JP2001191279A - Behavior control system, behavior controlling method, and robot device - Google Patents

Abstract

PROBLEM TO BE SOLVED: To easily realize cooperative behaviors of a plurality of robot devices. SOLUTION: This behavior control system is provided with a plurality of robot devices 1a, 1b, 1c deciding their behavior independently and a personal computer 33 serving as an external client for controlling behavior of the robot devices 1a, 1b, 1c. The personal computer 33 sends commands synchronously to the robot devices 1a, 1b, 1c respectively.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a behavior management system and a behavior management method for managing the behaviors of a plurality of robots, and to an autonomous robot that autonomously determines the behavior.

[0002]

2. Description of the Related Art Conventionally, there has been proposed and developed an autonomous robot apparatus which autonomously determines an action in response to a command from a user or a surrounding environment. For example, this type of robot device has a shape very similar to a multi-articulated quadruped such as a dog or a cat, and determines its own action autonomously based on an action pattern for action. Specifically, when the robot apparatus receives a voice command “push” from the user, the robot apparatus takes a “down” posture or “hands” in accordance with an operation in which the user puts his hand in front of his / her mouth. It has been made like that.

[0003]

Recently, there has been proposed a robot device capable of executing an arbitrary command by using a command device (remote controller). For example, there is a robot device that performs a predetermined operation according to a scale command output from a command device. Thus, the user can call and enjoy an interesting action that does not appear during the autonomous action other than playing as an autonomous robot apparatus by using the musical scale command. As described above, there is a need to call, watch, and enjoy a preprogrammed action like a karakuri doll, and such techniques have been proposed and developed.

[0004] Further, there is a need to make a plurality of robot apparatuses cooperate. Unlike the case where one robot device performs one action (trick), the cooperative behavior of a plurality of robot devices can make it appear as if the plurality of robot devices are linked. It can be said that such a cooperative action by a plurality of robot devices broadens the enjoyment of appreciation of the user.

[0005] Such a cooperative action by a plurality of robot devices can be realized by pre-programming each robot device in advance and linking a series of different action patterns by the plurality of robot devices. Occurs.

[0006] It is necessary to set a careful time plan and start the action at the same time. That is, it is necessary to know the time required for each action, and to carefully plan a timetable for synchronizing.

In the case of a multi-joint robot device, different behavior patterns require different postures. If it is desired to cause a plurality of robot devices to simultaneously exhibit an action, it is necessary to change the posture to the position where the action can be performed according to the immediately preceding posture. The design is also complicated.

Further, when it is desired to perform various kinds of cooperative actions (various action patterns), it is necessary to change the program or data of the robot apparatus for each action pattern.

[0009] It is also difficult to make the start times of the cooperative actions uniform, but if the actions are shifted for some reason after starting once, they will remain shifted to the end.

The present invention has been made in view of the above circumstances, and provides a behavior management system, a behavior management method, and a robot device capable of easily executing a cooperative behavior of a plurality of robot devices. The purpose is.

[0011]

SUMMARY OF THE INVENTION In order to solve the above-mentioned problems, an action management system according to the present invention provides an autonomous robot apparatus which independently determines an action by using an externally input command. And a behavior management system that manages the behavior of the autonomous robot device and has a behavior management unit that synchronously sends commands to a plurality of robot devices.

According to the behavior management system having such a configuration, a plurality of robot devices execute a synchronously transmitted command, and as a result, the behavior of the plurality of robot devices becomes a cooperative behavior.

Further, in order to solve the above-mentioned problems, the behavior management method according to the present invention has a function of determining an action by a command input from the outside to an autonomous robot apparatus which independently determines an action. A behavior management method for managing the behavior of the autonomous robot device, wherein the command is synchronously transmitted to the plurality of robot devices.

According to this behavior management method, a plurality of robot devices execute a synchronously transmitted command, and as a result, the behavior of the plurality of robot devices becomes a cooperative behavior.

Further, a robot apparatus according to the present invention is an autonomous robot apparatus for autonomously determining an action in order to solve the above-mentioned problem, and synchronizes a command with another plurality of robot apparatuses. And a behavior management means for sending out.

With the robot device having such a configuration, the other plurality of robot devices execute the synchronously transmitted command, and as a result, the actions of the plurality of robot devices become cooperative actions.

[0017]

Embodiments of the present invention will be described below in detail with reference to the drawings. In this embodiment, the present invention is applied to a behavior management system that manages the behavior of an autonomous robot apparatus that determines behavior independently. This behavior management system is configured to allow a plurality of robot devices to perform cooperative behavior.

As shown in FIG. 1, the robot device whose behavior is managed by this behavior management system has legs 2a, 2b, 2c, 2d driven for movement and a CCD (Charge).
Coupled Device) Head 3 with video camera etc.
And the body 4 and the like. As shown in FIG. 2, the robot apparatus 1 includes at least a CPU 15 as a control unit and a data storage unit 13 for storing various information, as shown in FIG. The data storage unit 13 includes, for example, a dynamic random access memory (DRAM).
y), information storage means such as a flash ROM (Read Only Memory) or a so-called memory stick.
Further, as shown in FIGS. 1 and 2, the robot apparatus 1 has a PC for mounting a PC card 200 such as a memory card for storing data or a wireless LAN card for transmitting and receiving data wirelessly. A card slot (PC card I / F) 14 is provided.

The robot device 1 can act and change emotions by a program that determines its own behavior based on an external factor or an internal factor. here,
A program for determining one's own behavior is constructed by a behavior model and an emotion model. Specifically, the robot apparatus 1 is configured to act autonomously based on a program for determining its own action in response to inputs from various sensors. For example, as shown in FIG. 3, the robot apparatus 1 performs a “hand” operation of placing the front leg 2 a on the user's hand 300 when the user says “hand”.

The robot apparatus 1 includes a PC card (memory card) 20 inserted in the PC card slot 14.
Predetermined data can be written to 0. The robot device 1 can perform data communication with an external device such as a personal computer using a PC card (wireless LAN card) 200 for wireless communication. The robot apparatus 1 performs data communication with a personal computer, which is an external client, by such a wireless communication function, and can perform a cooperative action in synchronization with another robot apparatus.

The robot apparatus 1 configured as described above has
The behavior is managed on the behavior management system as follows.

The behavior management system, as shown in FIG.
A personal computer that manages the behavior of the first to third robot devices 1a, 1b, 1c, and a relay device 34 that enables data communication between the personal computer 33 and the plurality of robot devices 1a, 1b, 1c. Have.

Personal computer 33 and repeater 34
Is connected to a local area network (LAN). For example, the repeater 34 forms a so-called access point. The local area network has
As other devices, for example, a workstation (WS)
May be connected.

The data communication protocol with the robot devices 1a, 1b, 1c in the local area network is used for network communication between the personal computer 33 and a workstation (not shown).
TCP / commonly used for Internet communications
IP (Transmission Control Protocol / Internet Proro
col) shall be used. That is, the behavior management system employs an existing network communication method and is configured to manage the behavior of the plurality of robot devices 1a, 1b, 1c.

Although the behavior management system shown in FIG. 4 describes a case where three behaviors of the first to third robot apparatuses 1a, 1b, 1c are managed, the present invention is not limited to this. That is, the behavior management system to which the present invention is applied can manage the behavior of at least two robot devices.

With such an action management system, the personal computer 3 sent via the repeater 34
Each of the robot devices 1a, 1b, and 1c performs an action based on a command transmitted in synchronization with the robot device 1. Therefore, since commands are sent from the personal computer 33 to the plurality of robot devices 1a, 1b, 1c at the same time, the overall behavior of the robot devices 1a, 1b, 1c is expressed as a cooperative behavior.

As described above, the behavior management system merely transmits commands to the plurality of robot devices 1a, 1b, 1c in synchronization with each other, and only transmits the commands to the robot devices 1a, 1b, 1c.
Since c performs cooperative action, such cooperative action is realized without being conscious of time management of the action of each robot device.

Further, for example, the personal computer 3
The content of the command (the content of the operation) from 3 may be different for each robot device or may be the same. Thereby, each of the robot devices 1a, 1b, and 1c can cause a cooperative action by the same operation or cause a cooperative action by a different operation. Further, since the external client that collectively controls the actions of the plurality of robot apparatuses 1a, 1b, and 1c is the common personal computer 33, the action pattern can be edited by one personal computer 33. In addition, a plurality of robot apparatuses 1a, 1b, 1
Since the behavior of c is managed, a cooperative behavior can be enabled between the plurality of robot apparatuses 1a, 1b, and 1c whose behaviors are managed, while making the programs and data identical.

Next, a plurality of robot apparatuses 1a, 1b, 1
A specific description will be given of a case where the action c is synchronized in the action management system and performs a cooperative action as a whole.

The personal computer 33 includes the repeater 3
4 through the robot device 1a, 1b, 1c.
Is transmitted in synchronization with the robot device 1a, 1b, 1
Although c is acting cooperatively, specifically, the command is written in the robot control script (script) RS held by the personal computer 33 in correspondence with each robot device. Specifically, the personal computer 33 sends a command to a corresponding robot device by a robot control script RS created according to a predetermined rule. The robot control script RS is configured from information that can control the timing of sending commands to the robot devices 1a, 1b, and 1c and detect the end timing of the action of each robot device.

The form of the robot control script RS includes, for example, a plurality of data strings D ₁ , D ₂ and D ₃ composed of command information and the like as shown in FIG. The first data sequence D ₁ corresponds to the first robot device 1a, the second data sequence D ₂ corresponds to the second robot device 1b, and the third data sequence D ₁ corresponds to the third robot device 1a. 1
c. In the present embodiment, since there are three robot devices, this data sequence
The data sequence of the robot control script RS is prepared according to the robot device that manages the behavior in the behavior management system.

Specifically, the robot control script RS
In the figure, the number of robot devices on the network to be controlled is described. Then, each data string D ₁ , D ₂ , D ₃
In each case, information on individual IP addresses and port numbers of the servers, abstract commands to be transmitted to the robot devices, commands for obtaining synchronization timing among the robot devices 1a, 1b, and 1c are described. Here, as described above, since the TCP / IP is used for the data communication between the personal computer 33 and the robot devices 1a, 1b, and 1c in the present embodiment, the IP address is assigned to the robot device on the network. Of the identification information.
For example, an IP address is written in advance in a data storage unit of the robot device 1.

These commands are described in the order of execution in the data strings D ₁ , D ₂ , and D ₃ , for example, from top to bottom.

Based on such a robot control script RS, the personal computer 33 sends a command to each of the robot devices 1a, 1b, 1c as follows to control its behavior.

The personal computer 33, prior to initiating communication between the robot apparatus 1a, 1b, and 1c, the thread 28a of the communication by the number of the robotic device by the main program PG _1, 28b, generates the 28c. First to third communication threads 28a, 28b, 2
8c is the first to third robot devices 1 for managing behavior
a, 1b, and 1c for data communication with the corresponding robot devices 1a, 1b, and 1c.

Then, the personal computer 33
The first to third communication threads 28a, 28b, and 28c generated prior to the start of the data communication as described above are assigned to the first to third communication threads RS corresponding to the robot control script RS.
Passing the data stream D _1, D _2, D _3, respectively. Each communication thread 28a, 28b, 28c extracts the IP address and port number from the passed information (data string),
As a client, each robot device is connected, whereby a data communication path between the communication threads 28a, 28b, 28c and each of the robot devices 1a, 1b, 1c is established.

Then, each communication thread 28a, 28
b and 28c transmit the commands described in the delivered data strings D ₁ , D ₂ and D ₃ to the respective communication threads 28a and 28c.
The transmission is performed synchronously between 28b and 28c.

The robot devices 1a, 1b, 1c receive commands transmitted from the corresponding communication threads 28a, 28b, 28c, respectively. Each robot device 1
a, 1b, and 1c execute the received command.

Thus, each communication thread 28a, 2
8b and 28c, the timing of sending commands is synchronized.
The actions 1b and 1c are based on cooperative actions as a whole.

Next, a specific description will be given using a robot control script RS described as shown in FIG.
The robot control script shown in FIG. 6 is for causing a plurality of robot apparatuses 1a, 1b, and 1c to perform a so-called "wave" action. The action of “wave” by the plurality of robot devices 1a, 1b, 1c is as follows.
A plurality of robot apparatuses 1a, 1b, 1c aligned as shown in a change from (A) to (C) in FIG.
However, in the posture of "sitting down," the coordinating behaviors are expressed by performing the actions of "banzai" in order.

The robot control script RS shown in FIG.
In the first line, the number of robot devices connected to the behavior management system is described. Specifically, the first column describes “NUMBER” as a menu, and the second column describes the number of connected robot apparatuses. In the case where the three robot apparatuses 1a, 1b, and 1c perform a cooperative action as in the present embodiment, the number column is "3".

From the second line onward, each communication thread 28a, 28b, 28c and each robot device 1a, 1b,
Information, commands, etc., for communicating with the device 1c are described. Specifically, the menu is written in the first column, and the communication threads 28a, 28b, and 28c correspond to the corresponding robot devices 1 in the second, third, and fourth columns.
Information for controlling the actions of a, 1b, and 1c is described.

In the second row, in the first column, "IP_ADDRESS"
And write the IP address of the server corresponding to each column. Alternatively, "HOSTNAME" may be written in the first column to describe the network host name of the server of each robot device.

The third line describes the port number of each host. In the present embodiment, “1000
Write "0".

In the fourth and subsequent rows, information for controlling the behavior of the robot device, such as commands, is described in each column corresponding to each robot device.

The commands are roughly classified into a synchronous command and an abstract command. It is described that the communication thread sends out the synchronous command and the abstract command in this order. Here, the abstract command is a command transmitted to the robot apparatus to cause the robot apparatus to perform an actual action, and the synchronous command is a cooperative action in which the action of each robot apparatus by such an abstract command is performed as a whole of the robot apparatus. Is to do so. Specifically, there are two synchronization commands, a SYNC instruction for synchronizing the command transmission timing and a WAIT instruction for synchronizing the beginning of the action of each robot device.

The SYNC instruction is provided with an ID number and an exponent (hereinafter referred to as a synchronization achievement exponent) as arguments. The ID number serves as identification information for the SYNC instruction.
It becomes identification information with other SYNC instructions. By this ID number
By identifying a SYNC instruction, appropriate processing can be performed without being confused with another SYNC instruction. The synchronization achievement index is calculated for each robot device 1a, 1
A certain value is given to each of b and 1c, and the sum thereof is set to a predetermined value. In the present embodiment, the total is set to “100”. Note that the synchronization achievement index is not limited to being set to “100”. That is, the synchronization achievement index is determined according to the application of the present invention.

This synchronization achievement index is added when the robot devices 1a, 1b, 1c are in a state where they can move to the next action. In other words, it is added when the robot device that has performed a certain action ends the action. Specifically, the synchronization achievement index is added by the following procedure.

For example, when the first robot device 1a completes the action corresponding to the command sent from the corresponding first communication thread 28a or is ready to move to the next action, Information (hereinafter referred to as standby information). When the corresponding first communication thread 28a receives the standby information transmitted by the first robot device 1a, the first communication thread 28a adds the synchronization achievement index to, for example, an area of the synchronization control global memory GM where the ID number matches (addition). In). At this time, the other communication threads perform exclusive control so as not to access this area. That is, when another robot device (here, one or both of the second robot device 1b and the third robot device 1c) is already in a state where the SYNC command corresponding to the ID number can be executed. Then, the synchronization achievement index that has just been sent is added to the synchronization achievement index of the other robot apparatus having the same ID number in the synchronization control global memory GM. Then, when it becomes possible to move to the next action of all the robot devices 1a, 1b, 1c, the total value of the synchronization achievement index becomes a predetermined value, in this example, "100".

By such processing, each of the communication threads 28a, 28b, 28c monitors the synchronization control global memory GM, and unless the total value of the synchronization achievement index reaches a predetermined value ("100"), Will not be read. Therefore, even if the robot apparatus can move to the next action, the standby state is maintained without causing the next action.

The plurality of robot devices 1a, 1b,
The personal computer 33 that manages the behavior of 1c is
By confirming that the total value of the synchronization achievement index has reached a predetermined value, the SYNC instruction corresponding to the ID number is completed in all the robot apparatuses 1a, 1b, 1c, and the robot apparatuses 1a, 1b. , 1c are ready for the next action. Thus, when the total value of the synchronization achievement index reaches a predetermined value, the personal computer 33 sets each communication thread 28a, 2
The next command is transmitted by 8b and 28c.

Next, in the command, a WAIT command is transmitted in the robot control script RS shown in FIG.

The WAIT command takes time (ms) as an argument, and the robot apparatus that has received the WAIT command forms information that waits for a specified time and then proceeds to the next command.

The arguments of the WAIT instruction are “0” and “100”, respectively.
In the embodiment having “0” and “2000”, the first robot device 1a that has received the WAIT command whose argument is “0”
Immediately executes the subsequently sent command, and upon receiving the WAIT command with the argument “1000”, the second robotic device 1b waits for one second for the subsequently sent command and executes it. Then, the third robot apparatus 1c that has received the WAIT command whose argument is “2000” waits for 2 seconds for the subsequently transmitted command and executes it. That is, for example, the WA received by the first to third robot devices 1a, 1b, 1c.
If the arguments of the IT command are all "0", the first to third robot apparatuses 1a, 1b, 1c simultaneously execute the contents of the command sent next.

In the present embodiment, the command transmitted after the WAIT instruction is a “BANZAI_SIT” command. The “BANZAI_SIT” command is a command for causing the robot apparatus to perform the “Banzai” operation in the “sitting” posture.

When such a WAIT instruction is transmitted, and subsequently, a “BANZAI_SIT” command is transmitted,
The first robot device 1a starts the command of "BANZAI_SIT" immediately after the total value of the synchronization achievement index reaches the predetermined value as described above, and the second robot device 1b starts the command of "BANZAI_SIT" one second after that. ", And the third robot apparatus 1c starts the command of" BANZAI_SIT "two seconds after that.

The following processing is roughly performed by the robot control script RS shown in FIG. 6 in which the SYNC command, the WAIT command, and the like are specified as described above.

When the robot apparatus 1a, 1b, 1c is ready to move to the next state by the first SYNC instruction (the fifth line), the synchronization achievement index is activated with ID = 1. ID
When the sum of the synchronization achievement indexes of “= 1” becomes “100”, the communication threads 28 a, 28 b, and 28 c
A command and a command of “BANZAI_SIT” are transmitted.

First data string (first robot apparatus 1a)
Since the WAIT instruction data string) D _1, which is corresponding to the (IP address 11.22.33.44) containing the 0 second ( "0"), the first robot apparatus 1a immediately "BANZAI_SIT Command, and as shown in FIG.
Perform the “Banzai” motion in the posture.

[0060] Further, the WAIT instruction D ₂ (data string is corresponding to the second robotic device 1b (IP address 11.22.33.45)) to the second data string second ( "1000") , The second robotic device 1b executes the command “BANZAI” one second after the first robotic device 1a executes the command.
_SIT ”command, and as shown in FIG.
Perform the “Banzai” operation in the “sitting” posture. Similarly third data string (data string is corresponding to the third robotic apparatus 1c (IP address 11.22.33.46)) D ₃ of WAI
Since two seconds ("2000") are included in the T command, the third robot device 1c executes the "BANZAI_SIT" command two seconds after the first robot device 1a executes the command. As shown in (C) of FIG. 7, the "banzai" operation is performed in the "sitting" posture.

When the execution of the command of “BANZAI_SIT” is completed, there is another SYNC instruction, and each robot device 1a,
It is detected that 1b and 1c are ready for the next action. As a result, the first robot apparatus 1a that has started executing first enters a waiting state, and the total of the action achievement indexes becomes “100”, so that each robot apparatus 1a,
1b and 1c are again the communication threads 28a, 28b,
The action as described above is executed again by the WAIT command and the second "BANZAI_SIT" command transmitted synchronously from 28c.

By the above processing by the robot control script RS shown in FIG. 6, the behavior management system changes the three robot apparatuses 1a, 1b, 1c so as to change from (A) to (C) in FIG. Is shifted one second at a time to execute banzai, and cooperative action of two “waves” is developed.

As described above, the behavior management system can easily realize the cooperative behavior by the plurality of robot apparatuses 1a, 1b, 1c by transmitting the command synchronously. Such a plurality of robot apparatuses 1a, 1
The cooperative actions by b and 1c increase the enjoyment of the appreciation of the user,
In addition to the pleasure of the robot device's autonomous action, it can be further expanded.

In the action management system, the actions of the plurality of robot devices 1a, 1b, 1c are managed by a common personal computer, so that the programs of the robot devices can be shared. On the other hand, each robot device 1 is controlled by one personal computer 33.
Managing the actions of a, 1b, and 1c is the same as grasping each robot device 1a, 1b, and 1c as a computer on the network. Thereby, in the behavior management system, the behavior management of the plurality of robot apparatuses 1a, 1b, 1c is facilitated, and the creation / editing of the behavior pattern is facilitated. Can be significantly shortened.

Next, a specific configuration of the above-described robot device will be described. The electrical circuit configuration of the robot device 1 is, for example, as shown in FIG.

The image data picked up by the CCD video camera 11 is supplied to a signal processing unit 12. Signal processing unit 1
2 performs predetermined processing on the image data supplied from the CCD video camera 11 and transfers the image data to the internal bus 18.
Through a DRAM (Dynamic Random Ac
cess Memory) 16.

The robot apparatus 1 has a memory stick interface 29 connected to a ROM interface 30 so that data can be recorded on and reproduced from a so-called memory stick 210.

Further, the robot apparatus 1 has a PC card slot (PC card I / F) 14. Accordingly, when the PC card 200 is a wireless LAN card, data communication with an external device (for example, the repeater 34) becomes possible, and when the PC card is a memory card, communication with the memory card is performed. Data can be recorded and reproduced.

Then, the CPU (Central Processing Uni
t) 15 reads an operation program stored in a flash ROM (Read Only Memory) 17 via the ROM interface 30 and the internal bus 18 and controls the entire system. The operation program of the CPU 11 stored in the flash ROM 17 is stored in an external personal computer (Perso
nal Computer (PC) 31.

The signals detected by the potentiometers 19a, 19b, 19c, the touch sensor 20, and the microphone 21 constituting the detecting means for detecting the external state are transmitted to the signal processing unit via the branching units 24a, 24b, 24c, 24d, 24e. 12 is supplied. The signal processing unit 12 includes the branch unit 2
The signals supplied from 4a to 24e are supplied to CPU 15 via internal bus 18. The CPU 15 controls the actuators 22a, 22b, 22 based on the supplied signals.
c, 22d (for example, the operation of the legs 2a to 2d and the head 3 of FIG. 1 driven thereby). Also, CP
U15 controls the sound output from the speaker 23.

Further, an infrared detector (IrDA) 26 is provided. The infrared detection unit (IrDA) 26 receives the command information output from the remote controller (not shown) by the operation of the user and transmits the command information to the signal processing unit 12 via the branching unit 24e.
To supply. The signal processing unit 12 supplies the signals to the CPU 15 via the internal bus 18, and the CPU 15 controls the actuators 22 a, 22 b, 22 c, 2 based on the supplied signals.
The operation of 2d is controlled to cause the robot device 1 to output an action according to the user's command.

Here, the potentiometers 19a to 19
c, touch sensor 20, microphone 21, actuators 22a to 22d, speaker 23, and infrared detector 2
6 and the like constitute the feet, ears, mouth, etc. of the robot apparatus 1, and collectively constitute a CPC (Configurable Physic).
al Component) device. The CPC device is not limited to this, and may be, for example, a measuring unit such as a distance sensor, an acceleration sensor, or a gyro.

FIG. 9 shows a configuration example of the signal processing unit 12. The DRAM interface 41 and the host interface 42 are connected to the DRAM 16 and the CPU 15, respectively, and are also connected to an external bus 44.
The bus controller 45 controls the external bus 44.
The bus arbiter 46 arbitrates the external bus 44 and the internal bus 47.

Parallel port 48 and serial port 5
0 is connected to a personal computer (PC) 31 as an external development environment, for example. The battery manager 49 manages the remaining amount of the battery 27 and the like.
The parallel port 48, the battery manager 49, and the serial port 50 are connected to the internal bus 47 via the peripheral interface 53, respectively.

The CCD video camera 11 supplies the captured image data to an FBK (Filter Bank) 56. FBK
Reference numeral 56 performs a thinning process on the supplied image data to create image data of various resolutions. The image data is transferred to the DMA (Direct Memory A
ccess) Supplied to the controller 51. The DMA controller 51 transfers the supplied image data to the DRAM 16 and stores it.

Further, the DMA controller 51
The image data stored in M16 is read out as appropriate,
It is supplied to PE (Inner Product Engine) 55. IPE5
5 performs a predetermined operation using the supplied image data. The calculation result is transferred to the DRAM 16 and stored in accordance with an instruction from the DMA controller 51.

The CPC device 25 is connected to the serial bus host controller 57. The CPC device 25 includes, for example, the potentiometers 19a to 1 described above.
9c, a touch sensor 20, a microphone 21, actuators 22a to 22d, a speaker 23, an infrared detector 26, and the like. The audio data supplied from the CPC device 25 is transmitted to a DSP (Digital Signal Processor) 52 through a serial bus host controller 57.
Supplied to The DSP 52 performs a predetermined process on the supplied audio data. A personal computer (PC) 32 or the like as an external development environment is connected to the USB interface 58. The timer 54 supplies time information to each unit via the internal bus 47.

The robot apparatus 1 is provided with a personal computer 33 for performing the cooperative action as described above.
As a structure related to wireless communication for performing data communication with the device driver 77, as shown in FIG.
It has a CP / IP protocol stack 78 and the like.

The device driver 77 is a software layer, and directly manages the PC card (wireless LAN card) 200 as an upper layer of the PC card (wireless LAN card) 200 to exchange data.

The TCP / IP protocol stack 78
Is in the software layer, supports TCP / IP services, and exchanges data between the device driver 77 and the semantics command server. The robot apparatus 1 has a unique IP address (for example, the IP address 11.22.
33.44 etc.) are written.

This TCP / IP protocol stack 7
A software module such as a semantics command server 79 is above the third. The semantics command server 79 is configured as a higher-level server that can receive a command given as a character string as described above.

In the semantics command server 79,
When the robot apparatus 1 is started, a server that receives a command is started using a predetermined port number (for example, 10000) of the TCP / IP service, and waits for a connection request from the personal computer 33 on the local area network. The connection request is made by specifying the IP address and the port number (for example, 11.22.33.34:10000) and transmitting the request.

Then, the semantics command server 79, which has established a connection with the external client (personal computer 33), starts accepting a command from the personal computer 33 and outputs the received command to the middleware layer 74. The command is sent to the semantics converter 101 to execute the action of the command content. The output semantics converter 101 can specifically recognize a command name as a character string, and thereby, the output semantics converter 10
1 controls a lower-order software module (object) according to the content of a command consisting of a character string.

The middleware layer 80 is a software group that provides basic functions of the robot device 1, and its configuration is set in consideration of the device and shape of the robot device 1. This middleware layer 80 is, specifically,
The recognition system (input system) is configured as shown in FIG.
Middleware layer 90 and output middleware layer 100
And is composed of, for example, an object group.

The middleware layer 90 of the recognition system recognizes information input from outside. As a result, the robot device 1 can autonomously determine an action according to information input from the outside. For example, the middleware layer 90 of the recognition system includes image data, sensor data,
It is composed of an object group that processes raw data of a device such as sound data and outputs a recognition result.

The objects for processing device data include, for example, a distance detecting unit 92, a touch sensor unit 93, a color recognizing unit 94, a scale recognizing unit 95, and a posture detecting unit 9.
6, a motion detection unit 97, and the like. Where, for example,
The distance detector 92 recognizes “there is an obstacle”, the touch sensor 93 recognizes “stroke” and “hit”, and the color recognizer 94 recognizes “the ball is red”.
Is recognized by the posture detecting unit 96, and “the ball is moving” is recognized by the motion detecting unit 97.

Then, in the input middleware layer 90, the input sentiment converter 91 passes the recognition information of the object as described above to the semantics command server 79 which is a higher order.

On the other hand, the output middleware layer 100 controls the device based on the information passed from the semantics command server 79. The information passed from the semantics command server 79 is, for example, a command based on internal information such as a device recognition result in the input semantics commander 91, or external information sent from the personal computer 33 in the behavior management system. Command and the like. The internal information includes, for example, the remaining battery power. The output middleware layer 100
Operates each unit by an object group configured for each operation function of the robot device, based on such a command.

As the object, the posture control unit 10
2, a motion reproducing unit 105, a fall return unit 106, a tracking unit 107, a walking module unit 108, an LED lighting unit 103, a sound reproducing unit 104, and the like. here,
For example, control relating to “motion reproduction” is performed by the motion reproducing unit 105, control relating to “return to fall” is performed by the fall return 106, control relating to “target following operation” is performed by the tracking 101, and “ The control relating to “walking” is performed. Note that “tracking” is an operation in which the user keeps looking at the moving object, specifically, an operation in which the user keeps turning his / her head toward the moving object. For example, when performing such an operation, the recognition information of the color recognition unit 94 and the motion detection unit 97 is directly used as information at the time of “tracking”. In addition, since these controls involve a change in the posture of the robot apparatus 1, the posture information is managed by the posture management 102. The sound output unit 104 controls the sound, and the LED lighting unit 103 controls the lighting of the eyes (LED).

In the output middleware layer 100, the contents of the command are interpreted by the output semantics converter 101 positioned above the object group for controlling the operation as described above, and the command is interpreted as described above according to the content. An object such as the motion reproduction unit 105 controls the device. Specifically, it generates and outputs a servo command value, output sound, and output light (eye LED) of each joint of the robot device 1 for each function. For example, if the command of "BANZAI_SIT"("Banzai") is sent as an abstract action command as described above, the object necessary for the action is controlled by the device necessary for the action of such command content. A signal is output to cause the action of "Banzai". Note that the abstract action command is not limited to this, but includes “forward”, “retreat”,
There are also commands such as "pleasure", "bark", "sleep", "exercise", "surprise", and "tracking", which are actions of animals.

Then, the output middleware layer 100 detects the operation status of the device due to the action (for example, a result of the device operation end), and the output semantics converter 101 returns the action status to the semantics command server 79. . Here, the information indicating that the operation has been successfully completed is information indicating that the robot apparatus can move to the next action.

The output middleware layer 100 as described above
By controlling each device based on the command, the robot apparatus 1 can express an action according to the command.

In the semantics command server 79,
Information about the end of execution of the output middleware layer 100 as described above is transmitted to the personal computer 33 (external client) via the PC card 200. Then, the personal computer 33 determines whether or not the robot apparatus 1 can move to the next action, as described above, in response to the notification of the end of the execution of the robot apparatus 1, and according to the determination result. To send the next command.

With the above-described configuration, the robot apparatus 1 can receive a command from the personal computer 33 and take action according to the content. Then, the end result of the command, that is, information indicating that it is possible to move to the next action, can be returned to the personal computer 33.

Note that, for example, the above-described abstract command (for example, a “Banzai” command) is realized by a combination of a plurality of basic commands, for example, a motion command, a sound command, and an LED output command. The robot apparatus 1 can execute the contents of the abstract command by storing the correspondence between such an abstract command and the basic command in a storage unit, for example, a memory stick. Have been. Further, by storing such correspondence information in the storage means as an editable file, it is possible for the user to change the operation of each unit executed by the abstract command according to his / her preference. ing.

The software layer of the robot device 1 is configured as shown in FIG. 12, for example.
The software layer is roughly divided into the application layer 12
0, middleware layer 80, manager object layer 1
30; a robot server object layer 140; and a device driver layer 150. Further, the manager object 130 has an object manager 131 and a service manager 132. For the robot server object,
It has a design robot 141, a power manager 142, a virtual robot 143, and a device driver manager 144. These function roughly as follows.

In the manager object layer 130, an object manager 131 manages activation and destruction of the application layer 120 and the middleware layer 80. A service manager 132 manages each object based on connection information between objects described in a connection file. Functions as a system object that prompts a connection.

The device driver layer 150 is an object that is allowed to directly access the device driver set 151 (for example, the above-mentioned hardware layer of the CPC device 25 or the like). That is, it constitutes a control unit immediately above which controls devices in the hardware layer. The device driver layer 150 performs processing in response to a hardware interrupt.

The robot server object layer 14
0, the designed robot 141 manages the configuration and the like of the robot apparatus 1, the power manager 142 manages the power supply, and the device driver manager 144 controls the externally connected devices such as a personal computer and a PC card. Manage access to.

Then, the robot server object layer 1
In 40, the virtual robot 143 forms a part that exchanges information between the middleware layer 80 and various device drivers under the communication protocol between the objects.

As described above, the output middleware layer 100 outputs a control signal to a lower device driver to actually control each device. Also, the above-described middleware layer 90 of the recognition system recognizes external information based on information from lower-level devices. The control of each device by the middleware layer 100 of the output system and the recognition of the status of each device by the middleware layer 90 of the recognition system are specifically performed through the virtual robot 110 as shown in FIG. Has been done.

The virtual robot 110 exchanges data between the middleware layer 90 of the recognition system and the middleware layer 100 of the output system, and device drivers constituting an input / output system to the outside. It functions as an object for bridging the object and an object operating based on the inter-object communication protocol. With the virtual robot 110, information is exchanged between the recognition middleware layer 90 and the output middleware layer 100 and the various device drivers under the communication rules between the objects.

Next, a function for the robot apparatus 1 to independently determine an action will be described. Robot device 1
Has an action model and an emotion model in order to realize a voluntary action. The behavior model and the emotion model change based on an external factor or an internal factor, whereby the robot device operates according to the output of the behavior model or the emotion model, and is configured as an autonomous robot device. The behavior model and the emotion model are constructed in the application layer 120 in the software layer shown in FIG. The emotion model 64 is constructed, for example, as shown in FIG.

First to third sensors 61, 62, 63
Detects a stimulus given from the outside of the user or the environment, converts the stimulus into an electric signal, and outputs the electric signal. This electric signal is supplied to the first and second input evaluation units 71 and 72. Here, the first to third sensors 61, 62, 63
Are potentiometers 19a to 19c, touch sensor 2
0, a microphone 21, and the like, a voice recognition sensor, an image color recognition sensor, and the like (not shown) that perform operations performed by the user to take care of the robot device 1 and voices emitted.
Converts to electrical signals and outputs. First to third sensors 6
The outputs of the first, second and first input evaluators 7
1,72.

The first input evaluator 71 evaluates the electric signals supplied from the first to third sensors 61, 62, 63 and detects a predetermined emotion. The predetermined emotion here is, for example, a feeling of joy. First input evaluation unit 71
Converts the detected emotion evaluation value into the first emotion module 73
To supply. Predetermined emotions are assigned to the first emotion module 73, and the emotion parameters increase or decrease based on the emotion evaluation values supplied from the first input evaluation unit 71. For example, when “joy” is assigned to a predetermined emotion, the parameter of the emotion of “joy” increases or decreases based on the evaluation value of the emotion supplied from the first input evaluation unit 71. . First emotion module 73
Supplies the emotion parameter to the output selection unit 75.

Similarly, the second input evaluator 72 evaluates the electric signals supplied from the first to third sensors 61, 62, 63 and detects a predetermined emotion. The predetermined emotion here is, for example, an angry emotion. The second input evaluation section 72 supplies the detected emotion evaluation value to the second emotion module 74. A predetermined emotion is assigned to the second emotion module 74, and the parameter of the emotion increases or decreases based on the evaluation value of the emotion supplied from the second input evaluation unit 72. For example, when “anger” is assigned to a predetermined emotion, the parameter of the “anger” emotion increases or decreases based on the evaluation value of the emotion supplied from the second input evaluation unit 72. Second emotion module 74
Supplies the emotion parameter to the output selection unit 75.

The output selector 75 determines whether the emotion parameters supplied from the first and second emotion modules 73 and 74 exceed a predetermined threshold, and outputs the emotion parameters exceeding the threshold. . Also, the output selection unit 75
Are the two from the first and second emotion modules 73 and 74.
If any one of the emotion parameters exceeds the threshold value, the one with the larger emotion parameter is selected and output.

The action generation section 65 converts the emotion supplied from the output selection section 75 into a command for instructing a specific action, supplies the command to the output section 66, and feeds it back to the output evaluation section 76.

The output evaluator 76 evaluates the action supplied from the action generator 65, and when the action is performed, controls to reduce the emotion parameter corresponding to the action.

The output section 66 outputs according to the action command from the action generation section 65. The output unit 66 is an output of the robot device 1, whereby the robot device 1 acts according to the action command from the action generation unit 65. That is, for example, the output unit 66 includes the legs 2a to 2d, the head 3, the body 4
Actuators 22a to drive members corresponding to
22d, a speaker 23, and the like. For example, a predetermined actuator is driven to rotate the head 3 or output a cry.

As described above, the robot apparatus 1 operates to indicate an emotional expression based on the emotional parameters of the emotional model. However, it is also possible to write predetermined data into the storage means based on the emotional parameters. For example, when the robot apparatus 1 performs an operation indicating such an emotion, the robot apparatus 1 writes a surrounding image or a surrounding voice as an external state in the storage unit. Here, the image is captured by a CCD video camera 1 serving as an external input unit that forms a part of a detection unit that detects an external state.
1, and voice is captured by a microphone 21 which is an external input means.

The above-described processing will be specifically described for the case where “joy” is assigned to the first emotion module 73 and “anger” is assigned to the second emotion module 74. Here, the following description will be made with the first sensor 61 as an image color recognition sensor, the second sensor 62 as a voice recognition sensor, and the third sensor 63 as a touch sensor 20.

The first input evaluation section 71 receives an electric signal corresponding to “yellow” from the image color recognition sensor (first sensor) 61 and a predetermined frequency (for example, from the voice recognition sensor (second sensor) 62). When an electric signal corresponding to “レ”) and an electric signal corresponding to “stroking state” are supplied from the touch sensor (third sensor) 63, the respective signals are evaluated and “joy” is provided. Is determined. The first input evaluation unit 71 supplies the evaluation value of “joy” to the first emotion module 73. The emotion module 73 increases an emotion parameter based on the evaluation value of “joy”. The emotion parameters are supplied to the output selection unit 75.

On the other hand, the second input evaluator 72 outputs an electric signal corresponding to “red” from the image color recognition sensor 61, an electric signal corresponding to a predetermined frequency (for example, “fa”) from the voice recognition sensor 62, When an electric signal corresponding to the “hitting state” is supplied from the touch sensor 63, each signal is evaluated, and an evaluation value of “anger” is determined.
The second input evaluation unit 72 supplies the evaluation value of “anger” to the second emotion module 74. Second emotion module 7
No. 4 increases the emotion parameter based on the evaluation value of “anger”. The emotion parameters are supplied to the output selection unit 75.

The output selection section 75 determines whether or not the emotion parameters supplied from the first and second emotion modules 73 and 74 exceed a predetermined threshold. here,
It is assumed that the emotion parameter of “anger” exceeds the threshold.

The action generation section 65 converts the emotion parameter of “anger” supplied from the output selection section 75 into a command for instructing a specific action (barking), supplies the command to the output section 66, and outputs the command to the output evaluation section. 76 is fed back.

The output section 66 outputs according to the action command (barking) from the action generation section 65. That is, the corresponding sound is output from the speaker 23. When the robot apparatus 1 barks, the “anger” is released, and the “anger” is released.
Is suppressed. In consideration of this, the output evaluation unit 76 reduces the emotion parameter of the second emotion module 74.

The emotion model of the robot device 1 has been described above. Next, an action model for determining an action of the robot device 1 based on various information will be described with reference to FIG.

As shown in FIG. 14, the behavior model determines a behavior output for causing the robot apparatus 1 to operate based on a sensor input. Here, the sensor input is an input from a sensor for acquiring external information such as the potentiometers 19a to 19c in the CPC device 25. Specifically, the information is information from the input semantics converter 91 that has obtained recognition information from the CPC device 25.

This behavior model M ₃ has a plurality of transition state tables having different behavior purposes as subsystems. More specifically, as shown in FIG. 15, the subsystems include system management F ₁ ,
Posture management F ₂ , which aims to manage posture, obstacle avoidance F ₃ , which aims to avoid obstacles,
Reflection F ₄ to the reflection behavior and action purposes, game F ₇ to emotional expression F _5, the autonomous behavior in general F _6, the action purpose execution of the game to the autonomous action action action purpose of general and action aimed at emotional expression , A performance F ₈ for the purpose of executing the performance, a soccer F ₉ for the purpose of soccer operation, and a record F _{10 for the} purpose of storing data.
Has a state transition table etc., the behavior model M _3, and determines the action output with the transition from the current state on the basis of such a state transition table to the state of interest.

For example, priorities are assigned to the state transition tables, and the status transition tables are associated with each other so that actions having high importance are executed with priority. In this example, the record F ₁₀ , the soccer F ₉ , the performance F ₈ , the game F ₇ , the general autonomous behavior F ₆ , the emotion expression F ₅ , the reflection F ₄ , the obstacle avoidance F ₃ , the posture management F _2, and the system management F _1. In this order, the priorities are higher in order of the system management F ₁ , the posture management F ₂ , the obstacle avoidance F ₃ , the reflection F ₄ , the emotion expression F ₅ , and the sensor input from the CPC device 25. autonomous behavior general F _6, game F _7, performance F _8, soccer F ₉ and recording F
It will be executed in priority order of ₁₀ .

For example, for the state transition table, the principle of an algorithm called a probabilistic automan that determines a state to transition stochastically based on the transition probability is used.

The probability automan is, as shown in FIG.
n (n is an integer) states are represented by nodes NODE _{0 to} NOD.
When expressed as E _n, the arc ARC of whether to transition from one node NODE ₀ to other nodes NODE ₁ ~NODEn, connecting each node NODE ₀ ~NODE _n
Is an algorithm determined by probabilistic based on the set transition probability P ₁ to P _n respectively _{1 ~ARC n.} Here, the arc is defined as a state realized in the device (robot device 1) in advance, and the operation of the device when transitioning between the states in order to transition the operation of the device between the defined states. It shows.

By applying the Stochastic Automan algorithm to the state transition table, when the current state is at the first node NODE ₀ , the current state and the CPC
The next node is determined based on information for state transition such as a sensor input of the device 25.

As described above, the action model is not limited to the action output based on the state transition table, and other means can be adopted. For example, a behavior model can be constructed using a neural network that refers to an information processing mechanism in a neural network.

It is needless to say that the subsystems constituting the behavior model are not limited to those described above.

In addition, when the action model M ₃ outputs an action, as shown in FIG. 14, an emotion value (emotion parameter) which is an output signal of the emotion model and an instinct value (instinct parameter) which is an output signal of the instinct model are output. Is referring.

In the emotion model M ₁ , the emotion parameter is
As described above, the value increases and decreases according to the input evaluation value based on the sensor input from the CPC device 25, and increases and decreases according to the output evaluation value obtained when an action is taken. In other words, the emotion model M ₁ is emotion parameters are updated by the input evaluation and output evaluation. The emotion model M ₁
Is based on a response, an internal state, or a change over time in response to an input from the outside world, and is based on sadness, fear, surprise, disgust, etc. in addition to the above-mentioned anger and joy.

Similarly, in the instinct model M ₂ , the instinct parameter increases and decreases according to an input evaluation value based on a sensor input from the CPC device 25 and increases and decreases according to an output evaluation value obtained when an action is taken. I do. That is, for even instinct model M _2, instinct parameters are updated by the input evaluation and output evaluation. Note that the instinct model M _2, primarily the internal state as a factor, as it gradually changes, based for example, appetite, movement desire, rest greed, affection greed, thirst for knowledge, mainly to desire such libido Model It is. For example, the instinct model of “appetite” can be obtained by referring to the remaining battery power.

Further, as shown in FIG. 14, the action model M ₃ is updated by the learning module M ₄ to update the action selection probability.

The learning module M ₄ is a module for reflecting past information in future actions and the like, and learns past actions, for example. For example, the learning module M ₄ based on the learning result, changing the priority of the sub-systems in the behavioral model M ₃ (the state transition table) (action selection probability). As a result, a subsystem in which past information is reflected is selected.

As described above, the behavior model M ₃ refers to the emotion value indicating the emotion parameter and the instinct value indicating the instinct parameter which change according to the input evaluation value and the output evaluation value, and furthermore, the learning module M ₄ The final action output is made by the subsystem whose degree is changed.

[0133] behavior selection module 160, so that the operation corresponding to the behavioral model M ₃ of the action output by the behavior object, and controls the CPC device 25, for example limbs, head,
The tail or the like is moved to complete the intended action. Then, the operation is fed back to the feeling model and M ₁ instinct model M ₂ described above is an output evaluation value described above. Note that the transfer of information between the CPC device 25 and the behavior module is specifically performed via the output semantics converter 101 described above.

As described above, the part that defines the emotion and behavior of the robot device 1 is configured. Robot device 1
According to the behavior model and the emotion model constructed as described above, the behavior is determined according to the change of the external factor caused by the external state or the internal factor caused by the internal state,
Cause the actual action. That is, the robot device is configured as an autonomous robot device that autonomously determines an action according to an external factor or an internal factor.

In the behavior management system, the autonomous robot device 1 performs a cooperative behavior with another robot device by receiving a command transmitted in synchronization. Thereby, the enjoyment of appreciation of the user is expanded.

In the above-described embodiment, a case has been described where the actions of a plurality of robot devices are managed by a personal computer. However, the present invention is not limited to this. As shown in FIG. 17, the robot device 1a can manage the actions of the other plurality of robot devices 1b. That is, a closed local area network is formed only between a plurality of robot devices. As a result, the behavior of another robot device can be managed by one robot device 1a. in this case,
The configuration of the personal computer 33 for managing the actions of a plurality of robot devices is the same as that of the robot device 1a.
Prepare for.

Further, in the above-described embodiment, a case has been described in which “wave” is performed as a cooperative action to be executed by a plurality of robot apparatuses. However, the present invention is not limited to this, and it is possible to cause a cooperative action with respect to actions that can be expressed, such as voice and motion. Further, the number of robot devices for behavior management is not limited to three as described above. More specifically, the following cooperative actions can be performed.

Each robot device emits a single sound. Thereby, a chord is generated by the cooperative action of the plurality of robot devices. Further, the robot devices may be bowed at the same time. As a result, a plurality of bowing robot devices perform cooperative actions such as generating a chord.

Further, a cooperative action of "vocalization practice" can be performed. More specifically, one robot device (a robot device capable of managing the behavior of another robot device) is a conductor, and a plurality of aligned robot devices are sequentially pointed by an arm. The pointed robot device utters while waving its arm. When the conductor's robot device sequentially points at the robot device, it utters in sequence accordingly. Finally, all the members utter at the signal of the conductor's robot device.

Further, a cooperative action of "ringing" can be performed. In this case, as in the case of the “vocalization practice”, the singing is performed while shifting the timing according to the timing pointed by the robot device of the conductor.

Further, it is also possible to cause a cooperative action of “gymnastics”. Specifically, for example, the robot device performs gymnastics in accordance with an instruction from a robot device of the instructor (a robot device that manages the behavior of another robot device). When there is a robot device that does not hear what is said, the robot device of the instructor stops the instruction to the whole and scolds only the robot device that does not listen to the instruction. For example, howling and scolding. The scolded robot device apologizes, and the instructor robot device gives instructions again. And this time, everyone will do gymnastics at the same time.

Since the cooperative action as described above is made possible by editing a script (script), the content of the cooperative action can be easily changed to the various contents described above.

Further, in the above-described embodiment, a case has been described in which a plurality of robot apparatuses constituting one group are caused to cooperate by the personal computer 33 for managing the action. However, the present invention is not limited to this, and the behavior management system can also manage cooperative behavior for a plurality of groups. Specifically, a first group of three robots is caused to perform a certain cooperative action, and a second group of four other robots is simultaneously subjected to a different cooperative action different from the first group. Let it.
Thereby, the enjoyment of the user's appreciation is widened.

In the above-described embodiment, a case has been described where the robot apparatus 1 voluntarily determines an action based on an emotion model or an instinct model. However, the present invention is not limited to this, and the action may be determined by another program or another algorithm.

Further, in the above embodiment, the appearance of the robot apparatus 1 as shown in FIG. 1 has been described. However, the present invention is not limited to this. As shown in FIG. 18, the appearance may be closer to a "dog".

In the above-described embodiment, the robot device has been described as having the shape of a "dog". However, the present invention is not limited to this. For example, other four-legged animals,
It can also be applied to a humanoid robot device.

[0147]

The behavior management system according to the present invention comprises an autonomous robot device for autonomously determining an action, and a function for determining an action by a command input from the outside. An action management system that manages the action of an apparatus, wherein a plurality of robot apparatuses are provided with an action management unit that transmits a command in synchronization, so that a plurality of robot apparatuses can perform cooperative actions.

Further, the action management method according to the present invention has a function of determining an action by a command input from the outside to an autonomous robot apparatus for autonomously determining an action. An action management method for managing an action, in which a command is synchronously transmitted to a plurality of robot apparatuses so that a plurality of robot apparatuses can perform a cooperative action.

Further, the robot apparatus according to the present invention is an autonomous robot apparatus which determines an action autonomously, and includes an action management means for synchronously sending a command to another plurality of robot apparatuses. Thereby, the plurality of robot devices can be caused to cooperate by the behavior management of the robot devices.

[Brief description of the drawings]

FIG. 1 is a diagram illustrating an external configuration of a robot device whose behavior is managed in a behavior management system according to an embodiment of the present invention.

FIG. 2 is a block diagram showing a circuit configuration of the robot device described above.

FIG. 3 is a perspective view showing the above-described robot device when the user automatically “hands”.

FIG. 4 is a block diagram showing a behavior management system according to the embodiment of the present invention.

FIG. 5 is a diagram showing a configuration of software and the like in a personal computer for managing the behavior of the robot device in the behavior management system described above.

FIG. 6 is a diagram showing a specific example of a robot control script for controlling actions of a plurality of robot devices.

FIG. 7 is a diagram illustrating a “wave” operation by a cooperative action of a plurality of robot devices.

FIG. 8 is a block diagram showing a specific circuit configuration of the robot device described above.

FIG. 9 is a block diagram illustrating a signal processing unit of the robot device described above.

FIG. 10 is a block diagram illustrating a portion that enables a wireless communication function in the robot device described above.

FIG. 11 is a block diagram showing a middleware layer of a software layer of the robot device described above.

FIG. 12 is a block diagram illustrating a configuration example of a software layer of the robot device described above.

FIG. 13 is a block diagram showing a part constituting an emotion model of the robot device described above.

FIG. 14 is a block diagram showing a part constituting a behavior model of the above-described robot apparatus.

FIG. 15 is a diagram showing a specific example of the behavior model described above.

FIG. 16 is a diagram used to explain a probabilistic automan which is an algorithm for determining a state transition in the above-described behavior model.

FIG. 17 is a diagram used to describe a robot device that manages the behavior of another robot device.

FIG. 18 is a perspective view showing a specific appearance of the robot device.

[Explanation of symbols]

1 robot apparatus, 15 CPU, 16 data storage unit, 33 personal computer, 34 repeater

 ──────────────────────────────────────────────────続 き Continued on the front page F term (reference) 2C150 AA14 BA06 CA02 DA05 DA24 DA26 DA27 DA28 3F059 AA00 BA00 BB06 BC06 FC01 5B089 GA23 GB02 HA11 JA35 JB10 KA13 KB04 KB06

Claims (10)

[Claims] 1. An action management system for autonomously determining an action, comprising a function of determining an action by a command input from the outside, and managing the action of the autonomous robot apparatus. A behavior management system, comprising: behavior management means for synchronously sending a command to the plurality of robot devices. 2. The behavior management means monitors the state of each robot device. The behavior management means detects that each robot device is ready for the next behavior, and synchronizes the command. 2. The behavior management system according to claim 1, wherein the action management system sends the information. 3. Each command is provided with identification information. The action management means monitors the state of each robot device based on the identification information. 3. The behavior management system according to claim 2, wherein each of the robot devices to which the command is sent is detected to be ready for the next action, and the command is sent synchronously. 4. Each command is provided with an index corresponding to each robot device to which each command is transmitted, so that the sum becomes a predetermined amount. Adding the index attached to the command sent to the robot device when the robot is ready to move to the state, and synchronously sending the command when the sum reaches the predetermined amount. Item 4. The behavior management system according to Item 2. 5. An action management method for autonomously determining an action, comprising a function of determining an action by a command input from the outside, and managing the action of the autonomous robot apparatus. A behavior management method, wherein a command is synchronously transmitted to a plurality of the robot devices. 6. The apparatus according to claim 5, wherein the status of each robot device is monitored, and it is detected that each robot device is ready to move to the next action, and the command is transmitted synchronously. Behavior management method. 7. An autonomous robot apparatus for autonomously determining an action, comprising: a behavior management means for transmitting a command to another plurality of robot apparatuses in a synchronized manner. 8. The behavior management means monitors the state of each of the other robot devices, and the behavior management means detects that the other robot devices are ready for the next behavior. 8. The robot apparatus according to claim 7, wherein said command is transmitted synchronously. 9. Each command is provided with identification information, wherein the behavior management means monitors the status of each of the other robot devices based on the identification information. Detecting, based on the identification information, that each of the other robot devices to which a command having the same identification information is transmitted is ready for the next action, and transmitting the command synchronously; Item 9. The robot device according to item 8. 10. Each command is provided with an index corresponding to each of the other robot devices to which the command is transmitted, so that the sum becomes a predetermined amount. When the robot device is ready to move to the next state, add the index attached to the command to be sent to the robot device and, when the sum reaches the predetermined amount, send the command synchronously. The robot device according to claim 8, wherein: