KR102348881B1 - The system and method for monitoring of semiconductor manufacturing process using air actuator motion image information - Google Patents

Abstract

The present invention relates to a system for monitoring a mechanism in a manufacturing process including a pneumatic cylinder as a manufacturing facility. More particularly, the present invention relates to a system for monitoring a manufacturing process using a pneumatic cylinder which may learn an operation of the pneumatic cylinder in a semiconductor manufacturing process by an artificial intelligence to monitor a manufacturing process using learning information, and to a method thereof. An existing system for monitoring a semiconductor manufacturing process using a camera uses a camera of high cost to increase a cost and have difficulty in configuring a camera system according to a semiconductor manufacturing process field. The present invention suggest a monitoring system for monitoring an operation of a pneumatic cylinder to effectively determine and manage whether or not instrument devices fail in a semiconductor manufacturing process facility, shoots a piston operation of the pneumatic cylinder by the camera, and learns artificial intelligence for the image information to be suitable for a current semiconductor manufacturing process.

Description

The semiconductor manufacturing process monitoring system and monitoring method using the motion image information of the pneumatic cylinder

The present invention relates to a motion monitoring system for a mechanical device in a manufacturing process including an air actuator as a manufacturing facility. In particular, the operation of a pneumatic cylinder in a semiconductor manufacturing process is learned through artificial intelligence, and the learning It relates to a manufacturing process monitoring system and monitoring method using a pneumatic cylinder so that the manufacturing process can be monitored using information.

Recently, a monitoring system for monitoring the manufacturing process using a variety of equipment has been configured at the automated production site. In particular, manufacturers of semiconductors and automobiles are equipped with a process monitoring system to check and monitor the process status for each process line. are doing

In this process monitoring system, a number of cameras are installed on site, and a number of digital video recorders suitable for the cameras in the general control room, a control terminal capable of controlling the digital video recorders, and a number of monitors are installed and the site is managed through this.

In the case of a large manufacturing facility, a digital video recorder that records images transmitted from multiple cameras must implement a mass storage function.

Therefore, the conventional process monitoring system not only requires many digital video recorders, matrix equipment, and equipment control systems, but also has a very difficult configuration in the general control room in the case of a system consisting of several hundred cameras or more, and it is necessary to add a camera In this case, a new program for controlling the camera connection with the system is required.

In particular, manufacturers of semiconductors and automobiles form large-scale production lines, and multiple vision equipment such as IP cameras are installed in each production line to perform defect rate or quality inspection for each process.

Republic of Korea Patent Publication No. 1619097, a process monitoring system in a semiconductor manufacturing process is installed in the semiconductor process apparatus, and constitutes a camera and a storage means for acquiring the first image data having a low resolution and the second image data having a high resolution to provide a process monitoring system in a semiconductor manufacturing process, including a management server that compares high-resolution data and standard image data to generate error information.

In the Republic of Korea Patent Publication No. 2014 0113130, an industrial vision integrated process management system, an API (Application Program Interface) and communication protocol suitable for each manufacturer's vision equipment or monitoring and analysis equipment through the Device Plug-in Framework It provides an industrial vision integrated process management system that can provide excellent compatibility performance by supporting all integrated management of connection and control of heterogeneous vision equipment or monitoring and analysis equipment.

However, as such a monitoring system using a camera is expensive as it uses an expensive camera, it is difficult to configure the camera system according to the semiconductor manufacturing process site.

In addition, in configuring the semiconductor manufacturing process monitoring system, a sensor for detecting an acceleration change (pattern) is required due to the characteristics of the semiconductor manufacturing process (deposition process).

In recent years, due to the rapid development of multimedia technology and information communication, the demand for semiconductor devices is increasing and the technology is developing. As the semiconductor device is required to operate at a high speed and to have a large-capacity storage capacity, there is a demand for technological improvement in the manufacturing technology of the semiconductor device, and thus, the manufacturing process monitoring system is a very important part.

A semiconductor device is generally manufactured by sequentially performing a series of unit processes for film formation, pattern formation, metal wiring formation, and the like.

In these unit processes, as semiconductor manufacturing technology requires higher precision, the importance of a chamber providing process conditions required for each process is being emphasized. The process conditions inside the chamber must be precisely controlled according to the process, and accurate opening and closing of the chamber door is an essential element in the control.

A pneumatic cylinder is used as a power source for the opening and closing of the chamber door. Due to the characteristics of the pneumatic cylinder, when the piston reciprocates, the piston packing mounted on the inner wall of the pneumatic cylinder tube and the piston, which serves as a seal, of the pneumatic cylinder They rub against each other by the stroke distance.

Therefore, if it is used for a long time, the piston packing, which is relatively weak in material, is worn out. When the piston packing begins to wear, the sealing effect inside the piston is reduced, and an internal air leak occurs.

The air leakage inside the pneumatic cylinder reduces the output of the pneumatic cylinder, and makes it impossible to accurately open and close the chamber door to be driven. These problems cause changes in process conditions and lead to process defects and product defects.

In addition, when an error occurs in the pneumatic cylinder due to leakage, the process is stopped and repair and replacement of parts for the pneumatic cylinder are required, and the repair work leads to a decrease in equipment utilization.

Therefore, it can be said that it is very advantageous to configure a monitoring system of a semiconductor manufacturing process to accurately monitor the operation of the pneumatic cylinder.

The present invention intends to propose a monitoring system that can monitor the operation of the pneumatic cylinder to more effectively determine and manage the abnormality of the mechanical device in the semiconductor manufacturing process manufacturing facility. This is to provide a monitoring system suitable for the current semiconductor manufacturing process by learning information by artificial intelligence.

The semiconductor manufacturing process monitoring system using the motion image information of the pneumatic cylinder of the present invention includes a camera installed in each mechanical device in a semiconductor manufacturing facility to acquire motion image information of the pneumatic cylinder in the mechanical device, and each mechanical device in the semiconductor manufacturing facility A sensor installed in the interior to detect the operating state of the pneumatic cylinder, and a terminal unit that provides image information acquired from the camera installed for each mechanical device and sensor detection information acquired from the sensor to a central monitoring and control unit, and from each terminal unit A central monitoring control unit that monitors and manages the operation of the mechanical device by referring to the learning image information learned by artificial intelligence according to the operation state of the pneumatic cylinder by receiving the transmitted image information and sensor detection information, and the central monitoring and control unit Provides operation state information Consists of including an external notification unit to receive and notify the administrator (user),

The central monitoring and control unit includes a terminal information receiving unit for receiving the image information and sensor detection information obtained from each terminal unit, and monitoring information for storing and managing the image information and sensor detection information received from the terminal information receiving unit in the information storage unit. Detects the piston operation of the pneumatic cylinder from the real-time image information provided from the management unit and the terminal information receiving unit, and according to the detection result and the sensor detection information provided from the terminal information receiving unit, the operating state of the pneumatic cylinder is registered and stored in the information storage unit. A monitoring control unit that judges whether there is an abnormal operation of the pneumatic cylinder with reference to the information, determines the operation of the mechanical device according to the judgment result, monitors and manages each mechanical device in the semiconductor manufacturing process, and provides monitoring information to an external notification unit; It is characterized in that it is configured to include an information storage unit for storing image information and image information received from the terminal information receiving unit, sensor detection information, and operation state information of the apparatus.

And in the semiconductor manufacturing process monitoring system using the operation image information of the pneumatic cylinder of the present invention, the monitoring and control unit registers the image information and sensor detection information of the real-time pneumatic cylinder provided from the terminal information receiving unit in the information storage unit, and refers to the learning image information stored in the pneumatic Monitoring and management of each mechanical device in the semiconductor manufacturing process by providing maintenance information that can inform when abnormal operation occurs to the operation judgment unit that determines the operation of the cylinder, and generates operation status information of the mechanical device according to the judgment result and an external notification unit and a monitoring unit.

The operation determination unit recognizes the piston of the pneumatic cylinder from the image information provided from the terminal information receiving unit and contrasts with the learning image information according to the piston motion detection means for detecting the recognized motion of the piston, and the piston motion detected from the piston motion detection means In contrast to the pattern information of the learning image information registered in the piston operation pattern generating means for generating a piston operation pattern for the information storage unit includes an operation determination means for determining the operating state of the pneumatic cylinder.

On the other hand, the central monitoring control unit is configured to further include an artificial intelligence learning unit that receives the operation state of the pneumatic cylinder determined from the real-time image information from the monitoring control unit, generates field learning image information, and stores it in the information storage unit, the artificial intelligence learning The unit is configured with means for generating field learning image information by performing artificial intelligence learning by receiving the motion detection information and the piston motion pattern information from the motion determination unit according to a predetermined algorithm, and generating with reference to the user's input information It is configured to further include means for generating field learning image information, including the user-guided learning information that becomes and the monitoring program of the computer terminal means or smart phone is characterized in that it is configured to include a field learning control process that provides a supervised learning input interface means for allowing a user to input supervised learning information.

The semiconductor manufacturing process monitoring method using the operation image information of the pneumatic cylinder of the present invention recognizes the piston operation of the pneumatic cylinder when the image information and sensor detection information are received from the terminal unit, detects the piston operation information, and confirms the identification information of the terminal unit The operation detection process of the pneumatic cylinder stored in the information storage unit and the operation information of the detected piston refer to the learning image information stored in the information storage unit to determine the operation abnormality of the mechanical device corresponding to the terminal unit. A monitoring management process in which the determination process and the operation state information determined through the operation abnormality determination process are stored in the information storage unit, and when an operation abnormality occurs, maintenance information is generated, provided to an external notification unit, and stored in the information storage unit And, when the user is connected and the user's access request information is input, the access is allowed by checking the visitor information stored in the information storage unit, and when the monitoring information inquiry request is input from the connected user's computer terminal means or smart phone, the input monitoring It is characterized in that it comprises a monitoring information providing process for providing the operation state information or pre-stored maintenance information stored in the information storage unit to the user's computer terminal means or smart phone according to the information inquiry request.

And, the semiconductor manufacturing process monitoring method using the operation image information of the pneumatic cylinder of the present invention is a field learning image by performing an artificial intelligence learning process according to a predetermined algorithm for the operation information of the pneumatic cylinder detected through the operation detection process of the pneumatic cylinder It is configured to further include a field learning image information generation process to generate information and store it in the information storage unit,

In the operation abnormality determination process, the operation information of the piston detected through the operation detection process of the pneumatic cylinder is referred to the learning image information and field learning image information stored in the information storage unit to determine whether the mechanical device corresponding to the terminal is abnormal. It characterized in that it consists of a process of determining, wherein the field learning image information generation process further comprises a process of generating field learning image information according to the supervised learning information input by the user when the user's supervised learning information is input do it with

According to the present invention as described above, by using the learning image information for the piston operation pattern of the pneumatic cylinder in the semiconductor manufacturing process to determine whether the operation of the pneumatic cylinder is abnormal and to monitor the operation of the corresponding mechanical device accordingly, expensive It is possible to construct an efficient monitoring system in a semiconductor manufacturing process without a camera (vision device) or a sensor for detecting acceleration.

1 is a block diagram showing the configuration of a first embodiment of the semiconductor manufacturing process monitoring system using the operation image information of the pneumatic cylinder of the present invention.
Figure 2 is a block diagram showing the configuration of the second embodiment of the semiconductor manufacturing process monitoring system using the operation image information of the pneumatic cylinder of the present invention.
3 is a diagram showing the comparison of the object recognition algorithm detection performance index
4 and 5 are views showing an open/close object recognition learning example of a pneumatic cylinder (vacuum valve), FIG. 4 is an open state, FIG. 5 is a closed state.
6 to 8 are views showing an example of object recognition learning by operation state of a pneumatic cylinder (gate valve), FIG. 6 is a 20% open state, FIG. 7 is a 40% closed state, and FIG. 8 is a 60% closed state.
9 and 10 are views showing an example of a pneumatic cylinder piston operation pattern for performing learning, FIG. 9 is a view showing a normal operation pattern of the piston, and the acceleration (a) is a view showing a constant speed within a certain time, FIG. 10 is a diagram showing the abnormal operation pattern of the piston.
11 is a diagram illustrating a convolutional neural network (CON) for artificial intelligence learning.
12 is a flowchart showing an embodiment of a method for monitoring a semiconductor manufacturing process using operation image information of a pneumatic cylinder of the present invention.
13 is a flowchart showing the execution process of another embodiment of the semiconductor manufacturing process monitoring method using the operation image information of the pneumatic cylinder of the present invention.

An embodiment of the semiconductor manufacturing process monitoring system using the operation image information of the pneumatic cylinder of the present invention will be described with reference to the embodiments shown in FIGS. 1 and 3 to 12 attached thereto.

A camera 100 installed in each mechanical device in the semiconductor manufacturing facility to acquire operation image information of the pneumatic cylinder in the mechanical device, and a sensor installed in each mechanical device in the semiconductor manufacturing facility to detect the operating state of the pneumatic cylinder ( 200), and a terminal unit 300 that provides the image information obtained from the camera 100 installed for each instrument and sensor detection information obtained from the sensor 200 to the central monitoring and control unit 400, and each terminal unit A central monitoring and control unit 400 that monitors and manages the operation of the mechanism by referring to the learning image information learned by artificial intelligence according to the operating state of the pneumatic cylinder by receiving the image information and sensor detection information transmitted from the 300, and central monitoring It is configured to include an external notification unit 500 for notifying an administrator (user) by receiving the operation state information and maintenance information from the control unit 400 .

The central monitoring and control unit 400 includes a terminal information receiving unit 410 for receiving the image information and sensor sensing information acquired from each terminal unit 300 , and image information and sensor sensing received from the terminal information receiving unit 410 . The monitoring information management unit 420 that stores and manages information in the information storage unit 440 and detects the piston operation of the pneumatic cylinder from the real-time image information provided from the terminal information receiving unit 410, and the detection result and the terminal information receiving unit ( 410) with reference to the learning image information registered and stored in the information storage unit 440 for the operation state of the pneumatic cylinder according to the sensor detection information provided from 410) to determine whether the operation of the pneumatic cylinder is abnormal, and the operation of the mechanical device according to the determination result Image information, sensor sensing information, and instrumentation received from the monitoring control unit 430 that judges and monitors and manages each instrument device in the semiconductor manufacturing process and provides the monitoring information to the external notification unit 500, and the terminal information receiver 410 is configured to include an information storage unit 440 including a first storage unit 441 in which operation state information and maintenance information are stored, and a second storage unit 442 in which learning image information is stored.

In the embodiment of the present invention configured as described above, the learning image information generated by learning the image information on the operation of the pneumatic cylinder configured in the mechanical device is interlocked in the semiconductor manufacturing process monitoring system to determine whether the operation of the pneumatic cylinder is abnormal, and the abnormality of the mechanical device to be able to monitor the presence of

In the present invention, a camera and a sensor are installed in each mechanical device in a semiconductor manufacturing process.

The camera 100 is a means for photographing the pneumatic cylinder configured in the mechanical device, and is installed to acquire an image of the piston motion of the pneumatic cylinder.

The camera 100 may be configured to photograph the piston operation of one or more pneumatic cylinders according to a location where the pneumatic cylinders are installed.

That is, considering that the general field of view range of the camera 100 is 71 degrees, the piston operation of the pneumatic cylinder can be photographed within the field of view range of the camera 100 at the position where the camera 100 is mounted.

The sensor 200 is a means for detecting the operating range of the piston, and may be configured as an optical sensor.

The terminal unit 300 may be installed in a position close to the mechanical device, and the image information and sensor detection information provided from the camera 100 and the sensor 200 configured in the corresponding mechanical device are transmitted to the central monitoring and control unit through a wired/wireless network. (400) means to transmit.

The central monitoring and control unit 400 uses the image information and sensor detection information transmitted from each terminal unit 300 to determine whether each mechanical device has an abnormal operation, and to manage the semiconductor manufacturing process system according to the determination result. It is a server tool.

The central monitoring and control unit 400 receives image information and sensor detection information from each terminal unit 300 and uses the artificial intelligence learning information generated by learning the operation of the mechanical device to determine whether or not the mechanical device operates abnormally from the information. As a means for determining and managing, it includes a terminal information receiving unit 410 , a monitoring information management unit 420 , a monitoring control unit 430 , and an information storage unit 440 .

The terminal information receiving unit 410 is a means for receiving image information and sensor detection information from each terminal 300, and each terminal 300 and the terminal information receiving unit 410 are configured through wired and wireless network means. can do.

The monitoring information management unit 420 is a means for storing and managing image information and sensor detection information received from the terminal information receiving unit 410 in the information storage unit 440 in units of each terminal unit 300 .

The monitoring control unit 430 registers real-time image information and sensor detection information provided from the camera 100 and the sensor 200 in the information storage unit 440 and determines the operation of the pneumatic cylinder by referring to the stored artificial intelligence learning image information. It is a means for judging and managing the presence or absence of abnormal operation of the corresponding mechanical device through this.

The monitoring control unit registers the image information and sensor detection information of the real-time pneumatic cylinder provided from the terminal information receiving unit 410 in the information storage unit 440 and determines the operation of the pneumatic cylinder with reference to the stored learning image information 431 ), a monitoring unit that monitors and manages each mechanical device in the semiconductor manufacturing process by generating operation status information of the mechanical device and providing maintenance information that can notify when an abnormal operation occurs to the external notification unit 500 according to the determination result (432) is included.

The operation determination unit 431 recognizes the piston of the pneumatic cylinder from the image information provided from the terminal information receiving unit 410 and detects the recognized operation of the piston. A piston operation pattern generating means for generating a piston operation pattern to contrast with the learning image information according to the pattern information of the learning image information registered in the information storage unit 440, In contrast, it includes an operation determination means for determining the operating state of the pneumatic cylinder.

The piston operation detection means of the operation determination unit 431 recognizes the piston of the pneumatic cylinder and applies a “YOLO” image detection algorithm to detect the operation of the piston.

This is because the operation speed of the pneumatic cylinder is relatively fast compared to other devices, so YOLO (You Only Look Once), which is known to be the fastest among the artificial intelligence object recognition algorithms, is applied for image detection.

3 is a diagram showing the comparison of the object recognition algorithm detection performance index.

In each object recognition target (VOC2007, Picasso, People Art), it can be seen that the recognition speed of YOLO is faster than that of other algorithms (R-CNN, DPM).

The YOLO algorithm is effective because the boundary between objects may be unclear due to the mutual reflection effect caused by lighting and the metal surface of each device in the semiconductor manufacturing process site.

In addition, the monitoring control unit 430 from the computer terminal means 510 of the external notification unit 500 or the connection management means for connection management of the smart phone 520, the connected computer terminal means 510 or the smart phone 520 and inquiry information providing means for providing information stored in the information storage unit 440 according to the input information inquiry request.

In addition, the monitoring control unit 430 computer terminal means 510 or the computer terminal means 510 in response to a request for registration of the learning image information registered in the second storage unit 442 of the information storage unit 440 from the smart phone 520. ) or by registering the learning image information transmitted from the smart phone 520 in the second storage unit 442 may further include a means for updating the learning image information.

The information storage unit 440 includes a first storage unit 441 in which information acquired and generated in real time is stored, and a second storage unit 442 in which learning image information is stored.

The first storage unit 441 is a means for storing image information received from the terminal information receiving unit 410 , sensor detection information, and operation state information of the instrumentation device generated by the monitoring and control unit 430 .

In addition, the second storage unit 442 stores maintenance information to be taken for each operation abnormality information.

The image information and sensor detection information stored in the information storage unit 440 are information transmitted from each terminal unit 300 and include terminal information.

The learning image information of the information storage unit 440 is information generated by performing artificial intelligence learning on the image image information acquired from the pneumatic cylinder installed on the semiconductor manufacturing process, and is information registered in advance in the second storage unit 442 . to be.

The learning image information registered in the second storage unit 442 detects the operation of the piston, and is information obtained by performing artificial intelligence learning on the detected piston operation information, and includes pattern information about the piston operation of the pneumatic cylinder.

The operation of the pneumatic cylinder installed in the mechanical device (vacuum valve, gate valve) in the manufacturing process facility is photographed with a camera (images generated at more than 28 frames per second), and artificial intelligence learning is performed by classifying normal operation and abnormal operation.

4 and 5 are views showing an example of learning to recognize an open/closed object of a pneumatic cylinder (vacuum valve), FIG. 4 is an open state, and FIG. 5 shows a closed state.

6 to 8 are views showing an example of object recognition learning by operation state of a pneumatic cylinder (gate valve), FIG. 6 shows a 20% open state, FIG. 7 shows a 40% closed state, and FIG. 8 shows a 60% closed state .

The segmentation elements to classify the piston motion required in AI learning are as follows.

(One). Motion speed (v) and acceleration (a)

(2). Change in operating speed (delta v)

(3). Load (in kg, N)

With respect to the above classification elements, learning is performed at the final stage of the image detection neural network of the "YOLO" algorithm, that is, when the mean average precision (mAP) is determined.

9 and 10 show an example of a pneumatic cylinder piston operation pattern for performing learning.

9 is a view showing a normal operation pattern of the piston, and showing a state in which the acceleration (a) is a constant speed within a certain time.

10 is a view showing an abnormal operation pattern of the piston.

Artificial intelligence learning applies a convolutional neural network (CON).

CNN is an artificial neural network model using convolution. As a model specialized in multidimensional array data processing, it is actively used in the image field (RGB channel is a three-dimensional array) because it can transmit the array data information to the next layer while maintaining the spatial information of the data.

In particular, it was applied because it clearly detects the features of the image.

11 shows a CNN.

CNN detects image features while repeatedly performing a convolution operation using a filter.

(a). Convolutional layer that extracts image features

(b). A pooling layer that extracts and collects image features

(c). Fully connected layer to classify images

In this way, it is composed of parts that extract features of images and classify images.

The external notification unit 500 receives monitoring information about the instrument device from the central monitoring and control unit 400, that is, operation state information and maintenance information of the instrument device, and a notification means such as a display or an alarm sound, and a computer terminal means of an administrator. 510 , and a smartphone 520 .

The external notification unit 500 may be configured with a notification means such as an electric billboard, and may be installed in the manufacturing process site so that a manager or an operator can easily check it in the field.

The computer terminal means 510 and the smart phone 520 receive the operation state information and maintenance information of the mechanical device from the central monitoring and control unit 400 and provide them to the administrator, and the administrator accesses the central monitoring and control unit 400 to centralize A monitoring program is installed so that the operation state information and maintenance information of the mechanical device can be provided through the monitoring and control unit 400 .

The monitoring program is a connection management process for managing a connection with the central monitoring and control unit 400, accessing the central monitoring and control unit 400 to view the operation state information and pre-generated maintenance information stored in the information storage unit 440. It includes an inquiry management process and a notification management process that generates and provides notification information to users according to maintenance information transmitted from the central monitoring and control unit.

In addition, the monitoring program accesses the central monitoring and control unit 400 to request an update of the learning image information, and transmits the learning image information to the central monitoring and control unit 400 to register the learning image in the information storage unit 440 . and an information update process.

In addition, a proximity user identification unit for identifying the smart phone 520 connected to the terminal unit 300 through the proximity communication means and the proximity communication means and transmitting the identified smart phone 520 information to the central monitoring control unit 400 . Including, the monitoring control unit 430 of the central monitoring control unit 400 according to the user information provided from the terminal unit 300 through the terminal information receiving unit 410 to the corresponding smart phone 520, the terminal unit 300 ) to provide operation state information and pre-stored maintenance information of the mechanical device corresponding to the It can be configured to include a user monitoring means in proximity.

The proximity communication means may be configured with Bluetooth.

The semiconductor manufacturing process monitoring method using the operation image information of the pneumatic cylinder of the present invention made in the central monitoring and control unit 400,

When image information and sensor detection information are received from the terminal unit 300, the piston operation of the pneumatic cylinder is recognized, the piston operation information is detected, and the identification information of the terminal unit 300 is checked and stored in the information storage unit 440. The operation detection process of the pneumatic cylinder,

An operation abnormality determination process of determining the operation abnormality of the mechanical device corresponding to the terminal unit 300 with reference to the learning image information stored in the information storage unit 440 for the detected piston operation information;

The operation state information determined through the operation abnormality determination process is stored in the information storage unit 440, and when an operation abnormality occurs, maintenance information is generated and provided to the external notification unit 500, and the information storage unit 440 The monitoring management process stored in the

When a user is connected and the user's access request information is input, the access information stored in the information storage unit 440 is checked and access is permitted, and monitoring information is viewed from the connected user's computer terminal means 510 or smart phone 520 . When a request is input, monitoring information is provided that provides operation state information or pre-stored maintenance information stored in the information storage unit 440 to the user's computer terminal means 510 or smart phone 520 according to the input monitoring information inquiry request comprised of the process.

In addition, when the user's smartphone 520 access is confirmed from the terminal unit 300, the corresponding terminal unit identification information is checked, and operation state information and pre-stored maintenance information of the corresponding terminal unit 300 stored in the information storage unit 440 are confirmed. It is configured to further include a process of providing monitoring information for a user in proximity to providing the access-confirmed smart phone 520 .

In addition, when a learning image information update request is input from the connected computer terminal means 510 or smart phone 520, the interface means for transmitting the learning image information to be updated is transferred to the computer terminal means 510 or the smart phone 520. The process of providing, and storing the learning image information input through the interface means provided to the computer terminal means 510 or the smart phone 520 in the second storage unit 442 of the information storage unit 440 to store the learning image information It is configured by further including the process of updating.

In the monitoring information providing process, the operation state information provided to the external notification unit 500 includes mechanical device information, the presence or absence of an operation state abnormality, and operation abnormality information and image information indicating the operation state when an operation abnormality occurs.

The operation of the embodiment of the present invention having such a configuration will be described as follows.

The terminal unit 300 installed for each mechanical device in the manufacturing process transmits the piston image information of the pneumatic cylinder acquired through the camera 100 to the central monitoring control unit 400, and the terminal information of the central monitoring control unit 400 The receiving unit 410 receives image information and sensor sensing information transmitted from the terminal unit 300 connected through a wired/wireless network.

When the image information and sensor detection information of the terminal unit 300 are received from the terminal information receiving unit 410 as described above, the monitoring information management unit 420 stores the received image information and sensor detection information in the information storage unit 440 of the terminal unit. It is stored in units of (300).

On the other hand, the monitoring control unit 430 refers to the learning image information stored in the information storage unit 440 with respect to the image information received from the terminal information receiving unit 410 and the sensor detection information to determine whether there is an abnormality in the operation of the mechanical device to determine the operating state The information is stored in the information storage unit 440 .

The operation determination unit 431 of the monitoring and control unit 430 recognizes the piston of the pneumatic cylinder, detects the operation of the piston, and determines whether the operation of the mechanical device is abnormal by referring to the learning image information to determine the operation abnormality of the pneumatic cylinder. will judge

At this time, if it is determined that the operation state is abnormal, maintenance information is generated with reference to the maintenance information for the learning image information stored in the information storage unit 440 and transmitted together with the operation state information to the external notification unit 500 .

The manager can check through the external notification unit 500 made of a large display means such as an electric billboard, or can check the current state of the appliance through the computer terminal means 510 or the smart phone 520 .

The manager (user) accesses the central monitoring and control unit 400 through the computer terminal means 510 or the smart phone 520, and the operation state information stored in the information storage unit 440 and the previously created and stored maintenance that has been transmitted to the user You can check the information by inquiring.

The monitoring unit 432 of the monitoring control unit 430 provides an interface means so that the information stored in the information storage unit 440 can be inquired by the connected computer terminal means 510 or smart phone 520, and the user can centrally monitor By accessing the control unit 400, it is possible to check the operation state information or maintenance information of the mechanical device.

And when the smartphone 520 is connected to the adjacent terminal unit 300, it transmits identification information to the terminal unit 300, and the terminal unit 300 receives identification information from the connected smartphone 520, the corresponding The identification information is transmitted to the monitoring control unit 430 of the central monitoring and control unit 400, and the monitoring and control unit 430 converts the operation state information and pre-stored maintenance information for the corresponding terminal unit 300 to the smart identification information corresponding to the identification information. It is transmitted to the phone 520 so that the manager can check the information of the mechanical device adjacent to his location.

Meanwhile, Figure 2 shows another embodiment of the present invention.

The embodiment of the present invention uses the learning image information generated through artificial intelligence learning for the pneumatic cylinder of the mechanical device installed in the semiconductor manufacturing process in conjunction with the system, whereas another embodiment of the present invention is the implementation of the present invention As in the example, learning image information is used, but information learned in the manufacturing process site can be continuously stored while the system of the present invention is installed, so that a system suitable for the semiconductor manufacturing process site in which the system is installed can be configured .

Another embodiment of the semiconductor manufacturing process monitoring system using the operation image information of the pneumatic cylinder of the present invention will be described with reference to the embodiments shown in FIGS. 2 and 13 of the accompanying drawings.

A camera 100 installed in each mechanical device in the semiconductor manufacturing facility to acquire operation image information of the pneumatic cylinder in the mechanical device, and a sensor installed in each mechanical device in the semiconductor manufacturing facility to detect the operating state of the pneumatic cylinder ( 200), and a terminal unit 300 that provides the image information obtained from the camera 100 installed for each instrument and sensor detection information obtained from the sensor 200 to the central monitoring and control unit 400, and each terminal unit A central monitoring and control unit 400 that monitors and manages the operation of the mechanical device in comparison with the learning image information learned by artificial intelligence according to the operating state of the pneumatic cylinder by receiving the image information and sensor detection information transmitted from the 300, and central monitoring It is configured to include an external notification unit 500 for notifying an administrator (user) by receiving the operation state information and maintenance information from the control unit 400 .

The central monitoring and control unit 400 includes a terminal information receiving unit 410 for receiving the image information and sensor sensing information acquired from each terminal unit 300 , and image information and sensor sensing received from the terminal information receiving unit 410 . The monitoring information management unit 420 that stores and manages information in the information storage unit 440 and detects the piston operation of the pneumatic cylinder from the real-time image information provided from the terminal information receiving unit 410, and the detection result and the terminal information receiving unit ( According to the sensor detection information provided from 410), the operation state of the pneumatic cylinder is compared with the learning image information registered and stored in the information storage unit 440 to determine whether the operation of the pneumatic cylinder is abnormal, and the operation of the mechanical device according to the determination result The monitoring and control unit 430 that judges and monitors and manages each mechanical device in the semiconductor manufacturing process and provides monitoring information to the external notification unit 500, and the monitoring and control unit receives the operation state of the pneumatic cylinder determined from the real-time image information from the monitoring and control unit. Artificial intelligence learning unit 450 that generates and stores learning image information in the information storage unit, image information received from the terminal information receiving unit 410, sensor detection information, operation state information of the mechanical device, maintenance information is stored The first storage unit 441 is configured to include an information storage unit 440 including a second storage unit 442 in which the learning image information and the field learning image information generated by the artificial intelligence learning unit 450 are stored. .

Another embodiment of the present invention configured as described above is to learn image information about the operation of a pneumatic cylinder at the site where the system configured in the instrument apparatus is installed to determine the operation of the instrument apparatus appropriately for the site. The same part will be described by omitting the operation process.

The central monitoring and control unit 400 uses the image information and sensor detection information transmitted from each terminal unit 300 to determine whether each mechanical device has an abnormal operation, and a server for managing the manufacturing process system according to the determination result As a means, the image information acquired from the camera 100 and the sensor detection information of the sensor 200 are received from each terminal unit 300 and stored in the information storage unit 440 and registered using the registered artificial intelligence learning information. As a means to determine and manage the abnormal operation of the mechanical device from It further includes an artificial intelligence learning unit 450 for generating.

The monitoring control unit 430 registers the image information and sensor detection information of the real-time pneumatic cylinder provided from the terminal information receiving unit 410 in the information storage unit 440 and determines the operation of the pneumatic cylinder with reference to the stored learning image information. The unit 431, according to the determination result, generates operation state information of the mechanical device and provides maintenance information that can notify when an abnormal operation occurs to the external notification unit 500 to monitor and manage each mechanical device in the semiconductor manufacturing process. It is configured to include a monitoring unit 432.

The operation determination unit 431 recognizes the piston of the pneumatic cylinder from the image information provided from the terminal information receiving unit 410 and detects the recognized operation of the piston. Piston operation pattern generating means for generating a piston operation pattern to contrast with the learning image information according to the pattern information of the learning image information registered in the information storage unit 440 and the piston operation pattern generated by the In contrast, it includes an operation determination means for determining the operating state of the pneumatic cylinder.

The artificial intelligence learning unit 450 is a means for generating field learning image information by receiving the motion detection information and the piston motion pattern information from the motion determination unit 431 according to a predetermined algorithm and performing artificial intelligence learning, It is a means to create artificial intelligence learning image information suitable for the field while the system is installed.

The artificial intelligence learning unit 450 is configured as a means for generating field learning image information generated by performing artificial intelligence learning according to a set algorithm, and field learning including learning information generated with reference to the user's input information means for generating image information.

In addition, the monitoring control unit 430 allows the user to inquire the information stored in the computer terminal means 510 of the external notification unit 500 or the connection management means for connection management of the smart phone 520 and the information storage unit 440 . Inquiry information providing means for providing an interface means to the connected computer terminal means 510 and smart phone 520 and providing information stored in the information storage unit according to an information inquiry request input through the interface means, and artificial intelligence learning It is configured to include a selection means so that the manager (user) can select whether to execute the artificial intelligence learning operation of the unit 450 and a learning control means for providing a supervised learning input interface so that the user can input learning information.

In addition, the monitoring control unit 430 computer terminal means 510 in response to a request for registration of learning image information registered in the second storage unit 442 of the information storage unit 440 from the computer terminal means 510 or the smart phone 520. ) or by registering the learning image information transmitted from the smart phone 520 in the second storage unit 442 may further include a means for updating the learning image information.

The information storage unit 440 includes a first storage unit 441 in which information acquired and generated in real time is stored, and a second storage unit 442 in which learning image information and field learning image information are stored.

The first storage unit 441 is a means for storing image information received from the terminal information receiving unit 410 , sensor detection information, and operation state information of the instrumentation device generated by the monitoring and control unit 430 .

In addition, the second storage unit 442 is a means for storing learning image information and field learning image information, which is learning information acquired in the field through the artificial intelligence learning unit 450, and maintenance information to be taken with respect to operation abnormality information. is saved

The image information and sensor detection information stored in the information storage unit 440 are information transmitted from each terminal unit 300 and include terminal information.

The learning image information of the information storage unit 440 is information generated by performing artificial intelligence learning on the image image information acquired from the pneumatic cylinder installed on the semiconductor manufacturing process, and is information registered in advance in the second storage unit 442 . to be.

The field learning image information stored in the second storage unit 442 is information performed by artificial intelligence learning through the artificial intelligence learning unit 450, and includes pattern information about the piston operation of the pneumatic cylinder.

The external notification unit 500 receives monitoring information about the instrument device from the central monitoring and control unit 400, that is, operation state information and maintenance information of the instrument device, and a notification means such as a display or an alarm sound, and a computer terminal means of an administrator. 510 , and a smartphone 520 .

The external notification unit 500 may be configured with a notification means such as an electric billboard, and may be installed in the manufacturing process site so that a manager or an operator can easily check it in the field.

The computer terminal means 510 and the smart phone 520 receive the operation state information and maintenance information of the mechanical device from the central monitoring and control unit 400 and provide them to the administrator, and the administrator accesses the central monitoring and control unit 400 to centralize A monitoring program is installed so that the operation state information and maintenance information of the mechanical device can be provided through the monitoring and control unit 400 .

The monitoring program is a connection management process for managing a connection with the central monitoring and control unit 400, accessing the central monitoring and control unit 400 to view the operation state information and pre-generated maintenance information stored in the information storage unit 440. The manager (user) selects whether to execute the inquiry management process, the notification management process that generates and provides notification information to the user according to the maintenance information transmitted from the central monitoring and control unit, and the artificial intelligence learning operation of the artificial intelligence learning unit 450 It is configured to include a field learning control process that provides a selection means for enabling and a supervised learning input interface means for a user to input learning information.

In addition, the monitoring program accesses the central monitoring and control unit 400 to request an update of the learning image information, and transmits the learning image information to the central monitoring and control unit 400 to register the learning image in the information storage unit 440 . and an information update process.

Further comprising a proximity user identification unit for identifying the smart phone 520 connected to the terminal unit 300 through the proximity communication means and the proximity communication means and transmitting the identified smart phone information to the central monitoring control unit 400, The monitoring control unit 430 of the central monitoring and control unit 400 corresponds to the terminal unit 300 with the corresponding smart phone 520 according to the user information provided from the terminal unit 300 through the terminal information receiving unit 410 . Closed user monitoring means for providing monitoring information of the adjacent mechanical device according to the location of the user, which provides the operation state information and pre-stored maintenance information of the mechanical device, and stores the history in the information storage unit 440; It can be configured to include

The proximity communication means may be configured with Bluetooth.

In another embodiment of the present invention made in the central monitoring and control unit 400, the semiconductor manufacturing process monitoring method using the operation image information of the pneumatic cylinder,

When image information and sensor detection information are received from the terminal unit 300, the piston operation of the pneumatic cylinder is recognized, the piston operation information is detected, and the identification information of the terminal unit 300 is checked and stored in the information storage unit 440. The operation detection process of the pneumatic cylinder,

Field learning image information generated by performing an artificial intelligence learning process according to a predetermined algorithm for the operation information of the pneumatic cylinder detected through the operation detection process of the pneumatic cylinder to generate field learning image information and storing it in the information storage unit 440 process and

The operation information of the piston detected through the operation detection process of the pneumatic cylinder is referred to the learning image information and field learning image information stored in the information storage unit 440 to determine whether the mechanical device corresponding to the terminal unit 300 has an abnormal operation. The process of judging whether there is an abnormality in the operation, and

The operation state information determined through the operation abnormality determination process is stored in the information storage unit 440, and when an operation abnormality occurs, maintenance information is generated and provided to the external notification unit 500, and the information storage unit 440 The monitoring management process stored in the

When a user is connected and the user's access request information is input, the access information stored in the information storage unit 440 is checked and access is permitted, and monitoring information is viewed from the connected user's computer terminal means 510 or smart phone 520 . When a request is input, monitoring information is provided that provides operation state information or pre-stored maintenance information stored in the information storage unit 440 to the user's computer terminal means 510 or smart phone 520 according to the input monitoring information inquiry request comprised of the process.

The field learning image information generating process is configured to further include, when the user's supervised learning information is input, generating field learning image information according to the supervised learning information input by the user.

In addition, the field learning image generation process enables it to be determined whether or not to execute the field learning image generation process according to the user's selection input information.

In addition, when the user's smartphone 520 access is confirmed from the terminal unit 300, the corresponding terminal unit identification information is checked, and operation state information and pre-stored maintenance information of the corresponding terminal unit 300 stored in the information storage unit 440 are confirmed. It is configured to further include a process of providing monitoring information for a user in proximity to providing the access-confirmed smart phone 520 .

In addition, when a learning image information update request is input from the connected computer terminal means 510 or the smart phone 520, the interface means for transmitting the learning image information to be updated is the computer terminal means 510 or the smart phone 520. The process of providing, and storing the learning image information input through the interface means provided to the computer terminal means 510 or the smart phone 520 in the second storage unit 442 of the information storage unit 440 to store the learning image information It is configured by further including the process of updating.

The operation state information provided to the external notification unit 500 includes mechanical device information, the presence or absence of an operation state abnormality, and operation abnormality information and image information indicating the operation state when an operation abnormality occurs.

The operation of another embodiment of the system of the present invention having such a configuration will be described as follows.

The terminal unit 300 installed for each mechanical device in the manufacturing process transmits the piston image information of the pneumatic cylinder acquired through the camera 100 to the central monitoring control unit 400, and the terminal information of the central monitoring control unit 400 The receiving unit 410 receives image information and sensor sensing information transmitted from the terminal unit 300 connected through a wired/wireless network.

When the image information and sensor detection information of the terminal unit 300 are received from the terminal information receiving unit 410 as described above, the monitoring information management unit 420 stores the received image information and sensor detection information in the information storage unit 440 of the terminal unit. It is stored in units of (300).

On the other hand, the monitoring control unit 430 refers to the learning image information stored in the information storage unit 440 with respect to the image information received from the terminal information receiving unit 410 and the sensor detection information to determine whether there is an abnormality in the operation of the mechanical device to determine the operating state The information is stored in the information storage unit 440 .

The operation determination unit 431 of the monitoring and control unit 430 recognizes the piston of the pneumatic cylinder, detects the operation of the piston, and determines whether the operation of the mechanical device is abnormal by referring to the learning image information to determine the operation abnormality of the pneumatic cylinder. will judge

At this time, in judging the abnormal operation of the pneumatic cylinder, the learning image information and field learning image information as described above are referred to, and the field learning image information is learned by the artificial intelligence learning unit 450 in a situation where the system is operating. information that has been acquired.

The field learning image information is information generated by performing artificial intelligence learning according to an algorithm determined for the detected operation of the piston in the artificial intelligence learning unit 450 .

In addition, such field learning image information may be generated by reflecting supervised learning information to the user.

Here, when it is determined that the operation of the mechanical device is abnormal in the field, the administrator can input operation abnormality information for the corresponding operation information, and when such user guidance learning information is input, the monitoring and control unit 430 checks the information and artificial intelligence learning This information is provided to the unit 450, and when user guidance learning information is input, the artificial intelligence learning unit 450 sets the corresponding motion information to a motion abnormality, generates learning image information according to this, and stores the information in the information storage unit 440. 2 Register and store the field learning image information generated in the storage 442 .

When it is determined that the operation determination unit 431 of the monitoring and control unit 430 is in an abnormal operating state, the maintenance information is generated with reference to the maintenance information for the learning image information stored in the information storage unit 440, and the external notification unit 500 ) along with the operation status information.

The manager can check through the external notification unit 500 made of a large display means such as an electric sign board, or can check the current state of the instrument through the computer terminal means 510 or the smart phone 520 .

The manager (user) accesses the central monitoring and control unit 400 through the computer terminal means 510 or the smart phone 520, and the operation state information stored in the information storage unit 440 and the previously generated and stored maintenance that was transmitted to the user You can check the information by inquiring.

The monitoring unit 432 of the monitoring control unit 430 provides an interface means so that the information stored in the information storage unit 440 can be inquired by the connected computer terminal means 510 or smart phone 520, and the user can centrally monitor By accessing the control unit 400, it is possible to check the operation state information or maintenance information of the mechanical device.

In addition, as described above, the user selects whether to perform the artificial intelligence learning process of the artificial intelligence learning unit 450 through the computer terminal means 510 or the smart phone 520 (on/off) to display the field learning image. It is possible to set whether to execute or not, and by inputting the user's instructional learning information for the piston operation information of the pneumatic cylinder, it is possible to generate learning image information suitable for the field.

As described in the embodiment of the present invention, the manager can automatically receive and view operation state information and previously generated maintenance information of the corresponding mechanical device at a location close to the terminal unit 300 .

As described above, the present invention allows to determine the operation of the mechanical device by judging the operation of the pneumatic cylinder using the operation pattern information more accurately than the learning image information about the piston operation of the pneumatic cylinder installed on the semiconductor manufacturing process, A more effective monitoring system can be built without equipment.

In addition, the present invention allows to add learning information at the site where the system is installed in addition to the learning image information that is made in advance and linked to the semiconductor manufacturing process monitoring system, so that a semiconductor manufacturing process monitoring system more suitable for the field can be built.

Claims (14)

A camera 100 installed in each mechanical device in the semiconductor manufacturing facility to acquire operation image information of the pneumatic cylinder in the mechanical device, and a sensor installed in each mechanical device in the semiconductor manufacturing facility to detect the operating state of the pneumatic cylinder ( 200), and a terminal unit 300 that provides image information obtained from the camera 100 installed for each instrument and sensor detection information obtained from the sensor 200 to the central monitoring and control unit 400, and the terminal unit ( The central monitoring and control unit 400, which receives the image information and sensor detection information transmitted from 300), and monitors and manages the operation of the mechanical device by referring to the learning image information learned by artificial intelligence according to the operating state of the pneumatic cylinder, and the central monitoring and control unit It is configured to include an external notification unit 500 for notifying the administrator (user) by receiving the operation state information from the 400,
The central monitoring and control unit 400 includes a terminal information receiving unit 410 for receiving the image information and sensor sensing information acquired from the terminal unit 300 , and image information and sensor sensing information received from the terminal information receiving unit 410 . Detects the piston operation of the pneumatic cylinder from the real-time image information provided from the monitoring information management unit 420 and the terminal information receiving unit 410 to store and manage the data in the information storage unit 440, and the detection result and the terminal information receiving unit 410 ) with reference to the learning image information registered and stored in the information storage unit 440 for the operation state of the pneumatic cylinder according to the sensor detection information provided from The monitoring control unit 430 that monitors and manages each mechanical device in the semiconductor manufacturing process and provides monitoring information to the external notification unit 500, the image information received from the learning image information and the terminal information receiving unit 410, sensor detection information , A semiconductor manufacturing process monitoring system using operation image information of a pneumatic cylinder, characterized in that it comprises an information storage unit 440 in which operation state information of the mechanical device is stored. The method of claim 1, wherein the monitoring and control unit 430 registers the image information and sensor sensing information of the real-time pneumatic cylinder provided from the terminal information receiving unit 410 in the information storage unit 440 and the stored learning image information of the pneumatic cylinder. The operation determination unit 431 that determines the operation, generates operation state information of the mechanical device according to the determination result, and provides maintenance information that can notify when an abnormal operation occurs to the external notification unit 500 to each in the semiconductor manufacturing process It is configured to include a monitoring unit 432 that monitors and manages the instrumentation,
The operation determination unit 431 recognizes the piston of the pneumatic cylinder from the image information provided from the terminal information receiving unit 410 and detects the recognized operation of the piston. A piston operation pattern generating means for generating a piston operation pattern to contrast with the learning image information according to Comprising an operation determination means for determining the operating state of the pneumatic cylinder in preparation, the information storage unit 440 includes a first storage unit 441 in which information acquired and generated in real time is stored, and learning image information is stored It includes a second storage unit 442,
The first storage unit 441 stores the image information received from the terminal information receiving unit 410, sensor detection information, and operation state information of the instrumentation device generated by the monitoring control unit 430,
The second storage unit 442 stores training image information and compensation information to be taken with respect to each operation abnormality information,
The image information and sensor detection information in the first storage unit 441 are information transmitted from each terminal unit 300 and include terminal information,
The learning image information of the second storage unit 442 is information generated by performing artificial intelligence learning on the image image information acquired from the pneumatic cylinder installed on the semiconductor manufacturing process, and detects the motion of the piston, and the detected piston A semiconductor manufacturing process monitoring system using motion image information of a pneumatic cylinder, characterized in that the motion information is information obtained by artificial intelligence learning, and includes pattern information about the piston motion of the pneumatic cylinder. According to claim 1 or 2, wherein the monitoring and control unit 430 is a computer terminal means 510 of the external notification unit 500 or a connection management means for connection management of the smart phone 520, the connected computer terminal means It is configured to include inquiry information providing means for providing information stored in the information storage unit 440 according to the information inquiry request input from the 510 or smart phone 520,
The external notification unit 500 connects to the central monitoring and control unit 400 through a wired/wireless network, and receives and displays monitoring information for the instrument from the central monitoring and control unit 400, that is, operation state information and maintenance information of the instrument. Or it is configured to include a notification means such as an alarm sound, the administrator's computer terminal means 510, and a smart phone 520,
The computer terminal means 510 and the smart phone 520 receive the operation state information and maintenance information of the mechanical device from the central monitoring and control unit 400 and provide them to the administrator, and the administrator accesses the central monitoring and control unit 400 to centralize A monitoring program is installed so that the operation state information and maintenance information of the mechanical device can be provided through the monitoring and control unit 400,
The monitoring program is a connection management process for managing a connection with the central monitoring and control unit 400, accessing the central monitoring and control unit 400 to view the operation state information and pre-generated maintenance information stored in the information storage unit 440. A semiconductor manufacturing process monitoring system using operation image information of a pneumatic cylinder, characterized in that it includes a query management process and a notification management process that generates notification information according to the maintenance information transmitted from the central monitoring and control unit and provides it to the user. The method according to claim 3, wherein the monitoring and control unit 430 responds to a request for registration of learning image information registered in the second storage unit 442 of the information storage unit 440 from the computer terminal means 510 or the smart phone 520. For registering the learning image information transmitted from the computer terminal means 510 or the smart phone 520 to the second storage unit 442, it further comprises an update means of the learning image information,
The monitoring program accesses the central monitoring and control unit 400 to request an update of the learning image information, and transmits the learning image information to the central monitoring and control unit 400 to register the learning image information in the information storage unit 440 . The semiconductor manufacturing process monitoring system using the operation image information of the pneumatic cylinder, characterized in that it further comprises an update process. The central monitoring and control unit 400 according to claim 1 or 2, wherein the terminal unit 300 identifies a smart phone 520 connected through a proximity communication means and a proximity communication means and transmits the identified smart phone 520 information. ) further comprising a proximity user identification unit to transmit to, the monitoring control unit 430 of the central monitoring and control unit 400 according to the user information provided from the terminal unit 300 through the terminal information receiving unit 410 corresponding to the smart phone According to the location of the user who provides the operation state information and pre-stored maintenance information of the mechanical device corresponding to the terminal unit 300 to 520 and stores the history in the information storage unit 440, A semiconductor manufacturing process monitoring system using operation image information of a pneumatic cylinder, characterized in that it further comprises a user monitoring means close to providing monitoring information of the instrument device. The method of claim 3, wherein the central monitoring control unit 400 receives the operating state of the pneumatic cylinder determined from the real-time image information from the monitoring control unit 430 to generate field learning image information and store it in the information storage unit 440 Consists of further comprising an artificial intelligence learning unit (450),
The artificial intelligence learning unit 450 receives the motion detection information and the piston motion pattern information from the motion determination unit 431 according to a predetermined algorithm, and performs artificial intelligence learning to generate field learning image information. becomes,
It is configured to further include means for generating field-learning image information, including user-guided learning information generated with reference to the user's input information,
The monitoring control unit 430 is configured to include a learning control means for providing a supervised learning input interface so that the user can input learning information,
The computer terminal means 510 or the monitoring program of the smart phone 520 is pneumatic, characterized in that it comprises a field learning control process that provides a supervised learning input interface means allowing the user to input the supervised learning information A semiconductor manufacturing process monitoring system using cylinder motion image information. According to claim 6, wherein the monitoring and control unit 430 further comprises a selection means so that an administrator (user) can select whether to execute the artificial intelligence learning operation of the artificial intelligence learning unit 450,
The monitoring program using the operation image information of the pneumatic cylinder, characterized in that it further comprises a process of providing a selection means so that the manager (user) can select whether to execute the artificial intelligence learning operation of the artificial intelligence learning unit 450 Semiconductor manufacturing process monitoring system. When image information and sensor detection information are received from the terminal unit 300, the piston operation of the pneumatic cylinder is recognized, the piston operation information is detected, and the identification information of the terminal unit 300 is checked and stored in the information storage unit 440. The operation detection process of the pneumatic cylinder,
An operation abnormality determination process of determining the operation abnormality of the mechanical device corresponding to the terminal unit 300 with reference to the learning image information stored in the information storage unit 440 for the detected piston operation information;
The operation state information determined through the operation abnormality determination process is stored in the information storage unit 440, and when an operation abnormality occurs, maintenance information is generated and provided to the external notification unit 500, and the information storage unit 440 The monitoring management process stored in the
When a user is connected and the user's access request information is input, the access information stored in the information storage unit 440 is checked and access is permitted, and monitoring information is viewed from the connected user's computer terminal means 510 or smart phone 520 . When a request is input, monitoring information is provided that provides operation state information or pre-stored maintenance information stored in the information storage unit 440 to the user's computer terminal means 510 or smart phone 520 according to the input monitoring information inquiry request A semiconductor manufacturing process monitoring method using operation image information of a pneumatic cylinder, characterized in that it comprises a process. The method according to claim 8, wherein when the user's smartphone 520 access is confirmed from the terminal unit 300, the corresponding terminal unit identification information is checked and the operation state information of the corresponding terminal unit 300 stored in the information storage unit 440 and A semiconductor manufacturing process monitoring method using operation image information of a pneumatic cylinder, characterized in that it further comprises a process of providing monitoring information for a user in proximity to provide the pre-stored maintenance information to the connected smartphone 520. The method of claim 8, wherein when a request for updating learning image information is input from the connected computer terminal means 510 or smart phone 520, the interface means for transmitting the learning image information to be updated is connected to the computer terminal means 510 or smart phone. The process of providing to the phone 520 and the learning image information input through the interface means provided to the computer terminal means 510 or the smart phone 520 are stored in the second storage unit 442 of the information storage unit 440 . The semiconductor manufacturing process monitoring method using the operation image information of the pneumatic cylinder, characterized in that it further comprises the process of updating the learning image information. [10] The method of claim 8 or 9, wherein the operation state information provided to the external notification unit 500 or the smart phone 520 in the monitoring information providing process or in the monitoring information providing process for a user close to the user is information on the device and the operation. A semiconductor manufacturing process monitoring method using operation image information of a pneumatic cylinder, characterized in that it includes information on the operation abnormality and image information indicating the operation state when there is an abnormality and when an operation abnormality occurs. According to claim 8, wherein the artificial intelligence learning process is performed according to a predetermined algorithm for the operation information of the pneumatic cylinder detected through the operation detection process of the pneumatic cylinder to generate field learning image information and stored in the information storage unit (440) It is configured to further include a field learning image information generation process,
In the operation abnormality determination process, the operation information of the piston detected through the operation detection process of the pneumatic cylinder refers to the learning image information and the field learning image information stored in the information storage unit 440. The mechanism corresponding to the terminal unit 300 A semiconductor manufacturing process monitoring method using operation image information of a pneumatic cylinder, characterized in that it comprises the step of determining whether there is an abnormality in the operation of the device. The pneumatic cylinder according to claim 12, wherein the field learning image information generating process further comprises the step of generating field learning image information according to the guidance learning information input by the user when the user's guidance learning information is input. A semiconductor manufacturing process monitoring method using the motion image information of The semiconductor manufacturing process using operation image information of a pneumatic cylinder according to claim 12, wherein in the field learning image information generating process, it is possible to determine whether to execute the field learning image generating process according to a user's selection input information. monitoring method.

Images