JP2012040655A - Method for controlling robot, program, and robot - Google Patents

Abstract

PROBLEM TO BE SOLVED: To reduce unnaturalness in the autonomous operation of a robot.SOLUTION: In a method for controlling a robot, the robot having a movable part detects a dialogue request to the robot based on voice data or image data outputted from a sensor, and outputs detected data indicating the detected dialogue request after adding a degree of the confidence of the detection. The robot operates the movable part based on the detected data when the degree of confidence is higher than a predetermined level, and when the degree of confidence is not higher than the predetermined level, operates the movable part based on preset operation data.

Description

  The present invention relates to control of a robot capable of autonomous operation.

  Conventionally, in remote communication, there is a method for realizing rich communication by transmitting a physicality in addition to voice and video by controlling a robot from a remote place. In such a remote robot media type communication, it is said that it is important to look at the participants.

  In remote robot media type communication, as an example of a method of directing the line of sight to the participant, the robot's line of sight can be given to the participant by operating the operation screen (GUI) on the PC from a remote place with a mouse, joystick, or game pad. There is a way to turn. As another example, by measuring the human body movement using a camera, microphone, or sensor in a remote place, and sending a control command to the robot based on this measurement result, the movement of the remote person is controlled. There is a method to reproduce faithfully. As another example, there is a method of making the robot autonomously operate, such as detecting a participant's utterance and its direction by providing a plurality of microphones in the robot and moving the line of sight in the detected direction ( For example, Patent Document 1).

JP 2004-198656 A

  However, the above method has the following problems. For example, operating an operation screen on a PC from a remote location has a problem in that it prevents a person at a remote location from speaking or listening, or is troublesome and cannot be operated. In addition, the method of measuring the body motion of a remote person using a camera, microphone, or sensor requires the introduction of the camera or sensor at a remote place, which increases the cost. Further, in the method of controlling the robot's line of sight according to the detection of the voice direction, there is a problem that the line-of-sight movement becomes inappropriate when detection errors increase in a noisy environment.

  As described above, when a remote robot is introduced, it is not easy to cause the robot to direct the line of sight to the speaker. As described above, if the robot is not operated because it is troublesome, or if an inappropriate gaze movement occurs, the person (communication participant) will not trust the gaze movement of the robot, and further look at the robot. It also causes the problem of communication using voice alone.

  In view of these situations, it is required to reduce the unnaturalness of the autonomous operation of the robot in the remote robot media type communication.

  In the robot control method according to one aspect of the present invention, the robot having a movable part detects a dialogue request to the robot based on voice data or image data output from the sensor, and indicates the detected dialogue request. If the reliability including the detection reliability is output to the data and the reliability is higher than a predetermined level, the movable part is operated based on the detection data. If the reliability is not higher than the predetermined level, the operation data registered in advance The movable part is operated based on the above.

  In the present invention, the “part” does not simply mean a physical means, but includes a case where the function of the “part” is realized by software. Also, even if the functions of one “unit” or device are realized by two or more physical means or devices, the functions of two or more “units” or devices are realized by one physical means or device. May be.

  According to the present invention, unnaturalness in autonomous operation of a robot can be reduced.

It is a figure which shows the structure of the remote communication system which is one Embodiment of this invention. It is a figure which shows the external appearance of a robot. It is a figure which shows an example of the user interface for a teleworker. It is a figure which shows the structural example of a robot including the function which controls the operation | movement of a robot. It is a figure which shows an example of the user interface at the time of setting operation | movement data. It is a flowchart of the process which registers the position of a person or a target object into an action data recording part. It is a flowchart which shows an example of the operation control process of the robot by an operation control part.

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

== System configuration ==
FIG. 1 is a diagram showing a configuration of a remote communication system according to an embodiment of the present invention. The system includes a robot 10, a network 12, and a terminal 14.

  The robot 10 is connected to a remote terminal 14 via a network 12 such as the Internet or a dedicated line.

  FIG. 2 is a view showing the appearance of the robot 10. In addition, the robot 10 shown here is an example and may be a robot of another form. As shown in FIG. 2, the robot 10 includes a head actuator 20, a body actuator 22, a video camera 24, a microphone array 26, and a speaker 28.

  The head actuator 20 is a movable part for rotating the head of the robot 10 in the vertical direction and the horizontal direction. The body part actuator 22 is a movable part for operating the body part of the robot 10, and is constituted by two independently operable wheels arranged in parallel to the bottom part of the body part.

  The video camera 24 is a sensor that detects an image around the robot 10 and outputs image data. For example, the video camera 24 is provided in a portion corresponding to the eyes of the head of the robot 10 and is configured by a CCD, a CMOS element, or the like. . The microphone array 26 is a sensor that detects voice around the robot 10 and outputs voice data, and includes, for example, a plurality of microphones. The speaker 28 is for outputting sound to the participants of the conference, and is provided, for example, on the front surface of the body part of the robot 10.

  Returning to FIG. 1, the terminal 14 is an information processing apparatus that receives input of commands and settings for operating the robot 10 by a teleworker participating in the conference from a remote location, and is configured by a personal computer, a portable information terminal, and the like. The Moreover, the terminal 14 is provided with a microphone and collects the voice of the teleworker.

  An example of communication using such a remote communication system will be described. For example, commands accepted by the terminal 14 and collected voice are transmitted to the robot 10 disposed in the conference room via the network 12. The robot 10 operates the head actuator 20 and the body actuator 22 based on the received command. Further, the robot 10 outputs the received sound from the speaker 28. Further, the robot 10 collects audio on the conference room side with the microphone array 26 and captures an image on the conference room side with the video camera 24. The collected sound and the captured image are transmitted to the teleworker side terminal 14 via the network 12, and the teleworker can see and hear the conference room side image and sound.

  FIG. 3 is a diagram illustrating an example of a user interface for a teleworker displayed on the terminal 14. As shown in FIG. 3, a robot camera video display window 30, a robot control operation window 32, and a shared information display window 34 are displayed on the screen of the terminal 14.

  In the robot camera image display window 30, a video image captured by the video camera 24 of the robot 10 transmitted via the network 12 is displayed.

  The robot control operation window 32 is provided with various buttons for transmitting commands for controlling the robot 10 to the robot 10 via the network 12. For example, the teleworker issues a command for controlling the head actuator 20 or the body actuator 22 of the robot 10 or a command for controlling the volume of the speaker 28 by pressing a button displayed on the robot control operation window 32. Can be sent. The command that can be transmitted is not limited to the above, and other types of commands can be transmitted. In this manner, the teleworker can control the robot 10 connected via the network 12 by operating the robot control operation window 32. That is, the teleworker moves the body or face of the robot 10 toward a specific person participating in the conference, or shakes the head of the robot 10 toward the person participating in the conference. Instead of this, the robot 10 can make a gesture.

  In the shared information display window 34, information to be shared between the conference room participant and the teleworker is displayed. For example, information displayed on the projector screen on the conference room side can be transmitted to the terminal 14 via the network 12, and this information can be displayed on the shared information display window 34. In addition, the teleworker displays the information stored in the terminal 14 in the shared information display window 34, so that this information is transmitted to the conference room via the network 12 and displayed on the projector screen on the conference room. You can also make it.

  FIG. 4 is a diagram illustrating a configuration example of the robot 10 including a function for controlling the operation of the robot 10. In addition to the above-described head actuator 20, body actuator 22, video camera 24, microphone array 26, and speaker 28, the robot 10 includes an operation data recording unit 40, a setting reception unit 42, a function for controlling operations, The operation receiving unit 44, the dialogue request detecting unit 46, the operation control unit 48, the control result recording unit 50, and the parameter changing unit 52 are configured.

  The motion data recording unit 40 records motion data for moving movable parts such as the head actuator 20 and the body part actuator 22 of the robot 10 to a plurality of motion positions. For example, the motion data recording unit 40 records motion data for directing the head or body of the robot 10 to a position such as a person on the conference room side or a projector display. By using the operation data stored in the operation data recording unit 40, it is possible to point the head or body of the robot 10 in the direction of the desired person or the image displayed on the projector screen. That is, it is possible to cause the robot 10 to perform a body motion for transmitting an intention such as paying attention to a desired person or a display image.

  The setting receiving unit 42 receives the setting of operation data recorded in the operation data recording unit 40 from the terminal 14 via the network 12. FIG. 5 is a diagram illustrating an example of a user interface when the teleworker sets operation data using the terminal 14. In the example shown in FIG. 5, the conference room side image taken by the robot 10 is displayed in the display area 70. Then, for example, the teleworker moves the pointer 72 on the screen to the display position of the person's face or the object with the mouse, the touch pad or the like, and then presses the registration button 74. Thus, data indicating the position corresponding to the selected person or object, that is, operation data for directing the head or torso of the robot 10 in the direction of the selected person or object is the operation data recording unit. 40. The setting reception unit 42 can assign a person / object ID (identifier) for identifying each person or object automatically or by input from a teleworker. . Then, the setting reception unit 42 records this person / object ID in the motion data recording unit 40 in association with the motion data.

  The operation accepting unit 44 accepts various commands transmitted via the network 12 according to the operation of the teleworker.

  The dialogue request detection unit 46 detects a dialogue request to the robot 10 and includes a feature data recording unit 60, a detection unit 62, and a detection event output unit 64. For example, the dialog request detection unit 46 detects a dialog request indicating that a person around the robot 10 has spoken to the robot 10 and the arrival direction of the spoken voice. The dialogue request to the robot 10 is not limited to the case of speaking only to the robot 10 but also includes the case where a conference participant speaks to a plurality of participants including the robot 10.

  The feature data recording unit 60 can record feature data (for example, feature vectors) indicating the voice and image features of conference participants. This feature data can be generated, for example, by photographing the participant's face with the video camera 24 or acquiring the participant's voice with the microphone array 26 before the start of the conference. In this case, the feature data recording unit 60 can record the feature data in association with the participant's person / object ID. Also, feature data indicating general features of a person's voice and face, not a participant of the conference, may be recorded in the feature data recording unit 60.

  The detection unit 62 detects an interaction request to the robot 10 based on image data output from the video camera 24 and audio data output from the microphone array 26. In addition, the detection unit 62 can detect the arrival direction of the human voice by using the difference in the arrival time of the sound to the plurality of microphones constituting the microphone array 26. The detection unit 62 can detect the position and direction of a person whose line of sight is facing the robot 10 and the movement of the person's face based on the image data output from the video camera 24.

  Further, the detection unit 62 detects the dialogue request based on the image data output from the video camera 24, the audio data output from the microphone array 26, and the feature data recorded in the feature data recording unit 60. Calculate reliability. For example, the detection unit 62 uses a conventional pattern recognition technique (for example, linear discriminant analysis) to calculate the similarity between feature data generated from audio data or image data and feature data recorded in the feature data recording unit 60. And can be used as reliability. If the feature data is recorded in the feature data recording unit 60 in association with the person / object ID, the detection unit 62 displays the person / object indicating the person corresponding to the detected dialogue request. ID can be specified.

  The detection event output unit 64 outputs a detection event (detection data) indicating the dialogue request detected by the detection unit 62. The detection event includes, for example, the calculated reliability in addition to data indicating the content of the dialogue request such as the direction of voice arrival, the direction of the person whose line of sight is facing the robot 10, and the person / object ID. It is.

  The motion control unit 48 controls the motion of the robot 10 according to the motion data recorded in the motion data recording unit 40 or the detection event output from the dialogue request detection unit 46. For example, the motion control unit 48 reads motion data corresponding to a command transmitted via the network 12 from the motion data recording unit 40, and controls the motion of the robot 10 according to the motion data. In addition, when the operation control unit 48 receives a detection event indicating that the voice arrival direction has been detected from the dialogue request detection unit 46, if the reliability is higher than a predetermined threshold (predetermined level), the operation control unit 48 sets the detection event. The operation of the robot 10 is controlled accordingly. For example, the operation control unit 48 controls the head actuator 20 and the body part actuator 22 so that the head part or the body part of the robot 10 faces the direction in which the voice comes.

  In addition, the operation control unit 48 detects the dialogue request for a certain period of time or more when the dialogue request detection unit 46 does not output the detection event for a certain period of time or when the detection event is output. When the motion control according to the above is not performed, the motion data recorded in the motion data recording unit 40 is taken out and the motion of the robot 10 is controlled. By performing such control, it can be avoided that the robot 10 does not operate for a long time and becomes stationary, and the presence of the robot 10 can be expressed. In addition, when the reliability of the detected event is low, an operation corresponding to the detected dialog request is not executed, and therefore an unnatural operation of the robot 10 due to an erroneous detection of the dialog request can be avoided.

  The control result recording unit 50 records data indicating the results of operation control of the robot 10 performed by the operation control unit 48.

  The parameter changing unit 52 can change various parameters for operation control in the operation control unit 48. For example, the parameter changing unit 52 can change the parameter by receiving a command for changing the parameter output from the terminal 14 via the network 12. The parameters that can be changed by the parameter changing unit 52 include, for example, a threshold when the reliability of the detected event is determined by the operation control unit 48 and an interval at which the operation control (action) is performed by the operation control unit 48. It is. Note that the parameters that can be changed are not limited to these parameters, and all parameters that change the behavior of the operation control of the robot 10 such as the operation speed of the robot 10 are targeted.

  The motion data recording unit 40, the control result recording unit 50, and the feature data recording unit 60 can be configured using, for example, a memory included in the robot 10. In addition, the setting reception unit 42, the operation reception unit 44, the operation control unit 48, the parameter change unit 52, the detection unit 62, and the detection event output unit 64 include, for example, a program in which a processor included in the robot 10 is stored in a memory. This can be realized by executing.

== Control processing ==
Next, control processing of the robot 10 in the remote communication system will be described.

  FIG. 6 is a flowchart of processing for registering the position of a person or an object in the motion data recording unit 40. For example, when the registration screen illustrated in FIG. 5 is displayed on the terminal 14, the setting reception unit 42 starts this process. In the following, the $ location registration number and $ person / object ID are variables.

  First, the setting reception unit 42 initializes the $ position registration number and the $ person / object ID to, for example, 0 (zero) (S601). Then, the setting reception unit 42 waits for an event for the setting reception unit 42 to arrive. Note that events received by the setting reception unit 42 include an end request event and a registration button press event. The end request event is an event for ending the registration process, and is generated, for example, when the registration screen shown in FIG. Further, the registration button pressing event is generated in response to pressing of the registration button on the registration screen shown in FIG. The setting reception unit 42 checks whether an end request event has been received (S602), and if it has been received (S602: Y), the process ends. If no end request event has been received (S602: N), the setting reception unit 42 checks whether a registration button press event has been received (S603). If not received (S603: N), an end request is received. The process returns to the event check process (S602).

When the registration button press event has been received (S603: Y), the setting reception unit 42 selects from the image center coordinates based on the coordinates of the position selected by the cursor 72 transmitted via the network 12. A displacement amount (Δx, Δy) up to the coordinates of the determined position is calculated (S
604). Here, the image center coordinates indicate the positions in the direction in which the head of the robot 10 is facing (robot head x direction position, robot head y direction position).

Therefore, the setting reception unit 42 selects operation data indicating the position (x-direction position, y-direction position) of the person or the object corresponding to the current $ person / object ID selected by the cursor 72 (robot head). (Δx, Δy) is added to (part x direction position, robot head y direction position), and is registered in the motion data recording unit 40 in association with $ person / object ID (S605).

  Thereafter, the setting reception unit 42 adds 1 to the $ location registration number and the $ person / object ID (S606), and repeatedly executes the above-described processing (S602 to S606) until an end request event is received. Note that the $ location registration count can be used, for example, to limit the number of persons or objects that can be registered, or to control the display of the location registration screen shown in FIG. For example, if the value of the $ location registration number is “2”, the “registration 1” and “registration 2” buttons are not selected from the “registration 1” to “registration 4” buttons displayed on the screen. It is possible to select the “Register 3” and “Register 4” buttons. If the person / object ID is a serial number, the same control can be performed using the $ person / object ID without using the $ position registration count.

  FIG. 7 is a flowchart illustrating an example of operation control processing of the robot 10 by the operation control unit 48. For example, this process is started when the robot 10 is activated. In the following, $ previous action execution time and $ previous event check time are variables and are recorded in the control result recording unit 50. The $ action execution interval and the $ event check interval are constants. The $ action execution interval and the $ event check interval can be changed via the parameter changing unit 52.

  First, the operation control unit 48 initializes $ previous action execution time and $ previous event check time. Specifically, the operation control unit 48 sets a time before (current time− $ action execution interval) to the $ previous action execution time (S701) and sets the $ previous event check time (current time− $ event check interval). ) Set the previous time (S702).

  Subsequently, the operation control unit 48 confirms whether it is time to perform an event check. Specifically, the operation control unit 48 checks whether or not (current time− $ previous event check time)> $ event check interval (S703). While this condition is not satisfied (S703: N), the process of confirming the event check timing (S703) is repeated.

  When it is time to perform the event check (S703: Y), the operation control unit 48 sets the current time to the $ previous event check time (S704), and checks whether an event has arrived (S705). The events received by the operation control unit 48 include an end request event, a remote operation event, and a detection event. The end request event is an event for ending the control process, and is generated, for example, by instructing the robot 10 to shut down via the network 12 or by a direct operation. The remote operation event is generated when the teleworker operates the operation window 32 for robot control. The detection event is generated when the dialog request detecting unit 46 detects a dialog request.

  If the event has arrived (S705: Y), the operation control unit 48 checks whether the event is an end request event (S706), and if it is an end request event (S706: Y), the process ends. If it is not an end request event (S706: N), the operation control unit 48 checks whether it is a remote operation event (S707). If the event is not a remote operation event (S707: N), the received event is a detection event, and the operation control unit 48 confirms whether the reliability included in the detection event is higher than a predetermined threshold (S708). . If the reliability of the detected event is not higher than the predetermined threshold (S708: N), the process returns to the process for checking the event check timing (S703).

  When the operation control unit 48 receives a remote operation event (S707: Y), or when it receives a detection event whose reliability is higher than a predetermined threshold (S708: Y), depending on the content of the event, Actions such as actuator control of the robot 10 are executed (S709). Specifically, the operation control unit 48 controls the head actuator 20 and the body actuator 22 of the robot 10 according to the contents of the event, adjusts the zoom of the video camera 24, and configures the microphone array 26. The sensitivity of the microphone to be adjusted is adjusted, and the volume of the speaker 28 is adjusted.

  For example, when the detection event includes the direction of voice arrival or the position or direction of a person whose line of sight is facing the robot 10, the operation control unit 48 makes the head of the robot 10 face the position or direction. The head actuator 20 and the body actuator 22 are controlled. Also, for example, when the person / object ID is included in the detection event, the operation control unit 48 acquires the position corresponding to the person / object ID from the operation data recording unit 40, and the head of the robot 10 The head actuator 20 and the body actuator 22 are controlled so as to face the person or the object corresponding to the person / object ID.

  Thereafter, the operation control unit 48 sets the current time as the $ previous action execution time (S710), and returns to the process of checking the event check timing (S703).

  When the arrival of the event is checked (S705), if the event has not arrived (S705: N), the operation control unit 48 checks whether it is time to perform some action. Specifically, the operation control unit 48 checks whether (current time− $ previous action execution time)> $ action execution interval is satisfied (S711). If this condition is not satisfied, the process returns to the process of checking the event check timing (S703).

  When it is time to perform some action (S711: Y), the motion control unit 48 randomly selects the person / object ID recorded in the motion data recording unit 40 (S712). Then, the motion control unit 48 acquires the position corresponding to the selected person / object ID from the motion data recording unit 40, and the head of the robot 10 is the person or object corresponding to the person / object ID. The head actuator 20 and the body actuator 22 are controlled so as to face the direction (S713).

  Thereafter, the operation control unit 48 sets the current time as the $ previous action execution time (S710), and returns to the process of checking the event check timing (S703).

  Here, in the process of FIG. 7, when the reliability of the detected event is not higher than the threshold, if the detected event with the higher reliability than the threshold does not arrive by the next action execution timing, it is recorded in the operation data recording unit 40. The robot 10 is controlled based on the motion data that has been recorded. That is, when the reliability of the detected event is not higher than the threshold value, as a result, the head actuator 20 and the body actuator 22 of the robot 10 are controlled based on the motion data recorded in the motion data recording unit 40. It becomes. In the process of FIG. 7, if the reliability of the detected event is not higher than the threshold (S708: N), the process is recorded in the operation data recording unit 40 without returning to the process of checking the event check interval (S703). It is good also as changing directly to the process (S711-S712) based on the operation data which exists.

  According to the above process, when the reliability included in the detection event is higher than the predetermined threshold, the operation of the robot 10 is controlled according to the detection event. Further, when the action corresponding to the detection event or the remote operation event is not executed for a predetermined time or longer, the operation of the robot 10 is controlled based on the operation data recorded in the operation data recording unit 40.

  As a result, when the reliability of the detected event is low, an operation corresponding to the detected dialogue request is not executed, so that an unnatural operation of the robot 10 due to an erroneous detection of the dialogue request can be avoided. Furthermore, it is possible to avoid the robot 10 from moving stationary for a long time, and to express the presence of the robot 10. In other words, the combination of the responsive autonomous motion control based on the detected event and the routine and stable autonomous motion control based on the motion data recorded in the motion data recording unit 40 is effective in the autonomous motion of the robot 10. Naturalness can be reduced.

  In the present embodiment, the parameter changing unit 52 can adjust various parameters of the operation control unit 48 to reduce unnaturalness in the autonomous operation of the robot 10. For example, when there is a lot of noise and the recognition event is unreliable, the operation of the robot 10 corresponding to the erroneous detection event can be reduced by setting a high threshold when determining the reliability of the detection event. . Further, for example, by periodically changing the action execution interval, it is possible to cause fluctuation in the operation of the robot 10 and realize a more natural robot operation. The parameter changing unit 52 may change the parameter based on a statistical value of reliability of the detected event or a predetermined rule without receiving a command for changing the parameter via the network 12.

  Note that this embodiment is intended to facilitate understanding of the present invention and is not intended to limit the present invention. The present invention can be changed / improved without departing from the spirit thereof, and the present invention includes equivalents thereof.

  For example, the motion data recording unit 40 can record motion data for operating the robot 10 in accordance with a specific motion (gesture) of a person or an object. In this case, the motion data can be recorded in the motion data recording unit 40 in association with the motion identifier for identifying the motion. When the detection unit 62 detects a predetermined operation based on the image data output from the video camera 24, the detection event output unit 64 outputs the operation identifier corresponding to the detected operation in the detection data. The motion control unit 48 can acquire motion data corresponding to the motion identifier included in the detection data from the motion data recording unit 40, and can control the motion of the robot 10 based on the acquired motion data.

  Further, for example, in the present embodiment, the number of operations according to the operation data recorded in the operation data recording unit 40 and the number of operations according to the detection event are controlled by monitoring the action execution interval. The method for controlling the number of operations is not limited to this. For example, the operation control unit 48 controls the number of executions per unit time of the operation according to the operation data recorded in the operation data recording unit 40 and the number of executions per unit time of the operation according to the detected event. By recording in the recording unit 50, the total number of operations of the robot 10 per unit time can be controlled to be substantially constant. For example, when the number of executions per unit time of the operation according to the detected event is large, the operation control unit 48 sets the number of executions per unit time of the operation according to the operation data recorded in the operation data recording unit 40. On the contrary, when the number of executions per unit time of the operation according to the detection event is small, the number of executions per unit time of the operation according to the operation data recorded in the operation data recording unit 40 is increased. It is good.

Part or all of the above-described embodiments can be described as in the following supplementary notes, but is not limited thereto.

  (Additional remark 1) It is a control method of the robot which has a movable part, Comprising: Based on the audio | voice data or image data output from a sensor, the interaction request | requirement with respect to the said robot is detected, and the detected data which shows the detected said interaction request | requirement are used. If the reliability is higher than a predetermined level, the movable part is operated based on the detection data. If the reliability is not higher than the predetermined level, it is registered in advance. A control method for a robot, wherein the movable unit is operated based on the obtained operation data.

  (Supplementary note 2) The robot control method according to supplementary note 1, wherein the robot includes a motion data recording unit that records a plurality of motion data corresponding to a plurality of motion positions of the movable portion, and When the degree is not higher than the predetermined level, the robot control method of operating the movable unit based on at least one operation data among the plurality of operation data recorded in the operation data recording unit.

  (Supplementary note 3) The robot control method according to supplementary note 1 or 2, wherein, according to a decrease in the number of operations per unit time of the movable unit based on the detection data, the movable unit based on the operation data A robot control method that increases the number of operations per unit time.

  (Supplementary note 4) The robot control method according to any one of supplementary notes 1 to 3, wherein a voice arrival direction is detected based on the voice data output from the sensor, and the detection data includes the detection data. A method for controlling a robot, which outputs the reliability and the arrival direction of the voice, and operates the movable unit in the arrival direction of the voice included in the detection data when the reliability is higher than the predetermined level.

  (Supplementary note 5) The robot control method according to any one of supplementary notes 1 to 4, wherein the position or direction of a person whose line of sight is directed to the robot based on the image data output from the sensor And outputs the detection data including the reliability and the direction of the person. When the reliability is higher than the predetermined level, the movable unit is operated in the direction of the person included in the detection data. Let the robot control method.

  (Supplementary note 6) The robot control method according to any one of supplementary notes 2 to 5, wherein the motion data is recorded in the motion data recording unit in association with a motion identifier for identifying a motion with respect to the robot. And detecting the dialogue request by a predetermined action based on the image data output from the sensor, and outputting the detection data including the reliability and the action identifier corresponding to the detected predetermined action. When the reliability is higher than the predetermined level, the movable unit is operated based on the operation data corresponding to the operation identifier included in the detection data, which is recorded in the operation data recording unit. , Robot control method.

  (Supplementary note 7) The robot control method according to any one of supplementary notes 1 to 6, wherein the robot includes a feature data recording unit in which feature data indicating a feature of a sound or an image is recorded, and the sensor A robot control method for calculating the reliability based on a similarity between the audio data or the image data output from the image data and the feature data recorded in the feature data recording unit.

  (Supplementary note 8) The robot control method according to any one of supplementary notes 2 to 6, wherein the robot has a plurality of feature data indicating voice characteristics of a plurality of persons, A feature data recording unit that is recorded in association with each other, wherein the motion data is recorded in the motion data recording unit in association with the person identifier, and is output from the sensor; and the feature data Based on the feature data recorded in the recording unit, the person identifier corresponding to the audio data is detected, and the detection data is output including the reliability and the detected person identifier, When the reliability is higher than the predetermined level, based on the motion data corresponding to the person identifier included in the detection data, which is recorded in the motion data recording unit Operating the movable part, the control method of the robot.

  (Supplementary note 9) A function of detecting a dialogue request to the robot based on voice data or image data output from a sensor to a robot having a movable part, and detection data indicating the detected dialogue request A function for outputting including reliability, a function for operating the movable part based on the detection data when the reliability is higher than a predetermined level, and a function when the reliability is not higher than the predetermined level. A program for realizing a function of operating the movable part based on registered operation data.

  (Supplementary Note 10) A movable part, a sensor that detects voice or image and outputs voice data or image data, a dialog request detection part that detects a dialog request based on the voice data or the image data, and the detection A detection data output unit for outputting the detection data including the reliability of the detection to the detection data indicating the interaction request, and if the reliability is higher than a predetermined level, the movable unit is operated based on the detection data, And a motion control unit that operates the movable unit based on motion data registered in advance when the reliability is not higher than the predetermined level.

DESCRIPTION OF SYMBOLS 10 Robot 12 Network 14 Terminal 20 Head actuator 22 Body part actuator 24 Video camera 26 Microphone array 28 Speaker 30 Robot camera image display window 32 Robot control operation window 34 Shared information display window 40 Operation data recording part 42 Setting reception part 44 Operation Reception unit 46 Dialog request detection unit 48 Operation control unit 50 Control result recording unit 52 Parameter change unit

Claims (10)

A control method for a robot having a movable part,
Based on voice data or image data output from the sensor, detects a dialogue request to the robot,
Outputting detection data indicating the detected dialogue request including reliability of the detection;
When the reliability is higher than a predetermined level, the movable unit is operated based on the detection data,
When the reliability is not higher than the predetermined level, the movable unit is operated based on operation data registered in advance.
Robot control method. The robot control method according to claim 1, comprising:
The robot includes an operation data recording unit that records a plurality of the operation data corresponding to a plurality of operation positions of the movable unit,
When the reliability is not higher than the predetermined level, the movable unit is operated based on at least one operation data among the plurality of operation data recorded in the operation data recording unit.
Robot control method. The robot control method according to claim 1, wherein:
In response to a decrease in the number of operations per unit time of the movable unit based on the detection data, the number of operations per unit time of the movable unit based on the operation data is increased.
Robot control method. It is a control method of the robot according to any one of claims 1 to 3,
Based on the voice data output from the sensor, the direction of voice arrival is detected,
Outputting the detection data including the reliability and the direction of arrival of the voice;
When the reliability is higher than the predetermined level, the movable unit is operated in the direction of arrival of the voice included in the detection data.
Robot control method. The robot control method according to any one of claims 1 to 4,
Based on the image data output from the sensor, detects the position or direction of a person whose line of sight is facing the robot,
Outputting the detection data including the reliability and the direction of the person,
When the reliability is higher than the predetermined level, the movable unit is operated in the direction of the person included in the detection data.
Robot control method. A method for controlling a robot according to any one of claims 2 to 5,
The motion data is recorded in the motion data recording unit in association with a motion identifier that identifies a motion with respect to the robot,
Based on the image data output from the sensor, detect the dialogue request by a predetermined operation,
The detection data is output including the reliability and the action identifier corresponding to the detected predetermined action,
When the reliability is higher than the predetermined level, the movable unit is operated based on the operation data corresponding to the operation identifier included in the detection data, which is recorded in the operation data recording unit.
Robot control method. A robot control method according to any one of claims 1 to 6,
The robot includes a feature data recording unit in which feature data indicating features of voice or image is recorded,
Calculating the reliability based on the similarity between the audio data or the image data output from the sensor and the feature data recorded in the feature data recording unit;
Robot control method. A robot control method according to any one of claims 2 to 6, comprising:
The robot includes a feature data recording unit that records a plurality of feature data indicating voice characteristics of a plurality of people in association with a person identifier that identifies the person,
The motion data is recorded in the motion data recording unit in association with the person identifier,
Based on the voice data output from the sensor and the feature data recorded in the feature data recording unit, the person identifier corresponding to the voice data is detected,
Outputting the detection data including the reliability and the detected person identifier;
When the reliability is higher than the predetermined level, the movable unit is operated based on the operation data corresponding to the person identifier included in the detection data, which is recorded in the operation data recording unit.
Robot control method. For robots with moving parts,
A function of detecting a dialogue request to the robot based on voice data or image data output from a sensor;
A function of outputting detection data indicating the detected dialogue request including reliability of the detection;
When the reliability is higher than a predetermined level, a function of operating the movable part based on the detection data;
When the reliability is not higher than the predetermined level, a function of operating the movable unit based on pre-registered operation data;
A program to realize Moving parts;
A sensor that detects sound or image and outputs sound data or image data;
A dialog request detection unit that detects a dialog request based on the audio data or the image data;
A detection data output unit for outputting the detection data indicating the detected dialogue request including the reliability of the detection;
When the reliability is higher than a predetermined level, the movable unit is operated based on the detection data, and when the reliability is not higher than the predetermined level, the movable unit is operated based on pre-registered operation data. An operation control unit to be operated;
Robot equipped with.