Anomaly notification system, anomaly notification method, and program

Abstract

To provide a technique for making a robot appear to have better conversation capability.SOLUTION: A conversation output system 1000 includes: first information acquisition means for acquiring first information which is information of at least one of information related to a user of a robot 100 and situation information which is information of a situation around the robot 100; creation means for creating conversation data for making the user recognize that the robot 100 and a predetermined target have a conversation matching at least the first information, based on the acquired first information; and output control means for controlling an output unit to output information based on the created conversation data, thereby making the user recognize that the robot 100 and the predetermined target have a conversation matching at least the first information, where the robot 100 does not have a function of executing a conversation of a level equal to or higher than a level of the conversation based on the conversation data.SELECTED DRAWING: Figure 1

Description

[[Technical Field]]

[0001] The present invention relates to Anomaly notification system , an abnormality notification method and a program. [[Background Art]]

[0002] Robots imitating animals have become widespread. Such robots are also called electronic pets and have an appearance that users will love. In addition, robots that can recognize users' voices, perform actions according to users' instructions, and even converse with each other have emerged. For example, Patent Document 1 discloses a robot system in which a plurality of robots converse with one another. [[Prior Art Literature]] [[Patent Literature]]

[0003] [[Patent Document 1]] Japanese Unexamined Patent Application Publication No. 2003-205483 [[Summary of the Invention]] [[Problem to be Solved by the Invention]]

[0004] Conventionally, a robot system that enables robots to converse with each other is disclosed, for example, in Patent Document 1. The content of conversations between the plurality of robots described in Patent Document 1 is based on a script with lines preset according to the number of conversing participants, resulting in fixed and rigid conversations. Although reasonably advanced conversations can be achieved by making full use of AI (Artificial Intelligence) technology using a supercomputer, it is not realistic to mount such technology on consumer products for general users. Therefore, there is room for improvement in the conversation technology between robots. Such a problem applies not only to the conversation technology between multiple robots, but also to the conversation technology between a robot and a predetermined target.

[0005] The present invention provides robot that can enhance the sense of being a living organism Anomaly notification systemThe purpose is to provide a method and program for notifying abnormalities. [Means for solving the problem]

[0006] To achieve the above objective, the abnormality notification system according to the present invention includes a robot that mimics a living creature, and when an abnormality is detected in the robot through self-diagnosis by the robot, the system notifies a designated service center of countermeasures for the abnormality. via wireless communication In addition to making the above inquiry Ta Content The conversation between the designated service center staff and the robot was expressed in natural language so that it would appear as such. The system comprises an output terminal that displays information on the location of the abnormality in a form corresponding to a part of a living organism. [Effects of the Invention]

[0007] According to the present invention, robot It is possible to enhance the biological realism Anomaly notification system We can provide an abnormality notification method and program. [Brief explanation of the drawing]

[0008] [Figure 1] This figure shows the functional configuration of a conversation output system according to Embodiment 1 of the present invention. [Figure 2] This figure shows an example of the appearance of the robot according to Embodiment 1. [Figure 3] This figure shows an example of data stored in the scenario storage unit according to Embodiment 1. [Figure 4] This figure shows an example of data stored in the user profile storage unit according to Embodiment 1. [Figure 5] This figure shows an example of data stored in the user status storage unit according to Embodiment 1. [Figure 6] This is a flowchart of the thread group for collecting status information of conversation processing according to Embodiment 1. [Figure 7] This is a flowchart of the thread group for collecting conversation requests in the conversation processing according to Embodiment 1. [Figure 8]It is a flowchart of a thread group for conversation content output of conversation processing according to Embodiment 1. [Figure 9] It is a flowchart of scenario selection processing according to Embodiment 1. [Figure 10] It is a diagram showing an example of conversation content output to one conversation output terminal in the conversation output system according to Embodiment 1. [Figure 11] It is a diagram showing an example of conversation content output to the other conversation output terminal in the conversation output system according to Embodiment 1. [Figure 12] It is a diagram showing an example of data stored in the scenario storage unit according to Modification 3 of Embodiment 1 of the present invention. [Figure 13] It is a diagram showing the functional configuration of the conversation output system according to Embodiment 2 of the present invention. [Figure 14] It is a diagram showing the functional configuration of the conversation output system according to Embodiment 3 of the present invention. [Figure 15] It is a diagram showing the functional configuration of the conversation output system according to Embodiment 4 of the present invention. [Figure 16] It is a diagram showing the functional configuration of the conversation output system according to Embodiment 5 of the present invention. [Figure 17] It is a diagram showing the functional configuration of the conversation output system according to Embodiment 6 of the present invention. [Figure 18] It is a flowchart of a thread group for pseudo chat according to Embodiment 6. Mode for Carrying Out the Invention

[0009] Hereinafter, a conversation output system according to an embodiment of the present invention will be described with reference to the drawings. In the drawings, the same or corresponding portions are denoted by the same reference symbols.

[0010] (Embodiment 1) The conversation output system according to Embodiment 1 of the present invention is a system that allows a robot, which cannot carry out a conversation in natural language by voice, to make it appear to the user as if it is having a conversation in natural language with another robot.

[0011] As shown in Figure 1, the conversation output system 1000 according to Embodiment 1 of the present invention comprises a robot 100, a conversation output terminal 200, and a conversation output server 300. The robot 100 and the conversation output terminal 200 are wirelessly connected and paired by short-range wireless communication. The conversation output system 1000 may include multiple pairs of robots 100 and conversation output terminals 200 (for example, robot 100 and conversation output terminal 200, robot 100' and conversation output terminal 200', robot 100'' and conversation output terminal 200'', etc.). The robot 100 can appear to the user as if it is conversing in natural language with other robots 100 via the conversation output terminal 200 and the conversation output server 300.

[0012] Robot 100 is a cute, pet-type robot, and its functional configuration includes a situation acquisition unit 110 and a short-range communication unit 120. These functional configurations will be described later. Although not shown in the figures, robot 100 may also be equipped with a clock that can acquire the current date and time.

[0013] Furthermore, as shown in Figure 2, the robot 100 is equipped with a human presence sensor 131 at the eye position, a camera 132 at the nose position, a microphone 133 at the ear position, and a speaker 134 at the mouth position. The robot 100 can express emotions by moving its legs 135 and tail 136. For example, flapping its legs 135 back and forth and side to side, or wagging its tail 136, indicates happiness. In addition, although not shown, the robot 100 can detect when it is being held by a user by being equipped with an acceleration sensor, a tactile sensor, etc. In addition, although not shown, it can acquire information about the surrounding environment by being equipped with a temperature sensor, a humidity sensor, etc.

[0014] The functional configuration of robot 100 will now be described. The situation acquisition unit 110 is equipped with a human presence sensor 131, a camera 132, a microphone 133, an acceleration sensor, a tactile sensor, a temperature sensor, a humidity sensor, etc., and acquires situation information, which is information about the surrounding environment of robot 100, using these. Situation information includes information about the user of robot 100 and information about robot 100. For example, information about the user of robot 100 includes information obtained from the human presence sensor 131 such as "the user approached robot 100", information obtained from the camera 132 such as "the user looked at robot 100", information obtained from the microphone 133 such as "the user spoke to robot 100", and information obtained from the acceleration sensor or tactile sensor such as "the user picked up robot 100". Information about robot 100 includes information obtained from the temperature sensor such as "the current temperature around robot 100", information obtained from the humidity sensor such as "the current humidity around robot 100", and information obtained from the clock such as "the current date and time in the area where robot 100 is located". Regarding user information, the frequency (number of occurrences) may be included in the status information. Here, the various sensors provided by the status acquisition unit 110 are collectively referred to as the first sensor.

[0015] Furthermore, the situation information may include only one of either information about the user of robot 100 or information about robot 100. For example, in the scenario selection process described later, if information about robot 100 is not used and only information about the user of robot 100 is used, then the situation information may include only information about the user of robot 100 and may not include information about robot 100.

[0016] The short-range communication unit 120 is equipped with a short-range wireless communication device (for example, a Bluetooth® communication device) and transmits the status information acquired by the status acquisition unit 110 to the conversation output terminal 200. Therefore, the short-range communication unit 120 functions as a status transmission unit.

[0017] Although robot 100 is equipped with a speaker 134, it does not have the ability to speak natural language. Robot 100 can output electronic sounds or sounds that mimic animal noises from the speaker 134. Therefore, as long as the user is listening to the sounds that robot 100 outputs, they cannot assume that robot 100 understands human speech. Here, robot 100 may have the ability to converse in a rudimentary way, like a baby or toddler, but it does not have the ability to perform conversations at a level higher than the conversational data described later.

[0018] Next, the conversation output terminal 200 will be described. The conversation output terminal 200 is a smartphone equipped with a display 270, and its functional configuration includes a short-range communication unit 210, an LTE (Long Term Evolution) communication unit 220, and a conversation output unit 231 (terminal control unit 230). Although not shown in the figures, the terminal control unit 230 also has a clock function and can obtain the current date and time. The display 270 is composed of, for example, an LCD (Liquid Crystal Display), an EL (Electroluminescence) display, etc., and displays the content instructed by the conversation output unit 231 (terminal control unit 230). The display 270 functions as an output unit.

[0019] The short-range communication unit 210 is equipped with a short-range wireless communication device that can communicate using the same communication standard as the short-range communication unit 120, and receives status information transmitted from the robot 100. Therefore, the short-range communication unit 210 functions as a status receiving unit.

[0020] The LTE communication unit 220 is equipped with a device that communicates using the LTE communication standard and communicates with the conversation output server 300. For example, the LTE communication unit 220 transmits status information to the conversation output server 300. The LTE communication unit 220 also receives data for the scenario described later that has been transmitted by the conversation output server 300. Since the LTE communication unit 220 communicates with the conversation output server 300, it functions as a server-to-server communication unit.

[0021] In some cases, depending on the robot 100 (for example, if it is not equipped with a clock or temperature sensor), the status information transmitted to the conversation output terminal 200 may not include information about the current date and time, weather, temperature, news, etc., for the area where the robot 100 is located. In this case, the conversation output terminal 200 adds the current date and time, weather, temperature, news, etc., for the area where the robot 100 is located to the status information and transmits the status information with this information added to it to the conversation output server 300 via the LTE communication unit 220. The conversation output terminal 200 can obtain the current date and time using the clock provided in the terminal control unit 230. The conversation output terminal 200 can also access the internet via the LTE communication unit 220 to obtain information such as weather, temperature, and news.

[0022] The terminal control unit 230 consists of a CPU (Central Processing Unit) and other components, and realizes the functions of the conversation output unit 231 by executing a program stored in a memory unit (not shown). For example, after purchasing the robot 100, a user can install the robot application on the conversation output terminal 200 (smartphone) by accessing the robot application's URL (Uniform Resource Locator) with the conversation output terminal 200. This robot application then realizes the functions of the conversation output unit 231. The robot application's URL will be listed, for example, in the robot 100's instruction manual.

[0023] The conversation output unit 231 controls the display 270 to display conversation content on the display 270 that appears to be a conversation in natural language with another robot 100', based on scenario data received from the conversation output server 300 via the LTE communication unit 220 (described later). The conversation output unit 231 can be implemented as SNS client software that accesses the SNS (Social Networking Service) provided by the conversation output server 300. The conversation output unit 231 functions as an output control means.

[0024] For example, consider a scenario where a screen is output that makes it appear as if robot 100 (wirelessly connected to conversation output terminal 200) and robot 100' (wirelessly connected to conversation output terminal 200') are having a conversation. In this case, the conversation output unit 231 receives only the content of robot 100's statements as scenario data from the conversation output server 300, and transmits the received scenario data as content to be posted to the SNS provided by the conversation output server 300. The conversation output unit 231 of the conversation output terminal 200', which is wirelessly connected to the conversation partner (robot 100'), also receives only the content of robot 100's responses as scenario data from the conversation output server 300, and transmits the received scenario data as content to be posted to the SNS provided by the conversation output server 300.

[0025] Then, due to the SNS function of the conversation output server 300, the conversation output unit 231, acting as SNS client software, outputs the content of what robot 100 said and the content of what robot 100' replied to the display 270 of the conversation output terminal 200.

[0026] Next, the conversation output server 300 will be explained. The conversation output server 300 is a cloud server located on the internet and provides a social networking service (SNS) for robots 100 to communicate with each other. By having the conversation output terminal 200, which is wirelessly connected to the robot 100, access the SNS provided by the conversation output server 300, the robot 100 can appear to the user as if it is conversing (chatting) with other robots 100.

[0027] The conversation output server 300 comprises, as a functional configuration, an LTE communication unit 310, a scenario selection unit 321 (control unit 320), a scenario storage unit 331 (storage unit 330), a user profile storage unit 332 (storage unit 330), and a user status storage unit 333 (storage unit 330).

[0028] The LTE communication unit 310 is equipped with a device that communicates using the LTE communication standard and communicates with the conversation output terminal 200. For example, the LTE communication unit 310 receives status information transmitted by the conversation output terminal 200. The LTE communication unit 310 also transmits scenario data to the conversation output terminal 200. In this way, the LTE communication unit 310 communicates with the conversation output terminal 200 and functions as a terminal-to-terminal communication unit.

[0029] The control unit 320 consists of a CPU and other components, and executes programs stored in the memory unit 330 to realize the functions of the scenario selection unit 321. The control unit 320 is also equipped with a clock (not shown) and can obtain the current date and time in the area where the robot 100 is located.

[0030] The scenario selection unit 321 selects scenario data stored in the scenario storage unit 331 based on the status information received by the LTE communication unit 310. When selecting scenario data to make it appear as if robot 100 and robot 100' are conversing, the scenario selection unit 321 uses not only the status information of robot 100 but also the status information of robot 100', as will be described later. Although both are status information, if we want to distinguish between the two, we will refer to the status information of robot 100 as the first status information and the status information of robot 100' as the second status information.

[0031] The memory unit 330 consists of ROM (Read Only Memory), RAM (Random Access Memory), etc., and functionally includes a scenario memory unit 331, a user profile memory unit 332, and a user status memory unit 333. The ROM stores the program executed by the CPU of the control unit 320 and the data necessary in advance for executing the program. The RAM stores data that is created or modified during program execution.

[0032] As shown in Figure 3, the scenario memory unit 331 stores, for each conversation topic, the content that robot 100 will say and the content that its conversation partner (referred to here as robot 100') will respond to, based on situational information, etc. This pair of statements made by robot 100 and its conversation partner is referred to here as a scenario.

[0033] For example, the top table in Figure 3 shows an example of scenario data when the topic is the level of intimacy in communication with the master (the respective users of robot 100 and 100'). In Figure 3, the scenario data (what robot 100 says and what robot 100' replies to it) is defined according to the level of intimacy of robot 100 and the level of intimacy of robot 100'. The level of intimacy represents the degree of the user's interaction with robot 100 (defined by comprehensively considering factors such as the number of times the user has spoken to it, the number of times it has looked at its face, the number of times it has been hugged, etc.).

[0034] For example, the intimacy level is 1 if there is no interaction at all (the user neither speaks to Robot 100 nor holds Robot 100), and 5 if there is frequent interaction (for example, the number of times spoken to + the number of times looked at + the number of times held is 10 or more). Also, the part labeled "[I]" in Figure 3 indicates where a personal pronoun should be inserted. In countries where personal pronouns differ by gender, this part changes depending on the gender of Robot 100. For example, if the country of residence is Japan, "[I]" will be "boku" if Robot 100 (the user) is male, and "watashi" if female. You can also put a pet's name here. For example, if the pet's name is "GladDog", it might say, "That's nice. GladDog's owner never plays with him."

[0035] Furthermore, the table in the middle of Figure 3 shows an example of scenario data when the topic is weather. In this example, if the weather included in the situation information of robot 100 (first situation information) is sunny and the weather included in the situation information of robot 100' (second situation information) is rainy, then the content of robot 100's statement is defined as "It's sunny and nice today," and the content of robot 100's response is defined as "That's nice. It's raining here and it's boring."

[0036] Furthermore, the table at the bottom of Figure 3 shows an example of scenario data when the topic is an anniversary. The current date and time of robot 100 and the current date and time of robot 100' are only different due to the time difference between their respective locations, so the difference in dates is at most one day. Therefore, in this example, the scenario data is defined while ignoring the date and time information included in the status information of robot 100'. Of course, taking the time difference into account, separate scenario data could be defined for cases where the date and time included in the status information of robot 100' is one day before or one day after the date and time included in the status information of robot 100.

[0037] As shown in Figure 4, the user profile storage unit 332 stores the ID and icon of the robot 100 on the SNS provided by the conversation output server 300, the user's country of residence, gender, birthday, hobbies, etc., and the profiles of the robot 100 and its user. In Figure 4, only the user's gender, birthday, etc. are stored, so the robot 100's gender, birthday, etc. are treated as the same as the user's. However, separate items for the robot 100's gender, birthday, etc. may be added to the user profile storage unit 332, allowing for separate gender and birthday settings from the user.

[0038] As shown in Figure 5, the user status memory unit 333 stores status information transmitted from each robot 100 (each conversation output terminal 200). The user status memory unit 333 is updated each time it receives status information from each conversation output terminal 200. In addition, counts such as "number of times spoken to" and "number of times stared at" are reset to 0 at regular intervals (for example, at midnight every day).

[0039] The control unit 320 creates conversation data to make the user perceive that the robot 100 is having a conversation that conforms to the first information by selecting scenario data stored in the scenario storage unit 331 based on user information stored in the user profile storage unit 332 and situation information stored in the user situation storage unit 333. The control unit 320 may also select scenario data stored in the scenario storage unit 331 based on only one of the user information stored in the user profile storage unit 332 or the situation information stored in the user situation storage unit 333. At least one of the user information stored in the user profile storage unit 332 and the situation information stored in the user situation storage unit 333 is called the first information. When acquiring the first information, the control unit 320 functions as a first information acquisition means.

[0040] Furthermore, in the conversation data created by the control unit 320, status information of the conversation partner of robot 100 (referred to as robot 100') and information about the user of robot 100' are also stored in the user status storage unit 333 and the user profile storage unit 332, respectively. At least one of the information regarding the user of robot 100' and the status information of robot 100' can also be called first information, but if we want to distinguish it from at least one of the information regarding the user of robot 100 and the status information of robot 100 (first information), then at least one of the information regarding the user of robot 100' and the status information of robot 100' is called second information. When acquiring second information, the control unit 320 functions as a second information acquisition means.

[0041] Next, we will describe the conversation process in which robot 100 converses with other robots 100 via the conversation output terminal 200 and the conversation output server 300. This conversation process consists of three thread groups: a thread group for collecting situation information, a thread group for collecting conversation requests, and a thread group for outputting conversation content. These three thread groups are not executed sequentially, but rather each is executed in parallel and simultaneously. The thread group for outputting conversation content transmits the scenario data stored in the scenario storage unit 331 (after being modified as necessary) to the conversation output terminal 200, which is wirelessly connected to robot 100. Based on this scenario data, the conversation output terminal 200 outputs a screen that appears to show robot 100 conversing (chatting) with a conversation partner (for example, robot 100') on the SNS provided by the conversation output server 300. First, we will explain the thread group for collecting situation information with reference to Figure 6.

[0042] When power is turned on to robot 100, the status transmission thread shown in the flowchart on the left side of Figure 6 is activated. Similarly, when power is turned on to conversation output terminal 200, the status relay thread shown in the flowchart in the middle of Figure 6 is activated. Furthermore, when power is turned on to conversation output server 300, the status update thread shown in the flowchart on the right side of Figure 6 is activated. In Figure 6, the dotted arrows indicate that data is sent and received between each thread.

[0043] In the status transmission thread, the robot 100 acquires status information with the status acquisition unit 110 (step S101). Then, it transmits the acquired status information to the conversation output terminal 200 via the short-range communication unit 120 (step S102), and returns to step S101. Note that in step S101, the robot 100 may continuously record (record audio or video) the acquired status information and execute step S102 as appropriate, or it may transmit the entire recorded (recorded audio or video) data to the conversation output terminal 200 for analysis.

[0044] In the status relay thread, the conversation output terminal 200 waits until status information is sent from the robot 100, and once it is sent, the short-range communication unit 210 receives the status information (step S111). Then, information such as the current date and time and weather is added to the received status information as needed, and it is sent to the conversation output server 300 via the LTE communication unit 220 (step S112), and the process returns to step S111.

[0045] Furthermore, in the status update thread, the conversation output server 300 waits until status information is sent from the conversation output terminal 200, and once it is sent, the LTE communication unit 310 receives the status information (step S121). Then, using the received status information, the status information stored in the user status storage unit 333 is updated (step S122), and the process returns to step S121.

[0046] If there are multiple pairs of robots 100 and conversation output terminals 200, the above processing is performed on each robot 100 and each conversation output terminal 200 that is communicating with the robot 100. Therefore, the user status storage unit 333 of the conversation output server 300 accumulates status information about the user of each robot 100.

[0047] The above describes the thread group for gathering situational information. Next, the thread group for gathering conversation requests will be explained with reference to Figure 7.

[0048] When power is turned on, the request transmission thread shown in the flowchart on the left side of Figure 7 is activated in robot 100. Similarly, when power is turned on, the request relay thread shown in the flowchart in the middle of Figure 7 is activated in conversation output terminal 200. Furthermore, when power is turned on, the request update thread shown in the flowchart on the right side of Figure 7 is activated in conversation output server 300. In Figure 7, the dotted arrows indicate that data is sent and received between each thread.

[0049] In the request transmission thread, robot 100 determines whether a user has approached using the human presence sensor 131 or the like provided by the status acquisition unit 110 (step S201). If no user has approached (step S201; No), the process returns to step S201. If a user has approached (step S201; Yes), robot 100 sends a conversation request packet to the conversation output terminal 200 via the short-range communication unit 120 (step S202). A conversation request packet is a packet indicating that robot 100 is requesting to converse with another robot 100', and it contains information about robot 100's ID.

[0050] Then, the robot 100 determines whether the user is still nearby using the human presence sensor 131 and other sensors provided in the situation acquisition unit 110 (step S203). If the user is still nearby (step S203; Yes), the process returns to step S203. If the user is no longer nearby (step S203; No), the robot 100 sends a conversation cancellation request packet to the conversation output terminal 200 via the short-range communication unit 120 (step S204). A conversation cancellation request packet is a packet sent by the robot 100 to cancel a conversation request packet it previously sent, and it contains the robot 100's ID. The process then returns to step S201.

[0051] Furthermore, in the request relay thread, the conversation output terminal 200 waits until a conversation request packet is sent from the robot 100, and once it is sent, the short-range communication unit 210 receives the conversation request packet (step S211). Then, it sends the conversation request packet to the conversation output server 300 via the LTE communication unit 220 (step S212). Then, the conversation output terminal 200 waits until a conversation cancellation request packet is sent from the robot 100, and once it is sent, the short-range communication unit 210 receives the conversation cancellation request packet (step S213). Then, it sends the conversation cancellation request packet to the conversation output server 300 via the LTE communication unit 220 (step S214), and returns to step S211.

[0052] Furthermore, in the request update thread, the conversation output server 300 continues to wait until a conversation request packet or a conversation cancellation request packet (hereinafter referred to as a "request packet") is sent from the conversation output terminal 200. Once a request packet is received, the LTE communication unit 310 receives the request packet (step S221). Then, using the received request packet, the information of the conversation-enabled members is updated (step S222), and the process returns to step S221.

[0053] The information of conversation-enabled members is a set of IDs of the robot 100 that sent the conversation request packet (i.e., the robot 100 that the user is near). When the conversation output server 300 receives a conversation request packet, it adds the IDs contained in the conversation request packet to the information of conversation-enabled members, and when it receives a conversation cancellation request packet, it removes the IDs contained in the conversation cancellation request packet from the information of conversation-enabled members.

[0054] If there are multiple pairs of robots 100 and conversation output terminals 200, the above processing is performed on each robot 100 and the conversation output terminal 200 that is communicating with that robot 100. Therefore, the conversation output server 300 will have the IDs of all robots 100 that the user is near at that time as information on members who can converse.

[0055] The above describes the thread group for collecting conversation requests. However, the above process is merely an example, and the robot 100 may send conversation request packets regardless of the user's approach (for example, periodically). Next, the thread group for outputting conversation content will be explained with reference to Figure 8.

[0056] When the conversation output terminal 200 is powered on, the conversation output thread shown in the left flowchart of Figure 8 is started. Similarly, when the conversation output server 300 is powered on, the scenario transmission thread shown in the middle flowchart of Figure 8 is started. The conversation output thread shown in the right flowchart of Figure 8 is the same as that in the left flowchart. It is assumed that the right flowchart is executed on the conversation output terminal 200' which is wirelessly connected to another robot 100' that is the conversation partner of robot 100 which is wirelessly connected to conversation output terminal 200. Also, in Figure 8, the dotted arrows indicate that data is sent and received between each thread.

[0057] First, in the scenario transmission thread, the conversation output server 300 selects conversation partners by arbitrarily choosing two robot IDs from the information of conversationable members that is being updated in the request update thread mentioned above (step S311). If there is only one robot ID in the information of conversationable members, a conversation between that robot 100 and a virtual (non-existent) robot 100 may be provided. Also, although the scenario storage unit 331 shown in Figure 3 only stores scenario data assuming a conversation between two robots 100, if it also stores scenario data assuming a conversation between three or more robots 100, three or more IDs may be selected in step S311 to select them as conversation partners.

[0058] The conversation partner may be selected based on recent conversation history. For example, users can pre-register their conversation partner preferences (such as whether they want their robot 100 to converse with a robot 100' belonging to someone they know, or with a robot 100'' belonging to someone they don't know) in the user profile storage unit 332. The conversation output server 300 can then select conversation partners so that robots belonging to users with the same conversation partner preferences converse with each other. The virtual (non-existent) robot 100 may be an AI-powered conversation engine on a social networking service. This conversation engine may be located within the conversation output server 300, or it may be implemented on another computer connected via the network.

[0059] Next, the control unit 320 of the conversation output server 300 sends a conversation start notification packet via the LTE communication unit 310 to the conversation output terminals 200 connected to each robot 100 selected in step S311 (step S312). The conversation start notification packet contains the ID of each robot 100 that will be the conversation partner and notifies each conversation output terminal 200 that scenario data indicating the content of the conversation will be sent. The conversation output server 300 then uses the SNS function it provides to start a conversation (chat) among the robots 100 corresponding to the IDs included in the conversation start notification packet.

[0060] Next, the scenario selection unit 321 performs a scenario selection process to select scenario data from the scenario storage unit 331 (step S313). Details of this scenario selection process will be described later with reference to Figure 9. Next, the control unit 320 transmits the scenario data to each conversation output terminal 200 via the LTE communication unit 310 (step S314). At this time, the control unit 320 functions as a transmission means.

[0061] The scenario data transmitted here differs depending on the recipient conversation output terminal 200. The conversation output terminal 200 of the person who speaks first receives scenario data corresponding to the "statement content" in the scenario storage unit 331 (Figure 3). Then, the conversation output terminal 200' of the person who responds receives scenario data corresponding to the "response content" in the scenario storage unit 331 (Figure 3).

[0062] Next, the control unit 320 determines whether or not to terminate the conversation (step S315). Normally, after selecting and transmitting two scenario data from the scenario storage unit 331, the control unit 320 terminates the conversation. If the conversation is not terminated (step S315; No), the process returns to step S313.

[0063] If the conversation is to be terminated (step S315; Yes), a conversation termination notification packet is sent to each conversation output terminal 200 via the LTE communication unit 310 (step S316), and the process returns to step S311. The conversation termination notification packet contains the ID of the robot 100 that was the conversation partner and notifies that the transmission of the scenario data that was being performed up to that point has ended.

[0064] The operation of the scenario transmission thread has been explained above. The operation of the corresponding conversation output terminal 200, the conversation content output thread (flowchart on the left in Figure 8), will be explained below. The operation of conversation output terminal 200' (flowchart on the right in Figure 8) is similar.

[0065] First, the terminal control unit 230 of the conversation output terminal 200 determines whether or not it has received a conversation start notification packet via the LTE communication unit 220 (step S301). If it has not received a conversation start notification packet (step S301; No), it returns to step S301. If it has received a conversation start notification packet (step S301; Yes), the terminal control unit 230 prepares to have a conversation (chat) with the ID contained in the conversation notification packet and receives the scenario data sent by the conversation output server 300 via the LTE communication unit 220 (step S302).

[0066] The conversation output unit 231 then transmits the scenario data to the SNS of the conversation output server 300 via the LTE communication unit 220. The conversation output unit 231, acting as the client software for the SNS, then outputs the data returned from the SNS to the display 270 of the conversation output terminal 200, so that the conversation content on the SNS is output (step S303). The SNS provided by the conversation output server 300 may have a function to translate the data returned to the client software based on the country of residence information of the accessing user. If it has this translation function, the conversation content in the natural language used by that user will be output to the display 270 of each conversation output terminal 200.

[0067] Furthermore, once the terminal control unit 230 has output the conversation content, it may notify the robot 100 via the short-range communication unit 210 whose statement was output. When the robot 100 receives notification via the short-range communication unit 120 that the conversation partner's statement has been output, it may, for example, wag its tail 136. Then, each time a message comes from the conversation partner, the robot 100 will wag its tail 136, making it appear to the user as if the robot 100 is reading the messages from the conversation partner. In this case, the short-range communication unit 120 functions as a receiving means.

[0068] The terminal control unit 230 then determines whether or not it has received a conversation end notification packet via the LTE communication unit 220 (step S304). If it has not received a conversation end notification packet (step S304; No), it returns to step S302. If it has received a conversation end notification packet (step S304; Yes), it terminates the conversation and returns to step S301.

[0069] The above describes the thread group for outputting conversation content. Through the above process, the conversation content with the conversation partner is output to the displays 270 of all conversation output terminals 200 (conversation output terminals 200 to which the robot 100, indicated by the ID selected in step S311 of the scenario transmission thread) selected by the conversation output server 300.

[0070] Next, the details of the scenario selection process in step S313 of the scenario transmission thread of the conversation output server 300 will be explained with reference to Figure 9.

[0071] First, the scenario selection unit 321 selects a topic for conversation (step S321). This selection method is arbitrary, but for example, it can be randomly selected from the topics stored in the scenario memory unit 331. However, recently selected topics (for example, the five most recent topics) are excluded from the selection to prevent the same scenario data from being selected again.

[0072] Next, the scenario selection unit 321 refers to the user status storage unit 333 and acquires status information transmitted from each robot 100 (step S322). For example, it acquires the status information of robot 100 (first status information) and the status information of robot 100' (second status information). Then, based on the selected topic and the acquired status information (first status information and second status information), it refers to the scenario storage unit 331 and acquires scenario data (step S323). Note that if a virtual (non-existent) robot 100 is set as the conversation partner in the conversation partner selection (step S311), the status information of the conversation partner may be set randomly in step S322.

[0073] Furthermore, in step S322, the scenario selection unit 321 may, in addition to or instead of the situation information, refer to the user profile storage unit 332 to obtain information about the user of each robot 100. In this case, in step S323, the scenario selection unit 321 refers to the scenario storage unit 331 to obtain scenario data based on at least one of the acquired situation information and user information and the topic selected in step S321. For example, in step S322, the scenario selection unit 321 obtains at least one of the information (first information) between the situation information of robot 100 and the information about the user of robot 100, and at least one of the information (second information) between the situation information of robot 100' and the information about the user of robot 100'. Then, based on the selected topic and the acquired first and second information, it refers to the scenario storage unit 331 to obtain scenario data (step S323).

[0074] The scenario selection unit 321 then modifies the acquired scenario data as needed (step S324). For example, if the scenario data contains the string "[I]", it replaces this part with "I" or "me" depending on the user's gender. It also performs translation if the language of the user's country of residence differs from the language of the scenario data. The scenario selection unit 321 obtains the user's country of residence information from the user profile storage unit 332 and the language of the scenario data from the scenario storage unit 331. This allows the scenario selection unit 321 to determine whether the language of the user's country of residence differs from the language of the scenario data. At this time, the scenario selection unit 321 also functions as a means of obtaining the country of residence. In step S321, the scenario selection unit 321 also adds an icon of the robot making the statement so that it can distinguish which robot 100 each statement belongs to. In this way, conversation data presented to the user is created. Alternatively, scenario data for each country may be stored in advance in the scenario storage unit 331, and when acquiring the scenario data in step S323, the scenario data written in the language of the user's country of residence may be acquired.

[0075] Then, the scenario selection process ends. Normally, the control unit 320 ends the conversation after selecting and transmitting two scenario data from the scenario storage unit 331, so the above scenario selection process is performed twice in one conversation. That is, in the first step S321, the first topic is selected, and based on the first topic, the first scenario data is acquired in step S323, and finally the first conversation data is created. Then, in the second step S321, the second topic is selected, and based on the second topic, the second scenario data is acquired in step S323, and finally the second conversation data is created. When executing the scenario selection process, the control unit 320 functions as a creation means.

[0076] Through the scenario selection process described above, the conversation output server 300 can send scenario data to the conversation output terminal 200 that can generate a conversation based on the selected topic and acquired situational information.

[0077] Next, we will explain the processing by the thread group for outputting conversation content (Figure 8) with a specific example. Here, we assume that a robot 100 exists near the conversation output terminal 200, and that this robot 100 has a user profile with the ID GladDog as shown in Figure 4. Also, we assume that a robot 100' exists near the conversation output terminal 200', and that this robot 100' has a user profile with the ID HappyCat as shown in Figure 4. Furthermore, we assume that users are approaching these robots 100 and 100', respectively, and that these robots 100 and 100' have sent conversation request packets.

[0078] Then, in step S311, robot 100 with ID GladDog and robot 100' with ID HappyCat were selected as conversation partners. And, the situational information was as shown in Figure 5. Furthermore, in the first step S321, "communication intimacy" was selected as the topic, and in the second step S321, "anniversary" was selected as the topic.

[0079] Then, in the first step S322, the intimacy level of each is calculated from the situation information shown in Figure 5. For example, if the intimacy level is represented by "number of times spoken to" + "number of times looked at" + "number of times hugged" (however, if these are greater than 5, it is set to 5, and if they are less than 1, it is set to 1), then GladDog's intimacy level is 5 and HappyCat's intimacy level is 1. Then, in step S323, referring to Figure 3, GladDog's statement becomes "I'm so happy you played with me so much," and HappyCat's response becomes "That's nice. My owner never plays with me."

[0080] Furthermore, since HappyCat is female, as shown in Figure 4, the scenario data "That's nice. My husband never plays with me." is modified in step S324 to "That's nice. My husband never plays with me." Also, according to the user profile storage unit 332 shown in Figure 4, the user of robot 100' with ID HappyCat is American, so the scenario data to be sent to the conversation output terminal 200' connected to robot 100' is translated into English.

[0081] Then, in the next step S322, the situation information shown in Figure 5 indicates that today is Christmas, so in step S323, referring to Figure 3, GladDog's statement becomes "Merry Christmas!" and a Christmas card is attached as an image file. HappyCat's response becomes "A Christmas card? Thank you!" In this case, in step S324, no correction of the first-person pronoun is necessary, but the scenario data to be sent to the conversation output terminal 200' is translated into English.

[0082] As a result of the processing described above, the screen displayed on the conversation output terminal 200 connected to robot 100 will be as shown in Figure 10, and the screen displayed on the conversation output terminal 200' connected to robot 100' will be as shown in Figure 11. In both figures, the conversation content is displayed along with an icon representing robot 100 or robot 100', making it possible to distinguish which is the content spoken by robot 100 and which is the content spoken by robot 100'. Furthermore, since the user of robot 100 resides in Japan, the conversation content shown in Figure 10 is displayed in Japanese, and since the user of robot 100' resides in the United States, the conversation content shown in Figure 11 is displayed in English.

[0083] As described above, in this embodiment, the displays of each user's conversation output terminals 200, 200' show text that appears as if robot 100 and robot 100' are having a conversation. Therefore, it is possible to make users believe that robots 100, 100' are equipped with high conversational capabilities.

[0084] (Modification 1 of Embodiment 1) As illustrated in the example above, if, in a conversation about the first topic, you (GladDog in the example above) are feeling happy but the other person (HappyCat in the example above) is feeling down, you may add a "transitional sentence" such as "Oh, I see, well, cheer up!" at the beginning of your statement when moving on to the next topic. By inserting such transitional sentences to connect topics, the flow of the conversation can be made to seem more natural. A modification 1 of Embodiment 1, which performs this processing, will now be described.

[0085] Modification 1 of Embodiment 1 includes a functional configuration of the conversation output server 300 in which the memory unit 330 is equipped with a "connecting sentence memory unit" (not shown) that stores connecting sentences according to the moods of the user and the respondent, and each scenario data in the scenario memory unit 331 also stores "mood" information indicating the moods of the user and the respondent. This "mood" information can be considered to be similar to the level and the level of the other person when the topic shown in Figure 3 is the level of intimacy of communication.

[0086] Then, in the scenario modification process for the second and subsequent topics (step S324), based on the moods of the user and the respondent in the scenario data of the previous topic, a transition sentence corresponding to that mood is retrieved from the transition sentence storage unit and added to the beginning of the utterance. However, depending on the moods of the user and the respondent, a transition sentence may not be necessary, in which case no transition sentence is added.

[0087] Through the above process, the statement "Merry Christmas!" made by GladDog in Embodiment 1 is modified to "Oh, I see, well, cheer up! Merry Christmas!" in Modification 1 of Embodiment 1.

[0088] Thus, in this modified example 1 of Embodiment 1, the flow between topics can be made to appear more natural. Therefore, it can be made to the user that the robots 100,100' have high conversational capabilities.

[0089] (Modification 2 of Embodiment 1) In Embodiment 1, the conversation output server 300 provides SNS functionality, and the conversation output unit 231 of the conversation output terminal 200 acts as client software for this SNS, outputting conversation content that appears to be a natural language conversation between robot 100 and other robots 100' to the display of the conversation output terminal 200. However, it is not essential for the conversation output server 300 to provide SNS functionality. Here, we will describe a modification 2 of Embodiment 1 in which conversation content that appears to be a natural language conversation between robot 100 and other robots 100' is output to the display of the conversation output terminal 200, even if the conversation output server 300 does not provide SNS functionality.

[0090] In Modification 2 of Embodiment 1, the processing of the conversation content output thread group (Figure 8) differs from Embodiment 1 in that the content transmitted by the conversation output server 300 as scenario data and the processing when the conversation output terminal 200 receives the scenario data. Furthermore, in this Modification 2, the conversation output server 300 does not need to provide SNS functionality. In the following description, we assume a case where a screen is output that makes it appear as if robot 100 (wirelessly connected to conversation output terminal 200) and robot 100' (wirelessly connected to conversation output terminal 200') are having a conversation.

[0091] In the modified version 2 of Embodiment 1, the conversation output server 300 transmits not only the content spoken by robot 100 but also the content of robot 100's response as scenario data to each conversation output terminal 200 during scenario transmission (step S314) of the scenario transmission thread (flowchart in the middle of Figure 8). At this time, the conversation output server 300 has previously translated this content spoken and the response into the language of the country of residence of the user of the destination conversation output terminal 200 during scenario modification (step S324) of the scenario selection process (Figure 9).

[0092] Furthermore, in the modified example 2 of Embodiment 1, the conversation output unit 231 receives not only the content of what robot 100 says, but also the content of robot 100's response as scenario data from the conversation output server 300 during scenario reception (step S302) of the conversation output thread (the flowchart on the left or right in Figure 8). Then, in conversation content output (step S303), the content of what robot 100 says and the content of robot 100's response received in step S302 are output sequentially to the display of the conversation output terminal 200.

[0093] Through this process, in Modification 2 of Embodiment 1, even if the conversation output server 300 does not provide SNS functionality, text is displayed on the displays of each user's conversation output terminals 200, 200' as if robot 100 and robot 100' were having a conversation. Therefore, it is possible to make users believe that robots 100, 100' possess high conversational capabilities. Furthermore, even if the SNS provided by the conversation output server 300 does not have a function to translate the data returned to the client software based on the information of the accessing user's country of residence, according to Modification 2, the conversation content in the natural language used by that user will be output to the display of each conversation output terminal 200.

[0094] (Modification 3 of Embodiment 1) A third modification of Embodiment 1, which guides a conversation between the user of robot 100 and the user of robot 100' through a conversation between robot 100 and robot 100', will be described.

[0095] In the third variation, the scenario memory unit 331 stores scenario topics that are related to the user of robot 100 (robot 100'), such as "Introduction of the owner's ID" and "Owner's hobbies," as shown in Figure 12. The user profile memory unit 332 also stores the name of the SNS used by the user and their ID on that SNS.

[0096] Then, as shown in Figure 12, if the topic of the scenario is "Introducing the husband's ID," the content of the statement is, for example, "My husband is on an SNS called [SNS name] with the ID [ID on that SNS]," and the response is, "Oh, really? My husband is on an SNS called [SNS name] with the ID [ID on that SNS]." This is stored in the scenario storage unit 331. Here, the [SNS name] and [ID on that SNS] parts are replaced in step S324 of the scenario selection process (Figure 9) with the name of the SNS used by each user and their ID on that SNS, which are stored in the user profile storage unit 332.

[0097] Also, as shown in Figure 12, when the topic of the scenario is "My husband's hobbies," the dialogue might include, for example, introducing one of the husband's hobbies (in Figure 12, "reading") as "One of my husband's hobbies is reading," and the response might be, if there is a common hobby (in Figure 12, "reading"), "My husband's hobby is also reading. It might be fun to introduce each other to our favorite authors and works," and if there is no common hobby, one of the other person's hobbies (in Figure 12, "tennis") might be introduced as "Oh really? My husband's hobby is tennis," and these responses are stored in the scenario memory unit 331.

[0098] By adopting the above configuration, in this modified example 3 of Embodiment 1, it is possible to demonstrate to the user that robots 100 and 100' possess high conversational capabilities. Furthermore, conversations between robots 100 and 100' can guide conversations between users, thereby expanding the opportunities for interaction among users.

[0099] (Embodiment 2) In Embodiment 1, the conversation output terminal 200 was a smartphone, and the conversation content was output as text. However, the format of conversation output is not limited to text. Embodiment 2, in which the conversation content is output as voice, will be described.

[0100] The conversation output system 1001 according to Embodiment 2 of the present invention comprises a robot 100, a conversation output terminal 201, and a conversation output server 300, as shown in Figure 13. The conversation output terminal 201 is a smart speaker (AI speaker) and outputs conversation content as voice. The functional configuration of the conversation output terminal 201 is that of the conversation output terminal 200, but with a speaker 280 instead of a display 270. The conversation output unit 231 synthesizes the conversation content into voice and outputs it from the speaker 280. The speaker 280 functions as an output unit.

[0101] It is desirable that the voice used when outputting the content spoken by robot 100 is different from the voice used when outputting the content spoken by the robot 100' that is conversing with it. Therefore, the conversation output terminal 201 has the function to output multiple types of voices. The user profile storage unit 332 according to Embodiment 2 also stores information on the type of voice for each robot, and the conversation output server 300 transmits the information on the type of voice corresponding to each robot 100 in the conversation start notification packet. The conversation output terminal 201 then speaks the content corresponding to each robot 100 using the type of voice notified in the conversation start notification packet.

[0102] As a result, the conversation output terminal 201 can speak as if multiple robots 100 were having a conversation by using different voices within a single unit.

[0103] As described above, in this embodiment, each user's conversation output terminal 200, 200' outputs audio that makes it appear as if robot 100 and robot 100' are conversing. Therefore, it is possible to make the user believe that robots 100, 100' are equipped with high conversational capabilities.

[0104] The conversation output terminal 201 may also include a display 270 in addition to the speaker 280. In this case, the conversation output unit 231 may output the conversation content as sound from the speaker 280 and also display it on the display 270. Furthermore, the output unit may be switchable so that, based on user instructions, it sometimes outputs the conversation content as sound from the speaker 280 and at other times displays it on the display 270.

[0105] (Modified version of Embodiment 2) In the above embodiment 2, even if the conversation output terminal 201 outputs different voices for the statements of different robots, it was sometimes difficult for the user to determine which voice corresponds to the statements of their own robot 100. Therefore, in order to make it easier for the user to understand what their robot 100's voice sounds like, when setting up the connection between robot 100 and the conversation output terminal 201, the conversation output terminal 201 may be configured to speak in robot 100's voice, such as "I am ○○." Alternatively, during this setup, the user may be able to configure the conversation output terminal 201 to speak a voice of their choice (for example, a male voice, a female voice, a child's voice (with gender distinctions), an elderly person's voice (with gender distinctions), an announcer's voice, a voice actor's voice, etc.) as the voice of robot 100.

[0106] Here, regarding the voices of announcers and voice actors, the voices of famous announcers and voice actors may be pre-registered in the conversation output terminal 201. Alternatively, voice data of less well-known announcers and voice actors may also be made available for download so that the conversation output terminal 201 can output their voices.

[0107] Furthermore, the conversation output terminal 201 may be configured to speak in the voice of the robot 100, such as "I am ○○," at any time (for example, when the user gives instructions to the robot 100 or the conversation output terminal 201 to confirm the voice (for example, by pressing a voice confirmation button provided on the robot 100 or the conversation output terminal 201, not shown in the diagram)).

[0108] Furthermore, when setting up the conversation output terminal 201 to connect robot 100 to the conversation output terminal 201, even if it does not speak in robot 100's voice, the conversation output terminal 201 may, at the beginning of the conversation, speak in the voice of robot 100, "I am robot ○○," and in the voice of robot 100', "I am robot △△," etc., to make the user perceive that robot 100 and robot 100' are actually having a conversation.

[0109] (Embodiment 3) In Embodiments 1 and 2, the conversation output server 300 was a separate device from the conversation output terminals 200 and 201. However, the functions of the conversation output server 300 may be provided by a specific conversation output terminal. Embodiment 3, which is such an embodiment, will be described.

[0110] The conversation output system 1002 according to Embodiment 3 of the present invention comprises, as shown in Figure 14, at least a robot 100 and a scenario conversation output terminal 250. The conversation output system 1002 may include multiple pairs of robots 100 and conversation output terminals 200, similar to Embodiment 1. The robot 100 can appear to the user as if it is conversing with other robots 100 via the scenario conversation output terminal 250 and the conversation output terminal 200.

[0111] The difference between Embodiment 3 and Embodiment 1 is that the conversation output server 300 is absent, and instead, a scenario conversation output terminal 250 is present. Otherwise, the configuration is the same as Embodiment 1, so the scenario conversation output terminal 250 will be described below.

[0112] The scenario conversation output terminal 250, like the conversation output terminal 200, is a smartphone equipped with a display, but it also has the functionality of a conversation output server 300. As shown in Figure 14, the scenario conversation output terminal 250 has the following functional configuration: a short-range communication unit 210, an LTE communication unit 220, a terminal control unit 230, and a storage unit 240. The terminal control unit 230 functions as a conversation output unit 231 and a scenario selection unit 232. The storage unit 240 has the following functional configuration: a scenario storage unit 241, a user profile storage unit 242, and a user status storage unit 243.

[0113] The short-range communication unit 210, the LTE communication unit 220, and the conversation output unit 231 are the same as those in the conversation output terminal 200. However, the LTE communication unit 220 in the scenario conversation output terminal 250 communicates with the LTE communication unit 220 in the conversation output terminal 200. For example, the LTE communication unit 220 in the scenario conversation output terminal 250 receives status information transmitted by the LTE communication unit 220 in the conversation output terminal 200. Also, the LTE communication unit 220 in the scenario conversation output terminal 250 transmits scenario data to the conversation output terminal 200. In this way, the LTE communication unit 220 in the scenario conversation output terminal 250 communicates with the conversation output terminal 200 and functions as a terminal-to-terminal communication unit.

[0114] The scenario selection unit 232 is the same as the scenario selection unit 321 provided in the conversation output server 300 of Embodiment 1. Furthermore, the scenario storage unit 241, user profile storage unit 242, and user status storage unit 243 are the same as the scenario storage unit 331, user profile storage unit 332, and user status storage unit 333 provided in the conversation output server 300 of Embodiment 1, respectively.

[0115] The same processing can be applied to the process by which robot 100 converses with other robots 100 via the scenario conversation output terminal 250 and conversation output terminal 200, with the exception of the following two points. First, the scenario conversation output terminal 250 performs the processing of the conversation output server 300 instead. Second, if a robot 100 connected to the scenario conversation output terminal 250 is selected as a conversation partner, the scenario conversation output terminal 250 performs the processing of conversation output terminal 200 and conversation output server 300 in parallel in a separate thread. In this case, communication between the status relay thread and the status update thread performed by the scenario conversation output terminal 250 is performed by inter-thread communication (LTE communication unit 220 is not used). Similarly, communication between the request relay thread and the request update thread, and communication between the conversation output thread and the scenario transmission thread, performed by the scenario conversation output terminal 250 are also performed by inter-thread communication (LTE communication unit 220 is not used).

[0116] With the above configuration and processing, in Embodiment 3, even without providing a conversation output server 300, the scenario conversation output terminal 250 can transmit scenario data based on the situation information received from each robot 100 to each conversation output terminal 200. Therefore, in this embodiment, even without providing a conversation output server 300, text is displayed on the displays of each user's scenario conversation output terminal 250 and conversation output terminal 200 as if robot 100 and robot 100' were having a conversation. Thus, it is possible to make the user believe that robots 100 and 100' are equipped with high conversational capabilities.

[0117] (Embodiment 4) In the above embodiment, communication between robots is enabled primarily using SNS functions. Users can then observe their own robots engaging in fluent conversations, as shown in Figures 10 and 11, and be impressed, thinking, "This robot has no voice conversation function, or only a limited one, yet it can converse so fluently," and enjoy watching the conversations. Since each robot has its own user, the user of robot 100 may become curious about the users of robots such as robot 100', which are conversation partners of robot 100.

[0118] In such cases, to enable interaction between users, in Modification 3 of Embodiment 1, the scenario memory unit 331 is equipped with scenario topics such as "Introduction of the Master's ID" and "Master's Hobbies" for each user of Robot 100 and Robot 100', and the user profile memory unit 332 is also equipped with the name of the SNS used by the user and their ID on that SNS. This makes it possible for the user of Robot 100 to communicate with other users who are users of Robot 100'.

[0119] However, in Modification 3 of Embodiment 1, the conversation output server 300 arbitrarily selects the scenario topics, so the conversation may not necessarily be about what the user wants to know when they want to know it. Also, some users may not want their personal information to be known to others. Therefore, Embodiment 4 will be described, which allows users to disclose their personal information only when they want to interact with other users.

[0120] The conversation output system 1003 according to Embodiment 4 of the present invention comprises a robot 101, a conversation output terminal 202, and a conversation output server 301, as shown in Figure 15. The robot 101 includes a user intention acquisition unit 130. The user intention acquisition unit 130 acquires information (user intention information) indicating whether or not the user wants to disclose their personal information (SNS or IDs they participate in, their hobbies, etc.). The user intention acquisition unit 130 may be, for example, a simple switch (set to ON if the user wants to disclose personal information, and OFF if they do not want to). The user intention acquisition unit 130 functions as a means for acquiring user intention information. Alternatively, the robot 101 may acquire user intention information by voice recognition. In this case, the user intention acquisition unit 130 will consist of a microphone 133 and a voice recognition unit (not shown).

[0121] Robot 101 transmits the user intention information acquired by the user intention acquisition unit 130 to the conversation output terminal 202 via the short-range communication unit 120. The conversation output terminal 202 then transmits the user intention information received from Robot 101 to the conversation output server 301. The conversation output server 301 then selects a scenario based on the received user intention information.

[0122] For example, consider a case where robot 101 and robot 101' are selected as conversation partners. If both the user intention information sent from robot 101 and the user intention information sent from robot 101' indicate "personal information cannot be disclosed," then when selecting a scenario, the system should avoid selecting scenarios that involve the disclosure of personal information, such as "introduce the owner's ID" or "the owner's hobbies." Conversely, if both user intention information indicates "personal information can be disclosed," then when selecting a scenario, the system should frequently select scenarios that involve the disclosure of personal information, such as "introduce the owner's ID" or "the owner's hobbies."

[0123] Furthermore, if only one of the robots 101 in the conversation has user intent information set to "Personal information cannot be disclosed," the conversation output server 301 sends a statement such as "[My] master wants to exchange personal information, is that okay?" to the conversation output terminal 202, thereby giving the user who has set "Personal information cannot be disclosed" an opportunity to allow the disclosure of personal information. Alternatively, the conversation output server 301 may choose a scenario in which only the personal information of users who have set "Personal information can be disclosed" (i.e., the user intent information obtained by the robot 101's user intent acquisition unit 130 is "I do not want to disclose personal information") is disclosed, without giving the user who has set "Personal information cannot be disclosed" (i.e., the user intent information obtained by the robot 101's user intent acquisition unit 130 is "I want to disclose personal information") an opportunity to allow the disclosure of personal information.

[0124] Furthermore, the conversation output terminal 202 is equipped with a user input unit 260. The user input unit 260 receives the content (text) of the user's statement, and the conversation output terminal 202 posts it to the SNS provided by the conversation output server 301. This allows the user of robot 101 to enter the SNS where robot 101 and other robots 101' are conversing, and to converse with the users of other robots 101'.

[0125] For example, the following conversation example (scenario example) is possible. GladDog: You played with me so much, it was really fun. HappyCat: That's great! Actually, my husband played with you a lot too. GladDog: My husband's hobby is tennis. HappyCat: Oh really? My husband's hobby is tennis too. Maybe we could play tennis together sometime. Here, the scenario shown in Figure 12 includes pre-defined scenarios based on assumed combinations of hobbies. However, it is not limited to these. For example, in the scenario selection process (Figure 9), a judgment step could be added to determine whether the conversation partner's hobbies match, and if they do, a scenario could be generated in which the two people mention this to each other. In the above conversation, since the hobby of tennis matches, a scenario like the following could be generated, for example. GladDog: We have the same hobbies! HappyCat: That's right. It's the same.

[0126] If the user of robot 101, after seeing this conversation, wants to interact with the user of robot 101', they inform robot 101's user intention acquisition unit that they want to disclose their personal information (for example, by turning on a switch that indicates they are okay with disclosing their personal information). This user intention information is then sent to the conversation output server 301 via the conversation output terminal 202. If the user of robot 101', the conversation partner of robot 101, also wants to disclose their personal information, robot 101's user intention acquisition unit 130 acquires user intention information indicating that they want to disclose their personal information, and this user intention information is sent to the conversation output server 301.

[0127] As a result, both the user intent information and the conversation output server 301 become "personal information can be disclosed," and when selecting a scenario, the server will frequently select scenarios that involve the disclosure of personal information, such as "introduction of the master's ID" or "master's hobbies." For example, if the topic "master's address" is selected, the following conversation will take place. GladDog: Where does your husband live? My husband lives in Tokyo. HappyCat: My husband lives in Seattle. For example, if the topic "my husband's email address" is selected, the following conversation will take place. GladDog: What is your husband's email address? My husband's email address is name@ne.jp. HappyCat:[My] husband's email address is name@net.

[0128] Here, user information is exchanged through conversations between robots, but if a user observes the conversation between robots 101 and becomes interested in the user of the other robot 101', they may initiate a conversation with the other robot 101'' user themselves via SNS from the conversation output terminal 202.

[0129] For example, the following conversation is also possible. GladDog: You played with me so much, it was really fun. HappyCat: That's great! Actually, my husband played with you a lot too. GladDog: My husband's hobby is tennis. HappyCat: Oh really? My husband's hobby is tennis too. Maybe we could play tennis together sometime.

[0130] After such a conversation takes place between the robots 101, users can also converse directly with each other as follows: Tarou: Hello. My name is Tarou and I'm a GladDog user. May I contact you directly? Lily: Nice to meet you. It's alright. Tarou: Thank you. I see you enjoy tennis. I enjoy tennis too. Lily: What a coincidence! In this way, users can communicate with each other through conversations with robot 101.

[0131] As a variation of Embodiment 4, the user intent acquisition unit 130 may be provided in the conversation output terminal 202 instead of the robot 101. In this case, the user can input whether or not to disclose personal information by operating a touch panel or the like provided in the conversation output terminal 202, and the user intent acquisition unit 130 acquires user intent information from the touch panel or the like.

[0132] (Embodiment 5) In Embodiment 1 described above, since the robot 100 does not have an LTE communication unit, the status information acquired by the status acquisition unit 110 is transmitted to the conversation output terminal 200 by the short-range communication unit 120, and the conversation output terminal 200 transmits the status information to the conversation output server 300 by the LTE communication unit 220. However, if the robot is equipped with an LTE communication unit, it becomes possible to transmit the status information directly to the conversation output server 300. Furthermore, in Embodiment 1, the status information was acquired by the status acquisition unit 110 equipped in the robot 100, but if the status information can also be acquired by a sensor provided separately from the robot, it becomes possible to utilize the status information from a sensor that the robot does not have. Embodiment 5, which is such an embodiment, will be described.

[0133] The conversation output system 1004 according to Embodiment 5 of the present invention, as shown in Figure 16, comprises a robot 102, a conversation output terminal 200, a conversation output server 300, and a sensor device 400. The sensor device 400 is usually attached to the robot 102 and supplements the function of the robot 102's status acquisition unit 110. However, the sensor device 400 does not necessarily have to be attached to the robot 102 as long as it can acquire information about the surrounding environment of the robot 102. The robot 102, the conversation output terminal 200, and the sensor device 400 are wirelessly connected by short-range wireless communication to form a triplicate. The conversation output system 1004 may include multiple triplicates of the robot 102, conversation output terminal 200, and sensor device 400 (for example, robot 102, conversation output terminal 200, and sensor device 400; robot 102', conversation output terminal 200', and sensor device 400', etc.).

[0134] The sensor device 400 includes a status acquisition unit 410 and a short-range communication unit 420, and transmits the status information acquired by the status acquisition unit 410 to the robot 102.

[0135] The situation acquisition unit 410 acquires situation information, which is information about the surrounding conditions of the robot 102, from various sensors (temperature sensor, humidity sensor, illuminance sensor, human presence sensor, etc.) provided by the sensor device 400.

[0136] The short-range communication unit 420 is equipped with a short-range wireless communication device (for example, a Bluetooth® communication device) and transmits the status information acquired by the status acquisition unit 410 to the robot 102. Although not shown in Figure 16, the short-range communication unit 420 may also communicate with the short-range communication unit 210 of the conversation output terminal 200 and transmit the status information acquired by the status acquisition unit 410 to the conversation output terminal 200.

[0137] Although not shown in the diagram, the sensor device 400 may also be equipped with an LTE communication unit, allowing the status information acquired by the status acquisition unit 410 to be transmitted directly to the conversation output server 300.

[0138] Furthermore, there may be multiple sensor devices 400. For example, there may be sensor device 400A equipped only with a temperature sensor, sensor device 400B equipped only with a humidity sensor, and sensor device 400C equipped only with an illuminance sensor, and each of the sensor devices 400A, 400B, and 400C may transmit situational information acquired by its own sensors to the robot 102.

[0139] Robot 102 includes an LTE communication unit 140 in addition to the configuration of robot 100 according to Embodiment 1. The LTE communication unit 140 is equipped with a device that communicates using the LTE communication standard and communicates with the conversation output server 300. For example, the LTE communication unit 140 transmits status information to the conversation output server 300.

[0140] Furthermore, the short-range communication unit 120 of the robot 102 communicates not only with the short-range communication unit 210 of the conversation output terminal 200, but also with the short-range communication unit 420 of the sensor device 400, and receives status information acquired by the sensor device 400.

[0141] Note that the robot 102 does not necessarily need to be equipped with a situation acquisition unit 110. In this case, all situation information, which is information about the surrounding environment of the robot 102, will be acquired by the sensor device 400.

[0142] The configuration of the conversation output terminal 200 and the conversation output server 300 is the same as in Embodiment 1, so we will omit their explanation. Also, since much of the processing content of the various threads executed by the robot 102, the conversation output terminal 200, and the conversation output server 300 is common to Embodiment 1, we will focus on explaining the differences below.

[0143] In the robot 102 according to Embodiment 5, in step S101 of the status transmission thread (left in Figure 6), it acquires not only the status information acquired by the status acquisition unit 110, but also the status information acquired by the status acquisition unit 410 of the sensor device 400 via the short-range communication unit 120.

[0144] Furthermore, in step S102 of the status transmission thread (left in Figure 6), the robot 102 may transmit the status information acquired in step S101 to the conversation output server 300 via the LTE communication unit 140. In this case, the conversation output terminal 200 does not need to execute the status relay thread (middle in Figure 6), and the conversation output server 300 waits in step S121 of the status update thread (right in Figure 6) until status information is sent from the robot 102, and then receives the status information with the LTE communication unit 310.

[0145] Furthermore, in the robot 102 according to Embodiment 5, if the sensor device 400 is attached to the robot 102, in steps S201 and S203 of the request transmission thread (left in Figure 7), the system may also use the situation information obtained from the sensor device 400 to determine whether a user has approached or whether a user is still nearby.

[0146] The processes other than those described above are the same as in Embodiment 1, so their explanation will be omitted. In Embodiment 5, the robot 102 directly transmits status information to the conversation output server 300, which reduces the load on the conversation output terminal.

[0147] Furthermore, in Embodiment 5, the situational information acquired by the sensor device 400 can also be used, which broadens the range of situational information that can be used in the scenario and expands the variations of the scenario. For example, the robot 102 may be equipped with only the minimum necessary sensors to reduce costs, but even in this case, the user can enjoy a wider variety of conversations by purchasing additional sensor devices 400 later.

[0148] In Embodiment 5, if the robot 102 does not need to directly transmit status information to the conversation output server 300 (i.e., transmits status information via the conversation output terminal 200), the robot 102 does not need to be equipped with the LTE communication unit 140. Also, in Embodiment 5, if there is no need to expand the variations of status information, the conversation output system 1004 does not need to be equipped with the sensor device 400.

[0149] (Embodiment 6) In the above embodiment, the conversation output server 300 selects scenario data based on situational information, and the conversation output terminal 200 outputs it, making it appear to the user that the robot 100 has high conversational capabilities. However, it is also possible to make the user appear that the robot 100 has high conversational capabilities without using scenario data. Embodiment 6, which describes such an embodiment, will be described next.

[0150] As shown in Figure 17, the conversation output system 1005 according to Embodiment 6 of the present invention comprises a robot 103, a conversation output terminal 203, and a conversation output server 302.

[0151] Robot 103 includes a self-diagnosis unit 150 in addition to the configuration of robot 100 in Embodiment 1. The self-diagnosis unit 150 diagnoses robot 103 itself, and if an abnormality is found, it transmits information about the abnormality as status information to the conversation output terminal 203 via the short-range communication unit 120. For example, if the legs 135 or tail 136 become unable to move or their movement becomes impaired, it transmits information indicating this as status information to the conversation output terminal 203.

[0152] The conversation output terminal 203, in addition to the configuration of the conversation output terminal 202 of Embodiment 4, includes a terminal control unit 230 which is equipped with a status notification creation unit 233. Furthermore, the conversation output terminal 203 stores contact information for a service center that undertakes repairs and other services for the robot 103, and can notify the service center of a message via the LTE communication unit 220.

[0153] In the conversation output terminal 203, when the short-range communication unit 210 receives status information regarding an abnormality of robot 103 from robot 103, the status notification creation unit 233 creates a natural language message to notify the robot of that status information. For example, if it receives status information that it is unable to move its legs 135, the status notification creation unit 233 creates a message such as, "I am a robot of model number ○○ and serial number △△, and this morning I suddenly became unable to move my legs. What should I do?"

[0154] The conversation output terminal 203 then notifies the service center of the message created by the status notification creation unit 233 via the LTE communication unit 220, and outputs the exchange with the service center regarding the message via the conversation output unit 231.

[0155] By outputting such messages and interactions with the service center, the conversation output system 1005 can make it appear to the user as if the robot 103 is conversing with a person at the service center. In this way, the conversation output terminal 203 can make it appear to the user as if the robot 103 is conversing with a person at the service center without using the user input unit 260. Therefore, if only this function needs to be provided, the conversation output terminal 203 does not need to have a user input unit 260.

[0156] The conversation output server 302 includes a text creation unit 322 in addition to the configuration of the conversation output server 300 in Embodiment 1. Note that if the conversation output server 302 does not perform processing for the scenario transmission thread (Figure 8), the conversation output server 302 does not need to include a scenario selection unit 321 or a scenario storage unit 331.

[0157] When the text generation unit 322 is given a utterance (a sentence entered by a user, etc.), it creates a response sentence that appears to be a reply to that utterance. The method of creating the response sentence is arbitrary, but for example, the following processing can be considered.

[0158] In advance, a large amount of conversational exchanges from SNS publicly available on the internet are acquired via the LTE communication unit 310 and stored in the storage unit 330. Then, when a message is given, the system searches the conversational exchanges stored in the storage unit 330 for the message that is closest to the message, and the message used as a response to the message that is closest to the message in that exchange is used as the response message. In the process of searching for the message that is closest to the message, it is not necessary to search for the exact closest message; for example, the system may search for a message that contains some keyword in the message, and the message used as a response to that message is used as the response message.

[0159] The conversation output server 302 then transmits the text created by the text creation unit 322 to the conversation output terminal 203 via the LTE communication unit 310, and the conversation output terminal 203 outputs the text using the conversation output unit 231. In this way, the user can enjoy a simulated chat with the robot 103. The threads for this simulated chat will be explained with reference to Figure 18.

[0160] When the conversation output terminal 203 is powered on, the sentence input / output thread shown in the left flowchart of Figure 18 is started. Similarly, when the conversation output server 302 is powered on, the sentence transmission thread shown in the middle flowchart of Figure 18 is started. The sentence input / output thread shown in the right flowchart of Figure 18 is the same as that in the left flowchart. It is assumed that the right flowchart is executed on the conversation output terminal 203' which is wirelessly connected to another robot 103' that is the conversation partner of robot 103, which is wirelessly connected to conversation output terminal 203. Also, in Figure 18, the dotted arrows indicate that data is sent and received between each thread.

[0161] The text input / output threads and text transmission threads operate in parallel with the conversation content output threads described with reference to Figure 8, if the conversation output server 302 is equipped with a scenario selection unit 321 and a scenario storage unit 331. Even in this case, the conversation start notification reception determination (step S301) and conversation end notification reception determination (step S304) are common processes for the text input / output threads (Figure 18) and the conversation output threads (Figure 8). Furthermore, the conversation partner selection (step S311), conversation start notification transmission (step S312), conversation end determination (step S315), and conversation end notification transmission (step S316) are common processes for the text input / output threads (Figure 18) and the conversation output threads (Figure 8).

[0162] Therefore, if the conversation output server 302 includes a scenario selection unit 321 and a scenario storage unit 331, steps S331 to S337 of the sentence input / output thread will operate in parallel with scenario reception (step S302) and conversation content output (step S303), and steps S341 to S344 of the sentence transmission thread will operate in parallel with scenario selection processing (step S313) and scenario transmission (step S314). Whether or not the conversation output server 302 includes a scenario selection unit 321 and a scenario storage unit 331, the processing of the steps in Figure 18 that share the same symbols as in Figure 8 is the same as the processing described in Figure 8, so the following will explain the processing of the parts that differ from the processing described in Figure 8.

[0163] First, in the input / output thread, upon receiving a conversation start notification packet (step S301; Yes), the terminal control unit 230 determines whether a message has been input from the user input unit 260 or whether a message has been received from the LTE communication unit 220 (step S331). If there is neither input nor reception of a message (step S331; No), the process returns to step S331.

[0164] If a message has been input from the user input unit 260 or if a message has been received from the LTE communication unit 220 (step S331; Yes), the terminal control unit 230 determines whether or not a message has been input from the user input unit 260 (step S332). If no message has been input from the user input unit 260 (step S332; No), the process proceeds to step S334.

[0165] If a message has been input from the user input unit 260 (step S332; Yes), the terminal control unit 230 transmits the message input from the user input unit 260 to the conversation output server 302 via the LTE communication unit 220 (step S333). Then, in the message transmission thread of the conversation output server 302, the control unit 320 receives the message transmitted from the conversation output terminal 203 via the LTE communication unit 310 (step S341), and transmits the received message to all conversation output terminals 203 that are conversation partners, including the source of the message (step S342).

[0166] Then, in the message transmission thread, the message creation unit 322 creates a reply message to the message (step S343), and sends the created reply message to all conversation output terminals 203 that are conversation partners (step S344).

[0167] Meanwhile, in the text input / output thread, the terminal control unit 230 receives the spoken text transmitted from the conversation output server 302 via the LTE communication unit 220 (step S334), and the conversation output unit 231 outputs the received spoken text (step S335).

[0168] Then, the terminal control unit 230 receives the reply message sent from the conversation output server 302 via the LTE communication unit 220 (step S336), and the conversation output unit 231 outputs the received reply message (step S337). The other processes are the same as those described in Figure 8 (indicated by the same reference numerals), so their explanation is omitted.

[0169] As described above, by performing the processing shown in Figure 18, the conversation output terminal 203 can output a response based on an actual conversation, such as one that took place on social networking services, to a statement input by a user, etc., so that the conversation output system 1005 can make the user perceive that the robot 103 is conversing in natural language.

[0170] In the embodiments described above, the LTE communication units 140, 220, and 310 were described as being equipped with devices that communicate using the LTE communication standard. However, the LTE communication units 140, 220, and 310 are not limited to the LTE communication standard. Any device that can communicate between each of the LTE communication units 140, 220, and 310 can be used.

[0171] Furthermore, although the above embodiments were described as performing data communication wirelessly, some or all of the data communication may be performed wirelessly via a wired connection. In particular, in Embodiment 3, as described above, the communication between the status relay thread and the status update thread, the communication between the request relay thread and the request update thread, and the communication between the conversation output thread and the scenario transmission thread performed at the scenario conversation output terminal 250 are performed by inter-thread communication, and the LTE communication unit 220 is not used in this communication.

[0172] (Effects of the embodiment) Users can see and hear their robots engaging in fluent conversations, displayed as text or spoken aloud on a conversation output terminal, making them feel as if their robots possess superior conversational abilities, even if those robots lack or have insufficient voice conversation capabilities. Users can also enjoy watching the robots converse with each other. Furthermore, users can communicate with other users through these conversations. And because the conversations output make it seem as if the robots have become friends without the user's knowledge, users can be inspired by the robots' conversations and expand their own circle of communication.

[0173] The functions of the conversation output terminals 200, 201, 202, the scenario conversation output terminal 250, and the conversation output servers 300, 301 can also be performed by a regular PC (Personal Computer) or other computer. Specifically, in the above embodiment, it was described that the programs executed by the terminal control unit 230 of the conversation output terminals 200, 201, 202, and the scenario conversation output terminal 250, and the control unit 320 of the conversation output servers 300, 301, are pre-stored in the ROM of the storage units 240, 330. However, a computer capable of realizing the above functions may be configured by distributing programs stored on computer-readable recording media such as flexible disks, CD-ROMs (Compact Disc Read Only Memory), DVDs (Digital Versatile Discs), and MOs (Magneto-Optical Discs), and then loading and installing those programs into a computer. Furthermore, although the above embodiment was described using a one-to-one conversation between robots 100 and 100' as an example, it is also possible to have multiple robots converse, similar to the group function of LINE (registered trademark). In this case, the scenario data will be more complex and larger in size, but basically it will only increase the number of patterns. Furthermore, although the scenario data was explained as being prepared in advance, it can also be configured to be updated or newly created sequentially using learning functions or AI. Furthermore, in the above embodiment, since the same conversation output server manages information for both robot 100 and robot 100', the conversation output server generates both statements from robot 100 and robot 100', and can create conversational text that is consistent with the scenario.

[0174] Although preferred embodiments of the present invention have been described above, the present invention is not limited to these specific embodiments, and the present invention includes the invention described in the claims and its equivalents. The invention described in the original claims of this application is listed below.

[0175] (Note 1) A first information acquisition means for acquiring first information which is at least one of the following: information about the user of the robot and situation information which is information about the surrounding conditions of the robot. A means for creating conversation data that causes the user to recognize that the robot and a predetermined object are having a conversation that conforms to at least the first information, based on the acquired first information; The system includes an output control means that controls the output unit to output information based on the conversation data created, thereby causing the user to recognize that the robot and the predetermined object are having a conversation that conforms to at least the first information, The robot does not have the ability to perform conversations at a level higher than the level of conversation based on the conversation data. Conversation output system.

[0176] (Note 2) The aforementioned robot, A first sensor that acquires the status information as the first information, The system includes a status transmission unit that transmits the status information, which is the first information acquired by the first sensor, to the conversation output system, The aforementioned conversation output system comprises a conversation output server and a conversation output terminal. The aforementioned conversation output server comprises a terminal communication unit, the first information acquisition means, and the creation means, The aforementioned conversation output terminal comprises the output control means, the output unit, the status receiving unit, and the server communication unit. The output control means of the conversation output terminal acquires the status information as first information transmitted by the status transmission unit of the robot via the status receiving unit, and transmits the received status information as first information to the terminal communication unit of the conversation output server via the server communication unit. The first information acquisition means of the conversation output server acquires the transmitted first information by receiving it via the terminal communication unit, and transmits the conversation data created by the creation means to the server communication unit of the conversation output terminal via the terminal communication unit. The output control means of the conversation output terminal receives and acquires the conversation data transmitted by the conversation output server via the server-to-server communication unit, and controls the output unit to output information based on the acquired conversation data, thereby causing the user to recognize that the robot and the predetermined object are having a conversation that matches at least the first information. The conversation output system described in Appendix 1.

[0177] (Note 3) The aforementioned conversation output server is configured to provide SNS (Social Networking Service), The output control means of the conversation output terminal is It functions as SNS client software that accesses the SNS provided by the aforementioned conversation output server, The acquired conversation data is transmitted to the terminal-to-terminal communication unit of the conversation output server via the server-to-server communication unit as content to be posted to the SNS. By controlling the output unit to output information based on the aforementioned conversation data, the user is made to recognize that the robot and the predetermined target are having a conversation on the SNS that matches at least the first information. The conversation output system described in Appendix 2.

[0178] (Note 4) The aforementioned creation means is Select a topic for conversation, The conversation data is created based at least on the selected topic and the first information. A conversation output system as described in any one of the appendices 1 to 3.

[0179] (Note 5) The aforementioned creation means is Select the first topic of conversation, Based on the first topic selected and the first information, first conversation data is created. After the output unit is controlled by the output control means to output information based on the first conversation data, a second topic different from the selected first topic is selected. Based on the second topic selected and the first information, second conversation data is created. The conversation output system described in Appendix 4.

[0180] (Note 6) The aforementioned creation means is A connecting sentence, which is a sentence that connects the first conversation data and the second conversation data, is obtained. The acquired connecting sentence is added before the second conversation data created above. The conversation output system described in Appendix 5.

[0181] (Note 7) The system further comprises means for obtaining the user's country of residence, The creation means translates the conversation data into the language of the acquired country of residence. A conversation output system as described in any one of the appendices 1 through 6.

[0182] (Note 8) The aforementioned predetermined target is the user, another robot other than the robot, or a person other than the user. A conversation output system as described in any one of the appendices 1 through 7.

[0183] (Note 9) The aforementioned predetermined object is a robot provided separately from the robot and belonging to a different user from the aforementioned user. The system further comprises a second information acquisition means for acquiring second information which is at least one of the information relating to the other user and situational information which is information relating to the surroundings of the other robot, The creation means creates, based on the acquired first and second information, data as conversation data to allow the user to recognize that the robot and the other robot are having a conversation that conforms to the first and second information. The output control means controls the output unit to output information based on the conversation data, thereby causing the user to recognize that the robot and the other robot are having a conversation that conforms to the first and second information. A conversation output system as described in any one of the appendices 1 through 7.

[0184] (Note 10) The aforementioned predetermined object is a virtual robot, The creation means sets second information relating to the virtual robot, and based on the first information and the set second information, creates data as conversation data to cause the user to recognize that the robot and the virtual robot are having a conversation that conforms to the first and second information. The output control means controls the output unit to output information based on the conversation data, thereby causing the user to recognize that the robot and the virtual robot are having a conversation that conforms to the first and second information. A conversation output system as described in any one of the appendices 1 through 7.

[0185] (Note 11) The system further comprises a means for acquiring user intent information, which acquires user intent information from the user indicating whether or not the user wishes to disclose personal information. The creation means, when the acquired user intention information indicates that the user wishes to disclose personal information, creates the conversation data based on the first information, which includes information about the user. A conversation output system as described in any one of the appendices 1 through 10.

[0186] (Note 12) The aforementioned robot does not have the ability to converse. A conversation output system as described in any one of the appendices 1 through 11.

[0187] (Note 13) The output unit includes a display, The output control means displays the conversation data created by the creation means on the display, thereby causing the user to recognize that the robot and the predetermined object are having a conversation that conforms to at least the first information. A conversation output system as described in any one of the appendices 1 through 12.

[0188] (Note 14) The output section includes a speaker, The output control means synthesizes the conversation data created by the creation means into speech and outputs it from the speaker, thereby causing the user to recognize that the robot and the predetermined object are having a conversation that conforms to at least the first information. A conversation output system as described in any one of the appendices 1 through 13.

[0189] (Note 15) Information acquisition means for acquiring first information which is at least one of the following: information about the user of the robot and situational information which is information about the surrounding conditions of the robot. A means for creating conversation data to make the user of the robot recognize that the robot and a predetermined object are having a conversation that conforms to at least the first information, based on the acquired first information. The system includes a transmission means for transmitting the created conversation data to a conversation output terminal provided separately from the robot, Conversation output server.

[0190] (Note 16) A conversation output method executed by a conversation output system, A step of acquiring first information which is at least one of the following: information about the user of the robot and situational information which is information about the surrounding conditions of the robot. Based on the acquired first information, the step of creating conversation data to allow the user to recognize that the robot and a predetermined object are having a conversation that is at least consistent with the first information, The steps include controlling an output unit, which is provided separately from the robot, to output information based on the conversation data created, thereby causing the user to recognize that the robot and the predetermined object are having a conversation that matches at least the first information, A method for outputting conversations that includes this.

[0191] (Note 17) On the computer of the conversation output server, A process for acquiring first information which is at least one of the following: information about the robot's user and situational information which is information about the surrounding conditions of the robot. Based on the acquired first information, a process is performed to create conversation data that causes the user of the robot to recognize that the robot and a predetermined object are having a conversation that conforms to at least the first information. The process involves controlling an output unit, which is provided separately from the robot, to output information based on the conversation data created, thereby causing the user to recognize that the robot and the predetermined object are having a conversation that matches at least the first information. To execute program. [Explanation of Symbols]

[0192] 100, 100', 100'', 101, 101', 101'', 102, 102', 103, 103', 103''... Robot, 110, 410... Situation acquisition unit, 120, 210, 420... Short-range communication unit, 130... User intention acquisition unit, 131... Human presence sensor, 132... Camera, 133... Microphone, 134, 280... Speaker, 135... Legs, 136... Tail, 140, 220, 310... LTE communication unit, 150... Self-diagnosis unit, 200, 200', 200'', 201, 202, 202', 202'', 203, 203', 203''... Conversation output terminal, 230... Terminal control 231...Conversation Output Unit, 232, 321...Scenario Selection Unit, 233...Status Notification Creation Unit, 240, 330...Storage Unit, 241, 331...Scenario Storage Unit, 242, 332...User Profile Storage Unit, 243, 333...User Status Storage Unit, 250...Scenario Conversation Output Terminal, 260...User Input Unit, 270...Display, 300, 301, 302...Conversation Output Server, 320...Control Unit, 322...Text Creation Unit, 400, 400A, 400B, 400C, 400'...Sensor Device, 1000, 1001, 1002, 1003, 1004, 1005...Conversation Output System

Claims

1. Robots modeled after living creatures, When the robot detects an abnormality through its self-diagnosis, it inquires about countermeasures for the abnormality to a designated service center via wireless communication, and displays the content of the inquiry in natural language on a display unit so that it appears as if the robot and the person in charge at the designated service center have had a conversation. Equipped with, The output terminal displays information indicating the location of the abnormality in a representation corresponding to a part of a living organism. An anomaly notification system characterized by the following:

2. The output terminal creates a message containing the content of the inquiry based on the information regarding the anomaly received from the robot. The abnormality notification system according to feature 1.

3. An anomaly notification method performed by an anomaly notification system comprising a robot modeled after a living creature and an output terminal, The aforementioned output terminal is If the robot detects an abnormality through its self-diagnosis, it will inquire about countermeasures for the abnormality with a designated service center via wireless communication, and the content of the inquiry will be displayed on the output terminal's display unit in natural language so that it appears as if the robot is having a conversation with the person in charge at the designated service center. The information indicating the location of the abnormality is displayed using a representation that corresponds to the part of the organism. An abnormality notification method characterized by the following.

4. In an anomaly notification system comprising a robot modeled after a living creature and an output terminal, the computer of the output terminal is, When the robot detects an abnormality through its self-diagnosis, the control means functions to inquire about countermeasures for the abnormality from a designated service center via wireless communication, and to display the content of the inquiry in natural language on the display unit of the output terminal so that it appears as if the robot is having a conversation with the person in charge at the designated service center. The control means displays information indicating the location of the abnormality in a representation corresponding to a part of the organism. A program characterized by the following features.

Images