment, the second and / or third computing unit calculates an accuracy rate, wherein the accuracy rate is a numerical ratio, in particular a decimal fraction, of suggested usage parameters to the usage parameters actually selected by the user, and user data is only transmitted to a third computing unit if an accuracy threshold is exceeded or fallen below. The accuracy rate can thus, or alternatively, be an estimable or learned action or reaction.

[0020] The usage parameters can include the usage time of a usage object during a day, the electricity consumption of a usage object during a day, or other consumption of the usage object during a day or a previously defined other operating time.

[0021] For example, if the user sets a desired usage time and / or a desired initial usage time in the second computing unit, and a decimal fractional ratio is calculated from this subsequently changed usage time to the usage time previously stored in the database, the corresponding accuracy is determined. If this value is below or above an accuracy threshold, the user data is transmitted to the third computing unit for control.

[0022] In the first and / or second and / or third processing unit, a target accuracy threshold interval can also be stored, within or outside of which the calculated target accuracy must lie.

[0023] It can therefore be stipulated that only after such a condition has been fulfilled by the first and / or the second computing unit will it be possible to transmit the respective user data to the third computing unit. The third computing unit is preferably the computing unit that controls and / or regulates the corresponding object of use.

[0024] According to at least one embodiment, in a first step of the method for user-dependent operation of at least one data processing system, at least a first computing unit is provided which, on the basis of user data of a user stored in a database, generates at least one user profile of usage objects of the user that is uniquely assigned to the user.

[0025] In a second step, a second computing unit is provided, which makes at least one usage suggestion to the user for the future use of a usage object, whereby the user can confirm, change and / or reject the usage suggestion, and furthermore, in the event of a change and / or rejection of the usage suggestion by the user, the second computing unit adapts the user profile to the then changed usage data.

[0026] In a third step, a third computing unit is provided, which receives the user data transmitted by the first computing unit, whereby the user data includes control and / or regulation data, by means of which at least one usage object is controlled and / or regulated.

[0027] According to at least one embodiment, the second and / or third computing unit calculates an accuracy, wherein the accuracy is a numerical ratio, in particular a decimal fraction, of suggested usage parameters to usage parameters actually selected by the user, and only if an accuracy threshold is exceeded or not met are the user data transmitted to the third computing unit.

[0028] It is also conceivable that the first computing unit, the second computing unit and / or the third computing unit are integrated into a single or single common computing unit, for example a single integrated chip.

[0029] System-on-a-chip (SOC) refers to the integration of all or some of the functions of a system onto a single chip (DIE), i.e., with integrated circuits (ICs) on a semiconductor substrate (also called monolithic integration). All the computing units described above can be monolithically integrated together, forming a single system-on-a-chip.

[0030] Such a chip can then perform the different tasks of the computing units described here.

[0031] According to at least one embodiment, if the threshold is undershot and / or exceeded, the second and / or third computing unit (2, 3) sends and / or triggers an error message and / or a correction response to the first computing unit (1).

[0032] In particular, an error message can be generated if the accuracy deviates by more than 10%, preferably more than 5%, from the accuracy threshold and / or the accuracy interval and its limits.

[0033] According to at least one embodiment, the computer units form a completely independent and preferably self-contained data system such that the establishment and operation of the communication link between the computer units is based exclusively on data communication between the computer units themselves. A local communication network may be created in this context.

[0034] In particular, this allows communication between individual computer units to be structured very independently and free from external network and / or internet interference. An external network component is therefore, for example, a component that is not assigned to the local network described here. This assignment can be made via appropriate IP addressing.

[0035] In other words, the above-described method and thus the above-described data processing system is completely decentralized, so that a complete decoupling of the above-described data processing system from a higher-level data processing system (for example, the Internet) is possible, which significantly increases private security through the above-described decoupling.

[0036] A local area network (LAN) as used in this application refers to a personal area network (PAN) that can be set up and taken down ad hoc by small devices such as PDAs or mobile phones (e.g., smartphones). PANs can therefore be established using various wired transmission technologies such as USB, FireWire, or Ethernet, or wireless technologies such as IrDA, Bluetooth, or WLAN (WPAN). PANs can be used for communication between devices, but they can also be used to communicate with a larger network (uplink). A PAN created using Bluetooth is called a piconet.

[0037] These devices then control, for example, other devices such as lamps (one or many), light switches (one or many), sockets (one or many), etc. - that they are assigned or will be assigned an ID (identification) (for example, by specifically entering the ID) and data for authentication with peers (other devices containing the invention), and / or - that data is reproduced from peers (as provided based on ID and authentication), and / or - the control of output signals is carried out based on this data.

[0038] According to at least one embodiment, at least one data record is transferred from one computer unit to another, or vice versa, by means of a separate input element and / or an input element integrated into one or more of the computer units. Alternatively, such data transmission can be performed without the need for separate input into such an input element, so that data transmission to at least one of the computer units can occur fully automatically, for example, after prior configuration.

[0039] In this context, it is conceivable that such a separate input element could be, for example, a mobile phone display, a light switch, or any other mechanical, electrical, or electromechanical element that generates a data record which is immediately forwarded to a computing unit associated with this input element, preferably one that is uniquely assigned. Within a computer of the computing unit, this data record generated by the separate input element is then converted into a data record or data packet that is at least partially written in HTML and also contains structured markups compatible with HTML-based language elements.This modified data packet is then forwarded from a temporary storage area of ​​the computer unit back to that computer, which then forwards this modified data packet (HTML data packet), for example in the form of a push notification, to at least one other data computer.

[0040] In particular, it is conceivable that a chain reaction occurs in this context, such that this further computing unit (for example, the second and / or third computing unit) forwards this, for example, modified data packet, or through a purely illustrative further modification of this further data packet, this even further modified data packet to an additional data computer (second and / or third computing unit) or any number of further data computers that constitute the private communication system. Therefore, one of the data records can be displayed on a further separate input element and / or separate output element of one of the computing units, or an output element of a data computer, which is different from the input element of a data computer, can be controlled or regulated using this data record that has been modified at least once.Therefore, this data transfer is preferably not, or not solely, a data transfer within the framework of a relay function. Preferably, the data fed into each computer unit is modified, among other things, by adding specific markups.

[0041] The term "modification" can refer either to a partial or complete change of the dataset, for example, with regard to its dataset architecture, or "modification" can simply mean that the dataset or data remain unchanged and that internal indexes and / or context files are appended and / or prepended to this unchanged dataset by each or one of the data servers, for example, for identification purposes. In this respect, it is conceivable that the HTML files generated by a data server remain identical across the entire system and can only be supplemented by the internal indexes and / or context files.

[0042] Furthermore, user access and / or entry can be controlled and monitored particularly easily with regard to the user themselves, the location, the time as well as generic (e.g. externally or internally generated) data.

[0043] Furthermore, it is possible that very large files, such as video files, can be exchanged between users via the private data processing system described above, particularly using an email exchange program, since this system does not have access to the overarching internet.

[0044] Address books from inbox containers, inbox recovery programs, etc., can also be transferred from one data server to another. This is made possible, among other things, by at least one of the servers acting as a central processing unit, which is selected by one or more users (for example, via IP addressing) to form a central storage unit on which at least some, preferably all, of the data is stored temporarily or permanently. This data is stored by this central processing unit and also forwarded from this central storage unit to the respective decentralized data servers of the private data communication system, for example, via push notifications. However, instead of data transmission based on push notifications, a pull-message data transmission is also possible.

[0045] According to at least one embodiment, the data packets comprise HTML data packets. Communication between the individual computer units and / or the user objects, and between the computer units themselves, can therefore take place on the basis of an HTML-based communication connection.

[0046] In a further step, an HTML-based communication connection is established between the data computers, whereby, after the communication connection is established, data is transferred from one data computer to at least one other data computer. "HTML-based communication connection" refers to any communication connection that is at least partially structured according to an HTML standard and / or involves the transmission of data packets that are at least partially structured as HTML data packets or at least partially contain such packets.

[0047] The term "HTML-based" simultaneously defines and / or establishes a communication layer within the meaning of the application, within which the data packets are moved. This HTML-based communication layer, as defined here, is therefore an application-oriented layer. For example, the data packets presented in this invention do not leave the communication layer defined by the HTML data packet.

[0048] This can mean, in particular, that the communication link is free of data packets that represent a "deeper" programming language or representation in the sense of the OSI standard compared to the HTML communication layer described above. Therefore, the present invention is fundamentally based, among other things, on the concept that the communication system described herein not only has a predefinable selection of well-defined users that can always be configured by the user, but is also a communication system which—after prior specific IP addressing by the user—then uses the HTML layer described above as its lowest communication layer in a particularly user-friendly manner.

[0049] According to at least one embodiment, at least one, for example exactly one, computer unit (central data computer) within the data processing system is a data storage device of the data processing system, wherein preferably all data packets are sent from the data computer (peripheral computer units) to this data computer and then managed and / or stored by this data computer.

[0050] According to at least one embodiment, data from a peripheral computing unit, i.e., at least one of the computing units described above, is restored by retrieving one of the data stored in the central computing unit (also one or more of the computing units described above).

[0051] According to at least one embodiment, a computing unit is a particularly mobile user terminal device, such as a toothbrush, a coffee machine, or similar.

[0052] For example, the mobile user device is a robot. A robot, as the term is used here, is an electromechanical machine that contains computer hardware and software enabling the robot to perform functions independently and without human assistance. Examples of robots currently available on the market include robotic vacuum cleaners and lawnmowers. A typical commercially available robotic vacuum cleaner, for instance, contains computer-executable instructions which, when executed by the robot's processor, cause it to automatically vacuum a specific area (e.g., a room) based on input received from the robot's sensors.A conventional, commercially available robotic lawnmower is configured with computer-executable commands which, when executed by a processor of the robotic lawnmower, cause such a robotic lawnmower to automatically mow grass in a specific area (e.g., in a homeowner's garden) without human intervention.

[0053] However, it is also possible that the mobile user device differs from a robot. In this case, a mobile user device is understood to be a device that operates semi-automatically rather than fully automatically. This can mean that a process can only be completed with the assistance of a user or another device. In this version, human intervention is therefore necessary to operate the mobile user device. Alternatively, the computing unit can be a non-mobile user device, such as permanently installed equipment.

[0054] Artificial intelligence (AI) as a scientific discipline is only a few decades old and thus significantly younger than the classical natural sciences. Nevertheless, it rightly claims the status of its own scientific sector. Its subject matter exerts a particular fascination. This may be because abstract thinking, creativity, and especially consciousness are widely considered to be generically human abilities. The fact that AI systems increasingly seem to acquire or at least imitate these abilities, and may soon surpass humans in them, represents a paradigm shift no less profound than the discovery of the heliocentric worldview or biological evolution. AI also plays an increasingly important role in technological development, and there is no indication that this trend will end anytime soon.

[0055] The decisive factor for access to patent protection is the technical nature of an invention. Whether inventions such as a computer program for machine translation or text processing, image recognition on satellite images, a computer-based medical diagnostic system, a learning tutorial for a foreign language, a legal tech or autonomous driving system, speech recognition software, or a chess computer are to be considered technical or not is not immediately obvious. Nevertheless, the distinction between technical and non-technical inventions, developed in detail by case law, also plays a central role in promoting innovation through patents in the new and dynamic field of AI. Artificial intelligence and machine learning are based on computational models and algorithms for classification, bundling, regression, and dimensionality reduction, such as neural networks, genetic algorithms, support vector machines, and k-means. Kernel regression and discriminant analysis. Such computational models and algorithms are inherently abstract mathematical in nature, regardless of whether they can be "trained" using training data. Therefore, the guidelines in G-II, 3.3 generally also apply to such computational models and algorithms.

[0056] Terms such as "Support Vector Machine", "Reasoning Engine" or "neural network" may, depending on the context, refer only to abstract models or algorithms and therefore do not necessarily imply the use of technical means in themselves.

[0057] Artificial intelligence and machine learning are used in various fields of technology. For example, the use of a neural network in a heart monitoring device to identify irregular heartbeats makes a significant technological contribution. The classification of digital images, videos, audio, and speech signals based on low-level features (e.g., edges or pixel attributes for images) is another typical technical application of classification algorithms.

[0058] If a classification method serves a technical purpose, the steps of “generating the training data set” and “training the classifier” can also contribute to the technical character of the invention if they support the achievement of that technical purpose.

[0059] According to the invention, the method for automatically adapting the behavior of a mobile user device to increase the probability that the mobile user device will successfully complete a task, which, for example, involves interaction with a person, wherein successful completion of the task includes, for example, causing the person to interact with an execution device, for example, a mobile user device, and / or causing the execution device to interact with the person, wherein the method is executed by a processor in the mobile user device.

[0060] According to the invention, the method comprises receiving sensor data, which includes a, in particular, a first signal, which is output by a first sensor, wherein the first signal indicates that a person is in an environment of the mobile user device, wherein the reception of the sensor data can also be carried out by pressing, heating or by automatic detection systems, in particular of a person.

[0061] According to the invention, the method comprises a second signal output by a second sensor, wherein the second signal indicates a first state of the mobile user device, wherein the first state is subject to control by the mobile user device and the first state is identified as relevant to the probability that the mobile user device will successfully complete the task.

[0062] According to the invention, the method comprises identifying an action to be performed by the mobile user device, wherein the action is identified in order to increase the probability of successful completion of the task.

[0063] According to the invention, the identified action is based at least partially on the received sensor data and a model that represents successes and failures of past attempts by the mobile user device to complete the task with other people.

[0064] According to the invention, a signal is transmitted to an actuator of the mobile user device in order to cause the mobile user device to perform the action as an attempt to complete the task.

[0065] According to the invention, an indicator identifies whether the user device has successfully completed the task, based on a human response to the mobile user device performing the action.

[0066] According to the invention, sensor data of the action performed and the identified display indicating whether the mobile user device has successfully completed the task are provided for model adaptation.

[0067] According to one embodiment, individual users are classified, for example, in the form of digital images, videos, audio and speech signals, particularly on the basis of low-level features (e.g., edges or pixel attributes for images).

[0068] This classification can be used to provide the user with a recommendation, which the user can then decide whether to follow or correct. The corrected data set then serves, among other things, to increase the probability of successfully completing the pending action.

[0069] According to at least one embodiment, the method for automatically adapting the behavior of a mobile user terminal increases the probability that the mobile user terminal will successfully complete a task, which, for example, involves interaction with a human, wherein successful completion of the task includes, for example, causing the human to interact with an execution device, for example, a mobile user terminal, and / or causing the execution device to interact with the human, wherein the method is executed by a processor in the mobile user terminal, and wherein the method comprises: Receiving sensor data comprising an initial signal output by a first sensor, the initial signal indicating that a human is in the vicinity of the mobile user device.

[0070] According to at least one embodiment, the method comprises a second signal output by a second sensor, wherein the second signal indicates a first state of the mobile user device, the first state being subject to control by the mobile user device and the first state being identified as relevant to the probability that the mobile user device will successfully complete the task, wherein the method is characterized in that it further comprises: Identifying an action to be performed by the mobile user device, wherein the action is identified to increase the probability of successful completion of the task, wherein the identified action is based at least partially on the received sensor data and a model representing successes and failures of past attempts by the mobile user device to complete the task with other people;

[0071] Transmitting a signal to an actuator of the mobile user device to cause the mobile robotic device to perform the action as an attempt to complete the task.

[0072] According to the invention, the method is characterized by identifying an indication of whether the user terminal has successfully completed the task, based on a human response to the execution of the action by the user terminal; and Providing the sensor data, the action performed, and the identified indication of whether the user device has successfully completed the task, for model customization.

[0073] According to at least one embodiment, at least one computing unit has an interface to a non-local computer network (Internet) such that a further computing unit connected to the non-local computer network is controlled to generate further use of a usage object and / or another usage object. According to at least one embodiment, at least one computing unit is designed as a system-on-a-chip.

[0074] The invention will now be described in more detail with reference to an exemplary embodiment and the accompanying figures. Fig. Figure 1 shows a schematic representation of the exchange of information within the data system.

[0075] In the Fig. Figure 1 shows that the method 1000 described here for the user-dependent operation of at least one data processing system comprises three computing units, each of which is labelled with the reference symbols 1, 2, 3.

[0076] In particular, computer unit 1 can be represented as a behavioral database that communicates with computer unit 2 and stores the respective user behavior. Computer unit 2 can perform a behavioral analysis (or user behavior analysis), whereby computer unit 3 can perform the prediction of behavior.

[0077] The individual tasks can also be exchanged between the computing units described here. In particular, it is shown that all three computing units communicate with each other and with each other. Computing unit 1 can communicate directly with computing unit 3, just as computing unit 3 can communicate directly with computing unit 2, and the same applies to communication between computing unit 1 and computing unit 2.

[0078] The computing units 1 to 3 can also be implemented as a single data communication element in the form of a "smart device". These can be integrated into a local household.

[0079] It is also evident that this smart device includes a control interface 4 and a user ID interface 5, so that the control interface 4 and the user ID interface 5 can communicate independently with the user interface 7 in order to connect various external usage objects and / or computer units.

[0080] Furthermore, corresponding user objects 110B are displayed around 11B. A user profile is stored primarily in the smart device, but alternatively also in the user object. These user objects can include mobile phones, tablets, notebooks, alarm clocks, smart toothbrushes, smart razors, smart motion sensors, smart mattresses, smart TVs, and smart microphones. This list is not exhaustive. The user objects are in direct data communication with the smart device and thus in direct data communication with computing units 1 to 3.

[0081] Both the usage objects and the smart device can communicate with an internal or external database 8.

[0082] The usage objects can therefore identify individual users, for example, household users. The usage objects and / or the individual computers can determine and learn the individual usage characteristics, that is, the individual preferences of each person, based on individual usage wishes, schedules, moods, or other household users.

[0083] The individual user objects and / or the individual computing units 1 to 3 then predict future usage patterns based on the user's individual wishes and expectations, taking into account personal and environmental factors. Based on the user's response, specific suggestions or behavioral patterns related to the user object or its use are predicted. If the prediction deviates from the thresholds and / or threshold intervals described above, the usage forecast is adjusted accordingly.

[0084] The user objects can only be made available to the user after personal confirmation or unlocking, in particular based on learned behavior.

[0085] Communication can either be rooted through a local gateway or can be carried out directly between the usage objects, i.e., the smart device and the usage objects.

[0086] As an example, consider the operation of a coffee machine in a household with several people.

[0087] The behavior of the residents (settings of the coffee program, such as type, intensity, etc., as well as times of use on different days) is recorded in the learning phase already described above (use of the devices and smart sensor data).

[0088] During a correction phase, the user is given suggestions regarding the objects of use and / or the smart device, which may include, for example, an offer for the preset coffee program at certain times of day or occasions.

[0089] If the user makes a different choice, this is first analyzed by the second computing unit and correlated with other external influences (such as the daytime temperature, ambient humidity, ambient temperature).

[0090] The presence of other people in the room or household may also depend on this.

[0091] During the implementation phase, the accuracy of the suggestions preferably exceeds a certain threshold that can be confirmed by the user. This allows the smart device to independently control certain actions, such as switching on the device at a specific time, preheating, selecting programs, etc. Ideally, the suggestions are prepared and offered automatically.

[0092] In particular, it may be stipulated that data exchange between the smart devices placed in the household and the smart sensors (level of use) with internally based external intelligence or databases is prevented.

[0093] Preferably, for example, such data exchange is not possible. The gateway can serve for local data storage and control. This gives the individuals concerned the ability to monitor and / or delete this database at any time. A connection to internet-based services, such as ordering consumables like groceries, detergents, etc., or repair services, can be established through a controlled and monitored interface as described above.

[0094] The Fig. Figure 2 shows in one embodiment a method 400 for automatically adapting the behavior of a mobile user terminal 102 to increase the probability that the mobile user terminal 102 successfully completes a task 408, which, for example, involves an interaction with a person 104, wherein successful completion of the task includes, for example, that the person 104 is caused to interact with an execution device, for example, a mobile user terminal 102, and / or that the execution device interacts with the person 104, wherein the method 400 is executed by a processor 220 in the mobile user terminal 102, and wherein the method 400 comprises: Receiving sensor data 302, comprising one, in particular a first, signal output by a first sensor 404, wherein the first signal indicates that a human 104 is in an environment 100 of the mobile user terminal 102.

[0095] Fig. 2 A second signal can also be obtained from a second sensor 406, wherein the second signal indicates a first state of the mobile user terminal 102, wherein the first state is subject to control by the mobile user terminal 102 and the first state is identified as relevant to the probability that the mobile user terminal 102 will successfully complete task 408, wherein the method 400 further comprises: Identifying an action to be performed by the mobile user terminal 102, wherein the action is identified to increase the probability of successful completion of the task, wherein the identified action is based at least partially on the received sensor data 302 and a model 306 representing successes and failures of past attempts by the mobile user terminal 102 to complete the task 408 with other people 104; Transmitting a signal to an actuator 410 of the mobile user terminal 102 to cause the mobile user terminal 102 to perform the action as an attempt to complete task 408.

[0096] Also, from the Fig.2. Identifying an indicator to recognize whether the robot 102 has successfully completed task 408, based on a response of the human 104 to the robot 102 performing the action; and providing the sensor data 302, the action performed, and the identified indicator of whether the mobile user device 102 has successfully completed task 408, for the purpose of adapting the model 306.

[0097] The invention is not limited by the description and the exemplary embodiment. Rather, the invention encompasses every new feature as well as every combination of features, which also includes in particular every combination of the patent claims, even if this feature or this combination itself is not explicitly disclosed in the patent claims or in the exemplary embodiment. Reference symbol list 1 computer unit 2 computer units 3 computer units 4 Control interface 5 User ID Interface 7 Interface 8 internal or external database 11 Database 11A User data 11B users 11C User Profile 100 Data processing systems 101 Environment 102 mobile user device 104 People 110B Usage objects 220 processor 302 sensor data 306 model 400 procedures 404 first sensor 406 second sensor 408 Task 410 Actuator 1000 procedures

Claims

[1] Method (400) for automatically adjusting the behavior of a mobile user terminal (102) to increase the probability that the mobile user terminal (102) will successfully complete a task (408) which, for example, involves an interaction with a human (104), wherein successful completion of the task includes, for example, causing the human (104) to engage with a robotic device (102) and / or causing the robotic device (102) to engage with the human (104), wherein the method (400) is executed by a processor (220) in the mobile user terminal (102), and wherein the method (400) comprises: Receiving sensor data (302) comprising a, in particular first, signal output by a first sensor (404), wherein the first signal indicates that a person (104) is in an environment (101) of the mobile user terminal (102), wherein the method (400) further comprises: Identifying an action to be performed by the mobile robot device (102), wherein the action is identified to increase the probability of successful completion of the task, wherein the identified action is based at least partially on the received sensor data (302) and a model (306) representing successes and failures of past attempts by the mobile robot device (102) to complete the task (408) with other people (104); Transmitting a signal to an actuator (410) of the mobile robot device (102) to cause the mobile robot device (102) to perform the action as an attempt to complete the task (408), wherein a second signal is output by a second sensor (406), the second signal indicating a first state of the mobile user device (102), the first state being subject to control by the mobile user device (102), and the first state being identified as relevant to the probability that the mobile user device (102) will successfully complete the task (408), the method comprising identifying an indication of whether the robot (102) has successfully completed the task (408) based on a human response (104) to the robot (102) performing the action;and providing the sensor data (302), the action performed, and the identified indication of whether the robot (102) has successfully completed the task (408) for the purpose of adapting the model (306).

Images