Speech recognition system for teaching industrial robot

Abstract

To provide a voice recognition system for teaching an industrial robot capable of surely teaching the movement to the industrial robot by the voice of a user.SOLUTION: A teaching speech recognition system 100 for an industrial robot that teaches a motion to an industrial robot 101 by speech according to the present invention includes a speech input device 122 that inputs a speech of a user, a general speech database 126 in which speech models of general words are registered, a dedicated speech database 128 in which speech models of dedicated words used for teaching the industrial robot are registered, a speech registration module 130 that registers a speech of a dedicated word input to the speech input device and pronounced by the user as a speech model in the dedicated speech database for each user before teaching the industrial robot, and a teaching speech recognition module 132 that recognizes the speech of the user input to the speech input device by referring to the general speech database and the dedicated speech database when teaching a motion to the industrial robot.SELECTED DRAWING: Figure 2

Description

[Technical Field]

[0001] The present invention relates to a voice recognition system for teaching an industrial robot, which teaches the industrial robot a movement by voice. [Background technology]

[0002] When teaching movements to industrial robots such as articulated robots, the teaching work is sometimes performed using a teaching pendant. However, in this type of teaching work, the operator (user) must operate the teaching pendant while watching the industrial robot, which places a heavy workload on the operator.

[0003] Therefore, in order to reduce the burden on the user in the teaching work, a robot teaching device is known that teaches movements to an industrial robot by voice, without using a teaching pendant (for example, Patent Document 1).

[0004] The robot teaching device of Patent Document 1 includes a voice recognition unit that recognizes the voice of an operator, a memory unit, a specific word extraction unit, and an operation command generation unit. The memory unit stores in advance specific words associated with the operation of the robot. The specific word extraction unit extracts specific words that match the recognized words. The operation command generation unit generates operation commands for the robot based on operation information associated with the extracted specific words. [Prior art documents] [Patent documents]

[0005] [Patent Document 1] Japanese Patent Application Publication No. 2019-126902 Summary of the Invention [Problem to be solved by the invention]

[0006] The robot teaching device of Patent Document 1 associates specific words with robot operations and stores them in a memory unit before teaching, and extracts the specific words from the user's pronunciation based on the memory unit during teaching. However, there are cases where the pronunciation of specific words varies slightly from the general pronunciation depending on the user. In such cases, even if the user pronounces the specific word during teaching, the specific word is not extracted, and it is not possible to create an operation command associated with the specific word, making it difficult to perform the intended teaching.

[0007] In particular, when teaching robots, there are many extremely short words, such as body part names like "J1, J2" or "right" and "left." Furthermore, because they are not sentences, it is difficult to infer and correct them from the context. Therefore, short words must be recognized solely from their pronunciation, making it difficult to improve recognition accuracy.

[0008] In view of the above problems, the present invention aims to provide a voice recognition system for teaching industrial robots that can reliably teach movements to industrial robots using a user's voice. [Means for solving the problem]

[0009] In order to solve the above problems, a typical configuration of a voice recognition system for teaching industrial robots according to the present invention is a voice recognition system for teaching industrial robots that teaches movements to industrial robots by voice, characterized in that it comprises: a voice input device for inputting a user's voice; a general voice database in which voice models of general words are registered; a dedicated voice database in which voice models of dedicated words used in teaching industrial robots are registered; a voice registration module that registers the voice of the dedicated words pronounced by the user input into the voice input device as a voice model in the dedicated voice database for each user before teaching the industrial robot; and a teaching voice recognition module that recognizes the user's voice input into the voice input device by referring to the general voice database and the dedicated voice database when teaching the industrial robot movements.

[0010] It is preferable that the training speech recognition module refers to the dedicated speech database in preference to the general speech database when recognizing the user's speech. [Effects of the Invention]

[0011] According to the present invention, it is possible to provide a voice recognition system for teaching an industrial robot that can reliably teach an industrial robot a movement using a user's voice. [Brief explanation of the drawings]

[0012] [Figure 1] 1 is a diagram illustrating an overview of a speech recognition system for teaching an industrial robot according to an embodiment of the present invention. [Figure 2] FIG. 2 is a block diagram showing the functions of the teaching speech recognition system of FIG. 1. [Figure 3] 3 is a flowchart showing the operation of the teaching voice recognition system. [Figure 4] FIG. 10 is a diagram showing an example of a screen displayed on a display device by a voice registration module. DETAILED DESCRIPTION OF THE INVENTION

[0013] Preferred embodiments of the present invention will be described in detail below with reference to the accompanying drawings. Dimensions, materials, and other specific values ​​shown in the embodiments are merely examples for facilitating understanding of the invention and, unless otherwise specified, do not limit the present invention. In this specification and drawings, elements having substantially the same functions and configurations are designated by the same reference numerals to avoid redundant explanation, and elements not directly related to the present invention are not shown.

[0014] 1 is a diagram illustrating an outline of a speech recognition system for teaching an industrial robot according to an embodiment of the present invention. The speech recognition system for teaching 100 is a system for teaching an industrial robot 101 a movement by voice, and includes a teaching terminal 102.

[0015] The teaching terminal 102 is an operation terminal for performing teaching work on the industrial robot 101. The teaching terminal 102 is a tablet, and is connected to the industrial robot 101 by wire via a robot control device 103. The robot control device 103 is connected to the industrial robot 101 by wire, and controls the operation of the industrial robot 101. However, the teaching terminal 102 may be configured to perform wireless communication with the robot control device 103.

[0016] The industrial robot 101 to be taught is, for example, a six-axis articulated robot, and includes a base 106 installed on a floor 104 of a factory or the like, a rotating frame 108, a first arm 110, a connecting arm 112, a second arm 114, and an end effector 116. The end effector 116 is attached to a wrist arm 118 attached to the tip of the second arm 114.

[0017] The swivel frame 108 rotates about the J1 axis. When the base 106 is placed on the floor 104, the J1 axis is supported in a direction approximately perpendicular to the floor 104. The first arm 110 rotates about the J2 axis. The J2 axis is supported on the swivel frame 108 in a direction approximately perpendicular to the J1 axis. The connecting arm 112 is rotatably connected to the first arm 110 via the J3 axis. The J3 axis is supported approximately parallel to the J2 axis at a tip 120, which is the end of the first arm 110 opposite the J2 axis.

[0018] The second arm 114 is connected to the connecting arm 112 via a J4 axis extending in the longitudinal direction of the second arm 114 so as to be rotatable around the J4 axis. A wrist arm 118 attached to the tip of the second arm 114 rotates relative to the second arm 114 via the J5 axis in a tilting direction rather than a twisting direction. The end effector 116 rotates via the J6 axis in a direction (twisting direction) substantially perpendicular to the rotation direction of the wrist arm 118. The J5 and J6 axes are supported by the wrist arm 118 in a direction substantially perpendicular to the J4 axis.

[0019] Therefore, when performing teaching work in which the industrial robot 101 is instructed to move by voice using the teaching voice recognition system 100, not only general words but also specialized words such as the names of the parts of the industrial robot 101, such as "J1-J6" mentioned above, are used.

[0020] Fig. 2 is a block diagram showing the functions of the teaching speech recognition system 100 of Fig. 1. The teaching terminal 102 of the teaching speech recognition system 100 includes a voice input device 122 such as a microphone for inputting the voice of an operator (user), and a display device 124 (which also serves as a manual input device) such as a touch panel shown in Fig. 1. Note that the teaching terminal 102 is not limited to the wired tablet shown in Fig. 1, and may be a teach pendant, an external PC connected to a microphone for voice input, or the like.

[0021] 2, the teaching terminal 102 further includes a general speech database 126, a dedicated speech database 128, a speech registration module 130, a teaching speech recognition module 132, a sentence analysis module 134, a robot operation command generation device 135, and a robot program recording device 136. In this application, a "module" refers to software (programs) that are realized by executing the software on a CPU (not shown).

[0022] The general voice database 126 stores voice models of general words. The dedicated voice database 128 stores voice models of dedicated words used to teach the industrial robot 101. The voice registration module 130, which will be described in detail later, registers (learns) the voice of the dedicated words pronounced by the user input to the voice input device 122 as a voice model for each user in the dedicated voice database 128 before teaching the industrial robot 101.

[0023] When teaching the industrial robot 101 a movement (during teaching), the teaching voice recognition module 132 recognizes the user's voice input to the voice input device 122 by referring to the general voice database 126 and the dedicated voice database 128, and converts it into text data.

[0024] The sentence analysis module 134 extracts text data related to the movement of the industrial robot 101 from the text data output by the teaching voice recognition module 132. Furthermore, the sentence analysis module 134 combines the extracted text data and outputs it to the robot operation command generation device 135. The robot operation command generation device 135 generates operation commands for operating the industrial robot 101 based on the text data, and records them as a robot program in the robot program recording device 136. Note that the operation commands output from the robot operation command generation device 135 may be recorded in the robot program recording device 136 as described above, but are not limited to this, and may be sent directly to the robot control device 103 to operate the industrial robot 101.

[0025] Fig. 3 is a flowchart showing the operation of the teaching speech recognition system 100 of Fig. 1. Fig. 4 is a diagram showing an example of a screen displayed on the display device 124 by the speech registration module 130.

[0026] In the teaching speech recognition system 100, first, before teaching the industrial robot 101, the user pronounces dedicated words to learn them (step S100). In detail, the voice registration module 130 lists dedicated words to be read out as shown in the example screen of FIG. 4. The user taps a dedicated word to enter an input waiting state. In the example of FIG. 4, "J2" is activated.

[0027] The voice of the dedicated words pronounced by the user is input to the voice input device 122. The voice registration module 130 registers the voice of the dedicated words (see FIG. 4) pronounced by the user, input to the voice input device 122, as a voice model in the dedicated voice database 128 for each "user name" shown in FIG. 4 (step S102). The dedicated words pronounced by the user include, as shown in FIG. 4, "J1-J6" which are body part names, "X, Y, Z" which indicate coordinate axes, "roll, pitch, yaw" which indicate rotation directions, as well as "movement," "tool" corresponding to a tool number, "tool name" such as a pickup tool, and "macro name" corresponding to a composite action.

[0028] 4, when a dedicated word is selected and in a waiting state for input, the voice registration module 130 displays a message such as "Please pronounce it" on the display device 124 to prompt the user to pronounce it. When the voice registration module 130 registers the voice of the dedicated word as a voice model in the dedicated voice database 128, it displays "OK", and if it is not registered, it displays "Not registered".

[0029] Next, the teaching voice recognition module 132 reads the general voice database 126 and the dedicated voice database 128 when recognizing the user's voice uttered during teaching of the industrial robot 101 (step S104). At this time, the teaching voice recognition module 132 refers to the dedicated voice database 128 with priority over the general voice database 126 to recognize the user's voice.

[0030] As described above, the teaching speech recognition system 100 according to the present invention can reliably recognize specialized words contained in a user's voice even if the pronunciation of the specialized words differs slightly from general pronunciation depending on the user or even if the specialized words are extremely short. In other words, the teaching speech recognition system 100 improves the speech recognition rate for each user, enabling safe teaching work.

[0031] Furthermore, when recognizing the user's voice spoken during teaching of the industrial robot 101, the dedicated voice database 128 is referred to preferentially over the general voice database 126, thereby enabling dedicated words to be reliably extracted and recognized from the user's voice.

[0032] While the preferred embodiments of the present invention have been described above with reference to the accompanying drawings, it goes without saying that the present invention is not limited to such examples. It is clear that those skilled in the art can conceive of various modifications and alterations within the scope of the claims, and it is understood that such modifications and alterations also fall within the technical scope of the present invention. [Industrial Applicability]

[0033] The present invention can be used as a voice recognition system for teaching industrial robots, which teaches industrial robots movements by voice. [Explanation of symbols]

[0034] 100...Teaching voice recognition system, 101...Industrial robot, 102...Teaching terminal, 103...Robot control device, 104...Floor, 106...Base, 108...Swivel frame, 110...First arm, 112...Connecting arm, 114...Second arm, 116...End effector, 118...Wrist arm, 120...Tip of first arm, 122...Voice input device, 124...Display device, 126...General voice database, 128...Dedicated voice database, 130...Voice registration module, 132...Teaching voice recognition module, 134...Sentence analysis module, 135...Robot operation command generation device, 136...Robot program storage device

Claims

1. In a speech recognition system for teaching an industrial robot, which teaches the industrial robot movements by voice, a voice input device for inputting a user's voice; A general speech database containing speech models of common words, a dedicated speech database in which speech models of dedicated words used for teaching the industrial robot are registered; a voice registration module that registers, as a voice model, the voice of the dedicated word pronounced by the user input to the voice input device in the dedicated voice database for each user before teaching the industrial robot; and a teaching voice recognition module that recognizes a user's voice input to the voice input device by referring to the general voice database and the dedicated voice database when teaching the industrial robot a movement.

2. 2. The teaching voice recognition system for industrial robots according to claim 1, wherein the teaching voice recognition module refers to the dedicated voice database in preference to the general voice database when recognizing a user's voice.

Images