JP2007000938A - Action-integrated robot device - Google Patents

Abstract

PROBLEM TO BE SOLVED: To enable a user to easily operate an information device in a robot apparatus having a bodily responsive and friendly body used as an interface between an advanced and complicated internet home appliance or information device and a user. An object of the present invention is to provide a robot apparatus that provides appropriate services to a user.
The present invention relates to an input unit that obtains information from a user and information about the surroundings, a recognition unit that recognizes user instructions and surroundings from the information from the input unit, and a recognition result of the recognition unit. An interpretation unit that combines and interprets the information; a first behavior generation unit that generates an action for performing a task based on the interpretation result of the interpretation unit; and a user's current behavior based on the recognition result of the recognition unit A second action generation unit that generates a simultaneous imitation action to imitate, an action integration unit that combines the actions of the first action generation unit and the second action generation unit, and an action integrated by the action integration unit And an action presentation unit for presenting.
[Selection] Figure 1

Description

  The present invention relates to a robot apparatus that provides a service to a user by interfacing between a device and a user by fusing a practical service behavior with respect to the user and a behavior that expresses affinity for the user.

  In recent years, entertainment robots that entertain users and robots useful to users have been researched and developed at home. In the former entertainment robot, the purpose is to entertain the user, and one of the challenges is not to get bored. As the latter useful robot, one that acts by listening to a user's command and one that recognizes the user's intention and performs an action suitable for the user have been proposed.

  As an entertainment robot, there is one that causes a robot apparatus to perform reinforcement learning for action acquisition so that the robot does not get tired of the user. In that case, a robot apparatus and a robot control method, a recording medium, and a program characterized by dynamically changing the learning rate of action value are cited (see Patent Document 1).

  As a useful robot, the user's operation information and the external situation when the operation is made in advance as a controller to control various devices according to the user's lifestyle and living environment or according to the user's preference An integrated controller, a device control method, and a device control program for storing an operation signal suitable for the user from the current user operation information and external situation information have been proposed. An adapted robot is cited (see Patent Document 2).

  In addition, there is a method for realizing face-to-face communication between a human and a robot by imitating human behavior by a robot (see Non-Patent Document 1). This means that in face-to-face communication between a human and a robot, voice information and joint angle information are acquired from a human action, and the robot imitates the person by reproducing the acquired information.

  On the other hand, with the spread of broadband, the construction of home networks including information devices is progressing at home. Furthermore, research on ubiquitous networks in which these information devices and various sensors are integrated is also active. However, the user interfaces of these information devices are not always easy to use. This is because each device has many useful functions and can be used in various ways, but because of its variety, complicated inputs are forced to execute the functions desired by the user. In order to solve this problem, an interface using a robot having a bodily body that responds by voice and is familiar is attracting attention as an interface that can be easily operated by mediating between a device and a user.

As an interface using the robot, in addition to the reliable execution (practicality) of the function that the user wants, it also understands the ambiguous request made by the user or communicates with the robot to clarify the request (affinity) Sex) is desired.
JP 2003-340760 A JP 2003-111157 A Basic experiment on human-robot communication based on imitation (Proceedings of the 19th Annual Conference of the Robotics Society of Japan (1J15)

  As described above, in order to keep the user from getting bored with the entertainment robot, the output operation pattern for the input signal to the robot is changed by reinforcement learning, and the useful robot records the user's past operation pattern. In addition, the operation that the user will perform next is pre-read from the current situation, and the robot tries to execute it instead of the user, and the user's behavior itself is not particularly reflected.

  Furthermore, in the face-to-face communication between a human and a robot, the face-to-face communication between a human and a robot is the purpose. In order for the robot to imitate a human, the acquired information and the output information must be of the same dimension (for example, acquired In order to use human joint angle information, there is a problem that the robot needs to have joints corresponding to the information.

  To make it easier to use complex home appliances and information devices, it is expected that instructions from the user can be input intuitively, such as voice and gestures, instead of input from the conventional keyboard and mouse. In recognition by voice or gesture, there are many misrecognitions, and it is necessary for the user to behave against the user's intention or to give instructions to the user over and over, and the user may be dissatisfied.

  In order to solve the above-described problems, the present invention is a robot device that is used as an interface between a sophisticated and complicated internet home appliance or information device and a user, and has a bodily body that responds by voice and is easily accessible. It is an object of the present invention to provide a robot apparatus that operates information equipment and provides appropriate services to users.

In order to solve the above problem, the present invention includes an input unit that obtains information from a user and information on the surrounding situation,
To perform a task based on a recognition unit that recognizes a user's instruction and surrounding situation from information from the input unit, an interpretation unit that combines and interprets information from a recognition result of the recognition unit, and an interpretation result of the interpretation unit A first action generation unit that generates an action of the user, a second action generation unit that generates a simultaneous imitation action that imitates a user's current action based on the recognition result of the recognition unit, and the first action generation And a second behavior generation unit, and a behavior presentation unit for presenting the behavior integrated by the behavior integration unit.

  According to the present invention, while the user is familiar with the action-integrated robot, the user can give an appropriate instruction according to the robot's recognition ability naturally, and can obtain the task he desires with a simple instruction. At the same time, it becomes possible to achieve pseudo-communication without getting bored.

  Hereinafter, embodiments of the present invention will be described with reference to the drawings.

  As for the embodiment of the present invention, FIG. 1 shows the configuration of a behavior-integrated robot apparatus using a robot that moves with wheels and has a camera head with two degrees of freedom (pan / tilt) as a head. Show.

  As shown in FIG. 1, the robot 100 includes an input unit 110, a recognition unit 120, an interpretation unit 130, a first behavior generation unit 140, a second behavior generation unit 150, a behavior integration unit 160, a basic control processing unit 170, The action presentation unit 180 is configured.

  The input unit 110 includes input devices for a camera 111, a microphone 112, an axle encoder 113 for measuring the position of the robot, and an ultrasonic sensor 114 for detecting an obstacle. The camera 111, the microphone 112, the axle encoder 113, and the ultrasonic sensor 114 of the input unit 110 obtain an instruction from the user to the robot and information obtained by detecting the surrounding situation. The recognition unit 120 uses the information obtained from the input unit 110 as, for example, image information obtained from the camera 111, user face orientation information and face recognition information, voice information obtained from the microphone 112, and user utterance words. The recognition information and chime detection results are converted from the wheel rotation speed obtained from the axle encoder 113 to the current position information of the robot, and from the reflected wave information obtained from the ultrasonic sensor to processed information such as obstacle information. .

  Interpreting unit 130 combines a plurality of pieces of information obtained from recognizing unit 120, interprets and holds the information for use in first action generating unit 140 and second action generating unit. This is information used for action selection and action generation in the first action generation part 140 and the second action generation part 150 using information from the recognition part 120 (for example, trigger information and user action information that the robot can imitate). Convert to

  The first behavior generation unit 140 and the second behavior generation unit 150 generate the first behavior and the second behavior, respectively, based on the information held by the interpretation unit 130. The behavior reconciliation unit 160 reconciles the behaviors generated by the two behavior generation units described above. The action presentation unit 180 includes an actuator control processing unit 181 and individual actuators (in this embodiment, movement 182, speech 183, gaze 184, home appliance operation 185). The actuator control unit 181 converts the behavior integrated by the behavior integration unit 160 so that each actuator (182 to 185) can execute, and each actuator operates accordingly. The basic control processing unit 170 directly controls the actuator based on information from the recognition unit 120 and performs processing necessary for the robot to behave autonomously, such as an obstacle avoidance operation.

  Next, the outline of the present embodiment will be described with reference to the schematic diagram of FIG. 2 and the sequence diagram showing the present embodiment of FIG. The situation is a setting in which the mother 220 is in the living room 210 and the child 220 comes home while doing housework. First, the robot 230 looks in the direction the mother 200 sees together (second action execution). Therefore, when the child 220 returns and rings a chime, the robot 230 copies the image of the entrance camera 250 on the display 240 provided in the room and shows it to the mother 200 (first action execution using the chime as a trigger). The mother 200 learns from the video that the child 220 is going home, instructs the robot 230 to go to the entrance 260 to pick up the child 220, and the robot 230 goes to the entrance (execution of the first action triggered by the user's instruction). ). At that time, the robot 230 extracts the keyword “greeting” from the instruction of the mother 200 “come to pick up” and moves to the entrance 260 while saying “greeting, picking up” (the first action is the first action). Modification by action 2 (simultaneous imitation of speech)). When the child 220 is found on the way to the entrance 260, the robot 230 moves with respect to the child 220 (the second action (the imitation action of turning around the child) takes precedence over the first action (moves to the entrance)). Execution).

  Processing for performing the above sequence will be described with reference to flowcharts of the first behavior generation unit 140, the second behavior generation unit 150, and the behavior reconciliation unit 160 shown in FIGS. The first behavior generation unit 140 and the second behavior generation unit 150 each independently perform processing based on the interpretation result held by the interpretation unit 130. The first action is recorded in advance in a first action database (not shown) in the first action generation unit 140 as the minimum action that is useful for the user by itself. An example of this content is shown in FIG. The first behavior database includes information serving as a trigger and first behavior corresponding one-to-one to the information. If a trigger is detected from the interpretation result (S401), the first action corresponding to the trigger is selected from the first action database (S402), and the first action is generated (S403). The first action continues until the new trigger is detected (S404) or until the action ends (S405). If no trigger is detected (S404), the first action is not output.

  Next, the second action will be described. In the present embodiment, the second action is a simultaneous copying action. Simultaneous imitation behavior is behavior that is said to be seen in the early stages of infants in developmental psychology. In the present embodiment, the second action generation is performed based on the interpretation result of the user's action for the second action such as the moving action, the utterance action, the gaze action, and the home appliance operation action that can be presented as the action by the action-consolidating robot device. The imitation behavior recorded in advance in a second behavior database (not shown) in the unit 150 is selected (S501), and the behavior to be imitated is generated (S502). However, if the user does not perform an action that can be copied in the second action database, the user does not do anything). Further, when copying, in order to clarify the difference in physicality between the user and the action-integrated robot apparatus and clarify the action, deformation may be performed according to a predetermined method. For example, when imitating an utterance, instead of imitating the recognition result of the user's word as it is, the keyword of the recognized result is repeated twice. The imitation target is a familiar user (human) captured by the input unit.

  The behavior reconciliation unit 160 will be described. The behavior reconciliation unit 160 receives the behaviors output by the first behavior generation unit 140 and the second behavior generation unit 150, and determines the presence or absence of the first behavior (S601). If there is no first action, the second action is output to the action presentation unit 180 and executed (S602). If there is a first action, it is determined whether it can be executed simultaneously with the second action (S603). If it can be executed simultaneously with the second action (such as when the same actuator is not used), the second action is executed so as to modify the first action (S604). If simultaneous execution is impossible, the priority order of the first action and the second action is determined based on a predetermined priority order (S605). Here, when the priority of the first action is high, the first action is switched and executed (S606), and when the priority of the second action is high, the first action is changed to the second action. The action is suppressed and executed (S607).

  In this way, the configuration of the present embodiment selects and executes imitation actions that can be executed simultaneously, so that it is possible to give appropriate instructions that naturally match the robot's recognition ability, and simplify the tasks that one wants. Will be able to be obtained with simple instructions.

  As for the embodiment of the present invention, for example, FIG. 8 shows a configuration of a behavior-integrated robot apparatus using a robot that moves by a wheel and has a camera head with two degrees of freedom (pan / tilt) as a head. Show.

  As shown in FIG. 8, the robot 800 includes an input unit 810, a recognition unit 820, an interpretation unit 830, a first behavior generation unit 840, a second behavior generation unit 850, a behavior integration unit 860, a basic control processing unit 870, An action presentation unit 880 and an accumulation unit 890 are included. Since the configuration other than the storage unit 890 is the same as that of the first embodiment, the description of the same configuration as that of the first embodiment is omitted.

  FIG. 9 shows the detailed configuration of the storage unit, and FIG. 10 and FIG. 11 show the contents of the database (DB) inside the storage unit. As illustrated in FIG. 9, the accumulation unit 900 includes a separation unit 901, a time series DB 902, a feature amount extraction unit 903, a feature amount DB 904, a feature amount non-dimensionalization unit 905, a non-dimensionalization feature amount DB 906, and an action classification processing unit 907. The user action classification DB 908 is configured.

The time series data recognized by the recognition unit 930 is surrounded by a dotted line as shown in FIG. 10A as user actions (U ₁ , U ₂ ,..., U _n ) having a certain meaning. Each segment is segmented by the segmenting unit 901. As a method of carving, a method of carving at regular intervals, a method of standardizing each data and carving it at a large deviation, a method of carving data when there is audio data, etc. are used.

For example, by using data from a sensor that detects the user's position and separating the position change amount in a large deviation, it is possible to separate the user action when the user is walking and the user action when the user is stopped. . And the recognition result value of each input part is recorded with a certain sampling period, and is recorded on time series DB902. The input value is the indoor coordinate value (X, Y) if it is position information, image data if it is a camera, and audio data if it is sound. Moreover, what can be recognized at an abstract level such as a word such as voice data is also recorded as the word. Each data becomes time-series data x _i (t) (where i is an input parameter) for each sampling period t of the sensor (abstract data such as words is, for example, data at the time of utterance).

Next, in the feature quantity extraction unit 903, each element time-series information x ₁ (t), x ₂ (t),..., X _l () of the user behavior (U ₁ , U ₂ ,..., U _n ). A feature p _j (n) (j is a feature number and n is a user action number) each represented by a single numerical value is extracted. Features include the average value, maximum value, minimum value, standard deviation, etc. of the time series data x _i (t) itself, the change amount x _l ′ (t) calculated from the time series data x _l (t), and the acceleration. The average value, maximum value, minimum value, standard deviation, or the like of the quantity x _i ″ (t), or the time of the user action itself is used. For example, if data from a sensor that detects the user's position is used, this feature p _j (n) is the average moving speed or the maximum moving speed of the user during the user's action, or if data from a microphone is used, The maximum volume or the pitch of sound calculated from audio data. As shown in FIG. 10B, the feature value p _j (n) of each user action extracted by the feature value extraction unit 903 is stored in the feature value DB 904.

Next, the feature quantity non-dimensionalization unit 905 calculates a standard deviation S _j (n) in each user action of each feature as shown in Expression 1, and converts the standard deviation S _j (n) to Expression 2 below. As shown, dimensionless and feature parameter P _j (n) is calculated.

Then, the feature quantity non-dimensionalization unit 905 performs dimensionless processing for each user action as shown in FIG. 10C and stores the feature parameter P _j (n) in the dimensionless feature quantity DB 906. Also, it expressed in the form of a position vector u _n a multidimensional space in which the user behavior itself and its characteristic parameters and the axis as indicated by the number 3.

Then, the action classification processing unit 907 determines the similarity of each behavior from short and long distance position vector u _n indicating the user behavior, similar classified for each action _{(U a, U b, ···} ) and Do. For example, as shown in FIG. 11A, the position vectors u ₁ and u ₃ indicating the user behavior are clustered and determined as the user behavior U _a . Similarly, the position vectors u ₂ and u ₄ are determined as user actions U _b and U _c , respectively. Further, the classification result of FIG. 11A is stored in the user behavior classification DB 908. Corresponding to the interpretation result 921 (A1, A2,...) Interpreted by the interpretation unit 920, the classification (U _a , U _b ,...) Classified by the behavior classification processing unit 907 is regarded as an arrears imitation behavior. It accumulates in the second behavior database 911 in the second behavior generation unit 910. Overdue imitation behavior is behavior that imitates the user's past behavior. For example, if the user previously viewed the entrance when the chime rang, the information that the user saw the entrance was The separation unit 901 divides the user behavior into “chimes sounded” and “seen the direction of the entrance”, and the interpretation result (A1) is displayed in the second behavior database 911 as shown in FIG. : “Chime”) and overdue imitation behavior (U _a : “I saw the direction of the entrance”).

FIG. 12 shows a flowchart of the second action generation unit 910 in the present embodiment. In this embodiment, an arrears imitation action is output as the second action generation. An arrears imitation behavior classification closest to the current interpretation result is selected from the second behavior database (S1201), and an action within the classification is generated (S1202). For example, the second behavior database 911 is searched, U _a is selected from A1, and the behavior of U _a is generated in the second behavior generation unit.

  In this way, the configuration of the present embodiment imitates the user's past behavior from the interpretation result, so that the task desired by the user can be obtained with a simple instruction while using with familiarity. Become.

  Regarding the embodiment of the present invention, for example, the configuration of a behavior-integrated robot apparatus using a robot that moves with wheels and has a camera head with two degrees of freedom (pan / tilt) as a head is taken as an example. The configuration of this embodiment will be described with reference to FIG. Except for the configuration of the second action generation unit 1300, the second embodiment is the same as the second embodiment. The second behavior generating unit 1300 includes three behavior generating units, a simultaneous imitating behavior generating unit 1301, an arrears imitating behavior generating unit 1302, and a moving object following behavior generating unit 1303. The simultaneous imitating behavior generating unit 1301 is an example. The first second behavior generation unit and the overdue imitation behavior generation unit 1302 are the same as the second behavior generation unit of the second embodiment, and thus description thereof is omitted. Although the moving body following behavior generating unit 1303 imitates the user's behavior in the simultaneous copying behavior generating unit 1301, the moving body following behavior generating unit 1303 generates an action that follows the moving object.

  Next, the selection method of these three action production | generation parts is demonstrated using FIG. It is determined whether the user is at a position recognized by the robot apparatus (S1401). When the user is at a position recognized by the robot apparatus, it is determined whether or not the user has a movement with a large amount of information (S1402). If the user has a large amount of information, the simultaneous copying action generation unit 1301 is selected (S1403). The amount of information here is based on the idea of Shannon's amount of information. Specifically, as shown in FIG. 15, an action at a sparse position in the feature parameter space calculated in the second embodiment is considered to have a high amount of information, and is located at a dense position in the feature parameter space. Think of actions as low information content. If the user is not at or near the position recognized by the robot but is taking a normal action (small amount of information), it is determined whether the amount of information of the detected object is large (S1404). Here, if the detected object has a movement with a large amount of information, the moving body following action generation unit 1302 is selected (S1405), and if not, the overdue imitation action generation unit 1303 is selected (S1406).

  Further, FIG. 16 shows the relationship between these three types of behaviors of simultaneous imitation behavior, arrears imitation behavior, and moving body following behavior and the presentation medium. As explained in the first embodiment, for example, the simultaneous imitation behavior is jointly watched so that the robot sees the same direction together when the mother sees it. Further, when the interpretation unit interprets that a chime has been sounded, the robot performs an arrears imitation behavior such as looking at the entrance. When the mother moves, the robot also follows the moving body following the mother. In the figure, there is no action by the corresponding medium. The parentheses indicate the general meaning of the heartbeat.

  As described above, according to the configuration of the present embodiment, since it is selected from the three behavior forms of the simultaneous imitation behavior, the overdue imitation behavior, and the moving body following behavior, it is possible to perform an appropriate instruction according to the recognition ability of the robot naturally. It becomes possible to get the task that he wants with simple instructions.

  An embodiment of the present invention will be described by taking, for example, a configuration of a behavior-integrated robot apparatus using a humanoid having arms and legs similar to those of a person. As shown in FIG. 17, the robot 1700 includes an input unit 81711, a recognition unit 1720, an interpretation unit 1730, a first behavior generation unit 1740, a second behavior generation unit 1750, a behavior integration unit 1760, a basic control processing unit 1770, The action presentation unit 1780 is configured, and the recognition unit 1720 also recognizes information from the input unit 1715 outside the robot. Since the configuration other than the input unit 1715 outside the robot is the same as that of the first embodiment, the description thereof is omitted.

  As for the embodiments of Example 1, Example 2 and Example 3 described so far, cameras 1821 and 1822 installed in the environment, chime ON / OFF detection sensors, user 1800 as shown in FIG. And a position measuring device 1830 using RFID tags (1801, 1811) for detecting where the robot 1810 is located, and a floor sensor 1840 for detecting where the user or robot is located may also be provided. As a unit, a gesture function 1786 may be provided, and simultaneous imitation behavior, arrears imitation behavior, and moving object tracking behavior may be performed, respectively.

    As described above, according to the configuration of the present embodiment, since it is selected from the three behavior forms of the simultaneous imitation behavior, the overdue imitation behavior, and the moving body following behavior, it is possible to perform an appropriate instruction according to the recognition ability of the robot naturally. It becomes possible to get the task that he wants with simple instructions.

The block diagram which shows the structure of embodiment. The schematic diagram which shows embodiment. The sequence diagram in the 1st, 2nd action of embodiment. The flowchart of the 1st action production | generation part in embodiment. The flowchart of the 2nd action production | generation part in embodiment. The flowchart of the action integration part in embodiment. The figure which showed the content of the 1st action database in embodiment. The block diagram which shows the structure of other embodiment. The block diagram which shows the structure of the storage part in other embodiment. The figure which showed the content of the accumulation | storage database in other embodiment. The figure which showed the content of the accumulation | storage database in other embodiment. The flowchart of the 2nd action production | generation part in other embodiment. The block diagram which shows the structure of other embodiment The flowchart of the action selection of the 2nd action production | generation part in other embodiment. The schematic diagram which shows the information content in other embodiment. The figure which showed the 2nd action classification | category in other embodiment. The block diagram which shows the structure of other embodiment. The schematic diagram which shows other embodiment.

Explanation of symbols

DESCRIPTION OF SYMBOLS 100 Robot 110 Input part 111 Camera 112 Microphone 113 Axle encoder 114 Ultrasonic 120 Recognition part 130 Interpretation part 140 1st action generation part 150 2nd action generation part 160 Action integration part 170 Basic control process part 180 Action presentation part 181 Actuator Control processing section

Claims (7)

An input unit for obtaining information from the user and surroundings;
A recognition unit for recognizing a user's instruction and surroundings from information from the input unit;
An interpretation unit that combines and interprets information from the recognition result of the recognition unit;
A first action generation unit for generating an action for performing a task based on an interpretation result of the interpretation unit;
A second behavior generation unit that generates a simultaneous imitation behavior that imitates a user's current behavior based on the recognition result of the recognition unit;
An action reconciliation unit that reconciles the actions of the first action generation unit and the second action generation unit;
And a behavior presentation unit for presenting the behavior integrated by the behavior integration unit. An input unit for obtaining information from the user and surroundings;
A recognition unit for recognizing a user's instruction and surroundings from information from the input unit;
An interpretation unit that combines and interprets information from the recognition result of the recognition unit;
An accumulating unit for storing the recognition result of the recognizing unit;
A first action generation unit for generating an action for performing a task based on an interpretation result of the interpretation unit;
A second behavior generation unit that generates an overdue imitation behavior that imitates a user's past behavior based on the interpretation result of the interpretation unit and the accumulation information of the accumulation unit;
An action integration unit for integrating the actions of the first action generation unit and the second action generation unit;
And a behavior presentation unit for presenting the behavior integrated by the behavior integration unit. An input unit for obtaining information from the user and surroundings;
A recognition unit for recognizing a user's instruction and surroundings from information from the input unit;
An interpretation unit that combines and interprets information from the recognition result of the recognition unit;
A first action generation unit for generating an action for performing a task based on an interpretation result of the interpretation unit;
Based on the interpretation result of the interpretation unit, a second behavior generation unit that generates a moving body following behavior that follows a moving object;
An action integration unit for integrating the actions of the first action generation unit and the second action generation unit;
And a behavior presentation unit for presenting the behavior integrated by the behavior integration unit. An input unit for obtaining information from the user and surroundings;
A recognition unit for recognizing a user's instruction and surroundings from information from the input unit;
An interpretation unit that combines and interprets information from the recognition result of the recognition unit;
An accumulating unit for storing the recognition result of the recognizing unit;
A first action generation unit for generating an action for performing a task based on an interpretation result of the interpretation unit;
A simultaneous imitation action generation unit that generates a simultaneous imitation action that imitates a user's current action based on the recognition result of the recognition unit;
Based on the interpretation result of the interpretation unit and the accumulation information of the accumulation unit, an overdue imitation behavior generation unit that imitates the user's past behavior;
Based on the interpretation result of the interpreting unit, a moving object following action generating unit that generates a moving object following action that follows a moving object;
Based on the interpretation result of the interpreter and the accumulated information of the accumulator, the second behavior is selected by selecting any one of the simultaneous imitation behavior generation unit, the arrears imitation behavior generation unit, and the moving body following behavior generation unit. A second action generation unit to generate;
An action integration unit for integrating the actions of the first action generation unit and the second action generation unit;
And a behavior presentation unit for presenting the behavior integrated by the behavior integration unit. 5. The behavior integration unit integrates the behavior generated by the second behavior generation unit so as to modify the behavior generated by the first behavior generation unit. A behavior-integrated robot system. 5. The behavior integration unit according to claim 1, wherein the behavior integration unit switches and integrates the behavior generated by the first behavior generation unit and the behavior generated by the second behavior generation unit. Type robot device. The behavior integration unit is presented while one of the behaviors generated by the first behavior generation unit and the behavior generated by the second behavior generation unit are contradictory. 5. The action-integrated robot apparatus according to claim 1, wherein the action-compatible robot apparatus is integrated.

Images