JP2006150562A - Robot, module selection device and control method of robot - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To select an optimum operation command in accordance with a residual quantity of a battery. <P>SOLUTION: A robot is equipped with a whole management part 2, a database part 3, a task planning part 4, a voice recognition part 5, a voice synthesis part 6, an image recognition part 7, an image generation part 8, and an external interface (I/F) part 9. Since the robot detects the current residual quantity of the battery, refers the power consumption data by each command, determines whether or not there is the residual quantity of the battery capable of executing each command, shows the determination result to a user, and changes the execution command, it securely prevents the problem that the robot stops due to the running out of the battery during operation to cause unintended consequences. <P>COPYRIGHT: (C)2006,JPO&NCIPI

Description

  The present invention relates to a robot that can execute various operations and a module selection device that selects at least one of a plurality of functional modules.

  In a mobile robot driven by a battery having some kind of moving mechanism, work interruption due to running out of battery during work and operation stoppage due to running out of battery become problems. Conventionally, with regard to this problem, the battery voltage and operating time are constantly monitored, and if the battery voltage drops below the voltage set during work or exceeds a predetermined time after starting battery operation, the work is interrupted. Thus, a countermeasure is taken such as moving to a charging place or a person who performs the charging operation.

Further, a method of shutting down the drive system in accordance with the remaining battery level (see Patent Document 1) has been proposed, but in this case, a task to be achieved is interrupted halfway. In addition, a method for correcting an action sequence when the remaining battery level is reduced from the amount of energy consumed for each action sequence has been proposed (see Patent Document 2). Therefore, there is a possibility that an unexpected problem as described in the present embodiment is caused at the end of the middle of the task.
JP 2001-328091 A JP 2002-144261 A

  The above countermeasures are used when the work is integrated with movement and work, such as cleaning the floor, mowing the lawn, or patrol around the work. Is valid. However, if the movement and work are discontinuous, such as looking at the designated place or bringing an object from the designated place, depending on the timing of the interruption, the movement or work up to that point may not be complete. It can be meaningless. Further, depending on the content of the work, such as taking beer from the refrigerator and interrupting it with the refrigerator door open, it may cause a serious problem for the environment.

  The present invention provides a robot, a module selection device, and a robot control method capable of selecting an optimal work instruction according to the remaining battery level.

  According to one aspect of the present invention, a remaining amount detection unit that detects a remaining battery level, a power consumption information storage unit that stores information on power consumption required to execute a plurality of work instructions for each work command, and Based on the remaining battery level detected by the remaining level detection means and the information stored in the power consumption information storage means, it is possible to determine whether there is a remaining battery level capable of executing a given work command. A determination means; a command presentation means for presenting a work command executable with the remaining battery level detected by the remaining amount detection means when the remaining battery level is determined by the executable determination means; And an instruction execution means for presenting a work instruction that can be executed in step (b).

  Also, according to one aspect of the present invention, attribute information including power consumption, identification information, and connection information with the robot body are associated with each of a plurality of functional modules that perform different operations. And an individual block information storage means for storing and a module selection means for selecting at least one functional block for constituting the robot based on the information stored in the individual block information storage means.

  According to the present invention, it is possible to select an optimal work command according to the remaining battery level.

  Hereinafter, an embodiment of the present invention will be described with reference to the drawings.

  FIG. 1 is a block diagram showing an internal configuration of a robot according to an embodiment of the present invention. The robot of FIG. 1 includes an upper system communication bus L1 and a lower system communication bus L2, and a trajectory generation / target value generation unit 1 is connected between the buses.

  The host system communication bus L1 includes an overall management unit 2, a database unit 3, a task planning unit 4, a speech recognition unit 5, a speech synthesis unit 6, an image recognition unit 7, an image generation unit 8, and an external An interface (I / F) unit 9 is connected. A speech preprocessing unit 10 is connected to the speech recognition unit 5, and an image preprocessing unit 11 is connected to the image recognition unit 7.

  A sensor signal processing unit 12 and a plurality of motor control units 13 are connected to the lower system communication bus L2. The plurality of motor control units 13 control the corresponding motors 14 respectively.

  The overall management unit 2 manages (controls) the entire robot shown in FIG. The task planning unit 4 controls the type and order of instructions executed by the robot. The voice recognition unit 5 and the voice preprocessing unit 10 perform recognition processing of a voice signal input from a microphone. The image recognition unit 7 and the image preprocessing unit 11 perform a recognition process on an image signal input from the camera. The voice synthesis unit 6 and the image generation unit 8 generate voice from a speaker that transmits information to the outside and an image of a display device. The external I / F 9 performs data communication with a communication device outside the robot, such as wireless communication such as a wireless LAN or a mobile phone, or a high-speed wired LAN. The trajectory generation / target value generation unit 1 is an arm tip trajectory of a robot mechanism unit (not shown in FIG. 1) for executing a command level task output from the task planning unit 4 or the overall management unit 2, and a target of the moving unit. Generate values, command values for each motor, etc. The overall management unit 2 has a function of detecting the remaining capacity of the battery used by the robot as a drive power source.

  The database unit 3 stores information on power consumption required for executing a plurality of work commands for the robot for each work command.

  Each unit connected to the higher-level communication bus L1 only needs to operate with a sampling time on the order of several ms at the shortest, so that the demand for real-time performance is low. For this reason, use of GbE (Gigabit Ethernet (registered trademark)), PCI-X, or the like is considered as the host communication bus L1.

  The sensor signal processing unit 12 performs signal processing necessary for an instantaneous avoidance or stop function when the robot operates, such as an ultrasonic sensor or a touch sensor. The motor control unit 13 performs servo control calculation for realizing the operation target value of the motor generated by the trajectory generation / target value generation unit 1 and outputs to the motor.

  Each unit connected to the lower system communication bus L2 requires high-speed sampling of 1 ms or less, and it is advantageous in terms of wiring that fewer signal objects are present. Therefore, the lower system communication bus L2 has a synchronous communication function. Use of serial communication such as USB2.0 or IEEE1394 is conceivable.

  FIG. 2 is a software configuration diagram of the robot according to the present embodiment. Each block shown in FIG. 2 corresponds to an individual functional module of software. The robot software according to the present embodiment includes a human interface unit 21, an image input / recognition unit 22, a voice input / recognition unit 23, an external sensor signal input / recognition unit 24 that detects external world information 20, and a command function. List providing unit 25, functional power consumption data providing unit 26, command power consumption data providing unit 27, battery remaining amount detecting unit 28, meta level task generating unit 29, meta level task analyzing unit 30, command level Work scheduling unit 31, command level task feasibility function review unit 32, command level task feasibility energy review unit 33, feasibility display unit 34, individual module unit 35, and usable module database 36 Have The individual module unit 35 includes a function profile 37, a module identifier 38, and a connection information signal 39.

  The human interface unit 21 acquires a state of a microphone, a camera, or a push button. For example, when a voice command “Please bring beer” is input to the microphone, the meta level task generation unit 29 recognizes it as a meta level task through the voice input / recognition unit 23. Further, based on the result of imaging the human arm and finger by the image input / recognition unit 22, a meta-level task “go to there” is generated. The same applies when a push button or the like is pressed, and some meta level task can be set in advance.

  It is conceivable that the meta level task is acquired not only from a command from the client of this robot but also from the external information by the robot itself. Meta-level tasks such as recognition of abnormal sounds, discovery of suspicious persons, and detection of obstacles are generated from input / recognition results from the image input / recognition unit 22, voice input / recognition unit 23, or external sensor input / recognition unit. It is also possible.

  The meta level task generated by the meta level task generation unit 29 is decomposed by the meta level task analysis unit 30 into the robot command level. As the decomposition method here, for example, a past task execution database or knowledge technology may be used. At this point, if it is impossible to disassemble the command level, it may be possible to notify the outside of the robot that the task cannot be executed.

  The robot of this embodiment is configured by combining a plurality of robot modules, and each robot module executes a task. The usable module database 36 in FIG. 2 is a database of functions possessed by each robot module. The command level task executability function examining unit 32 refers to the usable module database 36 to determine what kind of robot module is necessary and whether or not the task can be executed with the existing robot module. If the task cannot be executed, it may be possible to notify each part of the robot that the task cannot be executed at this point.

  Each robot module is connected to the power supply system and the signal system by a one-touch connector such as a USB standard. This connection is notified to the overall management module (not shown) by the connection information signal 39. Any of the modules shown in FIG. 2 may be substituted for this overall management module. The overall management module controls the operation inside the robot by software. The overall management module may be realized by combining a plurality of blocks shown in FIG. 2, but in this case, the plurality of modules have an overall management function. In this case, for example, priority such as first-come-first-served basis is determined, and any module may have the overall management function.

  Each robot module shown in FIG. 2 has a module identifier signal and profile data. The profile data includes power consumption information. These pieces of information are stored in the usable module database 36. The data stored in the usable module database 36 is continuously stored even when the corresponding robot module is removed from the robot.

  Thus, at the stage of considering which robot module should be used, if it is found that a certain robot module does not satisfy the functions required for task execution, the power consumption information etc. of the robot module that has already been connected can be obtained. It can be read from the available module database 36 and presented to the user.

  In addition, if function data of modules that have no connection experience is accumulated in the usable module database 36 via a network or a storage medium, a new function module to be prepared for task execution is presented to the user. It is also possible to do.

  The command level task feasibility review unit 32 uses the remaining battery level detected by the remaining battery level detection unit 28, the power consumption data classified by function, and the power consumption data classified by command to determine the analysis result of the meta level task analysis unit 30. Consider whether it is feasible with the current remaining energy.

When examining feasibility based on the remaining battery level, the feasibility study method changes depending on any of the following (a) to (e) for a given meta level task.
(A) A task where the battery may be used up during the work.
(B) A task that may be interrupted at any time to return to a chargeable state.
(C) A task that may interrupt the work at several stages and return to a chargeable state.
(D) A task that cannot be interrupted.
(E) A task whose work amount and work content cannot be predicted in advance.

  As the task (a), for example, entertainment tasks such as image output and audio output, and information presentation tasks can be considered. These tasks are not particularly accompanied by movement, and even if the function stops completely due to running out of battery, the user is not greatly affected. In this case, the robot command is executed regardless of the remaining battery level.

As the task (b), for example, tasks such as cleaning and mowing a garden can be considered. In these tasks, work can be interrupted during task execution, and the same task can be continued after recharging. In this case, it may be determined that the task can be executed when the following expression (1) is satisfied. In addition, Pr is the remaining battery level (Power Remaining), Ps is the power consumption required to move to the task start position (Power for moving to Start point), and Pm is a margin that is set in advance, taking into account various factors of the robot is there.
Pr> Ps + Pres + Pm (1)

The real time remaining battery level cPr during task execution is expressed by equation (2).
cPr> Pres + Pm (2)

  As the task (c), a room clean-up task, a dishwashing task, and the like can be considered. Since this type of task cannot enter the return operation while holding a tidy object or a dish, it is necessary to enter the return operation after completing at least the task related to the operation of gripping with the robot arm. However, as a whole task, there is no big influence even if returning midway under the above conditions, and the task can be continued after charging.

  FIG. 3 is a diagram showing an example of task scheduling of the “dishwashing work in the kitchen” that is a meta-level task. As shown in FIG. 3, the “dishwashing work in the kitchen”, which is a meta-level task, includes operations related to a plurality of commands. Specifically, voice command reception, command recognition, speaker confirmation, command reception response, kitchen sink movement, sink image recognition, work procedure creation, n-th dishwashing, return, end report, and command waiting state operations Consists of. Each operation includes image input, image recognition, image synthesis, image display, voice input, voice recognition, voice synthesis, voice output, external communication, sensor input, sensor signal processing, overall management, task planning, trajectory generation / target It consists of functions such as a value generator 1, motor control for movement, and motor control for arm control.

  FIG. 3 describes the required power consumption for each function, and describes the power consumption for each detailed command operation included in each function and the estimated time required to perform each command operation. Yes. “◯” in FIG. 3 means that the operation is included in the function.

As described above, “kitchen dishwashing” corresponds to a task in which the operation may be interrupted at several stages to return to a chargeable state. The number of plates is unconfirmed information, but even if one plate is washed, there is no problem even if it is interrupted and returned to an arbitrary position. Therefore, in the examination of the feasibility of step S5 in FIG. 3, it is determined that the following equation (3) is satisfied using the energy consumption Pmin for executing one dishwashing.
Pr> Pmin + Pothers + Ps + Pres + Pm (3)

  Here, “Pothers” is the total power consumption other than the start position movement and return work such as image recognition in the sink and work procedure creation in FIG.

In the case of “(c) Tasks where the work may be interrupted at several stages and returned to the chargeable state”, such as “washing dishes in the kitchen”, Pmin is one unit of work. It cannot be interrupted. Therefore, in the battery remaining amount check during work, the following dish washing operation is continued while the following equation (4) is satisfied, and when the equation (4) is not satisfied, one dish washing operation is finished. Thereafter, the feedback operation may be started.
cPr> Pmin + Pres + Pm (4)

  As the above-mentioned “(d) task incapable of interrupting work”, a patrol work or a care assisting work can be considered. In these tasks, if the work is interrupted in the middle, the meaning of the work may be lost, or the work object and the environment may be greatly affected.

  FIG. 4 is a diagram illustrating an example of task scheduling of “look-around work” that is a meta-level task. As shown in FIG. 4, the “look-around work” that is a meta-level task includes operations related to a plurality of commands. Specifically, it consists of voice command reception, command recognition, speaker confirmation, command reception response, roundabout route generation, roundabout operation, door opening, obstacle avoidance, suspicious object discovery, end report, and command waiting state operations. Each operation includes image input, image recognition, image synthesis, image display, voice input, voice recognition, voice synthesis, voice output, external communication, sensor input, sensor signal processing, overall management, task planning, trajectory generation / target It consists of functions such as a value generator 1, motor control for movement, and motor control for arm control.

In the case of “look-around work” in FIG. 4, it can be determined that the task can be executed when Expression (5) holds. Here, Pall is the power consumption required to execute this task.
Pr> Pall + Pm (5)

  In the case of the “(e) task whose work amount and work content cannot be predicted in advance” described above, shopping assistance (porter) work, work for searching for moving objects, and the like can be considered. In these works, the work time and work load are unpredictable, and if the work is interrupted halfway, the meaning of the work so far is lost. Therefore, in the case of the above-described task (e), the robot presents to the user that the content of the work is undecided, so that it cannot be determined whether the work can be completed, and asks the user for judgment. Here, if execution is instructed again, this task is executed.

In the tasks (d) and (e) described above, since it is necessary to complete the task to the end, even if the real-time battery remaining amount cPr during the work is monitored, the work cannot be interrupted appropriately. However, if the effect of stopping is larger, the power consumption Pstop for interrupting the work without affecting the work object or the environment is calculated from the object weight and the following equation (6) It is possible to continue the work only while satisfying the above. However, this eliminates the value of previous work.
cP> Pstop + Pres + Pm (6)

In the tasks (b) and (e) described above, the remaining power consumption Pres for interrupting the work and returning is set in advance. However, in Eq. (2), there are obstacles that were not envisioned in advance on the way back, and the power consumption Pm was used up to avoid this, or in the last one work unit in Eq. (4) When the external energy is consumed, or when there is more power consumption than expected in the interruption work in equation (6), the state of equation (7) can be assumed.
cPr <Pres (7)

  In this case, for example, among the functions used in the feedback command of FIG. 3, the voice input function is stopped, the image input for environment recognition for feedback and the image recognition are stopped, and only the infrared sensor with low energy consumption is used. Prepare a method such as setting a low power consumption emergency return operation mode or sounding an alarm notifying the user of the abnormality at that position.

  FIG. 5 is a diagram showing individual robot modules constituting the robot according to the present embodiment. FIG. 5 illustrates a robot module M1 serving as the top of the robot, a robot module M2 serving as the body, a robot module M3 serving as the moving unit, and a robot module M4 serving as the arm unit. These are merely examples, and other robot modules may be provided, or some of the robot modules in FIG. 5 may be combined into one.

  The robot modules shown in FIG. 5 can be used in combination as necessary. Therefore, a robot according to each application can be realized by arbitrarily combining desired robot modules from among a plurality of robot modules.

  Each of the robot modules M1 to M5 shown in FIG. 5 incorporates each component shown in FIG. 1 and the functional module shown in FIG. For example, the robot module M1 that is the top of the head includes the overall management unit 2, the task planning unit 4, the microphone, the voice preprocessing unit 10, the speaker, the voice synthesis unit 6, the camera, and the image preprocessing unit in FIG. 11, an image recognition unit 7, an image generation unit 8, a small display device, and an external I / F unit 9.

  The robot module M2 that is the trunk unit includes an image generation unit 8, a large display device, and a part of a sensor signal processing unit 12 (such as an ultrasonic distance sensor).

  The robot module M3, which is a moving unit, includes a trajectory generation / target value generation unit 1 for movement, a motor control unit 13, and a motor. The robot module M4 which is an arm unit includes an arm trajectory generation / target value generation unit 1 and a motor 14.

  Each of the robot modules in FIG. 5 is connected by a one-touch connector using, for example, the USB standard for both the power supply system and the signal system. This connection is notified to the overall management unit 2 by a connection information signal 39. The overall management unit 2 may be incorporated in any robot module shown in FIG. When a robot is configured by combining a plurality of robot modules, it is conceivable that a plurality of robots possess the overall management unit 2. In that case, for example, a priority order is determined between the plurality of robot modules. Alternatively, the functions of the overall management unit 2 may be shared by a plurality of robot modules.

  FIG. 6 is a diagram showing an example of a robot configured by arbitrarily combining the robot modules shown in FIG. 6A is a robot for an information terminal, FIG. 6B is a robot for looking around, FIG. 6C is a service robot that provides meals, and FIG. 6D is a service robot for work.

  The robot of FIG. 6A has only the robot module M1 at the top of the head. The robot of FIG. 6B has a robot module M1 at the top and a robot module M3 at the moving part. The robot shown in FIG. 6C includes a robot module M1 at the top, a robot module M2 at the body, and a robot module M3 at the moving unit. The robot shown in FIG. 6D has robot modules M1 to M4.

  FIG. 7 is a flowchart showing a processing procedure of the robot according to the present embodiment. First, an image input by a camera, an audio input by a microphone, and inputs from various sensors are acquired (step S1). Next, a meta level task is generated based on the acquired result (step S2). The processing in step S2 is performed by the meta level task generation unit 29.

  Next, the generated meta level task is analyzed and decomposed into a command level task (step S3). The processing in step S3 is performed by the meta level task analysis unit 30.

  Next, command level work scheduling is performed with reference to the command function list (step S4). The process of step S4 is performed by the command level work scheduling unit 31.

  Next, referring to the remaining battery level and the power consumption data for each command, the feasibility on function and the feasibility on power consumption are examined for each scheduled command (step S5). The process in step S5 is performed by the command level task feasibility function reviewing unit 32. The command level task feasibility function reviewing unit 32 determines what robot module is necessary and currently exists based on the usable module database 36 in which the functions of each robot module are stored in a database and the analysis result in step S3. It is determined whether the task can be executed by the robot module. If it can be executed, this is notified to the user (step S6). At that time, if the user is informed of the specific configuration of the combination of the robot modules, the convenience is enhanced.

  In addition, depending on the task, a configuration proposal of a plurality of robot modules can be considered. In this case, the power consumption for each function of each robot module is calculated, and the power consumption is the highest for any combination of robot modules. It is desirable to show the user whether there are few. Thereby, when assembling a robot capable of executing a certain function, it is possible to easily select a combination of robot modules that minimizes power consumption.

  As described above, in the present embodiment, the current remaining battery level is detected, the power consumption data for each command is referred to determine whether there is a remaining battery level that can execute each command, and the determination result Is displayed to the user or the execution command is changed, so that it is possible to reliably prevent a problem that the robot stops due to running out of battery during operation and causes an unintended result.

  2 and 6 described in the above-described embodiment may be configured by hardware or software. When configured by software, a program that realizes at least a part of the functions of the processing may be stored in a recording medium such as a floppy disk or a CD-ROM, and read and executed by a computer. The recording medium is not limited to a portable medium such as a magnetic disk or an optical disk, but may be a fixed recording medium such as a hard disk device or a memory.

Further, a program that realizes at least a part of the processes of FIGS. 2 and 6 may be distributed via a communication line (including wireless communication) such as the Internet. Furthermore, the program may be distributed through being stored in a recording medium via a wired line or a wireless line such as the Internet in a state where the program is encrypted, modulated, or compressed.

The block diagram which shows the internal structure of the robot which concerns on one Embodiment of this invention. The software block diagram of the robot which concerns on this embodiment. The figure which shows an example of the task scheduling of the "dishwashing work of a kitchen" which is a meta level task. The figure which shows an example of the task scheduling of "look-around work" which is a meta level task. The figure which shows each robot module which comprises the robot which concerns on this embodiment. The figure which shows an example of the robot comprised combining each robot module of FIG. 5 arbitrarily. The flowchart which shows the process sequence of the robot which concerns on this embodiment.

Explanation of symbols

2 General Management Unit 3 Database Unit 4 Task Planning Unit 29 Meta Level Task Generation Unit 30 Meta Level Task Analysis Unit 31 Command Level Task Scheduling Unit 32 Command Level Task Executability Function Review Unit 33 Command Level Task Executability Energy Review Unit 34 Executable Display unit 35 Individual module unit 36 Usable module database

Claims (8)

A remaining amount detecting means for detecting the remaining amount of the battery;
Power consumption information storage means for storing information on power consumption required to execute a plurality of work instructions for each work instruction;
Execution of determining whether there is a remaining battery level that can execute a given work command based on the remaining battery level detected by the remaining level detecting unit and the information stored in the power consumption information storage unit Possible judgment means,
And a command presentation unit that presents a work command that can be executed with the remaining battery level detected by the remaining level detection unit when it is determined by the executable determination unit that there is no remaining battery level. robot. Power consumption predicting means for predicting and calculating power consumption required to execute a work instruction scheduled to be executed while executing a work instruction,
The executable determination means determines whether or not there is a remaining battery capacity to execute all work instructions scheduled to be executed based on the result predicted by the power consumption prediction means,
The command presenting means changes at least a part of the work instructions that can be executed and determines a new one when the command presenting means determines that there is no remaining battery capacity to execute all the work commands scheduled to be executed. The robot according to claim 1, wherein a work command is presented.   3. The command presenting means preferentially presents a work command for returning to an arbitrary position when it is determined by the executable determining means that there is no remaining battery power. The robot described. Meta-level task generation means for generating a task indicating the operation of the robot based on at least one of image input, audio input and sensor input;
A scheduling means for analyzing the generated task and determining a detailed work order execution order, and
The executable determination means determines whether or not there is a remaining battery capacity capable of executing the work command determined by the scheduling means, based on information stored in the power consumption information storage means. The robot according to any one of claims 1 to 3. Individual block information storage means for storing attribute information including power consumption, identification information, and connection information with the robot body in association with each of a plurality of functional modules that perform different operations from each other,
A module selection device comprising: module selection means for selecting at least one functional block for constituting a robot based on information stored in the individual block information storage means.   6. The module selection unit, based on information stored in the individual block information storage unit, selects a functional block that minimizes power consumption when the robot performs an arbitrary operation. The module selection apparatus as described in.   The selection module assembly information presenting means for presenting a function module selected by the module selecting means and an assembly procedure of a robot configured using the selected function module is provided. The module selection apparatus of description. Information on power consumption required for the robot to execute a plurality of work instructions is stored in the power consumption information storage means for each work instruction;
Based on the information stored in the power consumption information storage means, it is determined whether there is a remaining battery capacity capable of executing a given work command,
A robot control method characterized in that, when it is determined that there is no remaining battery power, a work command executable by the remaining battery power detected by the remaining power detection means is presented by the command presentation means.

Images