JP7503741B2 - Smart Pendant - Google Patents

Description

The present invention relates to a smart pendant, which is one component of a robot system and is a device for teaching a robot.

There are two methods for registering instructions for the operation of a robot or a robot arm in a system: direct teaching and indirect teaching. An input display device called a pendant is usually used for teaching a robot. The pendant is equipped with a display unit that displays images for teaching the robot and images showing the robot's operating state, as well as operation keys for manually operating the robot and performing various input operations, and can be carried and operated by the worker.
Direct teaching and indirect teaching by a pendant, and prior art of the pendant, are disclosed, for example, in the following Patent Document 1. Problems with direct teaching and indirect teaching are also disclosed in Patent Document 1. The term "indirect teaching" does not appear in this Patent Document. However, what is not direct teaching is indirect teaching, and the term "text-based" is used in this Patent Document.

JP 2017-227864 A

Direct teaching is a teaching method in which a force sensor or torque sensor is attached to the tip of the robot's arm or to the joint, and the operator moves the arm directly. Motors are also considered to be part of the sensor. The torque of the joint can be estimated from the motor current of the motor attached to the arm, so the motor can also be used as a sensor. However, considerable skill is required to execute teaching solely or mainly using the motor current.
On the other hand, there are two types of indirect teaching: online and offline. Online teaching has been widely used in factory production sites. It is a method in which the worker inputs teaching data (teaching program) into the pendant. If the teaching data or teaching program is of high quality, it is a highly reliable teaching method. However, creating good teaching data is not easy. Providing a system that assists in how to easily create teaching programs, which are generally not easy to create, is where the skill of the pendant manufacturer can be shown, and also where the responsiveness and technical capabilities of the pendant users on-site are shown.
On the other hand, offline teaching has the advantage that it is resistant to changes in the robot's operating environment and teaching can be done without stopping the actual machine, but it has the problem that it is difficult to develop a program for offline teaching. Also, offline teaching is generally done using a personal computer rather than a pendant. Since robot pendants are carried and operated by the worker, it is common that the display area (screen) of the pendant does not have a sufficient area. For this reason, it is not easy to do offline teaching on a pendant. If offline teaching is not done on a pendant, it is necessary to prepare a personal computer every time teaching is done on site, which is a hassle. Even if a personal computer is prepared, it is not necessarily possible to provide better teaching than online teaching. In other words, it is not certain whether the effort of preparing a personal computer is worth it. If you try to do offline teaching on a pendant, it is more difficult to develop a program than online teaching.
The present invention has been made in consideration of the current state of conventional teaching methods, and aims to provide a clever pendant, or smart pendant, which is less dependent on the skill level of the operator and can improve the operability of teaching work in a pendant where the size of the screen is limited.

The smart pendant according to the present invention has the following configuration.
The invention of claim 1 is a portable and operable device for teaching a robot, comprising an input means and a display unit, a sensor input unit that receives information for implementing control by AI, and an AI control unit that refers to data from the sensor input unit and performs teaching of the robot by AI.

The invention of claim 1 has a non-AI control unit in a smart pendant for teaching a robot without AI control.

The invention of claim 1 is a smart pendant provided with an AI control implementation judgment unit that judges whether or not robot teaching by AI is appropriate and manages the operation of the AI control unit and the non-AI control unit.

The invention of claim 2 is characterized in that in the smart pendant of claim 1 , the AI control implementation judgment unit is provided with a means for learning how to combine AI teaching and non-AI teaching.

By teaching the robot pendant using AI, the operability and productivity of the worker who operates the pendant can be improved. Also, by not teaching using AI, it is possible to prevent the operability and productivity of the pendant from decreasing. Furthermore, since it is possible to operate the robot by combining AI control and non-AI control, the smart pendant allows both skilled and unskilled workers to operate the robot more appropriately than before.

FIG. 1 is a block diagram showing a first embodiment of the present invention; A block diagram showing a second embodiment of the present invention. A block diagram showing a third embodiment of the present invention. A block diagram showing an embodiment of AI control according to the present invention.

In the following, embodiments of the present invention will be described with reference to the drawings, from Examples 1 to 4. In the following examples, the smart pendant will be referred to as a pendant.

Fig. 1 is a block diagram showing a first embodiment of the present invention. The main components of the robot system are a robot 9 and a pendant 1. Although not shown, the robot has a drive mechanism, i.e., an arm. Although not shown, a so-called system controller for the robot may be provided anywhere in the system. A more convenient method for implementing the present invention is to have it built into the control unit 6 of the pendant 1.
In the present invention, a sensor input unit 4 is required. The sensor input unit may be electrically connected to a single sensor or multiple sensors. Fig. 1 shows a case where a camera 10a is used as the sensor. The location where the sensor is provided is free. The sensor may be built into the robot, the pendant, or it may be provided independently outside the pendant 1 or robot 9 as shown in Fig. 1.
The pendant 1 is a portable terminal device or operating device equipped with an input function (input means) and an output function (display means). The essential requirement is that it is capable of so-called teaching. The pendant 1 is equipped with a display unit 2 and an input unit 3. The input unit 3 may be any input means. For example, a mouse may be used, but a keypad is usually used since it is easy to implement on the pendant. The input unit 3 outputs signals to the pendant 1 or the control unit 6.
In the present invention, a tablet-type terminal is also considered to be a type of pendant 1. In other words, the display unit 2 may be a touch panel and the input keys (input unit 3) may be provided on the display unit 2. Generally, a notebook computer is not a pendant. However, a terminal that can be operated while being carried by a worker is a pendant as defined by the present invention. For example, a notebook computer with a strap attached is a pendant. Also, if a notebook computer is placed in a bag such as a shoulder bag and operated, it is a pendant.
The important characteristic components of the present invention are the sensor input unit 4 and the subsequent AI control unit 5. The AI control unit 5 is a hardware or software module for operating and teaching a robot using artificial intelligence (AI). Various AI algorithms are known. Any of them may be used in the present invention. Any algorithm or method that refers to information from a sensor and attempts to intelligently move an actuator (arm) is acceptable. However, deep learning, subsumption architecture, etc. are particularly recommended.
The AI control unit 5 may be provided anywhere in the pendant 1. In general, it is provided within the control unit 5 or is connected to the control unit 5 so as to be able to communicate with it.
By providing the pendant 1 with a sensor input unit 6 and an AI control unit 5, it becomes possible to teach a robot through AI control using the pendant 1, which is a portable terminal.
In one embodiment of the AI control of the present invention, an operator directly teaches the robot 9 by moving the tip of the arm of the robot 9, and the data obtained is converted into a program and stored in the AI control unit 5. It is also recommended to use AI for algorithms and modules that estimate torque from motor current.

FIG. 2 is a block diagram showing a second embodiment of the present invention.
As described above, the sensor 10 may be attached to the robot 9. Other than a camera, the sensor 10 that can be used is a microphone, a force sensor, a torque sensor, a force sensor, an acceleration sensor, an encoder, etc. Naturally, a motor (not shown) can also be used as a sensor. This is because the torque can be calculated from the motor current. Furthermore, key operation information (operation history) of the input unit 3 can also be used as sensor information. In this case, the input unit 3 will also have some of the functions of the sensor input unit 4.
1 and 2 show almost the same embodiment, but the major difference is that in Fig. 2, a non-AI control unit 8 is provided within the control unit 6. The non-AI control unit 8 may be provided separately from the control unit 6. The non-AI control unit 8 may also be provided within the AI control unit 5.
When an operator performs a specified key operation on the input unit, teaching using either AI control or non-AI control can be performed as appropriate.

FIG. 3 is a block diagram showing a third embodiment of the present invention.
This embodiment is configured by providing an AI control implementation determination unit 7 in addition to the embodiment shown in Fig. 2. The rest is the same as in Fig. 2.
The AI control implementation determination unit 7 may be provided separately from the control unit 6. The AI control implementation determination unit 7 may be provided within the AI control unit 5.
The AI control implementation judgment unit 7 is a module that judges the merits and demerits of AI control and non-AI control. If it is judged that control using AI is advantageous for teaching operation by evaluating the worker's proficiency, operation history, work content, and AI evolution (learning degree and willingness to learn), teaching of the robot 9 by AI is performed. If it is judged that teaching operation by AI control is not advantageous, teaching of the robot 9 by non-AI control is performed. Teaching by non-AI control itself is a conventional technology, and is typically indirect teaching online teaching. In the case of the present invention, the consideration is made to perform teaching by AI control. However, if the AI control implementation judgment unit 7 judges that it is disadvantageous, teaching by non-AI control is performed and the data is accumulated. The AI control implementation judgment unit 7 is a function specialized for evaluation, and the teaching operation itself is performed by the AI control unit 5 or the non-AI control unit 8.

4 is a block diagram showing an embodiment of AI control according to the present invention, which is defined as a fourth embodiment of the present invention.
As described above, the AI control implementation judgment unit 7 may be provided in the control unit 6 or in the AI control unit 5.
The AI control unit 5 has a module configuration as shown in FIG.
It is not necessary to separate the direct teaching module 5a and the indirect teaching module 5b. This is especially true when only one module is called. However, when the control calculation unit 5c combines, synthesizes, or compares the multiple modules called up, it is better to use the modules as shown in Figure 4, which is less likely to confuse both the pendant maker and the user.
The control calculation unit 5c is a means for learning how to combine AI teaching and non-AI teaching, and learns which robot control steps including teaching should be AI controlled, and which parts of a series of control operations should be AI controlled.
The results of this learning are processed by the AI control implementation judgment unit 7. The AI control implementation judgment unit 7 may be provided anywhere within the pendant.

By providing the pendant with a means for executing AI control, the operability of the robot pendant can be improved compared to the conventional art, thus providing one form of smart pendant.

REFERENCE SIGNS LIST 1 Pendant 2 Display unit 3 Input unit 4 Sensor input unit 5 AI control unit 5a Direct teaching module 5b Indirect teaching module 5c Control calculation unit 6 Control unit 7 AI control execution determination unit 8 Non-AI control unit 9 Robot 10 Sensor 10a Camera (sensor)

Claims (2)

A portable and operable device for teaching a robot, comprising:
Equipped with an input means and a display unit,
A sensor input unit that receives information for implementing AI control;
An AI control unit that refers to data from the sensor input unit and executes AI-based robot teaching ;
A non-AI control unit for teaching a robot without AI control;
A smart pendant comprising an AI control implementation judgment unit that judges whether or not robot teaching by AI is appropriate and manages the operation of the AI control unit and the non-AI control unit . 2. The smart pendant of claim 1 , wherein the AI control implementation judgment unit is provided with a means for learning how to combine AI teaching and non-AI teaching.

Images