Autonomous robots that act based on experience

Abstract

This technology provides robots with the ability to make behavioral choices while creating memories associated with places or objects. [Solution] The robot comprises a memory unit, a processor, and a drive mechanism. The robot's processor detects an object, recognizes the impression memory associated with the object, and determines the motion associated with the impression memory of the object. The drive mechanism executes the motion determined by the processor.

Description

Technical Field

[0001] The present invention relates to a robot that autonomously makes action selections according to internal states or external environments.

Background Art

[0002] People keep pets in search of healing. On the other hand, there are various reasons such as not being able to secure enough time to take care of pets, not having a living environment where pets can be kept, having allergies , finding it hard to bear the parting due to the death of a pet, etc., and many people give up on having pets. If there were a robot that could perform the role of a pet, it might be able to give healing to those who cannot keep pets so that they can experience what a pet gives. (See Patent Documents 1 and 2).

Prior Art Documents

Patent Documents

[0003]

Patent Document 1

Patent Document 2

Summary of the Invention

Problems to be Solved by the Invention

[0004] In recent years, although robot technology has been rapidly advancing, it has not yet achieved the presence as a companion like a pet. This is because people do not think that robots have free will. People observe the actions of pets as if they have free will, feel the existence of free will in pets, empathize with pets, and are healed by pets. [[ID=�8]]

[0005] One of the characteristics of living beings is "memory." Among these memories are associative memories that are linked to places and objects. For example, by experiencing the warmth of a window, The impression of "warmth" is associated with "by the window." Or, next to a red coat. When we have a pleasant experience, we may associate a positive image with the red coat. Living organisms form various memories based on their external environment and record them in their external environment. We make choices based on our memories.

[0006] This invention was completed based on the above-mentioned problem recognition, and its main purpose is to enable robots to make action choices while creating memories associated with places or objects. The objective is to provide the technology to enable this. [Means for solving the problem]

[0007] An autonomous robot in one aspect of the present invention comprises a memory unit, a processor, and a drive mechanism that executes motions selected by the processor. The processor recognizes the impression memory of an object and determines the motion for that object based on that impression memory. An autonomous robot in another aspect of the present invention includes a motion control unit that selects the robot's motion, a drive mechanism that executes the motion selected by the motion control unit, an image acquisition unit that acquires captured images, and a unit that acquires feature points from the captured images. The image feature acquisition unit obtains image feature information by extracting the captured image Environmental information acquisition unit acquires environmental information of the location where the image was acquired, and image feature information A map management unit generates environmental memory by associating it with environmental information, It is equipped with. The operation control unit refers to the environmental memory and selects environmental information that satisfies predetermined destination conditions. Set the corresponding point as the target destination for movement.

[0008] An autonomous robot in another aspect of the present invention includes a motion control unit that selects the robot's motion, and a unit that executes the motion selected by the motion control unit. A drive mechanism, an image acquisition unit that acquires captured images, and a feature amount of the captured image The image feature acquisition unit obtains image feature information by extracting it, and the captured image At the time the data is acquired, primary environmental information is obtained, and image feature information and primary The first environmental information acquisition unit registers environmental information in association with it, and newly acquired This corresponds to image feature information whose similarity to the first image feature information is above a predetermined threshold. The primary environmental information that is obtained corresponds to the secondary environmental information that corresponds to the first image feature information. It includes a second-level environmental information acquisition unit that identifies and identifies the following. The operation control unit performs the operation based on both the primary and secondary environmental information. Change the behavioral characteristics of the bot.

[0009] An autonomous robot in another aspect of the present invention includes a motion control unit that selects the robot's motion, a drive mechanism that executes the motion selected by the motion control unit, an image acquisition unit that acquires an image, and a feature amount of the image. The image features to be extracted are obtained from the image feature acquisition unit and the primary environment at the location where the captured image was acquired. The first environmental information acquisition unit acquires information, and the second unit takes feature quantities as input and outputs environmental information. The features of the newly acquired image are input to a predetermined inference model. By using this as a force, the environmental information output from the inference model becomes secondary environmental information. It includes a second-stage environmental information acquisition unit that acquires the information as follows. The operation control unit performs the operation based on both the primary and secondary environmental information. Change the behavioral characteristics of the bot.

[0010] In another aspect of the present invention, the autonomous mobile robot includes an operation control unit that selects the motion of the robot, a drive mechanism that executes the motion selected by the operation control unit, an object detection unit that detects an object, an environmental information acquisition unit that acquires environmental information at the time when the object is detected, and an object management unit that generates environmental memory by associating the object with the environmental information. The operation control unit changes the behavior characteristics based on the environmental memory. In another aspect of the present invention, the autonomous mobile robot includes an operation control unit that selects the motion of the robot, a drive mechanism that executes the motion selected by the operation control unit, an object detection unit that detects an object, and a state management unit that changes the emotion parameters when the object is detected. The operation control unit changes the behavior characteristics of the robot according to the emotion parameters. In another aspect of the present invention, the autonomous mobile robot includes an operation control unit that selects the motion of the robot, a drive mechanism that executes the motion selected by the operation control unit, an object detection unit that detects a plurality of objects, an environmental information acquisition unit that acquires environmental information at the time when the plurality of objects are detected, and a state management unit that sets impression memory for the plurality of objects based on the environmental information. When the plurality of objects are redetected, the state management unit changes the emotion parameters based on the impression memory associated with the plurality of objects, and the operation control unit changes the behavior characteristics of the robot according to the emotion parameters.

[0011] In another aspect of the present invention, the autonomous mobile robot includes an operation control unit that selects the motion of the robot, a drive mechanism that executes the motion selected by the operation control unit, an object detection unit that detects an object, and a state management unit that changes the emotion parameters when the object is detected. The operation control unit changes the behavior characteristics of the robot according to the emotion parameters. In another aspect of the present invention, the autonomous mobile robot includes an operation control unit that selects the motion of the robot, a drive mechanism that executes the motion selected by the operation control unit, an object detection unit that detects a plurality of objects, an environmental information acquisition unit that acquires environmental information at the time when the plurality of objects are detected, and a state management unit that sets impression memory for the plurality of objects based on the environmental information. When the plurality of objects are redetected, the state management unit changes the emotion parameters based on the impression memory associated with the plurality of objects, and the operation control unit changes the behavior characteristics of the robot according to the emotion parameters. In another aspect of the present invention, the autonomous mobile robot includes an operation control unit that selects the motion of the robot, a drive mechanism that executes the motion selected by the operation control unit, an object detection unit that detects a plurality of objects, an environmental information acquisition unit that acquires environmental information at the time when the plurality of objects are detected, and a state management unit that sets impression memory for the plurality of objects based on the environmental information.

[0012] In another aspect of the present invention, the autonomous mobile robot includes an operation control unit that selects the motion of the robot, a drive mechanism that executes the motion selected by the operation control unit, an object detection unit that detects a plurality of objects, an environmental information acquisition unit that acquires environmental information at the time when the plurality of objects are detected, and a state management unit that sets impression memory for the plurality of objects based on the environmental information. When the plurality of objects are redetected, the state management unit changes the emotion parameters based on the impression memory associated with the plurality of objects, and the operation control unit changes the behavior characteristics of the robot according to the emotion parameters. In another aspect of the present invention, the autonomous mobile robot includes an operation control unit that selects the motion of the robot, a drive mechanism that executes the motion selected by the operation control unit, an object detection unit that detects a plurality of objects, an environmental information acquisition unit that acquires environmental information at the time when the plurality of objects are detected, and a state management unit that sets impression memory for the plurality of objects based on the environmental information. When the plurality of objects are redetected, the state management unit changes the emotion parameters based on the impression memory associated with the plurality of objects, and the operation control unit changes the behavior characteristics of the robot according to the emotion parameters. In another aspect of the present invention, the autonomous mobile robot includes an operation control unit that selects the motion of the robot, a drive mechanism that executes the motion selected by the operation control unit, an object detection unit that detects a plurality of objects, an environmental information acquisition unit that acquires environmental information at the time when the plurality of objects are detected, and a state management unit that sets impression memory for the plurality of objects based on the environmental information. When the plurality of objects are redetected, the state management unit changes the emotion parameters based on the impression memory associated with the plurality of objects, and the operation control unit changes the behavior characteristics of the robot according to the emotion parameters. In another aspect of the present invention, the autonomous mobile robot includes an operation control unit that selects the motion of the robot, a drive mechanism that executes the motion selected by the operation control unit, an object detection unit that detects a plurality of objects, an environmental information acquisition unit that acquires environmental information at the time when the plurality of objects are detected, and a state management unit that sets impression memory for the plurality of objects based on the environmental information. When the plurality of objects are redetected, the state management unit changes the emotion parameters based on the impression memory associated with the plurality of objects, and the operation control unit changes the behavior characteristics of the robot according to the emotion parameters. In another aspect of the present invention, the autonomous mobile robot includes an operation control unit that selects the motion of the robot, a drive mechanism that executes the motion selected by the operation control unit, an object detection unit that detects a plurality of objects, an environmental information acquisition unit that acquires environmental information at the time when the plurality of objects are detected, and a state management unit that sets impression memory for the plurality of objects based on the environmental information.

Advantages of the Invention

[0013] According to the present invention, it becomes easier to generate memories in robots and to select actions based on those memories. [Brief explanation of the drawing]

[0014] [Figure 1] This is a conceptual diagram illustrating how robots form memories based on location. [Figure 2] Figure 2(a) is a front view of the robot. Figure 2(b) is a side view of the robot. [Figure 3] This is a cross-sectional view illustrating the structure of the robot. [Figure 4] This is a hardware configuration diagram of the robot in its basic configuration. [Figure 5] This is a functional block diagram of a robot system. [Figure 6] This is a functional block diagram of the robot in this embodiment. [Figure 7] This is a schematic diagram illustrating how to create keyframes. [Figure 8] This is a data structure diagram of the environment map information. [Figure 9] This is a data structure diagram of object information. [Figure 10] This is a schematic diagram illustrating a chain of keyframes. [Figure 11] This flowchart shows the process for updating environmental map information. [Figure 12] This is a schematic diagram illustrating how a robot selects a destination in response to an event. [Figure 13] This is a flowchart showing the event processing process. [Figure 14] This flowchart shows the details of the object detection process shown in S28 of Figure 11. [Figure 15] This is a data structure diagram of object environment information in a modified example. [Figure 16] This is a diagram illustrating the data structure of room impression information in a modified example. [Figure 17]In the modified example, this flowchart shows the processing steps when an object is detected. [Figure 18] This is a data structure diagram of behavioral experience information in another modified example. [Modes for carrying out the invention]

[0015] Figure 1 illustrates how robot 100 forms memories based on location. This is a conceptual diagram for that purpose. The robot 100 of this embodiment is equipped with a small camera. This camera periodically captures images of the surrounding area, thereby generating a large number of captured images. (Still images) are acquired. Additionally, robot 100 uses a microphone and temperature sensor. Environmental information is acquired using various sensors, such as sound and noise. Hey, the radio wave reception strength, temperature, user presence, and other detection values ​​from various sensors. It is a collection of these. Robot 100 stores memory based on captured images (hereinafter referred to as "images"). This forms memories based on environmental information (hereinafter referred to as "environmental memory") and memories based on environmental information. To achieve this, by associating these two types of memories, the robot can determine the location of 100. A memory of that location (hereinafter referred to as "spatial memory") is formed.

[0016] The image storage layer 400 in Figure 1 is a storage layer that conceptually represents image storage, and environmental memory Memory layer 402 is a memory layer that conceptually represents environmental memory. Image memory is a memory layer that contains multiple keys. - It is a collection of frames 404. Keyframe 404 is in the captured image. This is distribution information of feature points (feature quantities). The robot 100 of this embodiment is Graph-based SLAM (Simultaneous Localization and Image Feature Analysis) Mapping technology, more specifically ORB (Oriented Fast and Rotated BRI) EF) Keyframe 404 is formed using SLAM technology based on features (details (Details will be provided later.)

[0017] The robot 100 periodically forms keyframes 404 while moving. This results in a collection of 404 keyframes, or in other words, an image feature distribution. This forms the image storage layer 400. The robot 100 takes The acquired keyframe 404 and the numerous keyframes 404 already owned By comparing the image memory layer 400, the robot 100 can determine its own location within the image memory layer 400. In other words, the robot 100 performs "spatial recognition" by comparing the captured image it is currently viewing with previously viewed captured images (memories) and aligning its current situation with its past memories.

[0018] When acquiring images, the robot 100 also acquires environmental information. Memory is a collection of environmental information at each location. Robot 100 is an environmental The direction of movement is selected based on the information (environmental memory layer 402). For example, a robot When Net100 is at a high-temperature location P1, in a cool area with an average temperature of 20 degrees or less... Let's assume that the robot 100 has chosen to move to a point. First, the robot 100 must satisfy the above conditions. The environmental information is retrieved from the environmental memory layer 402, and the environmental information for location P2 satisfies the above conditions. Let's assume it was found that... Next, robot 100 will find the key corresponding to point P2. - Identify frame 404. Robot 100 identifies the image memory layer 400 (key Identify the movement path to point P2 by referring to the distribution of frame 404. With the above control method, the robot 100 will operate under the condition that the average temperature is 20 degrees or less. Move towards point P2, which satisfies the conditions.

[0019] The basic configuration of robot 100 will be explained below in relation to Figures 2 to 5. Next, we will explain the methods of memory formation and utilization in this embodiment. ru. Note that the basic configuration of robot 100 uses an external keyframe instead of keyframe 404. It is assumed that the position is recognized by the sensor. Keyframe 404 and External sensors may be used in conjunction.

[0020] [Basic configuration] Figure 2(a) is a front view of robot 100. Figure 2(b) is a front view of robot This is a side view of the 100. The robot 100 in this embodiment is based on the external environment and internal state. It is an autonomous robot that makes decisions about its actions. The external environment is controlled by cameras and sensors. It is recognized by various sensors such as motion sensors. The internal state is determined by robot 100. Emotions are quantified as various parameters that express them. Robot 100 Its range of activity is limited to the interior of the owner's home. The following concerns Robot 100 Those who use it are called "users."

[0021] The body 104 of robot 100 has an overall rounded shape, Formed from soft and elastic materials such as urethane, rubber, resin, and fibers. Including the outer skin. The robot 100 may be dressed in clothes. Total weight of the robot 100 They weigh approximately 5 to 15 kilograms and are about 0.5 to 1.2 meters tall. With its moderate weight, roundness, softness, and pleasant feel, it is a great choice for you. The idea is that the robot 100 is easy to pick up and makes you want to pick it up. The effect will be achieved.

[0022] Robot 100 has a pair of front wheels 102 (left wheel 102a, right wheel 102b) and , including one rear wheel 103. The front wheel 102 is the drive wheel, and the rear wheel 103 is the driven wheel. It is a wheel. The front wheel 102 does not have a steering mechanism, but its rotational speed and direction are individually controlled. It is said to be separately controllable. The rear wheel 103 is a caster, and robot 1 00 is rotatable to move forward, backward, left, and right. The rear wheel 103 is O Muni wheels are also acceptable.

[0023] The front wheels 102 and rear wheels 103 are driven by a drive mechanism (rotation mechanism, linkage mechanism) It can be completely stored in the body 104. Even when running, most of each wheel is in the body. Although hidden in body 104, when each wheel is fully retracted into body 104, the robot... Unit 100 becomes immobile. That is, due to the wheel retraction operation... Then the body 104 lowers and sits on the floor surface F. In this seated position, The flat seating surface 108 (ground contact bottom surface) formed at the bottom of the D 104 is on the floor surface F It comes into contact with.

[0024] The robot 100 has two hands 106. The hands 106 are capable of grasping objects. It does not have that function. Hand 106 can perform simple actions such as raising, shaking, and vibrating. The two hands 106 can also be controlled individually.

[0025] The eye 110 is capable of displaying images using liquid crystal elements or organic EL elements. Robot 100 is equipped with a microphone array capable of identifying the direction of the sound source and ultrasonic sensors. It is equipped with various sensors, including a built-in speaker, and can also emit simple voice messages.

[0026] Horns 112 are attached to the head of robot 100. Because Bot100 is lightweight, users can grab the horn 112 to access the robot. It is also possible to lift the T100. A panoramic camera is attached to the Horn 112. It can be mounted to capture images of the entire upper area of ​​robot 100 at once.

[0027] Figure 3 is a schematic cross-sectional view showing the structure of robot 100. As shown in Figure 3, the body 104 of the robot 100 is attached to the base frame 3 08, main frame 310, pair of resin wheel covers 312 and outer Includes skin 314. The base frame 308 is made of metal and body 104 It forms the core and supports the internal mechanism. The base frame 308 is The upper plate 332 and the lower plate 334 are connected to multiple side plates 33 It is constructed by connecting the upper and lower sections by 6. Passing between the multiple side plates 336 Sufficient spacing is provided to allow for ventilation. The unit houses the battery 118, control circuit 342, and various actuators. It is accepted.

[0028] The main frame 310 is made of resin material, and the head frame 316 and the torso Includes frame 318. The head frame 316 is hollow and hemispherical, and the robot It forms the head skeleton of T100. The torso frame 318 has a stepped cylindrical shape. This forms the torso skeleton of the robot 100. The torso frame 318 is a base frame It is fixed together with frame 308. The head frame 316 is attached to the torso frame 3 It is assembled to the upper end of 18 so as to be displaceable relative to it.

[0029] The head frame 316 has a yaw axis 320, a pitch axis 322, and a roll axis. It is provided with three axes 324 and actuators 326 for rotating each axis. The actuator 326 has multiple servos to drive each axis individually. Includes motor. The yaw axis 320 is driven for the swivel motion, and for the nodding motion The pitch axis 322 is driven, and the roll axis 324 is driven for the head tilting motion. It is driven.

[0030] A plate 325 supporting the yaw axis 320 is located on the upper part of the head frame 316. It is fixed in place. Plate 325 has multiple compartments to ensure ventilation between the top and bottom. A number of ventilation holes 327 are formed.

[0031] A metal support is provided to support the head frame 316 and its internal mechanism from below. A base plate 328 made of cross is provided. The base plate 328 is cross It is connected to the plate 325 via the link mechanism 329 (pantograph mechanism) Meanwhile, via joint 330, upper plate 332 (base frame It is connected to Mu308.

[0032] The fuselage frame 318 houses the base frame 308 and the wheel drive mechanism 370. The wheel drive mechanism 370 includes a pivot shaft 378 and an actuator 379. This includes the lower half of the body frame 318 between the front and the wheel cover 312. The width is made narrow in order to form a storage space S for the wheel 102.

[0033] The outer shell 314 is made of urethane rubber, and the main frame 310 and wheels The hand 106 covers the outer skin 314 from the outside. The hand 106 is integrally molded with the outer skin 314. An opening 390 for introducing outside air is provided at the upper end of the outer shell 314. ru.

[0034] Figure 4 is a hardware configuration diagram of robot 100. Robot 100 has an internal sensor 128, a communication device 126, a storage device 124, Includes a processor 122, a drive mechanism 120, and a battery 118. The sass 122 and the memory device 124 are included in the control circuit 342. Each unit is Power line 130 and signal line 132 are connected to each other. Battery 11 8 supplies power to each unit via power line 130. Each unit is Control signals are sent and received via line 132. Battery 118 is lithium-ion It is a rechargeable battery and the power source for robot 100.

[0035] The internal sensor 128 is a collection of various sensors built into the robot 100. Specifically, it includes a camera (spherical camera), a microphone array, and a rangefinder. Infrared sensor, thermal sensor, touch sensor, accelerometer, odor sensor These include sensors. The touch sensor consists of an outer casing 314 and a main frame 310. It is installed in between and detects user touch. The odor sensor is a known sensor that applies the principle that electrical resistance changes due to the adsorption of odor-causing molecules.

[0036] The communication device 126 is a communication module that performs wireless communication with various external devices. The memory device 124 is composed of non-volatile memory and volatile memory. It is configured and stores computer programs and various configuration information. (Processor) 122 is a means for executing a computer program. The drive mechanism 120 is It includes multiple actuators and the wheel drive mechanism 370 described above. It will also be equipped with a display and speakers.

[0037] The drive mechanism 120 mainly consists of wheels (front wheels 102) and a head (head frame). 316) is controlled. The drive mechanism 120 controls the direction of movement and movement of the robot 100. In addition to changing the speed, the wheels (front wheel 102 and rear wheel 103) are raised and lowered. It is also possible. When the wheels are raised, they are fully retracted into the body 104. The robot 100 makes contact with the floor surface F on its seating surface 108, and assumes a seated position. The drive mechanism 120 controls the hand 106 via the wire 134.

[0038] Figure 5 is a functional block diagram of the robot system 300. The robot system 300 consists of a robot 100, a server 200 and multiple Includes external sensor 114. Components of robot 100 and server 200 This involves calculations performed by the CPU (Central Processing Unit) and various coprocessors. Devices, memory and storage devices, and wired or wireless connections between them. Hardware including communication lines and memory, and processing instructions for the arithmetic unit. This is achieved by software that supplies it. Computer programs are Device drivers, operating systems, and the layers above them Various application programs, and functions common to these programs. It may be composed of the libraries provided. Each block is described below. This represents a functional block, not a hardware-level configuration. Some of the functions of robot 100 may be implemented by server 200, Even if some or all of the functions of server 200 are implemented by robot 100 good.

[0039] Multiple external sensors 114 are pre-installed inside the house. Server 20 The position coordinates of the external sensor 114 are registered in 0. Inside the robot 100 Based on the information obtained from sensor 128 and multiple external sensors 114, The server 200 determines the basic actions of the robot 100. The external sensor 114 It is intended to augment the sensory organs of robot 100, and server 200 is robot This is to enhance the brain of Robot 100. Robot 100's communication device 1 26 communicates periodically with external sensor 114, and server 200 communicates with external sensor 11 The position of the robot 100 is determined by step 4 (see also Patent Document 2).

[0040] (Server 200) Server 200 includes a communication unit 204, a data processing unit 202, and a data storage unit. Includes 206. The communication unit 204 performs communication processing with the external sensor 114 and the robot 100. Responsible for this. The data storage unit 206 stores various types of data. Data processing unit 20 2 is the data acquired by the communication unit 204 and stored in the data storage unit 206. The data processing unit 202 performs various processes based on the data received. It also functions as an interface for the signal unit 204 and the data storage unit 206.

[0041] The data storage unit 206 consists of a motion storage unit 232 and a personal data storage unit 21 Includes 8. Robot 100 has multiple motion patterns. The 6 shakes, swerving as it approaches the owner, who tilts his head. Various motions are defined, such as gazing at something.

[0042] The motion storage unit 232 defines the motion control content. Stores the "motion file". Each motion is identified by a motion ID. The motion file is also stored in the motion storage unit 160 of the robot 100. The download will begin. Server 200 will decide which motion to execute. Sometimes it's decided by Robot 100, and sometimes it's decided by Robot 100.

[0043] Many of the robot 100's motions are complex, involving multiple unit motions. It is composed of motions. For example, when robot 100 approaches its owner, it has a unit motion of turning to face the owner, and a unit motion of approaching while raising its hand. Position motion, approaching while shaking the body, unit motion, raising both hands It may also be expressed as a combination of unit motions for sitting. By combining four motions, "approach the owner and then move your hand." The motion involves lifting up, then shaking the body before finally sitting down. The motion file contains information about the actuators installed on the robot 100. Rotation angle and angular velocity are defined in relation to the time axis. According to the (actuator control information), each actuator By controlling ETA, various motions can be expressed.

[0044] The transition time when changing from one unit motion to the next. It's called an "interval." An interval is the time required for a change in unit motion. It should be defined according to the content of the pause or motion. The length of the interval is adjustable. It is adjustable. Below, we will discuss when and which motion to choose, and the various steps involved in realizing the motion. Settings related to the control of robot 100's actions, such as adjusting the output of the cutter. These are collectively referred to as "behavioral characteristics." The behavioral characteristics of robot 100 are: motion selection Defined by selection algorithm, motion selection probability, motion file, etc. It will be done.

[0045] The motion storage unit 232 stores motion files as well as various events. A motion selection table that defines the motion to be executed when an event occurs. Stores the following. In the motion selection table, there must be 1 or more for each event. The motion and its selection probability are associated with each other.

[0046] The personal data storage unit 218 stores user information. Specifically, Master information indicating the level of familiarity with the user and the user's physical and behavioral characteristics Store the data. Other attribute information, such as age and gender, may also be stored.

[0047] Robot 100 has an internal parameter called "intimacy level" for each user. Robot 100 will pick you up, talk to you, and generally be kind to you. When a user recognizes actions that indicate goodwill, their level of intimacy with that user increases. Users who do not interact with Robot100, users who behave violently, and users who are encountered infrequently. The level of intimacy with users will decrease.

[0048] The data processing unit 202 includes a position management unit 208, a recognition unit 212, and an operation control unit 2 22. Includes a closeness management unit 220 and a status management unit 244. The position management unit 208 determines the position coordinates of the robot 100. State management unit 244 is a variety of physical parameters such as charge level, internal temperature, and processing load of processor 122. It manages various internal parameters such as the state. In addition, the state management unit 244 controls the robot A variety of emotions, including 100 different feelings (loneliness, curiosity, need for recognition, etc.) Manage the parameters. These emotional parameters are constantly fluctuating. Emotions The target location for robot 100 changes depending on the parameters. For example, When loneliness is heightened, Robot 100 moves to the user's location. Set it as a moving target point.

[0049] Emotional parameters change over time. Furthermore, various emotional parameters also change depending on the interaction described later. For example, when the owner asks to be held... When this happens, the emotional parameter indicating loneliness decreases, and the owner continues to feel lonely for a long time. When the loneliness parameter is not visible, the emotional parameter indicating loneliness gradually increases.

[0050] The recognition unit 212 recognizes the external environment. This includes temperature and humidity. Recognition of weather and seasons based on light levels and temperature, and recognition of shaded areas (safe zones) based on light levels and temperature. A wide variety of recognition is included. The recognition unit 156 of the robot 100 uses an internal sensor 1 Various environmental information is obtained by 28, and after initial processing, server 200 This is then forwarded to the recognition unit 212.

[0051] Specifically, the recognition unit 156 of the robot 100 identifies moving objects from an image, particularly The system extracts image regions corresponding to people and animals, and then extracts moving objects from the extracted image regions. A "feature vector" is extracted as a set of feature quantities that represent the physical and behavioral characteristics of a person. To output. Feature vector components (feature quantities) quantify various physical and behavioral characteristics. These are the numerical values. For example, the width of a human eye is quantified as a number in the range of 0 to 1. Form one feature vector component. Extract feature vectors from captured images of people. The method used is an application of known facial recognition technology. Robot 100 sends feature vectors to server 200.

[0052] The recognition unit 212 of the server 200 uses the built-in camera of the robot 100 to capture images. The feature vector extracted from the image and the personal data storage unit 218 have been pre-stored By comparing it with the feature vectors of registered users (clusters), The system determines which person the captured image represents (user identification process). Furthermore, the recognition unit 212 recognizes the user's facial expressions through image recognition, It estimates emotions. The recognition unit 212 recognizes moving objects other than people, such as pets. The system also performs user identification on cats and dogs.

[0053] The recognition unit 212 recognizes the various interaction actions performed by the robot 100. The recognition unit 212 also recognizes the actions of the robot 100. By recognizing the owner's response to this, it can be classified into positive and negative reactions. do. Pleasant and unpleasant behaviors are those in which the user's actions are perceived as biologically pleasing. It is determined by whether it is unpleasant. For example, being held by a robot For Robot 100, it was a pleasant experience, and being kicked was a pleasant experience for Robot 100. This is an unpleasant act. Positive and negative responses indicate whether the user's response is pleasing to the user. It is determined by whether it indicates a positive feeling or an unpleasant feeling. Being kicked is a positive reaction that shows the user's pleasant feelings, while being kicked is an unpleasant reaction for the user. This is a negative reaction that expresses emotion.

[0054] The motion control unit 222 of the server 200 is the motion control unit 150 of the robot 100. In collaboration with them, the motion of robot 100 is determined. (Operation of server 200) The control unit 222 determines the target location for the robot 100's movement and the movement route for that location. The operation control unit 222 creates multiple movement routes, and then, You may choose either of the travel routes.

[0055] The motion control unit 222 selects from multiple motions in the motion storage unit 232 Select a motion for Bot 100. Each motion has a selection option depending on the situation. The rates are associated with each other. For example, when the owner performs a good deed. This defines a selection method such as executing motion A with a 20% probability, and executing motion B with a 5% probability when the temperature exceeds 30 degrees.

[0056] The intimacy management unit 220 manages the intimacy level for each user. As described above. The level of intimacy is registered as part of the personal data in the personal data storage unit 218. When a pleasurable act is detected, the intimacy management unit 220 will... Increases intimacy. Intimacy decreases when unpleasant behavior is detected. Furthermore, the level of familiarity with an owner who hasn't been seen for a long period of time gradually decreases.

[0057] (Robot 100) Robot 100 includes a communication unit 142, a data processing unit 136, and a data storage unit 1 48, including an internal sensor 128 and a drive mechanism 120. The communication unit 142 corresponds to the communication device 126 (see Figure 4), and the external sensor 114 It is responsible for communication processing with server 200 and other robots 100. The storage unit 148 stores various types of data. The data storage unit 148 stores various types of data. This corresponds to 24 (see Figure 4). The data processing unit 136 is controlled by the communication unit 142. Based on the acquired data and the data stored in the data storage unit 148 The data processing unit 136 then executes various processes. This corresponds to a computer program executed by processor 122. The data processing unit 136 includes the communication unit 142, the internal sensor 128, and the drive mechanism 120. It also functions as an interface for the data storage unit 148.

[0058] The data storage unit 148 defines the various motions of the robot 100. Includes a suction storage section 160. The motion storage unit 160 of robot 100 contains the motion data of server 200. Various motion files are downloaded from the storage unit 232. The motion is identified by motion ID. The front wheel 102 is retracted and seated. By lifting hand 106 and rotating the two front wheels 102 in opposite directions, By rotating only one of the front wheels 102, the robot 100 performs a rotational action. To make it vibrate, rotate the front wheel 102 while it is retracted. To express various motions, such as pausing and turning around when moving away from the user, the operation of various actuators (drive mechanism 120) Timing, duration, and direction of movement are all chronologically represented in the motion file. It is defined. The data storage unit 148 also stores various data from the personal data storage unit 218. You may download it.

[0059] The data processing unit 136 includes a recognition unit 156 and an operation control unit 150. The motion control unit 150 of the robot 100 controls the motion control unit 222 of the server 200. We will work with them to determine the motions for Robot 100. The first motion is decided by server 200, and the other motions are decided by robot 100. It may be determined. Also, the robot 100 determines the motion, but the robot When the processing load on server 100 is high, server 200 will make the motion decision. It is also acceptable. The base motion is determined on server 200, and the robot Additional motions may be determined in step 100. How to divide the workload between server 200 and robot 100 is It should be designed according to the specifications of Bot System 300.

[0060] The motion control unit 150 of the robot 100 drives the selected motion to the drive mechanism 12 The command to execute is given to 0. The drive mechanism 120 will, according to the motion file, It controls each actuator.

[0061] The motion control unit 150 will "hold" when a user with a high level of closeness is nearby. The motion involves raising both hands 106 as a gesture of begging for something. It can also be used as is, and when you get tired of "carrying" it, you can keep the left and right front wheels 102 retracted. The motion of a dog refusing to be held is expressed by repeatedly alternating between reverse rotation and stopping. It is also possible. The drive mechanism 120 operates according to the instructions of the motion control unit 150. By driving the front wheels 102, hands 106, and neck (head frame 316), Make Bot100 perform various motions.

[0062] The recognition unit 156 of the robot 100 receives external information from the internal sensor 128. Interpret the information. The recognition unit 156 performs visual recognition (visual unit) and odor recognition (olfactory unit). The auditory and tactile parts of the brain are capable of recognizing sound and touch.

[0063] The recognition unit 156 extracts feature vectors from the captured image of the moving object. As shown above, feature vectors represent the physical and behavioral characteristics of a moving object. It is a set of lamesters (feature quantities). When a moving object is detected, odor is used. Physical and behavioral characteristics can also be extracted from sensors, built-in microphones, temperature sensors, etc. These features are also quantified and become feature vector components. Recognition Unit 1 56 is based on known technology described in Patent Document 2, etc., and uses a feature vector to generate a Identify the 'za'.

[0064] Of the series of recognition processes including detection, analysis, and judgment, the recognition unit of the robot 100 156 selects and extracts the information necessary for recognition, while interpretation processing such as judgment is handled by S. The recognition process is performed by the recognition unit 212 of server 200. This can be done using only the recognition unit 212, or only the recognition unit 156 of the robot 100. You can either proceed, or, as mentioned above, both parties can divide the roles and carry out the recognition process described above. You may do so.

[0065] When a strong impact is applied to the robot 100, the recognition unit 156 taps The sensor and acceleration sensor recognize this, and the recognition unit 2 of the server 200 User 12 perceives that a “violent act” has been committed by a nearby user. Even when -za grabbed Horn 112 and lifted Robot 100, it was a violent act. It is acceptable to recognize this. A user facing directly towards robot 100 at a specific volume When a voice is uttered in a specific region and frequency band, the recognition unit 212 of the server 200 They may perceive this as an act of "speaking to them." Also, body temperature When a certain temperature is detected, it is recognized that a "contact action" has been taken by the user. And when upward acceleration is detected while contact is recognized, the "hugging" action is performed. It is recognized that the physical contact when the user lifts body 104 It may be used for sensing, or the load on the front wheel 102 may decrease. It is acceptable to recognize this. In summary, the robot 100 acquires the user's actions as physical information using its internal sensor 128, and the recognition unit 212 of the server 200 determines whether the user is pleased or displeased. The recognition unit 212 of the server 200 also performs user identification processing based on feature vectors.

[0066] The recognition unit 212 of the server 200 receives various responses from the user to the robot 100. Recognize the pair. Some typical response behaviors among various response behaviors include pleasant or An unpleasant, positive, or negative response is associated with it. Generally, it is a response that is considered pleasant. Most responses are positive, while most responses that result in unpleasant behavior are negative. The response is as follows: Pleasant and unpleasant behaviors are related to intimacy, and positive and negative responses are related to the robot 10 This affects the choice of action 0.

[0067] The intimacy level of the server 200 is determined according to the response behavior recognized by the recognition unit 156. The management unit 220 changes the level of intimacy with the user. In principle, pleasant actions The level of intimacy with the user who performed the action increased, and the level of intimacy with the user who performed the offensive action increased. The density decreases.

[0068] Based on the above basic configuration, next, the robot 100 in this embodiment Regarding the implementation, in particular, the features and purpose of this implementation and the differences from the basic configuration are explained. I will explain it in terms of feelings.

[0069] [Implementation of memory for places and things] Figure 6 is a functional block diagram of the robot 100 in this embodiment. In this embodiment, the robot 100 uses a key frame instead of an external sensor 114. The robot recognizes the location based on M404. Robot 100 uses keyframe 40 Location may be recognized based on both 4 and the external sensor 114, but In the implementation configuration, the location is recognized based solely on keyframe 404. I will explain it.

[0070] (Server 200) The location management unit 208 of the data processing unit 202 includes a map management unit 168 and an object management unit 174. The map management unit 168 manages information that associates the image storage layer 400 and the environment storage layer 402 (hereinafter referred to as "environment map" or "environment map information"). The object management unit 174 manages object information that associates events with objects. An event is an occurrence that is recognized as having occurred when the environment information satisfies predetermined event conditions (described later).

[0071] The data storage unit 206 further contains the map storage unit 170 and object information. Includes storage unit 172. Map storage unit 170 stores environmental map information. The object information storage unit 172 stores object information. Details will be described later. However, the state management unit 244, based on environmental information and events, determines the emotional state. Change the meter.

[0072] (Robot 100) In addition to the recognition unit 156 and the operation control unit 150, the data processing unit 136 has a data processing unit 136. It includes an image acquisition unit 162. The image acquisition unit 162 is used by the robot 100. The camera onboard acquires captured images. The image acquisition unit 162 periodically For example, images are acquired at a frequency of twice per second.

[0073] The recognition unit 156 includes an image feature acquisition unit 164 and an environmental information acquisition unit 166. The image feature acquisition unit 164 extracts image features from the captured image, thereby - Generate frame 404. For information on the image feature extraction method, see Figure 7 below. This will be explained in relation to the above. The environmental information acquisition unit 166 acquires the captured image at the time of acquisition. Acquire environmental information. Keyframe 404 (image feature information) and environmental information This is transmitted to server 200 by communication unit 142. The map of server 200 The management unit 168 associates keyframe 404 with environmental information, thereby Create a boundary map.

[0074] Figure 7 is a schematic diagram showing how to create keyframe 404. As described above, the image acquisition unit 162 periodically acquires captured images. The image 412 in Figure 7 shows a chair 406 and a shelf 408. The image feature acquisition unit 164 selects the edge points of these objects as feature points 410. Extract the keyframes. Keyframe 404 is a set of feature points 410. The number and distribution of feature points 410 included in 404, in other words, the image Image features characterize keyframe 404. The captured image 412 itself. Instead, image memory based on keyframe 404 (image feature information) is data One advantage is that it can reduce the amount of waste.

[0075] Figure 8 shows the data structure diagram of the environmental map information 420. The environment map information 420 is stored in the map storage unit 170 of the server 200. Keyframe 404 is the keyframe ID (hereinafter referred to as "FID"). It is identified by ( ). The image feature acquisition unit 164 acquires keyframes from the captured image. Each time a 404 is formed, an FID is assigned. The environmental information acquisition unit 166 is - Environmental information is acquired at the same time as frame 404 (captured image). (The following describes the environment that is sensed at the same time as the acquisition of keyframe 404.) The information is also called "primary environmental information." Map management unit 168 is key Frame 404 and environmental information are associated and registered in the environmental map information 420. Environmental map information 420 links image memory and environmental memory.

[0076] In this embodiment, the environmental information includes received signal strength, user detection rate, average volume, and average temperature. Received signal strength refers to the received signal strength of the radio waves transmitted from the server 200. This indicates that the value changes depending on the distance between robot 100 and server 200. Environmental Information The acquisition unit 166 measures the received signal strength and ranks it from A to C in descending order of strength. According to Figure 8, keyframe 404 (hereinafter referred to as "keyframe") has FID=F01. (It is written as "F01") and the received signal strength is "A". The point corresponding to keyframe 404 (F01) (hereinafter referred to as "point (F01 The notation ")" indicates a location with good communication. On the other hand, - Frame 404 (F04) has a communication strength of "C", therefore location (F04 ) is a location with poor communication. In this embodiment, a location is defined as a location between 0.1 and 1 meter. It is defined as an area with a radius of approximately one unit.

[0077] The user detection rate indicates how often users are detected at each location, or in other words, the likelihood of encountering users at each location. This is defined for each user. User X1 detection rate indicates the detection rate for user X1. For example, if robot 100 was present at location (F01) 10 times If user X1 is detected twice in 10 visit opportunities, then The management unit 168 calculates the detection rate of user X1 as 20%. According to Figure 8... Location (F04) is more likely to encounter users X1 and X2, and location (F05) has a high probability of encountering user X2, nearly 50%. On the other hand, location (F07) has never encountered users X1 or X2.

[0078] The environmental information acquisition unit 166 detects the sound volume at each location using a microphone. The output is then ranked from highest to lowest average volume as A, B, and C. Based on the average volume, Quiet and noisy areas are identified. Similarly, the environmental information acquisition unit 166 The temperature at each location is detected by a temperature sensor, and the average temperature is calculated. The average temperature identifies hot and cool areas. This environmental information... These may be further classified according to other parameters such as time of day or season.

[0079] In addition to this, various other sensor information can also be included as environmental information. Yes, for example, even if odors detected by odor sensors are included in the environmental information. Okay, and you can also include information about whether or not there is a "ceiling," such as under the table, in the environmental information. Good. Not only the user detection rate (visibility), but also the "probability of someone touching it." User interaction experiences such as "the probability of being held" and "the probability of being hit." It may be defined as environmental information. Environmental information is, at a minimum, image feature information. Any external information is acceptable, and can be identified as tactile, auditory, or olfactory information such as temperature and wind direction. Any information is acceptable.

[0080] Server 200 refers to the environment map information 420 and determines the predetermined destination conditions. Set the target location for robot 100 to satisfy the condition. For example, loneliness The value of the emotional parameter that indicates this (hereinafter referred to as "emotional value (loneliness)"). When the detection rate of any user exceeds a threshold, the operation control unit 222 will A point where the value exceeds a predetermined threshold, for example, point (F05), is set as the target point for movement. Set it. The destination conditions can be set arbitrarily. For example, the current location When the temperature is 25 degrees or higher, locations with environmental information indicating an average temperature of 20 degrees or lower It is also acceptable to assume that the destination condition is met.

[0081] The status management unit 244 acquires the first status of the robot 100 at its current location. The emotional parameters are changed based on the following environmental information (hereinafter referred to as "(environmental information (This is called "primary reflection"). Also, the status management unit 244 of the robot 100 The emotional parameters are also changed based on environmental information acquired in the past at the current location (hereinafter, the environmental information sensed at the time of acquisition of keyframe 404 and prior to that will also be called "secondary environmental information") (hereinafter referred to as "(secondary reflection based on environmental information)"). In other words, the state management unit 244 changes the emotional parameters not only based on the current primary environmental information but also based on stored past secondary environmental information. The changes in emotional parameters based on environmental information will be described in detail in relation to Figure 11.

[0082] Figure 9 is a data structure diagram of object information 430. Object information 430 is stored in the object information storage unit 17 of the server 200. It is stored in 2. Various events occur around robot 100. The recognition unit 212 of server 200 determines that the primary environment information meets predetermined event conditions. When the conditions are met, it is determined that an event has occurred. The event conditions are determined by the user's perspective. Confirmed, held by the user, heard a sound above a certain volume, had water splashed on it The recognition unit 212 can be set arbitrarily, such as "received". The event conditions are set. The event is identified by its event ID.

[0083] When the recognition unit 212 detects the occurrence of an event, the communication unit 204 tells the robot An event detection signal is sent to 100. The image feature acquisition unit 164 detects the event When a detection signal is received, an object is detected from the most recent captured image. The objects can be arbitrarily set to things like coffee cups, chairs, shelves, dolls, etc. In this embodiment, the image feature acquisition unit 164 acquires multiple types of objects. Regarding the ct, multiple objects whose shape and color characteristics are defined in advance It has a ct pattern, and based on the object pattern, objects The image recognition unit 164 identifies any of the objects. When this happens, an object ID is assigned and sent to server 200. The event management unit 174 associates the event ID with the object ID. Object recognition can be achieved by applying known image recognition technologies.

[0084] The object management unit 174 manages object information 430. In the object information 430, the strength of the relationship between events and objects The degree of relevance is defined as an index value that indicates this relationship. The degree of relevance between an event (E01) and an object (B01) is an index that expresses the strength of the associative and causal relationship between the event (E01) and the object (B01), based on the likelihood that the object (B01) will be recognized when the event (E01) occurs.

[0085] When an object (B01) is detected when event (E01) occurs The object management unit 174 handles events (E01) and objects (B01 ) increases the relevance. On the other hand, when event (E01) occurs, If no object (B01) is detected, the object management unit 174 will This reduces the correlation between the element (E01) and the object (B01). In this state, the detection rate of objects at the time of event occurrence is used as the relevance. Define. For example, in the last 10 occurrences of event (E01), When object (B01) is detected 3 out of 10 times, the object tube The science department 174 sets the correlation between event (E01) and object (B01) to 30. Set to (%).

[0086] When event (E01) occurs, there is a high probability that object (B01) will be found. If detected, an event (E0) will occur near object (B01). 1) can be assumed to have a "causal relationship" that makes it more likely to occur. The degree of association is relative to the causal relationship. It can also be said that this is a parameter that indicates the degree of confidence that something will happen. For example, when a baby is being held Let's assume that a sofa (object) is detected at this location with a high probability. In that case, if you go near the sofa, there's a high chance you'll be picked up. This makes such predictions possible.

[0087] The state management unit 244 changes the emotion parameter based on the event. Hereinafter referred to as "(initial reflection by the event)". The event represents the first environmental information Since it is also a type of report, the initial reflection due to an event is usually in the first environmental information. This occurs simultaneously with the initial reflection. The status management unit 244 monitors the current status of the robot 100. Based on events associated with objects detected at the location This also changes emotional parameters (hereinafter referred to as "secondary reflection (due to events)" and (Call). The state management unit 244 changes the emotion parameter not only based on events that are actually occurring, but also on events associated with the object. Changes in emotional parameters based on events are related to Figures 13 and 14. I will elaborate on this.

[0088] Figure 10 is a schematic diagram showing a sequence of keyframes 404. The image feature acquisition unit 164 periodically acquires keyframes 404. Figure 1 In case 0, when acquiring keyframe 404 (F01), environmental information The acquisition unit 166 acquires environmental information (e01). The map management unit 168 acquires environmental information. In boundary map information 420, keyframe 404 (F01) and environment information ( Register e01) in association. After a certain period of time has elapsed, the robot 100 will - Obtain frame 404 (F02) and environmental information (e02). Map The management unit 168 compares keyframe 404 (F02) and environmental information (e02). In response, registration will be made, and if location (F01) and location (F02) are adjacent The information that this is happening will also be recorded.

[0089] After acquiring keyframe 404 (F02), robot 100 then further... Obtain frame 404 (F03). Map management unit 168 retrieves location (F0 2) Register the information that location (F03) is adjacent to location (F03) in the map storage unit 170. As a result of the above process, the robot system 300 will move from point (F01) By passing through point (F02), it is possible to reach point (F03). Learn about the spatial relationships in Rehm 404, or in other words, the spatial configuration.

[0090] After keyframe 404 (F03), keyframe 404 (F04), Keyframes 404 (F05) and 404 (F06) were acquired. Let's assume that the map management unit 168 is similar to keyframe 404 (F06). Detect keyframe 404 (F02). Specifically, keyframe 4 04 (F06) and keyframe 404 (F02) each have 410 feature points. The similarity of the numbers and distributions is calculated, and if the similarity is above a predetermined threshold, then they are considered similar. The map management unit 168 determines that keyframe 404 (F02) and key Frame 404 (F06) is determined to be keyframe 404 of the same location. The map management unit 168 updates keyframe 404(F02) using keyframe 404(F06) in the method described later. The map management unit 168 also recognizes that keyframe 404(F05) and keyframe 404(F02) are adjacent. In this way, the positional relationships of multiple keyframes 404 are registered in the map storage unit 170 as a graph-type database.

[0091] Figure 11 is a flowchart showing the update process for the environment map information 420. be. The update process shown in Figure 11 is performed periodically in the robot system 300. First, the image acquisition unit 162 acquires the captured image (S10). The image feature acquisition unit 164 extracts image features from the captured image and creates keyframes. Generates a 404 error (S12). The environmental information acquisition unit 166 acquires (first) environmental information The report is obtained (S14). The state management unit 244 adjusts the emotional parameters according to the environmental information. The meter is changed (S16). This is the "primary reaction based on environmental information" mentioned above. It is "visual." How to change emotional parameters in response to environmental information is It can be set arbitrarily. For example, if no user is detected in the captured image, If the undetected state continues for a specified period of time or longer, the emotional value (loneliness) will increase. It is also possible that the recognition unit 212 has environmental information that soothes loneliness. The location is set as the target location for movement. The motion control unit 150 determines the emotional value (loneliness) If the frequency of action selection by robot 100 increases, it may be possible to suppress the frequency of action selection by specific motions. You may choose option with a high probability. Environmental information influences emotional parameters, and Emotional parameters influence behavioral characteristics.

[0092] When the user picks up the character, their emotional state (loneliness) may decrease. When a volume within a predetermined range is detected, the emotional value (curiosity) may be increased. As mentioned above, the intimacy management unit 220 changes not only the emotional value but also the intimacy level. You may allow it.

[0093] The recognition unit 212 determines the similarity between the keyframe 404 extracted in S12 and the keyframe 404. Is there a keyframe 404 where the value is greater than or equal to a predetermined value? In other words, is there a keyframe 404 where the value is greater than or equal to a predetermined value? Whether or not the location is a location that has already been visited (hereinafter referred to as a "revisited location") Check if it is a revisited location (Y in S18). If it is a revisited location, the map management unit 168 updates the keyframe 404 (image feature information) for that revisited location. (S20).

[0094] The captured images include some that are stable, such as bookshelves and sofas, as well as ( Below, we call them "stable features," which are temporary, like coffee cups or documents. Some objects that are simply indistinct may also be displayed (hereinafter referred to as "unstable features"). Map Management Department 1 68 is the keyframe acquired the first time at a certain revisited location (F05). 404(F05-1) and the second keyframe 404(F05- 2) Compare keyframe 404 (F05-1) and keyframe 404 (F05-2) is almost identical overall, but the unstable features are different. This will result in a difference. For example, in keyframe 404 (F05-1) This includes the image features (unstable features) of the note, and keyframe 404 (F05 -2) may not contain the image features of the note. This is because the location This is because the notebook was removed between the first and second visits to P5. At this time, the map management unit 168 sets the keyframe 404 of point (F05). Remove feature point 410, which exhibits unstable characteristics, from (F05). Such control Depending on the method, keyframe 404 containing only stable features, in other words, spatial Keyframe 404 containing only image features that are easily used as clues for recognition. It can form. As a simpler method, the map management unit 168 can use the keyf to find revisited locations. As frame 404, the most recently acquired keyframe 404 has been replaced by the previous keyframe You can overwrite the 404 error.

[0095] The status management unit 244 has previously associated (second-order) rings with the revisited location. The emotional parameters are changed based on the environmental information (S22). This is as described above. This is a "secondary reflection based on environmental information." For example, if environmental information is available at location (F05) Report (e05-1) is associated with location (F05). When the site is revisited, it is assumed that environmental information (e05-2) has been obtained. State Management Section 244 is not only the first environmental information (e05-2) (S16), but also the previous The emotional parameters are also changed based on the second environmental information (e05-1). According to this control method, the impressions obtained in the past at a revisited location are used in the current situation. This can be reflected in the emotional state. The status management unit 244 can acquire new information through revisits. The environmental map information 420 will be updated using the acquired environmental information (e05-2). S24)

[0096] As an example, the location (F07) is a "noisy location" with a large average volume (see Fig. 8). When the robot 100 revisits the location (F07), assume that no loud sound is heard. In this case, based on the past information of the "noisy location", for example, the state management unit 244 increases the emotional value (curiosity). At the same time, the map management unit 168 decreases the average volume, which is a type of environmental information of the location (F07).

[0097] On the other hand, if it is not a revisited location (N in S18), the recognition unit 212 determines that it is a newly discovered location, and the state management unit 244 registers the newly acquired key frame 404 and environmental information in the environmental map information 420 (S26). Thereafter, the object determination process is executed (S28). The details of the object determination process will be described later in connection with Fig. 14.

[0098] For example, at time t1, a sofa exists at the location (F04), and then the sofa is moved so that at time t2, the sofa does not exist at the location (F04). In this case, the key frame 404 acquired at time t1 (F04-1) and the key frame 404 acquired at time t2 (F04- 2) may be so different that they cannot be regarded as the same. In this case, the robot 100 moves to the location (F04) relying on the image memory that the sofa exists, but since the sofa does not exist, the location (F04) stored in the image memory cannot be found. Even though the location (F04) is revisited, the location (F04) cannot be found. Therefore, at this time, the recognition unit 212 reaches a newly discovered location because the location (F04) cannot be found. It recognized this (N in S18) and set keyframe 404, where the sofa does not exist, to environment Register a new entry for Top Information 420.

[0099] The map management unit 168 has acquired keyframes 40 that have been acquired for more than a predetermined time. By deleting item 4, the memory of a location that has not been revisited for some time may be erased. Such a control method can represent forgetting by robot 100. Both are also effective in reducing the amount of data stored in the map storage unit 170. Also, at location (F04), there is an old and useless description that "a sofa exists". It can efficiently remove outdated information.

[0100] Figure 12 shows how the robot 100 selects a destination in response to an event. This is a schematic diagram to explain the method. In Figure 12, when robot 100 is at point P5, the neighboring point P Event E3 occurred at location 6. Event E3 is "User X6 at location P It is stated that "there exists in 6". The recognition unit 212 of server 200 is user When you do not remember a user with similar physical characteristics to X6, The X6 is identified as an unknown user. At this time, the operation control unit 2 of server 200 22 sets a point that meets the evacuation conditions as the target point for movement. Evacuation clause The setting is optional. Specifically, the appearance of users whose intimacy level is above a certain value. This may be a location where the rate exceeds a predetermined value, or a location with an obstruction such as a ceiling or wall. It may be any point. The operation control unit 222 determines the event and the evacuation destination conditions. Set up in advance a location that meets the evacuation conditions when the event occurs. This is detected from the environment map information 420. In Figure 12, the conditions for evacuation are met. Points P7 and P8 have been identified.

[0101] The motion control unit 222 moves either point P7 or point P8 to the target point Select as such. In Figure 12, the operation control unit 222 is located near point P5. Select point P7 as the target location for movement. For example, point P7 is the parent Let's assume that this is a location with a high detection rate for user X2, which has a high density. As mentioned above, Robot 100 moves to location P7 while retrieving image memories. 100 moves to location P7, and if it can detect user X2 as expected, then Stay at point P7 for a while. Even if you move to point P7, you cannot detect user X2. If so, the motion control unit 222 sets point P7 to point P8 as the moving target point. The system will be changed. According to this control method, "the user is surprised by event E3." The robot 100 can be made to express confused behavior such as "moving to location P7 in search of X2, and then moving to another safe location P3 because user X2 is not there."

[0102] When a specific event occurs, the operation control unit 222 performs the actions described above. A location that meets the evacuation conditions is set as the target location for movement. Status Management Unit 24 4 changes the emotional parameters in response to events. For example, high frequency When the sound is detected for a predetermined period of time or longer, event E4 is triggered, and the state management unit 2 44 may increase the emotion value (curiosity) depending on event E4. Environmental The report is a numerical representation of the external environment based on the detection values ​​of various sensors. A vent is an event in environmental information that meets certain event conditions. However, this is a special case of environmental information.

[0103] Figure 13 is a flowchart showing the event processing process. Every time (primary) environmental information is detected at S14 in FIG. 11, the recognition unit 212 determines whether the event condition is satisfied. The event processing shown in FIG. 13 is an event, or rather, it is executed when the (primary) environmental information meets any of the pre-set event conditions.

[0104] When an event is detected, the imaging image acquisition unit 162 acquires a peripheral image (S30). The peripheral image in this embodiment is two or more imaging images targeting two or more directions. The imaging image acquisition unit 162 usually acquires one imaging image with a camera targeting one direction, and when an event occurs, it may acquire a peripheral image with an omnidirectional camera capable of wide-angle shooting of the periphery of the robot 100.

[0105] The image feature acquisition unit 164 extracts image features from the peripheral image, or rather, from each of the plurality of imaging images (S32). The state management unit 244 changes the emotion parameters according to the event (S34). This is the "primary reflection by event" described above. How to change the emotion parameters for each event is pre-set.

[0106] Next, the image feature acquisition unit 164 detects an object from the image features of the <000098​​​​​​​​If object (B01) is detected, an event (E0 1) is more likely to be associated with this. If the object does not exist, (N in S36) S38 is skipped.

[0107] When the event is an evacuation event, in other words, if it is a specific event that requires evacuation from the current location (Y in S40), the operation control unit 222 will determine the evacuation destination. Set a location that meets the conditions as the target location for movement (S42). Evacuation event Otherwise, the S42 process will be skipped.

[0108] Figure 14 is a flowchart showing the details of the object determination process shown in S28 of Figure 11. —This is a chart. The image feature acquisition unit 164 uses the captured image acquired in S10 to select an object The system detects the problem (S50). The state management unit 244 refers to the image storage and checks the past Determine whether the same object has been detected before (S52). It is identical to an object that has been detected in the past, or more precisely, it is similar to an object that has been detected in the same location. When an object exceeding a certain value is detected (Y in S52), the state management unit 244 It changes the emotion parameter depending on the object (S54). This is, This is the "secondary reflection due to an event" mentioned above. If the object was not detected... When this happens, or when a new object is detected (N in S52), S54 is S It will be uploaded.

[0109] The state management unit 244 manages events associated with an object, and The emotion parameter changes depending on the degree of relevance between the object and the event. For example, object (B01) is involved in event (E03) and event ( It is associated with E04), and the object (B01) and event ( The relevance of E03) is 20%, and the relationship between objects (B01) and events (E0 Let's assume the relevance of (4) is 3%. In this case, the state management unit 244 changes the emotion value in response to the event with the highest relevance (E03). For example, if event (E03) is set to "increase the emotion value (request for approval) by 30%", the state management unit 244 takes the relevance of 20% into consideration and increases the emotion value (request for approval) by 6% (=30% × 20%). With this control method, even if event (E03) is not actually occurring, the emotional parameter is affected because the memory of event (E03) is stimulated by an object (B01) that is associated with event (E03), and the change in the emotional parameter can be expressed as a change in behavioral characteristics.

[0110] Based on the embodiments described above, robot 100 and robot including robot 100 I explained Bot System 300. Living organisms store environmental information obtained through sensory organs, especially the eyes, as memories. It is thought that this allows for a better perception of space. According to this embodiment, a method similar to that of living organisms is employed. Spatial recognition is possible. The external sensor 114, shown as an example of the basic configuration, is a required component. To prevent this, the external sensor 114 is placed in advance within the planned range of action of the robot 100. This eliminates the need for pre-installation work.

[0111] Robot 100 acts based on keyframe 404. Therefore, If the internal state of the space in which robot 100 belongs changes, for example, the desk or When the position of an object such as a chair changes, the robot 100 temporarily This could potentially lead to a loss of spatial awareness. On the other hand, new environments can create new possibilities. -The image memory is updated by obtaining frame 404. Like living organisms, they may be temporarily confused by changes in the environment, but will eventually adapt. Behavioral characteristics can be realized.

[0112] When Robot 100 entered the new space, it moved around while making noise. It accumulates 404 data and forms a sense of place (territory). Frame 404 is stored in the map storage unit 170 of server 200. The storage section 170 is formed in a non-volatile recording medium such as a hard disk. This is preferable. There is no need to walk around the space and rebuild the spatial memory each time the robot 100 is powered on.

[0113] Robot 100 changes its emotional parameters according to environmental information, The behavioral characteristics are changed according to the information parameters. Also, robot 100 The (primary) environmental information obtained at the "current location / current time" is insufficient. Emotions are not determined by (secondary) environmental information at the "current location / past time point". The parameters are changed. With this control method, past memories are used Because behavioral characteristics change, changes in emotions based on spatial memory can be expressed through behavior. Furthermore, environmental information is accumulated in association with the features obtained from captured images. Therefore, robot 100 is in a different place from the place where it formed its spatial memory. They can also perform actions that utilize their existing spatial memory.

[0114] Robot 100 changes its emotional parameters in response to events. Furthermore, Robot 100 is not just about the event itself, but also about the event. The emotional parameters also change depending on the object being used. In particular, the emotional parameters If you set an event that strongly affects the meter, only the event itself will be affected. Rather, even objects that evoke the event influence emotions. They can express their actions to describe the situation.

[0115] The present invention is not limited to the above embodiments or modifications, and can be materialized by modifying the components without departing from the spirit of the invention. By appropriately combining the multiple components disclosed in the form and modified versions, Various inventions may be formed. Also, the entire structure shown in the above embodiments and modifications You may remove some of the components from the set of elements.

[0116] One robot 100, one server 200, and multiple external sensors 114 Although it was explained that the robot system 300 is composed of robot 10 Some of the functions of 0 may be implemented by server 200, or server 200 Some or all of the functions may be assigned to robot 100. The controller 200 may control multiple robots 100, or multiple sub-controllers The 200 may cooperate to control one or more robots 100.

[0117] A third device other than robot 100 or server 200 may perform some of the functions. The collection of functions of robot 100 and server 200 described in Figures 5 and 6 can be broadly understood as a single "robot". How to distribute the multiple functions necessary to realize the present invention to one or more hardware devices should be determined in consideration of the processing capacity of each hardware device and the specifications required for the robot system 300.

[0118] As mentioned above, "robot in the narrow sense" does not include server 200. Regarding Robot 100, "robots in a broad sense" refers to robotic systems. This refers to the Tem300. Many of the functions of the Server 200 will be robotic in the future. It's possible that it will be integrated into the T100.

[0119] The status management unit 244 was described as being installed on the server 200, but the status The functions of the control unit 244 may be mounted on the robot 100. Similarly, The functions of the top management unit 168 and the object management unit 174 are also available on server 200. It could also be implemented in Robot 100.

[0120] The environmental information acquisition unit 166 of robot 100 acquires primary environmental information. It functions as the "primary environmental information acquisition unit" and the status management unit for the robot system 300. 244 may include a "Secondary Environmental Information Acquisition Unit" that acquires secondary environmental information. Of course, the functions of the second-stage environmental information acquisition unit are installed in the robot 100. This is also acceptable. The first environmental information acquisition unit acquires information at the current location P1 of the robot 100. (Assuming that image feature information R1 is also acquired at the same location), Primary environment Information E1 is acquired, and image feature information R1 and primary environmental information E1 are associated. You may register them together. Correspond this image feature information R1 to the primary environmental information E1. The location information to be assigned is the location information format of the robot 100 or server 200. It is stored in the storage unit (not shown). The second environmental information acquisition unit then determines the current location Image feature information R1 acquired at point P1, and the acquired image feature information set. By comparing these, we detect image feature information R2 whose image similarity exceeds a predetermined threshold. The second environmental information acquisition unit acquires environmental information associated with image feature information R2. Report E2 is identified as the secondary environmental information E2 at the current location P1. The motion control unit 222 may change the behavioral characteristics of the robot 100 located at location P1 based on both the primary environmental information E1 and the secondary environmental information E2, as described above.

[0121] If image feature information R1 and image feature information R1 are similar, then image feature information R1 There is a high probability that the acquisition point P1 of the image and the acquisition point P2 of the image feature information R1 are the same. i. On the other hand, the image feature information (landscape) of points P1 and P2, which are actually separate locations. It's possible they just happened to look similar. Therefore, the scenery at the current location P1 is The recognized robot 100 has a landscape similar to that of point P1, and point P2 (over Recalling the second environmental information E2 that was recognized at a different location visited previously, This enables a process where emotional parameters are influenced by secondary environmental information E2. For example, at point P1 where there is a chair C2 that resembles Dad's chair C1: This is based on the image of "Point P2 where Dad's chair C1 is located" that I saw previously. It is possible to achieve complex behavioral characteristics that stimulate the memory. Robot 100 is at point When P2 has a positive image, the operation control unit 222 controls point P By keeping the robot 100 in place for a predetermined time or longer in step 1, the robot 10 You can indirectly express your affection for chair C1.

[0122] The second environmental information acquisition unit acquires image features similar to the image feature information R1 of location P1. If no characteristic information is detected, the image acquisition unit 162 is instructed to take an image. Change direction. For example, robot 100 takes a picture of the front and keyframe When acquiring Mu404, the shooting direction was shifted 30 degrees to the right and modified. The image feature information R3 may be acquired first. Then, the second image information acquisition unit Alternatively, you may search for previously acquired image feature information that is similar to image feature information R3. Robot 100 can change the orientation of its built-in camera or its body orientation. Alternatively, the extraction of keyframe 404 from a panoramic image taken with a panoramic camera. You may change the direction of output. In this way, by changing the shooting direction as appropriate... Then, from the image feature information already acquired, search for image feature information similar to the current location. It becomes easier to extract.

[0123] A prediction model may be prepared that takes the features of the captured image as input and outputs environmental information. This prediction model takes the features f1, f2, ..., ft as input. A neural network that outputs environmental information e1, e2, ..., eu. It could also be a qua-model. For example, feature f1 is keyframe 404. The number of included feature points is 410, and the feature quantity f2 is the feature in keyframe 404. It may also be the number of nearly perpendicular lines formed by point 410. Even if environmental information e1 is the received signal strength and environmental information e2 is the detection rate of user X1 Good. The second environmental information acquisition unit acquires information at the current location P1 of the robot 100. The obtained features are input into the inference model, and its output is taken as secondary environmental information. It may be beneficial. The secondary environmental information output by the inference model is derived from the features. This is hypothetical environment information that is inferred. Even with this processing method, the current location Robot 100, having recognized the landscape at point P1, uses the image features of point P1 to... Secondary environmental information is imagined, and emotional parameters are determined by secondary environmental information. This allows for the implementation of a process that is affected by something.

[0124] The first environmental information acquisition unit extracts feature quantities from the image captured at the current location P1. When outputting and acquiring primary environmental information, these are used as input / output values. In other words, you can update the inference model using the training data.

[0125] When shooting keyframe 404, use the TOF (Time-of-Flight) method. A camera or depth camera such as a stereo camera may be used. The images captured include depth information. The image feature acquisition unit 164 captures depth Feature quantities may be extracted for each degree range. The first environmental information acquisition unit is a depth range The image feature information for each image may be registered in association with the primary environmental information E1. The depth range may be defined as the absolute distance range from the camera, It may also be determined as a relative distance range based on a certain feature point. If five depth ranges are defined, then the image feature information for each of the five depth ranges will be... The common primary environmental information E1 is registered in association with it. Also, image feature acquisition is performed. Section 164 defines a combination of feature points originating from the same object as an image feature. The information may be extracted and registered in association with the primary environmental information E1. The second-level environmental information acquisition unit may acquire the second-level environmental information E2 using feature quantities for each depth range. In this way, feature quantities are extracted according to the depth range. This makes it easier to identify the visual features that contribute to creating a particular environment. This increases the probability that noisy feature points will be removed, and the form used with the inference model This will increase the learning effect.

[0126] For example, consider a situation where a person is near TV T1. (First-order ring) The boundary information acquisition unit acquires feature quantities for each depth range and registers information for multiple locations. This location information mainly consists of image feature information, which is a feature point cloud of TV T1. This includes information that is associated with the presence of people in the environment. (To be continued) Then, keyframe 404, including TV T2, was captured, and features were observed for each depth range. When the quantity is acquired, it contains image feature information consisting mainly of feature points related to "television". Assume that the following was extracted. Assume there are no people near TV T2. Second environment When the information acquisition unit acquires location information similar to that image feature information, TV T Based on the image feature information from step 1, secondary environmental information indicating "a person is present" is obtained. Therefore, the place where I had the experience of "there are people near the TV" was actually Even if it is in a different location, image feature information containing many feature points originating from television If you can obtain that, you can recall the experience of "having people there," and that there are people there. Even without it, they can choose learned behaviors such as staying longer.

[0127] In this embodiment, it was explained that one image is acquired periodically, and multiple images are acquired when an event occurs. However, this is not limited to this, Under normal circumstances, n images are acquired; when an event occurs, m images are acquired. Alternatively, we can obtain images where (m>n). The event is robot 100 Because this is a significant event that greatly influences behavioral characteristics, when the event occurs... Therefore, it is desirable to obtain more detailed image information in the surrounding area. By acquiring multiple images, the object at the time of the event It can detect with high accuracy. When an event occurs, robot 100 will Multiple images may be acquired by pointing the mounted camera in several directions. In some cases, the behavior of being surprised by an event and checking one's surroundings can also be expressed.

[0128] When an event occurs, it may be possible to acquire images at a higher resolution than usual. This is done while suppressing the image processing load during normal operation, when an event occurs. In this case, the external environment can be recognized in more detail. The image feature acquisition unit 164 is high By enlarging the image captured in the image, small objects can be seen in the enlarged image. It is acceptable to acknowledge this.

[0129] When an event occurs, features may be extracted for each depth and objects may be detected. Specifically, a depth camera may be used to capture the depth in multiple stages. It may be separated. From the robot 100, within a predetermined range, for example, 30 centimeters. The first range is defined as the area within 30 centimeters to 1 meter. The first range is defined as the second range, and anything over 1 meter is defined as the third range. Then, the image characteristics of the first range are defined. Even if the similarity between the characteristic information and the existing image feature information for the first range is determined, Good. In this case, even if they are not similar at a distance, they may be similar at close range. Since it determines similarity based on similar image feature information, event generation Even when acquiring high-resolution images in real time, image processing can be performed at high speed.

[0130] When an event occurs, the detection sensitivity of various sensors such as microphones may be increased. Alternatively, under normal circumstances, the detection sensitivity of various sensors such as odor sensors may be increased. The sensors in the unit will not be used, and these sensors will only be used when an event occurs. This may enhance the environmental information available when an event occurs. Humans, Under normal circumstances, they act unconsciously, but when they perceive danger or become interested in something... When we recognize an event, in an emergency, we try to accurately perceive the surrounding environment. Engage your awareness strongly. When an event occurs, in other words, in an emergency, By improving the detection capabilities of the system, we can better understand the external environment when an event occurs. It will become cheaper.

[0131] The image acquisition unit 162 continuously captures moving images and periodically takes images from the moving images. Still images may be extracted. The image feature acquisition unit 164 extracts from the moving image. Keyframe 404 may be extracted. The recognition unit 212, when an event occurs, In this case, objects are selected from video footage of a predetermined period before and after the recognition of the event. It may be possible to detect a ct.

[0132] The object detection process shown in Figure 14 may be executed at a different time than the update process shown in Figure 11. The object detection process may be executed at a high frequency. In this case, the emotion parameter changes each time an object is detected. This is one way to suppress excessive changes in emotional parameters, The object detection process may be performed at a lower frequency than the processing time.

[0133] In this embodiment, the causality between events and objects is defined as the degree of relevance, and the emotion parameter is described as being changed based on the object. In addition to visible objects, other environmental factors such as specific sounds and smells can also be detected. The information can be associated with an event. For example, the scent of perfume and female user X4. By linking the event of going out, when the scent of perfume is detected... This is associated with female user X4 going out, and the state management unit 244 sets the emotional value (loneliness). ) may be increased.

[0134] The object management unit 174 determines the strength of the association between events and objects. The degree of relevance expressed may decrease over time. According to the law, the association between events and objects weakens over time. The concept of "forgetfulness" can be expressed in robot 100. Object Management Unit 1 74 reduces the time decay rate of relevance when the relevance exceeds a predetermined threshold. You may allow this. With this control method, between events and objects When a strong association (degree of relevance) is established, such associations are quite... This will no longer be resolved. In other words, it will be an emotional expression similar to a "strong preconception" that robot 1 It can be represented in 00.

[0135] Some events may be designated as special events. For example, a special event could be defined as the occurrence of a sound exceeding a certain threshold, such as thunder or an explosion. The object management unit 174 inspected the objects that were seen at the same time as the special event. For each item, a predetermined degree of relevance is set, for example, a relevance of 90%. You can set a strong impression like "trauma" for Robot 100. In addition, when user X1 hits it hard enough to cause it to malfunction, Object Management Unit 174 handles the special event of being punched and user X1( You can also set the relevance of the object to 90%. With this control method, you can express "dislike" through parameters other than intimacy.

[0136] Event conditions can be set arbitrarily. These conditions are not limited to external events; they could also be triggered by an emotional value (loneliness) exceeding a predetermined value, or by the battery charge level exceeding a predetermined value. The following mechanical events, such as malfunctions, may be recognized as events. When the emotional value (loneliness) exceeds a predetermined value, the operation control unit 222 will As a destination for Bot 100, the probability of encountering a user with a friendship level above a certain value is high. While selecting a location as a candidate, you can also have the system wander around the room looking for this user. stomach.

[0137] The image feature acquisition unit 164 acquires a total of 10 keyframes 4 at point P6. Let's say we obtained 04. At this point, out of the 10 keyframes 404, A fixed number of images, for example, 7 or more keyframes 404, all share a common image characteristic. The feature may be registered as keyframe 404 (image feature information) of location P6. This control method also helps in recognizing the external environment, such as documents and clothing. Unstable features that are not easily used as landmarks, in other words, noise information in location recognition, are effectively utilized. It can be removed effectively.

[0138] Environmental information may be categorized. For example, environmental information related to temperature may be classified into two or more categories such as "warm" and "cold." Multiple environmental information categories are used for each location, such as "safe," "dangerous," "fun," and "lonely." It may be classified into categories. With this control method, keyframe 4 Because environmental information linked to 04 can be reduced, it reduces memory capacity and also reduces environmental This reduces the processing load associated with detecting boundary information.

[0139] The strength of the wireless signal received from server 200 determines the territory of robot 100. You may also express "ri". Robot 100 is within communication range with server 200. If you leave, you will no longer receive support from server 200. Unit 222 refers to object information 430 and determines if the received signal strength is B or higher. A set of locations may be set as the range of action. This prevents robot 100 from moving out of the communication range.

[0140] The robot 100 may define its territory based on the signal strength received from other devices, not just the server 200. For example, in the location where an external sensor 114 is installed. In that case, the operation control unit 222 will communicate wirelessly with any of the external sensors 114. The range within which the robot can perform its actions may be set as the range of action. Robot 100 It periodically communicates with server 200, and when the received signal strength falls below a predetermined value, He may restrict his own actions to prevent him from moving any further away from server 200. On the other hand, when a specific event occurs, such as welcoming a user, the operation control unit... 222 may set a mobile target location even if the signal strength is low.

[0141] The robot 100 is periodically charged using a charging device (not shown). During operation, the server 200 may suppress communication with the robot 100. The communication unit 142 of the robot 100 opens the charging device when charging begins. A start signal is sent to the server 200, and a charge complete signal is sent when charging is complete. Trust. When the server 200 receives the charging start signal, the robot 10 You may stop communicating with robot 100 until you receive a charging complete signal from 0. Alternatively, while robot 100 is charging, server 200 and robot The frequency of periodic communication with T100 may be reduced. According to this, by suppressing communication when robot 100 is not active and is charging, This reduces power consumption associated with communication between the robot 100 and the server 200.

[0142] This invention can be applied to applications other than robot 100. For example, navigation In the 404 keyframe system, the concept of location memory is also applicable. It is applicable. By periodically imaging the outside world, it can be used to "memorize" the scenery. Create a map. Link environmental information to it. Navigation system Based on environmental map information, the system guides users to walks they have never taken before. You could suggest a route, or a walking route that includes places where you've had enjoyable experiences. You could suggest a walking route that avoids places where you've had a dangerous experience. You could suggest S.

[0143] Environmental maps, or environmental map information, are what are commonly known as "maps" and It does not need to be expressed in that way. Environmental map information only needs to be information that links geographic information and environmental information, and does not need to be visually represented. Geography The information could be a 404 keyframe, or various geographical landmarks. Objects are also acceptable.

[0144] In this embodiment, keyframe 404 is associated with environmental information, and key Environmental information associated with frame 404 (notes regarding keyframe 404) Based on memory, the emotional parameters are changed (secondary reflection), and the emotional parameters Based on this, the behavioral characteristics of robot 100 were explained as being changed.

[0145] As a variation, an environmental memory may be formed by associating objects with environmental information. When the robot 100 detects an object, Based on environmental information (memory about the object) associated with the object You may change your behavioral characteristics.

[0146] Robot 100 changes its emotional parameters based on the type of object. The behavioral characteristics may be altered based on changes in emotional parameters.

[0147] Furthermore, the robot 100 can navigate a room (space) containing multiple objects. The robot stores its impressions and uses those impressions of the room (described later) to create a memory. By changing 100 emotional parameters, behavioral characteristics are determined based on these emotional parameters. You may change it.

[0148] In the following, we will discuss changes in emotional parameters based on memories of the room, changes in emotional parameters based on environmental information associated with objects, and A modified example will be given illustrating the changes in the behavioral characteristics of robot 100 that result from this.

[0149] The image feature acquisition unit 164 of robot 100 is an "object detection unit". It may function. As described above, the image feature acquisition unit 164 has multiple types of objects For an object, multiple objects whose shape and color characteristics are defined in advance. It has an object pattern, and based on the object pattern, the object The object may be recognized by image recognition. For example, the image feature acquisition unit 164 may recognize the object Based on the feature point distribution of ct(B01), object(B01) is "Tele It may be determined that it is a "television". The image feature acquisition unit 164 may also determine that another object (B04) is a "television" based on the feature point distribution of that object (B04). The image feature acquisition unit 164 or the recognition unit 212 can recognize that even though object (B01) and object (B04) are different objects, they belong to the same category of "television". The image feature acquisition unit 164 uses a neural network that takes the feature point distribution as input. The category of an object may be determined by the C-model.

[0150] Below, objects like object (B01) are treated as unique objects. When referring to it, use the term "inherent object," and specify the type of object, like a television. When referring to them, they are called "object categories." Inherent objects are objects Identified by the Object ID, the object category is determined by the Category ID. They are identified. Also, when no particular distinction is made, they are simply called "objects". Bu.

[0151] Robot 100 detects unique objects by methods other than image recognition. You may also specify the object category. For example, robot 100 is "Object detection" detects objects by communicating with unique objects. It may also include a "project detection unit". The home appliance can transmit a category ID. In this case, the object detection unit detects the category ID transmitted from the home appliance. By doing so, the object category of home appliances (unique objects) You may make a judgment. In the following modified example, the object detection unit of robot 100 detects an object It is described as detecting an effect and identifying an object category. do.

[0152] Figure 15 shows the data structure diagram of object environment information 440 in a modified example. be. The object environment information 440 is stored in the object information storage unit 172. Object environment information 440 includes object ID, category ID, It is associated with environmental information and room ID. As mentioned above, the object The Category ID identifies a unique object, and the Category ID indicates the object's category. For example, when both unique object (B01) and unique object (B04) are "televisions," the Category IDs of these two unique objects will be the same.

[0153] The object detection unit identifies the object ID and category ID when it detects a unique object. The environmental information acquisition unit 166 identifies the unique object When an event is detected, environmental information is obtained. Object management unit 174 Based on the environmental information at the time of unique object detection, object environment information 4 Update 40. The update of object environment information 440 is explained in relation to Figure 8. It is basically the same as the method I explained.

[0154] The recognition unit 212 detects walls and doors, thereby recognizing the room as a closed space. Recognize. Each room is identified by its room ID. Object detection unit Each unique object detected is managed in association with a room ID. Room ID=R01 indicates the "living room," and Room ID=R02 indicates the "study." This shall be shown. According to Figure 15, the unique object (B01: television) is It is installed in the room (R01: living room), and the unique object (B04: television) is It is installed in room (R02: study). For example, in room (R01: living room) Near the television (B01) installed there, there is a high probability of 70% that one of the following will occur. The user was detected, and the television (B0) installed in the room (R02: study) Near 4), it is remembered that users have rarely been detected. ru. Below, the object category (C01) is television, object category Gori (C02) is a printer, and Object Category (C03) is a table. The explanation assumes that the object category (C04) is a sofa.

[0155] Figure 16 is a data structure diagram of room impression information 450 in a modified example. In this modified example, the data storage unit 206 of the server 200 is further configured as follows: Stores room impression information 450. Room impression information 450 is room (room ID ), associates category ID and impression memory. "Impression memory" is related to the room. This information represents the impression of the robot 100, and is shown numerically (hereinafter referred to as "impression memory value"). The "impression" is the event experienced in the room, the robot's This is an index that increases or decreases depending on external and internal conditions, representing the robot's emotions (100). Parameters, the amount of movement of each drive unit, and speaker volume can be managed as numerical values. It is determined based on the available parameters.

[0156] For example, in the living room, Robot 100 has many opportunities to interact with its owner. The room has ample space to move around freely. The space is smaller than the living room, and there are fewer opportunities for interaction. If we define an index in terms of the amount of movement of the movable parts in a bot, then in a living room... The index of activity volume is greater than the index of activity volume in the bedroom.

[0157] If we were to set an index from the perspective of output from the speakers, in the living room the owner and Since there are many opportunities for physical contact and speaking, the speaker index is, The speaker index will be higher than that of a bedroom.

[0158] If we were to create an index based on the pleasant actions that Robot 100 receives from its owner, In the living room, there are more opportunities to be held, and in the bedroom, there are even more opportunities to be held than in the living room. Because the amount of pleasure decreases, the index of pleasure activities is higher in the living room than in the bedroom. Define such an index and determine the actual values ​​of the parameters that form the basis of the defined index. By aggregating data for each room under predetermined conditions over a set period, trends can be observed for each room. This reveals that these indices are managed as impressionistic memories.

[0159] For example, Robot 100 is active in the living room and also has opportunities to speak. When you have the experience of receiving many pleasant acts, your performance in the living room The values ​​"Action Volume Index," "Speaker Index," and "Pleasure Index" are high. Based on these actual results, it is acceptable to move around actively in a room like a living room. The impression is formed that "it's okay to make noise, and you often receive pleasurable acts." It can be done.

[0160] These indices also change depending on the time of day. For example, during the daytime on weekdays... They rarely receive such treatment, and may receive more pleasurable treatment on weekday evenings. In this case, the time slot would be weekday afternoons (for example, 11:00 to 13:00). Various indices are associated with this. As a result, impression memory values ​​also change between weekday afternoons and weekday evenings. Figure 16 illustrates the index of pleasant experiences among these impression memories.

[0161] When an event that occurs in a room (R01: living room) is a pleasant action, state management Section 244 increases the impression memory value of the room (R01: living room). Meanwhile, event When the behavior is unpleasant, the state management unit 244 monitors the impression of the room (R01: living room). The memory value is reduced. This control method allows the room (R01: living room) to The impression of robot 100 on the other side is expressed numerically. According to Figure 16, the room (R0 1: Living room) The impression memory value is "+10", so robot 100 is room ( I have a positive impression of the living room (R01). One example of a positive impression is when Robot 100 is in the living room, and the owner... Based on the experience of often receiving pleasant acts such as being held. To be nurtured.

[0162] When the owner scolds him in his room (R02: study) for (unpleasant behavior), the robot The impression memory value for room 100 (R02: study) decreases. Science Department 244, for example, if robot 100 is in a room (R02: study) If any owner is not detected for a certain period of time, for example, 10 minutes or more If it is not present, it will reduce the impression memory value of the room (R02: study). This is also acceptable. In this way, not only events that occur in the room, but also events that occur in the room The impression memory value was determined based on whether an event occurred or the frequency of the event. You may change it.

[0163] In this modified example, robot 100 assumes that the room (R01) is a "living room". It does not recognize it. The recognition unit 212 of robot 100 recognizes the room's ty The property is identified by the multiple object categories contained in the room, and each room Assign a room ID to it.

[0164] According to Figure 16, in room (R01: living room), television (C01), A table (C03) and a sofa (C04) have been detected. Robot 100 This refers to a space containing these object categories, such as a room (R01: Identify the living room. A high impression memory value for the room (R01: living room) means, in other words, that robot 100 has a positive impression of a room that meets the conditions of including a television (C01), a table (C03), and a sofa (C04). It means that it possesses. Robot 100 obtains environmental information in the room. In particular, the impression and memory value of the room changes depending on the events experienced within the room. The robot 100 selects the room that is the target of the impression memory value and includes multiple rooms within the room. Identify by object category.

[0165] Assuming the room impression information 450 shown in Figure 16, the robot 100 The room includes a TV (C01), a table (C03), and a sofa (C04). In contrast, it has a positive impression. The status control unit 244 said that when robot 100 first entered the room... When you enter (hereinafter, the room where robot 100 is located will be referred to as the "room with robots") (referred to as "room"), based on multiple types of object categories contained within the occupied room. Next, identify the room type. For example, a room that is known to be occupied (R0 1) When it is similar to the above, the status management unit 244 records the impression of the room (R01) The emotional parameters of robot 100 are changed based on the estimated value. The room in question is It can be the same as the room (R01), or it could happen to be the same as the TV (C01), the TV Another item that is similar in that it includes a table (C03) and a sofa (C04). It can be a room. With this control method, televisions, tables, etc. Robot 100, which has a positive impression of a room (R01: living room), Even when visiting an unfamiliar home for the first time, if you have a room similar to the room (R01: living room) You can feel "happy" when you find it. Therefore, against a known room Based on the impression it gives, the emotions of 100 robots when they enter an unfamiliar room are changed. It can be transformed. When in a room similar to a known room, the state The state management unit 244 may lower the value of the emotion parameter indicating loneliness. When you are in a room that does not resemble any of the known rooms, the state control The 244th component of the science department increases the emotional parameter indicating loneliness, and the emotional parameter indicating curiosity. You may also increase the lameter.

[0166] For example, the state management unit 244 of the robot 100 has a high impression memory value and is known When you enter a room similar to, or the same room as, you feel a sense of loneliness. You may lower the emotional parameter. The higher the impression memory value, the more "loneliness" is indicated. You may significantly lower the emotional parameter.

[0167] Furthermore, for example, an index of activity volume is defined as an impression memory, and the index of activity volume in room (R01: living room) is the same as the index of activity volume in room (R02: study). If it is greater, that is, if the activity index in the study is smaller than the activity index in the living room In that case, the amount of movement the robot 100 performs when it is in the room corresponding to the study will be It can be suppressed more effectively than when you are in the room in between.

[0168] In summary, Robot 100 is designed to detect the shadows of various external environments in each room. While receiving the feedback, it makes complex changes to emotional parameters and action levels. The behavioral records are recorded as various indices. Furthermore, based on these behavioral records... Subsequently, impressions such as "being able to move actively" and "being easily receptive to pleasurable activities" are formed. And when the room in which the robot 100 is located resembles a known room... It changes its behavioral characteristics based on the impression and memory of a known room. The behavioral record based on experience influences actual behavioral characteristics, thus creating "experience-based behavior." "Dynamic characteristics" are realized.

[0169] As mentioned above, a room is a combination of multiple object categories. It is more easily identified. The status management unit 244 identifies all unique individuals detected in the occupied room. It is not necessary to identify rooms based on objects.

[0170] The object management unit 174, based on environmental information, processes unique objects. You may also set impression memory values. Specifically, object management unit 174 The environmental information when a unique object is detected is the event condition (as described above). When this condition is met, the impression memory value of the unique object changes in response to the event. For example, when a pleasurable act is detected near a television (B01), the object Even if the effect management unit 174 increases the impression memory value for television (B01) Good. In other words, when a specified event occurs, the object management unit 174 identifies a unique object within a predetermined range from the robot 100. And, it changes the impression memory value for that unique object. Whether or not an object is within a predetermined range may be determined by measuring the distance between each object, or by basing the determination on the size of the objects included in the captured image. It may be assumed that this is the case.

[0171] Furthermore, the object management unit 174 is located in the frontal direction of the robot 100. In addition, it identifies unique objects within a predetermined range, and identifies those unique objects The impression memory value for an object may be changed. Also, object management Section 174 is not an intrinsic object (B01), but an object category. An impression memory value may be set for (C01). For example, around the television Since there are people present, events such as being held are more frequent. By setting an impression memory value for the category "television," initially Even in that location, based on memory, the action of moving closer to the television You can choose.

[0172] The object management unit 174 detects multiple unique objects in occupied rooms. Among the units, occupied rooms may be identified based on certain unique objects. For example, the object management unit 174 has particularly high impression memory values ​​and special For example, the top 3 unique objects and the bottom 3 6 objects. You may identify the unique objects of these six unique objects. Identify each object category, and identify the identified object • Occupied rooms may be identified as a combination of categories.

[0173] The status management unit 244 is responsible for the object category group detected in the occupied room and This corresponds to each of the multiple rooms registered in Room Impression Information 450. The group of objects to which the category is assigned is compared. The state management unit 244 is Similarity may be determined based on the number of matching object categories. For example, , of the three object categories included in the room (R01: living room) When a room is also detected as occupied, the status management unit 244 determines whether the room is occupied or not (R The similarity of 01: Living Room can be determined to be 66% (=2÷3). This indicates a high probability that the room in question is the same as a known room (R01: living room). It tastes good. The status management unit 244 detects object categories in occupied rooms. By comparing the room impression information (450) with the room impression information, the room with the highest similarity and a similarity of 70% or more may be determined to be the same type of room as the occupied room.

[0174] Figure 17 shows the processing steps when an object is detected in a modified example. This is a flowchart. The process shown in Figure 17 is performed periodically in the robot system 300. The environmental information acquisition unit 166 first acquires environmental information of the occupied room (S6 0). The state management unit 244 changes the emotion parameters based on environmental information. (S62). This is related to the "primary reaction based on environmental information" explained in relation to Figure 11. It is the same as "Image". The object detection unit detects unique objects. S64). At this time, the object detection unit detects the object of the unique object. Identify the ct category.

[0175] Compare the detected multiple object categories with 450 pieces of room impression information. By doing so, the status management unit 244 identifies the type of room occupied (S66). As described above, the status management unit 244 detects objects from occupied rooms. The categories and the room-specific items registered in Room Impression Information 450 Rooms whose similarity between a group of objects and categories exceeds a predetermined value are defined as "similar to known rooms". These rooms are identified as "known rooms" (hereinafter simply referred to as "known rooms").

[0176] When the occupied room is a "known room", the status management unit 244 will enter the known room The emotional parameters are changed based on pre-associated impression memory values. (S68). For example, for a room (R01: living room), the impression memory value is "+1 When "0" is associated, and the occupied room is similar to the living room The state management unit 244 controls the emotional parameter value indicating loneliness and the emotional parameter value indicating curiosity. The parameter values ​​may be reduced. In this way, the known rooms, more precisely For example, pre-associating with the object categories included in a known room Based on the impression memory values ​​that have been recorded, the robot's feelings (emotional parameters) The (data) can be changed. This is related to the "ring" explained in Figure 11. This is similar to "secondary reflection based on border information."

[0177] On the other hand, when it is an unknown room (N in S66), secondary reflection is not performed. Subsequently, based on the environmental information acquired in S62, the status management unit 244 Update impression information 450 (S70).

[0178] Robot 100 senses based on the impression memory value of a known room that is similar to the room it is in. Change the emotional parameters and change behavioral characteristics based on the changes in the emotional parameters. Robot 100 also makes it perform emotional parameters based on impression memory values. Behavioral characteristics may be changed without using a meter. For example, impression memory. In rooms with low values, the motion control unit 150 of the robot 100 100 movement speed, amount of movement (distance that can be traveled in a unit of time), momentum ( The amount of movement or range of movement of various actuators may be suppressed. In rooms with low memory values, the motion control unit 150 generates robot 100. The upper limit of the volume to be allowed may be suppressed. With such a control method, the impression In rooms with low memory values, robot 100 becomes docile, and impression memory values Robot 100 behaves actively in high-rise rooms. (Impression memory of the room) Based on this, the feelings of 100 robots can be expressed through actions. You can set the initial impression memory value.

[0179] When Robot 100 enters a room for the first time, it tries to determine what kind of room it is. We will make a judgment by referring to the room impression information 450. And, based on past experience... Next, determine the behavioral patterns such as movement speed. Then, object environment information 44 By referring to 0, based on past experience, objects in the room Explore and set it as the target location.

[0180] Also, when in a new home or feeling lonely, it looks for a "fun-looking room (a room that makes a good impression)." Specifically, the operation control unit 222 records the impression Set the room with the highest estimated value as the target destination.

[0181] The motion control unit 222 controls rooms containing object categories with high impression memory values. Search for... For example, robot 100 includes a TV, a table, and a sofa. Let's assume that the robot has a positive impression of the house. In this case, robot 100 will Assuming that a room including a TV, table, and sofa is a "fun room," I'm looking for a room that meets these conditions.

[0182] The motion control unit 222 of the robot 100 refers to the room impression information 450, The room ID of the room with the highest impression memory value is identified. Next, the operation control unit 222 Identify multiple object categories contained within the identified room. The control unit 222 controls multiple types of objects included in this "fun room". Among the object categories, those containing a predetermined number or more of object categories. Select the room as your destination.

[0183] Robot 100 selects a room to be the target location based on its experience. This is possible. For example, owner A can bring robot 100A to user B's house. Let's assume we proceed. User B's house is unknown to robot 100A. 450 room impression information formed at owner A's home for 100A. Based on this, the object categories included in rooms with high impression memory value are selected. Determine. Then, the robot 100 will determine the identified object category group. Search for rooms containing the same elements in User B's house. For example, rooms similar to room (R01: living room) When robot 100 finds room RB in user B's house, it moves to room R Enter B, and based on the impression memory value of the room (R01: living room), the emotional parameters This changes behavioral characteristics. With this control method, the experience at home Based on this, we make Robot 100 behave the same way it does at home, even when visiting a house for the first time. It is possible.

[0184] The motion control unit 222 of the robot 100 further determines which objects should approach within the room. Select a unique object. The operation control unit 222 selects the unique object. Based on the environmental information provided, approach in the same manner as in this embodiment. Select the specific object that should be selected. For example, robot 100 shows loneliness. When the emotional parameter value is high, the user detection rate is high for unique data. You may approach the object. The state management unit 244 is a ring of unique objects. Emotional parameters may be changed based on environmental information. For example, hugging When you are near a sofa, the event that is likely to occur is a robot T100 lowers the emotional parameter that indicates loneliness even without actually being held. It is also possible to change the emotional parameters and behavioral characteristics of the robot 100 based on environmental information associated with keyframe 404, or based on environmental information associated with an object (a unique object or object category). An impression memory value may be set in advance for the object category. When 100 detects a unique object, it may change its behavioral characteristics based on the impression memory value associated with that object category.

[0185] A room that makes a good impression includes a television, a table, and a sofa. After entering the room, the motion control unit 222 moves the robot 100 closer to the television. According to this control method, the robot 100 will perform actions based on impression memory. Then, find a "fun-looking room," and once inside, look for an object that makes a good impression. This will bring you closer to the TV. As a result, you may like the living room, or 100 robot preferences regarding spaces and objects, such as liking television. They can express this through actions.

[0186] Similarly, robot 100 may avoid rooms with low impression memory values. For example, when Robot 100 enters the bathroom, the owner chases it away. When any unpleasant behavior is detected, robot 100 will remove an object in the bathroom. For example, reduce the impression memory value for combinations of towels, washbasins, washing machines, etc. This can lead to disliking the bathroom or the washing machine, etc. It can express behavioral characteristics.

[0187] Impression-based memory is not limited to pleasant / unpleasant. For example, noisy / quiet, hot / cold. You may also set impression memory values ​​based on other evaluation axes such as, , and . Room (R 01: When the impression memory of "lively" is associated with the living room, the smell When the emotional parameter indicating "loneliness" increases, robot 100 You may set Room (R01: Living Room) as your target location. 1: Living room) is a place where people tend to gather, so this control method will prevent loneliness. When this happens, they exhibit behavioral characteristics such as approaching the room (R01: living room) and searching for someone. This can be represented. Unique objects do not have to be fixed objects; they can be portable, like a towel, or autonomously mobile, like a pet or other robot 100.

[0188] Figure 18 is a data structure diagram of behavioral experience information 460 in another modified example. . The object information storage unit 172 may also store behavior experience information 460. The behavior experience information 460 associates a category ID with a motion ID and an impression memory. For example, suppose the robot 100 selects a motion with motion ID = M01 within a predetermined range from an object (B01: television) and is praised by the owner. Since being praised is a pleasant behavior, the object management unit 174 increases the impression memory value for the combination of object category (C01: television) and motion (M01). This updates the behavioral experience information 460. Meanwhile, object (B01:Te When I selected motion (M03) near Levi, I got scolded by the owner. Let's assume that being scolded is an unpleasant act, so the object management unit 174 is A combination of project category (C01: television) and motion (M03). Reduces the perceived memory value of the combination.

[0189] The motion control unit 222 refers to the behavioral experience information 460 and, when near the television, Increases the selection probability of motion (M01) and increases the selection probability of motion (M03) The rate is reduced. With this control method, the user praises and scolds This action can change the behavioral characteristics of robot 100. For example, by scolding Robot 100, who is sitting near the cupboard, This can stop Robot 100 from sitting near the cupboard. As a result, users can "train" robot 100. ru.

[0190] The state of the object may also be taken into consideration when changing the impression memory. For example, when the vacuum cleaner is off, nothing happens when you approach it, but when the vacuum cleaner is off... Let's say when he approached the vacuum cleaner during the time, the owner chased him away. Since being chased away by the owner is an unpleasant experience, robot 100 learns the behavioral norm that "do not approach the vacuum cleaner when it is on." The on / off status of the vacuum cleaner can be determined by the presence or absence of noise it makes. In addition to the state of the object, the impression memory may also be altered by considering the time of day. For example, if robot 100 enters the kitchen while dinner is being prepared, the owner might chase robot 100 away (an unpleasant experience). As a result, robot 100 can learn the behavioral norm that "do not approach the kitchen where the refrigerator and cupboards are located in the evening (do not disturb the owner who is in the kitchen)."

[0191] Even if you associate impression memories with combinations of rooms (room IDs) and motions Good. In this case, for example, it is permissible to make loud noises in the living room, but not in the study. The robot 100 was taught the behavioral rule that it should not make loud noises. Similarly, you can move around quickly in the living room, but not in the study. We sometimes learn that we need to limit our activity levels.

[0192] When robot 100 recognizes an unpleasant act, it identifies the owner who performed the unpleasant act. It reduces intimacy with the owner. However, if the intimacy level is above a certain value, When the intimacy management unit 220 recognizes unpleasant behavior, it will not decrease the intimacy level. Alternatively, the decrease in intimacy level may be made smaller compared to normal conditions. According to such a control method, a strong relationship of trust has been established with the robot 100. Naa can scold Robot 100 with peace of mind.

[0193] As mentioned above, objects may be associated with environmental information, or spaces (room IDs) may be associated with environmental information. The state management unit 244 manages the environmental information In particular, impression memory may be altered based on events. Alternatively, the memory of impressions of space may be set initially. For example, If you set a high impression memory value for the damp living room, Robot 100 will be born with This can express the behavioral characteristic of preferring the living room. Also, the owner is smart The impression memory values ​​for the living room are sent to the server 200 via a mobile device such as a phone. The object management unit 174 then determines the specified impression memory value and stores it in the room impression information. You can also set it to 450.

[0194] Environmental information may be associated with combinations of object categories. Even if a room is recognized by considering its layout and the arrangement of objects, Good. For example, "The TV is by the window, and the sofa is placed a little distance away from the window." The room may be identified as "the room that will be used".

[0195] Multiple robots 100 collect environmental map information 420 and object information. Share the information shown in 430, room impression information 450, and behavioral experience information 460. This is also good. According to this control method, when robot 100A prefers the living room Robot 100B also came to prefer the living room, and the behavior of multiple robots 100 The characteristics can be linked together.

[0196] Robot 100A, as a fun room, only selects areas with high memorable value. You can also share this with Bot100B. For example, among multiple rooms, impression memory Rooms with a value higher than average, or rooms with an impression memory value that ranks among the top 3, are targeted. As an example, room impression information 450 may be shared. Alternatively, impression memory value This target rooms with a score below average, or rooms with an impression memory score in the bottom three. Room impression information 450 may be shared as such.

[0197] Robot 100A and Robot 100B target all rooms There is no need to share impression information 450; only some rooms are included. The impression information 450 may be shared. According to such a control method, for example Robot 100A transmits information to Robot 100B for some rooms. To prevent this from happening, I'll create a secret room that only Robot 100A likes. It is possible.

[0198] The room where robot 100A was first held by owner A, owner A Rooms with special conditions (memories), such as rooms where someone will hold the baby for a long time. Therefore, it is acceptable for robot 100A not to share information with robot 100B.

[0199] Impression memories may be visualized. Server 200 is the arrangement of objects and An impression map generator that produces images (impression maps) that show the impression memory of a subject. It may also have a component. The communication unit 204 of the server 200 may transmit an impression map to the owner's mobile terminal. For example, the area around objects with high impression memory values ​​may be displayed in blue, and the area around objects with low impression memory values ​​may be displayed in red. By doing so, the owner can see what kind of object or what You can visually see what kind of room they like.

[0200] The impression map generation unit may display changes in impression memory values ​​for an object or room on the impression map. For example, an object that was initially disliked The gradual development of liking To can be shown through a video display of an impression map.

Claims

1. Memory unit and, Processor and The system comprises a drive mechanism that performs the motion selected by the processor, The aforementioned processor, Recognize the impression memory of an object, A robot that determines motion for an object based on the impression memory it has.

2. In the robot according to claim 1, The aforementioned processor, When a first object is detected, a first impression memory of the first object is stored in the storage unit. When a second object is detected, a second impression memory, which is the impression memory of the second object, is recognized based on the similarity between the second object and the first object and the first impression memory. A robot configured to determine motion toward the second object based on the second impression memory.

3. In the robot according to claim 1, The aforementioned processor, A map representing the space is displayed, and objects present in that space are displayed on the map. A robot that determines the display format for each object based on its perceived image and memory of that object.

Images