JP2005131713A - Communication robot - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To provide a communication robot capable of performing behavior according to the past behavior of an interactive mate. <P>SOLUTION: A surrounding state is observed by a surrounding state observing device 16 arranged in an environment; a state from a view point of a human being is observed by a human being observing device 18 installed on the human being P2; and a state from a view point of the robot is observed by a robot body 12. An infrared tag 14 is installed on an exhibit B1 and the human being observing device 18, to thereby acquire data including identification information in the respective observing devices. Thus, interaction between human beings and exhibits in such an environment is observed, and a behavior history of the human being is successively recorded on a database. The past behavior data on the interactive mate is analyzed; behavior such as a guide and recommendation to be presented to the interactive mate is determined on the basis of its state; and the robot body 12 performs the behavior. This robot can present the behavior corresponding to the past behavior of the interactive mate. <P>COPYRIGHT: (C)2005,JPO&NCIPI

Description

  The present invention relates to a communication robot, and more particularly to a communication robot that observes behaviors to be interacted with and communicates using body movements such as gestures and voices.

An example of this type of conventional communication robot is disclosed in, for example, Patent Document 1 by the present applicant. In the robot of Patent Document 1, for example, an action is performed to guide a conversation partner to view an exhibit such as a poster using gestures and voices.
US Pat. No. 6,604,021

  In the prior art of Patent Document 1, guidance can be given to a conversation partner, but only a uniquely programmed guide method is executed regardless of past actions of the conversation partner.

  Therefore, a main object of the present invention is to provide a communication robot that can execute an action corresponding to a past action of a conversation partner.

  The invention according to claim 1 is a communication robot that observes and communicates with the behavior of the dialogue target, includes a sensor that acquires the behavior history of the dialogue target, a recording unit that records the behavior history acquired by the sensor, and a dialogue target. A partner determination unit that determines a conversation partner, a behavior determination unit that determines a behavior of the robot based on a behavior history regarding the conversation partner recorded in the recording unit, and an execution unit that executes the behavior determined by the behavior determination unit , A communication robot.

  According to the first aspect of the present invention, communication is performed by observing the behavior of the robot to be interacted with, such as a human in a predetermined space. The sensor acquires the action history of the conversation target, and the acquired action history is recorded by the recording unit. The partner determination unit determines a conversation partner from among the conversation targets, and the behavior determination unit determines a behavior to be presented by the robot based on the behavior history regarding the conversation partner. And an execution means performs the determined action with respect to the dialogue partner. Therefore, according to invention of Claim 1, the action according to the past action can be shown with respect to a dialogue partner.

  The invention according to claim 2 is a communication robot according to claim 1, wherein the sensor acquires at least one of identification information for identifying a conversation target and identification information for identifying a target object of the interaction of the conversation target. To do.

  According to the second aspect of the present invention, identification information for identifying a dialogue target and identification for identifying an interaction or action target (for example, an exhibit, a robot body, a stuffed animal, etc. in the embodiment) by a sensor. Acquire at least one of the information. In the case of acquiring identification information for identifying a conversation target, the sensor is provided in a place other than the conversation target, for example, in another conversation target, a target object, or a structure constituting an environment. Therefore, when the identification information of the dialogue target is acquired, the dialogue target to which the identification information is assigned may interact with another dialogue target, a target, or a structure provided with the sensor. Be grasped. Moreover, when acquiring the identification information for identifying a target object, the sensor is provided in places other than the target object, for example, a dialog target, another target object, or a structure constituting an environment. Therefore, when the identification information of the object is acquired, it is understood that the conversation object with the sensor is interacting with the object to which the identification information is attached. The recording means records at least one of an action history described by identification information that can identify a conversation partner and an action history described by identification information that can identify an object. When the identification information of the conversation target is acquired, the behavior determination unit can easily identify and refer to the behavior history related to the conversation partner from the behavior history of the recording unit based on the identification information of the conversation partner. Moreover, when the identification information of a target object is acquired, the action history acquired by the sensor provided in the conversation partner is the action history regarding the conversation partner. As described above, by acquiring the identification information of at least one of the conversation object and the object by the sensor, the action determination unit can easily determine the action for the conversation partner.

  The invention according to claim 3 is a communication robot according to claim 2, wherein the action determining means analyzes the state of the action history related to the conversation partner based on the positional relationship between the conversation partner and the object of interaction, and the analysis Determine actions based on results.

  In the invention according to claim 3, the behavior determining means analyzes the behavior history data related to the conversation partner based on the positional relationship between the conversation partner and the object of the interaction. The positional relationship between the conversation partner and the interaction object can be grasped by the action history described by the identification information of the interaction object, the action history described by the identification information of the interaction object, or a combination thereof. . By this analysis, it is possible to grasp the state of the action history related to the conversation partner. Therefore, it is possible to easily determine the action to be presented based on the grasped past action state.

  The invention according to claim 4 is a communication robot according to any one of claims 1 to 3, further comprising rule storage means for storing a rule defining an action to be presented by the robot corresponding to a past action. The action determining means determines the action based on the action history and rules regarding the conversation partner.

  In the invention of claim 4, the rule storing means stores a rule that defines the action to be presented by the robot corresponding to the past action. Then, the behavior determining means determines the behavior to be presented based on the behavior history regarding the conversation partner and the rules of the rule storage means. Therefore, it is possible to easily determine the action to be presented to the conversation partner by comparing the past action with the rule.

  The invention according to claim 5 is a communication robot according to any one of claims 1 to 4, wherein the action determined by the action determining means is an operation for performing guidance or recommendation related to the interacted object of the conversation partner. Including.

  In the invention of claim 5, since the operation of performing guidance or recommendation related to the interacted object of the conversation partner is executed, the conversation partner can supplement or deepen the knowledge obtained by the interaction, which is beneficial. Information can be obtained.

  According to this invention, since the action to be presented to the conversation partner is determined based on the past action data of the conversation partner, the action according to the past action of the conversation partner can be presented.

  The above object, other objects, features and advantages of the present invention will become more apparent from the following detailed description of embodiments with reference to the drawings.

  Referring to FIG. 1, a communication robot (hereinafter, also simply referred to as “robot”) 10 of this embodiment observes the behavior of one or more persons who are conversation objects and communicates. Specifically, the robot 10 is for presenting an action corresponding to a past action of the person to a person who is a conversation partner in a predetermined space. The robot main body 12 which presents an action using is included. In this embodiment, as an example, it is assumed that guidance is provided to visitors P1, P2, P3, etc. in an exhibition hall where various exhibits are arranged.

  The robot 10 relates to an interaction between at least one person who is an object of interaction of the robot body 12 in the space and an object (for example, an exhibit or the robot body 12) that is an object of interaction (interaction) of the person. Data, that is, data relating to the behavior of the dialogue target is sequentially recorded and accumulated. Then, based on the accumulated action history, the action to be presented to the conversation partner is determined, and the robot body 12 performs, for example, guidance or recommendation information, information on the interacted object, etc. Present by (action). For this reason, various sensors for detecting the interaction of a visitor or the like who is the subject of dialogue are provided in the exhibition hall, and an infrared tag 14 is attached to the detection target in this embodiment.

  For example, a plurality of surrounding state observation devices 16 are installed at predetermined positions such as a ceiling or a wall, which is a structure constituting the environment, and an interaction around each installation location is observed. In addition, visitors P2 and P3, the instructor Q1, and the like are each equipped with a human observation device 18, and the interaction from the viewpoint or position of each person is observed. In addition, one or more stuffed animal observation devices 20 are also arranged at predetermined locations in the environment, and the interaction from the viewpoint or position of the stuffed animal is observed. The robot body 12 also functions as a device for observing the interaction from the viewpoint or position.

  The infrared tag 14 is attached to the object itself, the vicinity thereof, or a ceiling or wall in the environment. In the example of FIG. 1, the display board or posters B1, B2, B3, B4... And the description computers M1, M2, M3. The robot body 12 is attached to, for example, the chest. Each infrared tag 14 transmits identification information (ID number or the like) for identifying the object by blinking infrared rays. Similarly, an infrared tag (not shown) is attached to the human observation device 18 attached to the visitors P2 and P3 and the instructor Q1, and the objects of interaction such as the visitor and the instructor are identified. Identification information is transmitted. Thus, the interaction data observed and recorded includes data described by the identification information of the target object and data described by the identification information of the dialogue target.

  The infrared tag attached to the infrared tag 14 and the human observation device 18 includes an infrared LED, a drive circuit, an internal battery, and the like, and an ID number uniquely assigned to an object to which the infrared tag is attached is identified. Control the blinking of the infrared LED as possible. As the infrared LED, for example, a high-power light-emitting diode for optical communication (DN311 manufactured by Stanley Co., Ltd.) or the like is applied, and those having a low directivity and about 800 nm close to visible light can be suitably used. The drive circuit is composed of a microcomputer or the like, and for example, a 4 MHz drive microcomputer AT90S2323 manufactured by Atmel can be used. Specifically, the drive circuit repeatedly transmits an ID number (6 bits) and a parity bit encoded by the Manchester encoding method, a start bit (1 bit), and an end bit (2 bits) by blinking at a cycle of 200 Hz.

  FIG. 2 shows an example of the configuration of the robot 10. Referring to FIG. 2, the robot 10 includes an ambient condition observation device 16, a human observation device 18, a stuffed animal observation device 20, a data acquisition server (processing computer) 22, a robot body 12, an interaction database 24, and a robot operation database. 26, a guide rule database 28, and the like. In FIG. 2, for ease of illustration, only one ambient condition observation device 16, human observation device 18, stuffed animal observation device 20, and robot body 12 are shown. It is needless to say that 16 is provided at one or a plurality of observation positions, the human observation device 18 is provided, for example, for each instructor and visitor, and the stuffed animal observation device 20 and the robot body 12 are provided as many as necessary.

  The ambient condition observation device 16 includes, for example, an infrared camera 30, a CCD camera 32, a microphone 34, a stationary computer 36, and the like. The surrounding state observation device 16 observes the interaction in the surroundings, and acquires identification information and position information of the infrared tag 14 and the like within the detection range, and also acquires video data, audio data, and the like.

  The infrared camera 30 is for detecting infrared rays emitted from the infrared tag 14 located in the photographing range or the infrared tag attached to the human observation device 18, and is installed in the stationary computer via the image processing device 38. 36. The infrared camera 30 includes, for example, an infrared filter, a lens, a CMOS image sensor, and the like.

  The infrared filter mainly transmits only near infrared rays, and the lens images the near infrared rays on the CMOS image sensor. As the infrared filter, for example, a plastic IR pass filter manufactured by Edmond that blocks visible light and passes near infrared light can be used. The angle of view of the lens is, for example, about 60 degrees. In this case, the light collection rate per pixel of the CMOS image sensor is high, and the infrared tag 14 and the like located at a long distance can be easily found.

  The CMOS image sensor captures a near-infrared image composed of the near-infrared image formed by the lens and outputs it to the image processing device 38. As the CMOS image sensor, for example, an artificial retina LSI (M64283FP) manufactured by Mitsubishi Electric Corporation or the like can be used, and the resolution in this case is 128 × 128 pixels.

  The image processing device 38 performs control of the CMOS image sensor, data processing, and the like. For example, an infrared tag is detected from a near-infrared image, an ID number is detected from the blinking state of the detected infrared tag, and an XY coordinate of the infrared tag on the infrared image is detected. Then, the detected ID number and data such as XY coordinates are output to the stationary computer 36 in accordance with a data transmission standard such as RS232C. As the image processing device 38, for example, a 49 MHz drive microcomputer C8051F124 manufactured by Cygnal can be used.

  In this case, the CMOS image sensor is driven with a clock of 115200 Hz, and after imaging (shutter opening), the brightness of 1 pixel is serially output as an analog value every clock. For this reason, the shortest frame rate at the time of photographing all pixels is (shutter speed) + (128 × 128 × clock speed). For example, 8 × 8 pixels out of 128 × 128 pixels are set as the detection area, and the shutter speed is 500 Hz. When the image is picked up with a frame rate of 400 Hz, a frame rate of 400 Hz can be realized, and the reading speed can be increased. As described above, since reading is performed at a frame rate (400 Hz) that is twice the blinking cycle (200 Hz) of the infrared tag 14 or the like, asynchronous communication can be performed using a single LED.

  Here, the infrared tag detection process will be described. This infrared tag detection processing is performed by executing a detection processing program stored in advance in the image processing device 38, and the same processing is performed in the human observation device 18 and the robot body 12 described later.

  First, the image processing device 38 initializes an infrared camera (CMOS image sensor) 30 and captures an infrared image of a full screen (for example, 128 × 128 pixels). Next, the image processing device 38 extracts a light spot of a predetermined size from the infrared image, for example, a light spot of 1 pixel as an infrared tag, and eliminates a light spot larger than the predetermined size. Thus, since the infrared tag can be detected by a simple process of detecting a light spot of a predetermined size from the infrared image, the infrared tag process can be speeded up.

  Next, the image processing device 38 determines, for example, an 8 × 8 pixel area centered on the extracted light spot as a detection area, and sets the detection area by a CMOS image sensor a predetermined number of times, for example, (((transmission bit number + start bit). (Number + number of end bits) × 2 × 2) times, and the ID number is detected by detecting the blinking state of the infrared tag from the read infrared image, the parity check is performed, and the determination process of the read data is performed.

  As described above, the detection region including the light spot is determined from the infrared image, and the blinking state of the infrared tag is detected using only the infrared image of the detection region, so that the infrared image to be processed is minimized. The infrared tag detection process can be speeded up. As a result, it is possible to sufficiently follow the movement of a person, and processing with high calculation costs such as motion prediction can be omitted. If the parity check is correct, the image processing device 38 outputs the ID number and XY coordinates of the infrared tag. If the parity check is not correct, the image processing device 38 reads the detection area again and detects the above-described infrared detection processing. For all light spots.

  The CCD camera 32 is for taking an image of surrounding visible light, includes a lens (not shown), etc., and takes a visible light image and outputs a video signal to the stationary computer 36. As the CCD camera 32, for example, a small CCD camera (horizontal angle of view 44 degrees) manufactured by Keyence Corporation having an analog video output can be used. Here, the optical axis of the lens is aligned with the optical axis of the lens of the infrared camera 30, and therefore, not only the object located in the axial direction within the imaging range is identified, but also a visible light image of the object. Can also shoot at the same time.

  The microphone 34 is used to detect ambient sounds, and an omnidirectional one is applied and connected to the stationary computer 36 via the sound processing circuit 40. The microphone 34 collects sound and outputs it to the sound processing circuit 40, and the sound processing circuit 40 outputs the recorded sound signal to the stationary computer 36.

  The stationary computer 36 specifies time information for specifying the observation time for each data such as the input ID number of the infrared tag 14 or the like, the XY coordinates in the infrared image of the infrared tag 14, the visible light image, or the sound. And the like, and each information is transmitted to the data acquisition server 22. By giving this time information, the time synchronization of each interaction data observed in a distributed manner by each observation device 16, 18, 20, 12 and the like is taken. As a method of synchronizing in time, for example, all devices perform an observation operation in accordance with the world clock, or each device has an internal clock independently and is transmitted from the data acquisition server 22, for example. The offset may be adjusted based on the synchronization signal. The stationary computer 36 may be connected to the data acquisition server 22 by radio.

  The human observation device 18 includes, for example, an infrared camera 42, a CCD camera 44, a microphone 46, a biological sensor 48, a portable computer 50, and the like. As can be seen from FIG. 1, the human observation device 18 is configured as an ear-necked neckband headset, for example, and is attached to the heads of the explanation staff Q1... And visitors P2, P3. For example, the infrared camera 42 and the CCD camera 44 are integrated in a rectangular parallelepiped housing so that the camera axis is directed in the direction of the line of sight of the person, and the microphone 46 is disposed near the mouth of the user. An infrared tag for identifying a person to whom the human observation device 18 is attached is integrally fixed to a side surface of the casing. The biosensor 48 is attached to a finger or the like of an instructor or a visitor, and the portable computer 50 is used on the back of the instructor or a visitor. The human observation device 18 observes the interaction from the human viewpoint and the like, and the identification information and position information of the infrared tag 14 and the like within the detection range are acquired, and the video data and audio data and the device 18 are obtained. Biometric data and the like of the person wearing the are acquired.

  The infrared camera 42 is connected to the portable computer 50 via the image processing device 52. The infrared camera 42 and the image processing device 52 are configured and operate in the same manner as the infrared camera 30 and the image processing device 38 of the ambient condition observation device 16 described above. However, the angle of view of the lens is set to about 90 degrees, for example, and in this case, the infrared tag 14 and the like located in a wide range at a relatively short distance can be easily detected in a face-to-face conversation state or the like. The infrared camera 42 and the like detect the ID number and XY coordinates of the infrared tag 14 and the like existing within the imaging range in the line-of-sight direction, and output data such as the detected ID number and XY coordinates to the portable computer 50.

  The CCD camera 44 captures a visible light image in the direction of the person's line of sight and outputs a video signal to the portable computer 50. The CCD camera 44 is also configured and operates in the same manner as the CCD camera 32 of the ambient condition observation device 16 described above. Also in this human observation device 18, the optical axis of the lens of the CCD camera 44 is aligned with the optical axis of the lens of the infrared camera 42. Therefore, not only the object located in the line-of-sight direction is identified but also the object A visible light image of an object can also be taken at the same time.

  The microphone 46 is for detecting the voice of an explanation person or a visitor or ambient sounds, and is connected to the portable computer 50 via the voice processing circuit 54. The audio processing circuit 54 outputs the audio signal output from the microphone 46 to the portable computer 50.

  The biological sensor 48 is connected to the portable computer 50 via the biological data processing circuit 56. The biometric sensor 48 and the like are composed of, for example, a biometric data recording module (Procomp +) including three sensors of human pulse, hand surface conductivity (perspiration), and temperature. The biological sensor 48 detects the pulse, sweating state, and body temperature of the instructor or the visitor, and the biological data processing circuit 56 calculates, for example, an average value of each detected data every several seconds, and the biological data is obtained in real time. A / D converted and transmitted to portable computer 50.

  The portable computer 50 specifies the observation time for each data such as the input ID number of the infrared tag 14 or the like, the XY coordinates in the infrared image of the infrared tag 14, the visible light image, sound, or biological data. For example, predetermined information such as addition of time information is performed, and each information is wirelessly transmitted to the data acquisition server 22.

  The stuffed toy observation device 20 has, for example, visual, auditory, and tactile sensations, and the explainer and the visitor hold the stuffed toy observation device 20 as if playing with a normal stuffed toy. The stuffed animal observation device 20 observes the situation of the instructor and visitors from the viewpoint of the stuffed animal itself, and obtains interaction data such as video data, audio data, and tactile data. The stuffed toy observation device 20 includes, for example, a CCD camera 58, a microphone 60, a touch sensor 62, a portable computer 64, and the like. The CCD camera 58 captures a visible light image in the sight line direction of the stuffed animal and outputs a video signal to the portable computer 64. The microphone 60 is for detecting the voice of an explanation person or a visitor and ambient sounds, and is connected to the portable computer 64 via the voice processing circuit 66. The audio processing circuit 66 outputs the audio signal output from the microphone 60 to the portable computer 64. The tactile sensor 62 is, for example, a contact sensor or a pressure sensor that detects contact of an explanation person or a visitor, and outputs tactile data to the portable computer 64 via a processing circuit (not shown). Then, the portable computer 64 performs predetermined processing such as addition of time information for specifying the observation time on the input data such as visible light image, sound, and tactile data, and stores each information. It transmits to the data acquisition server 22 by radio. Note that the stuffed toy observation device 20 may also detect, for example, posture data when the stuffed animal is held.

  The robot body 12 is, for example, a humanoid type autonomous movement type having perception, hearing, touch, and the like, and observes the situation of the exhibition hall, the situation of the instructors and visitors, etc. from the viewpoint of the robot body 12 itself. Then, time information and the like are added to the observed ID number and XY coordinates of the infrared tag 14 and the like, interaction data such as video data, audio data and tactile data, and transmitted to the data acquisition server 22 by radio.

  Further, the robot body 12 can execute a communication action with a conversation partner using gestures and voices. For example, run or face against that person by detecting the presence of a dialogue partner, further or actively speak as "Hello", also, a variety of communication behavior such or face against the contact point if it is touched it can. In this embodiment, for example, body motion data is received from the data acquisition server 22 that also functions as a processing computer external to the robot body 12, and the body motion data is executed, so that past actions of the conversation partner can be performed. It is also possible to present actions such as guidance.

  An example of the robot body 12 will be described in detail with reference to FIGS. 3 and 4. As shown in FIG. 3, the robot body 12 includes a carriage 68, and wheels 70 for autonomously moving the robot body 12 are provided on the lower surface of the carriage 68. The wheel 70 is driven by a wheel motor 72 (see FIG. 4), and the carriage 68, that is, the robot body 12 can be moved in any direction, front, back, left, and right.

  Although not shown in FIG. 3, a collision sensor 74 (see FIG. 3) is attached to the front surface of the carriage 68, and the collision sensor 74 detects contact of a person and other obstacles with the carriage 68. That is, when contact with an obstacle is detected while the robot body 12 is moving, the driving of the wheels 70 is immediately stopped and the movement of the robot body 12 is suddenly stopped.

  In this embodiment, the height of the robot body 12 is about 100 cm so as not to intimidate people, particularly children. However, this height can be changed.

  A polygonal column sensor mounting panel 76 is provided on the carriage 68, and an ultrasonic distance sensor 78 is mounted on each surface of the sensor mounting panel 76. The ultrasonic distance sensor 78 measures the distance between the sensor mounting panel 76, that is, the person around the robot main body 12, mainly a person.

  Further, on the carriage 68, the lower part thereof is surrounded by the sensor mounting panel 76 so that the body of the robot body 12 stands upright. This body is composed of a lower body 80 and an upper body 82, and the lower body 80 and the upper body 82 are connected to each other by a connecting portion 84. Although illustration is omitted, the connecting portion 84 has a built-in lifting mechanism, and the height of the upper body 82, that is, the height of the back of the robot body 12 can be changed by using this lifting mechanism. The lifting mechanism is driven by a waist motor 86 (see FIG. 4), as will be described later.

  One omnidirectional camera 88, one infrared camera 90, and one microphone 92 are provided in the approximate center of the upper body 82. The omnidirectional camera 88 captures the surroundings of the robot body 12 and is distinguished from an eye camera 94 described later. As this omnidirectional camera 88, for example, a camera using a solid-state imaging device such as a CCD or a CMOS can be adopted. The infrared camera 90 is for detecting the infrared rays emitted from the infrared tag 14 and the like, similar to the infrared camera 30 or 42 described above. The microphone 92 captures ambient sounds, particularly human voice. The installation positions of the omnidirectional camera 88, the infrared camera 90, and the microphone 92 are not limited to the upper body 82, and can be changed as appropriate.

  For example, an infrared tag 14 that emits an ID number for identifying the robot body 12 is attached to the chest of the upper body 82.

  Upper shoulders 98R and 98L are provided on both shoulders of upper body 82 by shoulder joints 96R and 96L, respectively. The shoulder joints 96R and 96L each have three axes of freedom. That is, the shoulder joint 96R can control the angle of the upper arm 98R around each of the X axis, the Y axis, and the Z axis. The Y axis is an axis parallel to the longitudinal direction (or axis) of the upper arm 98R, and the X axis and the Z axis are orthogonal to the Y axis from different directions. On the other hand, the shoulder joint 96L can control the angle of the upper arm 98L around each of the A axis, the B axis, and the C axis. The B axis is an axis parallel to the longitudinal direction (or axis) of the upper arm 98L, and the A axis and the C axis are axes orthogonal to the B axis from different directions.

  Further, forearms 102R and 102L are provided at the tips of the upper arms 98R and 98L via elbow joints 100R and 100L, respectively. The elbow joints 100R and 100L can control the angles of the forearms 102R and 102L around the W axis and the D axis, respectively.

  In the X axis, Y axis, Z axis, W axis, A axis, B axis, C axis, and D axis that control the displacement of the upper arms 98R and 98L and the forearms 102R and 102L, “0 degree” is the home position, respectively. In this home position, as shown in FIG. 3, the upper arms 98R and 98L and the forearms 102R and 102L are directed downward.

  Although not shown in the drawings, the shoulder portion including the shoulder joints 96R and 96L of the upper body 82, the upper arms 98R and 98L, and the forearms 102R and 102L described above are each shown by touch sensors (shown comprehensively in FIG. 4). : 104), and these touch sensors 104 detect whether or not a person has touched each part of the robot body 12.

  Spheres 106R and 106L corresponding to hands are fixedly provided at the tips of the forearms 102R and 102L, respectively. However, if a finger or palm function is required, a “hand” in the shape of a human hand can be used.

  A head 110 is provided above the center of the upper body 82 via a neck joint 108. The neck joint 108 has three degrees of freedom, and the angle can be controlled around each of the S, T, and U axes. The S-axis is an axis that extends from the neck directly upward (vertically upward), and the T-axis and the U-axis are axes that are orthogonal to the S-axis in different directions. The head 110 is provided with a speaker 112 at a position corresponding to a human mouth. The speaker 112 is used for the robot body 12 to communicate with people around it by voice or sound. However, the speaker 112 may be provided in another part of the robot main body 12, for example, the trunk.

  The head 110 is provided with eyeball portions 114R and 114L at positions corresponding to the eyes. Eyeball portions 114R and 114L include eye cameras 94R and 94L, respectively. Hereinafter, the right eyeball portion 114R and the left eyeball portion 114L may be collectively referred to as the eyeball portion 114, and the right eye camera 94R and the left eye camera 94L may be collectively referred to as the eye camera 94.

  The eye camera 94 captures a human face approaching the robot body 12 and other parts or objects, and captures a corresponding video signal. As the eye camera 94, a camera similar to the omnidirectional camera 88 described above can be used.

  For example, the eye camera 94 is fixed in the eyeball unit 114, and the eyeball unit 114 is attached to a predetermined position in the head 110 via an eyeball support unit (not shown). The eyeball support unit has two degrees of freedom and can be controlled in angle around each of the α axis and the β axis. The α axis and the β axis are axes provided with respect to the head 110, the α axis is an axis in a direction toward the top of the head 110, the β axis is orthogonal to the α axis and the front side of the head 110 ( It is an axis in a direction perpendicular to the direction in which the face is facing. In this embodiment, when the head 110 is in the home position, the α axis is set to be parallel to the S axis, and the β axis is set to be parallel to the U axis. In such a head 110, when the eyeball support portion is rotated around each of the α axis and the β axis, the tip (front) side of the eyeball portion 114 or the eye camera 94 is displaced, and the camera axis, that is, the line-of-sight direction is changed. Moved.

  In the α axis and β axis that control the displacement of the eye camera 94, “0 degree” is the home position. At this home position, the camera axis of the eye camera 94 is the head 110 as shown in FIG. The direction of the front side (face) is directed, and the line of sight is in the normal viewing state.

  FIG. 4 is a block diagram showing an electrical configuration of the robot main body 12. With reference to FIG. 4, the robot main body 12 includes a CPU 116 for controlling the whole. The CPU 116 is also called a microcomputer or a processor, and is connected to the memory 120, the motor control board 122, the sensor input / output board 124, and the audio input / output board 126 via the bus 118.

  Although not shown in the figure, the memory 120 includes a ROM and a RAM. The ROM stores a control program for the robot body 12 in advance, and the RAM is used as a work memory and a buffer memory. The control program includes, for example, a program for executing a communication action, a program for observing an interaction to acquire data, a program for adding time information and the like to the acquired data, and a data acquisition server (processing computer) 22 And a program for executing received physical movement data. The memory 120 also needs body movement data including voice data or voice data (speech synthesis data) to be generated from the speaker 112 when performing the communication action and angle data for presenting a predetermined gesture. Is stored according to

  The motor control board 122 is configured by a DSP, for example, and controls the driving of each axis motor such as each arm, head, and eyeball. That is, the motor control board 122 receives the control data from the CPU 116, and shows two motors for controlling the angles of the α axis and β axis of the right eyeball portion 114R (in FIG. 4, collectively referred to as “right eyeball motor”). .) Control the rotation angle of 128. Similarly, the motor control board 122 receives the control data from the CPU 116, and controls two angles of the α axis and β axis of the left eyeball portion 114L (collectively, “left eye motor” in FIG. 4). Controls the rotation angle of 130.

  In addition, the motor control board 122 receives control data from the CPU 116, and controls the angles of the X axis, Y axis and Z axis of the right shoulder joint 96R and the W axis angle of the right elbow joint 100R. The rotation angle of a total of four motors (one collectively shown as “right arm motor” in FIG. 4) 132 including one motor to be controlled is adjusted. Similarly, the motor control board 122 receives the control data from the CPU 116 and determines the angles of the three motors for controlling the angles of the A-axis, B-axis, and C-axis of the left shoulder joint 96L and the D-axis angle of the left elbow joint 100L. The rotation angles of a total of four motors (one collectively shown as “left arm motor” in FIG. 4) 134 with one motor to be controlled are adjusted.

  Furthermore, the motor control board 122 receives control data from the CPU 116 and controls three motors for controlling the angles of the S-axis, T-axis, and U-axis of the head 110 (collectively, “head motor” in FIG. 4). The rotation angle of 136 is controlled. Furthermore, the motor control board 122 receives control data from the CPU 116 and controls the rotation angle of two motors 72 (hereinafter collectively referred to as “wheel motors”) 72 that drive the waist motor 86 and the wheels 70. To do.

  In this embodiment, stepping motors or pulse motors are used for the motors other than the wheel motor 72 in order to simplify the control. However, like the wheel motor 72, a DC motor may be used.

  Similarly, the sensor input / output board 124 is also constituted by a DSP, and takes in a signal from each sensor and gives it to the CPU 116. That is, data relating to the reflection time from each of the ultrasonic distance sensors 78 is input to the CPU 116 through the sensor input / output board 124. A video signal from the omnidirectional camera 88 is input to the CPU 116 after being subjected to predetermined processing by the sensor input / output board 124 as necessary. Similarly, the video signal from the eye camera 94 is also input to the CPU 116.

  Further, the image signal from the infrared camera 90 detects an infrared tag from an infrared image in the same manner as the image processing device 38 of the ambient condition observation device 16 and the image processing device 52 of the human observation device 18, and the infrared tag. The ID number is detected from the blinking state of, and the XY coordinates of the infrared tag on the infrared image are detected. Data such as an infrared tag ID number and XY coordinates are input to the CPU 116.

  In addition, signals from the plurality of touch sensors described above (collectively indicated as “touch sensor 104” in FIG. 4) are provided to the CPU 116 via the sensor input / output board 124. Further, the signal from the above-described collision sensor 74 is also given to the CPU 116 in the same manner.

  Similarly, the voice input / output board 126 is also configured by a DSP, and voice or voice in accordance with voice synthesis data provided from the CPU 116 is output from the speaker 112. Also, the voice input from the microphone 92 is taken into the CPU 116 via the voice input / output board 126.

  The CPU 116 adds time information for specifying the observation time to the interaction data such as video data, audio data, tactile data, infrared tag ID number, and XY information captured via the sensor input / output board 124. .

  The CPU 116 is connected to the communication LAN board 138 via the bus 118. The communication LAN board 138 is configured by a DSP, gives transmission data sent from the CPU 116 to the wireless communication device 140, and transmits the transmission data from the wireless communication device 140 through a network such as a wireless LAN, although not shown. To the data acquisition server 22. Further, the communication LAN board 138 receives data via the wireless communication device 140 and gives the received data to the CPU 116. That is, the acquired interaction data or the like is transmitted to the data acquisition server 22, and the body movement data transmitted from the processing computer 22 is received.

  Returning to FIG. 2, the data acquisition server 22 includes, for example, a ROM, a CPU, a RAM, an external storage device, and the like, and is connected to the interaction DB 24. The data acquisition server 22 stores each interaction data transmitted from each observation device in the interaction DB 24. In this embodiment, the interaction DB 24 includes database servers such as an SQL (Structured Query Language) server 24a and an AV data server 24b. Among the interaction data, the ID number and XY coordinates of the infrared tag 14 and the like, biometric data, and the like are stored. The SQL server 24a is provided and stored, and the video data and audio data are supplied to the AV data server 24b and stored.

  The SQL server 24a includes a ROM, a CPU, a RAM, an external storage device, and the like, and stores interaction data such as ID numbers, XY coordinates, and biometric data to which time information is given in a machine-readable state as a database. In addition, each interaction data is stored so as to be able to identify which observation device has been acquired.

  The AV data server 24b includes a ROM, a CPU, a RAM, an external storage device, and the like, and stores interaction data such as video data and audio data as a database in a machine-readable state. Since it is not realistic to make one session into one huge video file, the AV data server 24b stores, for example, video data as separate files every minute and uses the interaction DB 24. At this time, the index data of each video data is managed by the SQL server 24a so that it is not necessary to be aware that the files are divided every minute.

  The data acquisition server 22 is also connected to the robot operation DB 26 and the guide rule DB 28. The robot motion DB 26 stores a plurality of body motion data to be executed by the robot body 12 in advance. The body motion data includes angle control data of each axis motor or the like when the body motion includes gestures, and includes speech synthesis data when the body motion includes speech.

  The guide rule DB 28 stores in advance a plurality of data relating to rules that define the behavior to be presented by the robot body 12 corresponding to the past behavior of the conversation partner. That is, for example, a predetermined action to be executed when the condition is satisfied or not satisfied is defined in association with a predetermined condition related to a past action. In this embodiment, according to the past action of the conversation partner, for example, guidance or recommended action for further explanation or the like is performed in connection with the object that the person has seen or visited, etc. The presentation of information about the interacted object itself is defined. For example, the robot body 12 talks about what the conversation partner has seen, and provides information on anything that should be seen in connection with it. Therefore, for the conversation partner, it is possible to supplement or deepen the knowledge and experience obtained by the interaction, and to obtain useful information. As an example of the guide rule, for example, when the conversation partner has seen the exhibit B1 in the past but has not yet seen the related exhibit B4, the utterance is recommended to see the exhibit B4. Rules are defined. In addition, for example, when the conversation partner plays with the stuffed toy observation device 20, it is prescribed that the information about the stuffed toy is uttered or the information about the object when the biological data changes is presented. The

  In this robot 10, the situation in the environment such as the exhibition hall, specifically, the environment, by the ambient condition observation device 16, the human observation device 18, the stuffed animal observation device 20, the robot body 12, and the like as described above. Interaction between one or a plurality of humans (explanator, visitor) to be interacted with and exhibits, stuffed animals 20, robot main body 12 and the like, which are human interaction objects, is observed. Since sensors are provided everywhere as described above, the environment such as the exhibition hall in this embodiment can be called a ubiquitous sensor environment. In such a ubiquitous sensor environment, each interaction data is acquired in real time and sequentially recorded in the interaction DB 24 by the data acquisition server 22.

  For example, a visitor P1 or the like who has entered the field of view of the instructor Q1 is identified by the human observation device 18 as an object of interaction, and the instructor Q1 is identified as an object by the ambient condition observation device 16, and the ambient condition observation is performed. The visitor P1, the robot body 12, and the like are identified as objects around the device 16. In this way, in addition to the ambient condition observation device 16 and the like that are ubiquitous in the exhibition hall, the same event can be performed by using the human observation device 18 worn by the instructor and the visitor who is the subject of the interaction. It is possible to record from a local viewpoint and an overall viewpoint from multiple observation devices. In addition, the interaction can be recorded from an objective and local viewpoint by the stuffed toy observation device 20, the robot body 12, or the like. Further, by automatically recognizing the ID numbers of the persons and objects that have entered the field of view of the human observation device 18 and the surrounding situation observation device 16, it is possible to index the accumulated interaction data in real time. Therefore, the instructor and the visitor can unconsciously add an index to the interaction with another person or the object by “seeing”. In addition, the ID number of the detected person and object and the other measured interaction data are stored in a state where each observation time can be specified, so that the person and object in each interaction data are identified. It is possible to analyze human interaction from the relationship between humans and objects, and systematically accumulate non-verbal common sense (non-linguistic information) that humans handle casually and make it machine-readable. Can be dictionaryd.

  In this embodiment, the robot body 12 presents an action (in this embodiment, guidance, recommendation, etc.) according to the interaction data to a person in the environment. Specifically, the interaction data of the conversation partner stored in the interaction DB 24 is analyzed in the data acquisition server 22, and the action to be executed by the robot body 12 is determined based on the guide rules stored in the guide rule DB 28. The Then, the data acquisition server 22 transmits data for executing the determined action to the robot body 12, and the robot body 12 executes the action accordingly. Therefore, the robot body 12 can present a predetermined action such as guidance adapted to the past action to the conversation partner.

  As described above, the data acquisition server 22 not only acquires and records the observed interaction data, but also functions as an external processing computer that controls the behavior of the robot body 12.

  FIG. 5 shows an example of a specific operation related to the guide presentation of the processing computer 22. This guide presenting process is started at a predetermined time period during, for example, recording of interaction data.

  First, in step S1, the CPU of the processing computer 22 determines whether there is an interaction target. For example, the SQL server 24a is inquired whether the interaction data observed by the infrared camera 90 of the robot body 12 at the latest observation time of the current time has an ID number of a visitor or the like. Here, when determining whether or not there is a conversation target, the robot body 12 may be moved around the robot body 12 to move the visual field. If “NO” in the step S1, that is, if there is no conversation target, the guide presenting process is ended.

  Note that the presence / absence of a conversation target may be determined based on interaction data detected by an observation device other than the robot body 12. For example, even when the ID number of a visitor or the like is not detected in the data from the infrared camera 90 of the robot body 12, the interaction data in which the ID number of the robot body 12 is detected is searched and the visitor detected in the interaction data is detected. Since the visitor or the like can be regarded as being present near the robot body 12, the ID number such as may be set as a conversation target. In this case, the visitor can be captured by causing the robot body 12 to investigate the surroundings and moving the field of view.

  Or even if it is the interaction data from which the ID number of the robot main body 12 is not detected, if the ID number of a visitor or the like is detected, the ID number may be set as a conversation target. In this case, the observation device that observed the interaction data in which the ID number is detected is specified. If this observation device is, for example, the ambient state observation device 16 installed in the environment, the installation location of the device is determined. The robot main body 12 may be moved in the direction to capture the visitor or the like, or the ID number of the human observation device 18 may be captured even when the specified observation device is the human observation device 18. The robot body 12 may be moved as appropriate.

  On the other hand, if “YES” in the step S1, the ID number of the conversation target is acquired from the SQL server 24a of the interaction DB 24 and determined as the conversation partner in a step S3.

  Subsequently, in step S5, the past action data is acquired from the interaction DB 24 for the conversation partner. That is, for example, the interaction data in which the ID number of the conversation partner is detected is referred to the SQL server 24a, and the corresponding interaction data is acquired from the SQL server 24a and the AV data server 24b. Since the interaction data includes data described by an ID number that can identify the conversation partner, it can be easily obtained and referenced.

  In step S7, the behavior data acquired from the interaction DB 24 is analyzed. In this analysis process, for example, the positional relationship between the conversation partner and the person or the object is specified using the ID numbers and XY coordinates of the person and the object, and the interaction data of the interaction data is determined based on the specified positional relationship. Determine the state of the event that means it.

  FIG. 6 schematically shows an example of event types. All events are simple elements that cannot be further simplified in the sense that the ambient condition observation device 16, the human observation device 18, the robot body 12, etc. capture the infrared tag 14, but the ambient condition observation device 16. Various meanings can be interpreted by a combination of the human observation device 18 and the robot main body 12 or the like and the object to which the infrared tag 14 or the like is attached. The event state is not particularly limited to the example of FIG. 6, and various changes and additions are possible.

  For example, as shown in FIG. 6 (A), an infrared tag 14 or the like (black circle display) of another person B is observed by a human observation device 18 (white circle display) worn by a person A, and at the same time, When the infrared tag 14 of the person A is observed by the human observation device 18 worn by B, this interaction state can be determined to be a state in which the person A and the person B are interacting with each other. .

  Moreover, when the infrared tag 14 etc. which were given to the person by the surrounding condition observation apparatus 16 installed in the environment side are observed, it can determine with the person staying in the said installation area, FIG.6 (B) As shown in FIG. 5, when the infrared tags 14 of a plurality of people A and B are observed simultaneously by the same ambient condition observation device 16 (white circle display), those people A and B coexist in the same area. It can be determined as a state.

  Further, as shown in FIG. 6C, when the infrared tag 14 (indicated by a black circle) attached to a certain object C is observed by the human observation device 18 worn by the person A, It can be determined that the person A is gazing at the object C. Although illustration is omitted, when the infrared tag 14 of the same object is simultaneously observed by a plurality of human observation devices 18, those people are paying common attention to the same thing. It is considered to be a state. Furthermore, when the number of people participating in joint attention increases, the object being noted is considered to be in a state of significant social events.

  Further, as shown in FIG. 6D, the ambient state observation device 16 installed on the environment side simultaneously observes the infrared tag 14 attached to a certain object C, the infrared tag 14 of a certain person A, and the like. If it is, it can be interpreted that the person A is visiting the place where the object C is installed.

  As shown in FIG. 6E, the infrared tag 14 of another person B is observed by the human observation device 18 worn by a person A, and the event continues for a predetermined time or more. If it is, it can be interpreted that the person A is staring at the person B. Alternatively, as shown in FIG. 6C, the infrared tag 14 attached to a certain object C is observed by the human observation device 18 worn by the person A, and the event continues for a predetermined time or more. If it is, the person A may be interpreted as meaning a state of staring at the object C.

  In addition, as shown in FIG. 6F, the infrared tag 14 attached to a certain object C is observed by the human observation device 18 worn by the person A, and is simultaneously installed on the object C. Further, when the infrared tag 14 or the like of the person A is being observed by the surrounding state observation device 16, it may be interpreted as meaning that the person A is gazing at the object C.

  If the interpretation of these event states overlaps, a priority order may be set in advance, and a higher priority order may be adopted.

  In this way, the interaction data of the conversation partner is analyzed in step S7 of FIG. 5, and the past behavior history is grasped by, for example, the event state as described above. Therefore, the action to be presented can be easily determined based on the grasped past action state.

  For example, biometric data or the like can be used. In other words, for example, an increase or a change in the pulse, sweating state, body temperature, etc. in the interaction with the person or object of the conversation partner is detected, and the event state of the interaction data is determined based on the detected change. May be.

  Subsequently, in step S <b> 9, the CPU of the processing computer 22 acquires a guide rule related to the analyzed behavior data from the guide rule DB 28. For example, when the dialogue partner is visiting the exhibit B1, a rule including that the exhibit B1 is visited as a condition is searched to obtain a matching rule.

  In step S11, an action to be executed (guide content in this embodiment) is determined based on the analyzed past action data and the acquired guide rule. Since the guide rule defines an action to be executed corresponding to a past action, the action can be easily determined by comparing the past action with the guide rule. For example, when the conversation partner visits the exhibit B1 but does not visit the exhibit B4, he / she makes a pointing operation toward the exhibit B4 and says “You should also see the exhibit B4”. It is decided to make a recommendation. Further, for example, if the visitor B2 visits an easy-to-understand exhibit B2 who has received an explanation directly from the instructor Q1, but does not interact with the instructor Q1, “explain the exhibit B2 is explained to the explainer Q1. It ’s good to hear from you. ” When a plurality of rules defining the presentation of actions are acquired in the previous step S9, the order in which the rules are applied is determined, for example, randomly or according to priority.

  Subsequently, in step S13, body motion data regarding the determined guide content is acquired from the robot motion DB 26, and the body motion data is transmitted to the robot body 12 in step S15. When the robot body 12 receives the body movement data, the CPU 116 of the robot body 12 executes a behavior process based on the body movement data. Therefore, behaviors such as guides and recommendations according to the past behavior are presented to the conversation partner. In other words, in one of the previous examples, for example, the robot body 12 outputs an utterance voice from the speaker 112 that “you should also see the exhibit B4”.

  In this way, a predetermined action corresponding to appropriate guidance or recommendation according to the past action of the conversation partner is presented by the body movement including at least one of gesture and voice of the robot body 12 having a communication function. As a result, you will feel very familiar to the person you are talking to, and you can feel secure by knowing your actions. Therefore, compared to the case where voice guidance is simply provided from a portable radio or mobile phone, for example, it can be expected that the guidance provided by the conversation partner will be heard properly and the conversation partner will follow the guidance. The possibility of taking can also be increased.

  In step S17, it is determined whether there is any other guide content to be presented. For example, if a plurality of rules that define the presentation of actions are acquired in step S9 and rules that are not applied remain, it is determined “YES” in step S17, and in this case, the process returns to step S11. Then, the process is repeated for the remaining rules.

  If “NO” in the step S17, it is determined whether or not there is another dialogue target in a succeeding step S19. For example, with respect to the SQL server 24a of the interaction DB 24, the visitor's ID number (the ID number previously determined as the conversation partner) is added to the interaction data observed by the infrared camera 90 of the robot body 12 at the latest observation time of the current time. Inquire whether or not there is.) If “YES” in this step S19, the process returns to the step S3, the ID number is acquired from the interaction DB 24, an ID number different from the previous ID number is determined as the conversation partner, and the other conversation partner is determined. On the other hand, the processing from step S5 is repeated.

  On the other hand, if “YES” in the step S19, since there is no conversation target, the guide presenting process is ended.

  According to this embodiment, it is possible to observe human interaction within a predetermined environment and record human behavior, and based on the past interaction data regarding the recorded conversation partner, The action to be presented to the conversation partner by the robot body 12 can be determined. Therefore, it is possible to present information such as appropriate guidance and recommendation according to the past action of the conversation partner. For this reason, it is possible to make the conversation partner feel familiar or feel secure, and for example, it can be expected that the presented guidance or the like can be effectively used by the conversation partner.

  Further, when a conversation partner who has presented guidance in the past appears again as a conversation target later, in the action data analysis process (step S7 in FIG. 5), the conversation partner behaves according to the guidance of the robot body 12. Whether or not it can be grasped. That is, for example, first, in step S15 in FIG. 5, the processing computer 22 transmits body motion data to the robot body 12, and the robot body 12 presents guidance to the conversation partner, and then the body motion presented with the conversation partner. Is recorded in the interaction DB 24 in association with the information related to. When the conversation partner appears again, it is determined in step S7 whether or not the subsequent interaction data of the conversation partner includes an interaction or event state in accordance with the presented guidance. For example, when it is recommended to view the exhibit B4, it is determined whether or not there is an event state as shown in FIGS. 6C, 6D, or 6F with the exhibit B4. Thereby, it is possible to grasp whether or not the conversation partner is a person who listens to the guidance of the robot body 12, or whether or not the person who trusts the robot body 12. Further, by registering a guide rule corresponding to the result in the guide rule DB 28 in advance, for example, the guide content presented by the robot main body 12 with respect to the conversation partner, the way of presentation, the manner of contact, etc. are changed. Can be made.

  In the above-described embodiment, the guide presentation process as shown in FIG. 5 is executed by the processing computer 22 outside the robot body 12. However, a program or the like for executing this guide presentation process is stored in the robot. This process may be executed by the robot body 12 so as to be provided inside the body 12. However, in this case, in step S15, the execution of the body movement data acquired from the robot DB 26 is processed. In this case, the robot motion DB 26 and the guide rule DB 28 may be provided in the robot body 12.

  In each of the above-described embodiments, identification information for identifying a person or an object is acquired using an infrared tag, an infrared camera, or the like. However, the method for acquiring identification information can be changed as appropriate. For example, the identification information may be acquired using a wireless tag (RFID) and a wireless tag reader. For example, when a wireless tag reader is provided in connection with an object and a person is attached with a wireless tag, or when a wireless tag reader is attached to a person and a wireless tag is provided in association with an object, etc. Can be determined that there is an interaction between the person and the object in the analysis of the interaction data detected by the RFID by entering the detection range of the RFID tag reader.

  Further, in each of the above-described embodiments, identification information that can individually identify a person and an object is set, and individual interaction data is recorded, for example, an action corresponding to the action of each individual person is presented. However, as long as the identification information can be acquired by sensors, the identification information may identify a predetermined unit when a person or an object is classified based on some characteristic or commonality. Good. That is, for example, in the case of human beings, differences in colors such as clothes or hair and age groups such as elderly or young may be used as identification information. In such a case, the interaction data for each unit is recorded, and a predetermined action corresponding to the past action for the unit to which the conversation belongs is presented to the conversation partner. That is, the action to be presented by the robot body 12 is determined based on the action history of people having the same attribute. Note that the above-described color identification information can be obtained directly from a person itself by image data, for example, without providing an identification tag or the like. The age group can also be detected by, for example, mounting a motion capture marker on a human and acquiring and analyzing motion data such as when walking using a motion capture system.

It is an illustration figure which shows the outline | summary of the communication robot of one Example of this invention. It is a block diagram which shows the structure of the robot of FIG. 1 Example. It is an illustration figure (front view) which shows the external appearance of the robot main body of FIG. It is a block diagram which shows the internal structure of the robot main body of FIG. It is a flowchart which shows an example of operation | movement of the guide presentation process of the processing computer of FIG. It is an illustration figure which shows an example of the state of the event determined in the analysis of action data.

Explanation of symbols

DESCRIPTION OF SYMBOLS 10 ... Communication robot 12 ... Robot main body 14 ... Infrared tag 16 ... Ambient condition observation device 18 ... Human observation device 20 ... Stuffed observation device 22 ... Processing computer 24 ... Interaction database 26 ... Robot motion database 28 ... Guide rule database 116 ... CPU

Claims (5)

A communication robot that observes and communicates with the behavior of a conversation,
A sensor for acquiring an action history of the conversation target;
Recording means for recording the action history acquired by the sensor;
Partner determining means for determining a conversation partner from the conversation target;
A communication robot comprising: an action determining unit that determines an action of the robot based on an action history related to the conversation partner recorded in the recording unit; and an executing unit that executes the action determined by the action determining unit.   The communication robot according to claim 1, wherein the sensor acquires at least one of identification information for identifying a conversation target and identification information for identifying an object of the interaction target interaction.   The behavior determination means analyzes a state of an action history related to the conversation partner based on a positional relationship between the conversation partner and the object of the interaction, and determines the behavior based on the analysis result. Communication robot. Rule storage means for storing a rule defining the action to be presented by the robot corresponding to the past action;
The communication robot according to any one of claims 1 to 3, wherein the behavior determining unit determines the behavior based on a behavior history and the rules regarding the conversation partner.   The communication robot according to any one of claims 1 to 4, wherein the action determined by the action determining means includes an operation of performing guidance or recommendation related to an object with which the conversation partner interacts.

Images