Robot device and behavior control method for robot device

• US 20030060930A1
• Filed: 09/24/2002
• Published: 03/27/2003
• Est. Priority Date: 10/13/2000
• Status: Active Grant

• Current Assignee: Sony Corporation (Sony Group Corp.)
• Original Assignee: Sony Corporation (Sony Group Corp.)
• Inventors: Minamino, Katsuki, Yokono, Jun, Fujita, Masahiro, Takagi, Tsuyoshi, Horinaka, Rika, Shimomura, Hideki, Costa, Gabriel
• Granted Patent: US 6,718,232 B2
• Time in Patent Office: Days
• US Class Current: 700/245
• CPC Class Codes: B25J 9/1602 characterised by the control system, structure, architecture
• CPC Class Codes: G05B 2219/33053 Modular hardware, software, easy modification, expansion, generic, oop
• CPC Class Codes: G05B 2219/33056 Reinforcement learning, agent acts, receives reward, emotion, action selective
• CPC Class Codes: G05B 2219/35453 Voice announcement, oral, speech input
• CPC Class Codes: G05B 2219/39254 Behaviour controller, robot have feelings, learns behaviour
• CPC Class Codes: G06N 3/008 based on physical entities controlled by simulated intelligence so as to replicate intelligent life forms, e.g. based on robots replicating pets or humans in their appearance or behaviour

Abstract

A robot apparatus causes the emotion in a feeling part (130) to be changed based on the information acquired by a perception part (120) to manifest the behavior of information acquisition as autonomous behavior. The robot apparatus includes a behavior control part for causing the robot apparatus to manifest a language acquisition behavior and a meaning acquisition part. The robot apparatus also includes a control part for performing the behavior control of pointing its object of learning. The robot apparatus causes changes in internal states, which are ascribable to the object, to be stored in a memory part in association with the object.

Claims

1. A robot apparatus performing autonomous behavior, comprising:

behavior control means for causing the robot apparatus to perform information acquisition behavior as being among the autonomous behavior.

2. The robot apparatus according to claim 1 wherein the information acquisition behavior is the language acquisition behavior of acquiring the language as the information.

3. The robot apparatus according to claim 1 performing autonomous behavior based at least on internal state;

said behavior control means causes the robot apparatus to perform the information acquisition behavior based on the constancy of the internal state.

4. The robot apparatus according to claim 3 further comprising:

an information acquisition model having, as parameter the desire for information acquisition as a change factor of said internal state;

said behavior control means manifesting said information acquisition behavior when the parameter of said desire for information acquisition is of a preset threshold value.

5. The robot apparatus according to claim 4 wherein the parameter of said desire for information acquisition is changed with lapse of time.

6. The robot apparatus according to claim 4 wherein the parameter of said desire for information acquisition represents the hungry feeling for learning.

7. The robot apparatus according to claim 4 wherein the parameter of said desire for information acquisition is the amount of the acquired information.

8. The robot apparatus according to claim 4 wherein the parameter of said desire for information acquisition is the learned amount of the novel information.

9. The robot apparatus according to claim 8 wherein the amount of learning is the amount of learning decreasing with lapse of time;

and

wherein if said amount of learning is lower than a preset threshold value, said behavior control means manifests said information acquisition behavior.

10. A method for controlling the behavior of a robot apparatus comprising:

causing the robot apparatus to perform the information acquisition behavior as being among the autonomous behavior.

11. A robot apparatus comprising:

meaning acquisition means for acquiring the meaning of an object.

12. The robot apparatus according to claim 11 wherein said meaning acquisition means acquires changes in said internal state, which occur on said robot apparatus taking an action on said object, as the meaning of said object.

13. The robot apparatus according to claim 12 further comprising:

internal state monitor means for detecting said internal state;

said meaning acquisition means acquiring changes in said internal state, as detected by said internal state monitor means, on said robot apparatus taking an action on said object, as the meaning of said object.

14. The robot apparatus according to claim 11 further comprising:

associative storage means for associatively storing the meaning of the acquired object in association with said object.

15. A method for controlling the behavior of a robot apparatus comprising:

acquiring the changes in said internal state, produced when the robot apparatus acting based on its internal state has taken an action on an object, as the meaning of said object.

16. A robot apparatus comprising:

speech inputting means;

a plurality of models characterizing a plurality of word sequences, said models having been classified based on characteristic values of said word sequences prevailing when the word sequences have been uttered;

speech input evaluating means for evaluating speech inputs, made by said speech inputting means, in accordance with said word sequence characterizing models; and

word sequence specifying means for specifying the word sequences of said speech input based on evaluation values of said speech input evaluating means.

17. The robot apparatus according to claim 16 wherein said speech input evaluating means includes a characteristic value detecting part for detecting the characteristic values pertinent to the word sequence of the speech input made by said speech inputting means and an evaluating part for evaluating the characteristic values of the speech input, as detected by said characterizing value detecting part, based on said word sequence characterizing models.

18. The robot apparatus according to claim 17 further comprising model registration means for registering a characteristic value of a word sequence of a speech input as detected by said characterizing value detecting part as a new speech sequence characterizing model when the evaluation of the characteristic value of said speech input is low.

19. The robot apparatus according to claim 16 wherein said word sequence characterizing model has a characteristic value of a word sequence acquired by phoneme learning.

20. A behavior control method for a robot apparatus, said method comprising:

a speech inputting step;

a speech input evaluating step of evaluating speech inputs, made at said speech inputting step, in accordance with a plurality of models characterizing a plurality of word sequences, said models having been classified based on characteristic values of said word sequences prevailing when the word sequences have been uttered; and

a word sequence specifying step of specifying the word sequences of said speech input based on evaluation values obtained at said speech input evaluating step.

21. A robot apparatus acting autonomously, comprising:

control means for behavior control of pointing an object of learning by the robot apparatus.

22. The robot apparatus according to claim 21 comprising:

control means for causing the robot apparatus to act with an object specified based on external stimuli as a preferential object of learning.

23. The robot apparatus according to claim 21 further comprising:

pointing the object of learning thereof as a behavior indicating the curiosity.

24. The robot apparatus according to claim 21 formed to an appearance which is like an animal provided at least with a limb;

said control means performing behavior control of pointing the object being learned with said limb.

25. A behavior control method for a robot apparatus, acting autonomously, said method comprising:

controlling the behavior of the robot apparatus so that the robot apparatus will point the object for learning thereof.

26. A robot apparatus, acting autonomously, comprising:

a sensor for detecting an object;

a perception evaluating part for evaluating input signals form said sensor;

an internal state management part fed with results of evaluation of said perception evaluating part for managing fake internal states changed based on said results; and

storage means for storing the relationship between said object and changes in said internal state which are based on said object;

the changes in said internal state which are based on said object as detected and said object being associatively stored in said storage means.

27. The robot apparatus according to claim 26 further comprising:

an emotion part for generating the fake emotion based on changes in said internal state;

said object and the emotion-related information for the object being stored in said storage means.

28. A robot apparatus according to claim 26 further comprising:

a behavior generating part;

said internal state management part supervising said internal state so as to maintain the constancy of said internal state, said internal state management part sending a first signal to said behavior generating part when the constancy is disturbed;

said behavior generating part generating the behavior to maintain said constancy based on said first signal.

29. The robot apparatus 1 according to claim 28 wherein changes in said internal state and the behavior for maintaining said constancy are associatively stored in said storage means.

30. The robot apparatus 1 according to claim 28 wherein said internal state management part has information acquisition desire variables and transmits a second signal to said behavior generating part based on the value of said information acquisition desire variables;

and

wherein said behavior generating part generating an information acquisition behavior based on said second signal.

31. A behavior control method for a robot apparatus, acting autonomously, comprising:

a perception evaluating step of evaluating input signals from a sensor which detects an object;

an internal state management part for managing fake internal states changed based on results of evaluation at said perception evaluating step; and

a storage step of storing the relationship between said object and changes in said internal state which are based on said object;

the changes in said internal state which are based on said object as detected and said object being associatively stored in said storage means.