TWI833168B - Method of diagnosing abnormalities - Google Patents

Abstract

An abnormal diagnosis method is applied to a machine. The machine has a processing room for processing objects. The method includes: capturing an image related to the process room by a photography device of the machine; preprocessing the image; transmitting the preprocessed image to a determination system outside the machine; determining, by the determination system, whether the preprocessed image has a feature; and issuing an alarm from an alarm module according to the determination result of the determination system.

Description

abnormal diagnosis method

The present disclosure relates to an anomaly diagnosis method, particularly an anomaly diagnosis method for a machine.

The current method of detecting machine anomalies in the semiconductor industry generally uses manual regular measurements. Manual measurement takes at least one day, and the machine cannot be operated during the measurement process, which will further affect the overall production efficiency of the machine. At the same time, the measurement time may be delayed, which may increase the probability of product scrapping or product yield decline. However, with the high degree of automation in the semiconductor industry, investment in labor costs has dropped, and there is a real lack of manpower to perform regular measurements.

Therefore, how to come up with an abnormality diagnosis method that can solve the above problems is one of the problems that the industry is currently eager to invest in research and development resources to solve.

In view of this, one purpose of the present disclosure is to propose an anomaly diagnosis method that can solve the above problems.

This disclosure relates to an abnormality diagnosis method, which is applied to machines. The machine has a processing space for processing objects. The abnormality diagnosis method includes: capturing images of the processing space through the camera device of the machine; performing image preprocessing on the images; transmitting the processed images to a judgment system outside the machine; and judging the processed images through the judgment system Whether it has characteristics; and an alarm is issued through the alarm module based on the judgment result of the judgment system.

In some current implementations, image preprocessing includes image sharpening.

In some current implementations, when the determination system determines that the processed image has characteristics, it determines that the image is abnormal and issues an alarm through the alarm module.

In some current implementations, the step of determining whether the processed image has characteristics through the determination system includes: comparing the processed image and the comparison image, where the comparison image is a normal state image.

In some current implementations, the feature includes a mechanism configuration of the processing space in the image that is different from a mechanism configuration of the processing space in the comparison image, and the alarm issued by the alarm module includes a mechanism abnormality detection result.

In some current implementations, the comparison image includes a pipeline, and the comparison image defines the liquid level in the pipeline in the image, the feature includes the image and the liquid level position of the comparison image is different, and when the liquid level is located inside the pipeline When, the alarm issued by the alarm module includes the pipeline back suction detection results.

In some current implementations, the comparison image includes a pipeline, the feature includes bubbles inside the pipeline in the processed image, and when there are bubbles inside the pipeline, the alarm issued by the alarm module includes the pipeline splash detection result.

In some current implementations, the comparison image includes a pipeline, the feature includes liquid flowing out of the pipeline in the processed image, and when the liquid flows out of the pipeline, the alarm issued by the alarm module includes a pipeline drip detection result or a pipeline splash. Detection results.

In some current implementations, when the liquid travels along a linear trajectory, the alarm issued by the alarm module includes the pipeline drip detection result.

In some current implementations, when the liquid travels along a non-linear trajectory, the alarm issued by the alarm module includes the pipeline splash detection result.

In summary, in the abnormality diagnosis method disclosed in the present disclosure, the status of the internal processing space of the machine can be remotely monitored and collected in real time through a photography device. In addition, with the assistance of the judgment system, abnormal conditions can be classified without relying on manpower, thereby improving the maintenance efficiency of personnel and reducing the labor cost of monitoring machines.

These and other aspects of the present disclosure will become apparent from the following description of preferred embodiments taken in conjunction with the accompanying drawings, but may be made without departing from the spirit and scope of the novel concepts of the present disclosure. changes and modifications.

The following disclosure will now be described more fully herein with reference to the accompanying drawings, in which some exemplary embodiments are shown. The present disclosure may be implemented in different forms and should not be limited by the embodiments mentioned below. However, these embodiments are provided to facilitate a more complete understanding of the disclosure and to fully convey the scope of the disclosure to those skilled in the art. The same reference numbers will be used throughout the text to refer to similar elements.

Figure 1 is a flowchart illustrating an abnormality diagnosis method M1 according to an embodiment of the present disclosure. Please refer to Figure 1, an abnormality diagnosis method M1 is applied to the machine 100. The machine 100 has a processing space 110 for processing objects. The abnormality diagnosis method M1 includes: capturing an image of the processing space through the camera device of the machine (step S100); performing image preprocessing on the image (step S200); transmitting the processed image to a judgment system outside the machine (step S200). Step S300); determine whether the processed image has features through the determination system (step S400); and issue an alarm through the alarm module according to the determination result of the determination system (step S500).

Figure 2 is a schematic diagram illustrating the machine 100 and the processing space 110 according to an embodiment of the present disclosure. Please refer to Figure 1 and Figure 2. In some embodiments, the machine 100 may also have multiple processing spaces 110. Each processing space 110 may have a different mechanical configuration, and each processing space 110 does not interfere with each other. Specifically, the processing space 110 of the machine 100 needs to be closed to the outside due to its requirements on the processing environment during operation, such as making the processing space 110 present a dust-free processing environment or causing the presence of specific gases in the processing space 110 . , but this disclosure is not limited to this. The operator can use the photography device 120 in the machine 100, such as a surveillance camera, to remotely monitor equipment abnormalities in the processing space 110. Specifically, the number of photography devices 120 installed in the processing space 110 can be increased or decreased according to different institutional configurations inside the processing space 110 . In some embodiments, the photography device 120 can change the shooting angle through remote control, but the disclosure is not limited thereto.

FIG. 3 is a schematic diagram illustrating a normal state image of the processing space 110 according to an embodiment of the present disclosure. Please refer to Figure 1 and Figure 3. The configuration of the processing space 110 of the embodiment shown in Figure 3 will be described below. The processing space 110 has a plurality of pipelines (including three pipelines 112, 114, and 116 in FIG. 3) and a liquid tank 118. The pipelines 112, 114, and 116 can individually input liquid into the liquid tank 118, and the input liquid amounts of the pipelines 112, 114, and 116 are controlled via the automation system. In some embodiments, the pipelines 112, 114, and 116 under normal conditions will be completely filled with liquid, and the liquid levels are exactly aligned with the liquid outlets 112a, 114a, and 116a of the pipelines 112, 114, and 116. The liquid outlets 112a, 114a, and 116a are located in the opening 118a of the liquid tank 118.

Figure 4 is a functional block diagram illustrating an abnormality diagnosis method M1 according to an embodiment of the present disclosure. Referring to FIGS. 1 to 4 , in step S100 , the photography device 120 captures images including the pipelines 112 , 114 , 116 and the liquid tank 118 . However, the processing space 110 can still have other different configurations, and multiple camera devices 120 can be used to obtain a sufficient monitoring field of view according to monitoring requirements. In step S200, the photography device 120 transmits the image to the image preprocessing device 130 for image preprocessing. In some embodiments, image preprocessing includes image sharpening, but the disclosure is not limited thereto. Specifically, the image preprocessing device 130 is used to clarify the images captured by the photography device 120 so as to identify and compare the images. Any appropriate image processing can be added to the image preprocessing process.

Please refer to Figure 1, Figure 2 and Figure 4. In step S300, the processed image is sent to the determination system 140 outside the machine 100 by the image pre-processing device 130. Specifically, the determination system 140 and the image preprocessing device 130 can be independent of the machine 100 , for example, a device that includes both the image preprocessing device 130 and the determination system 140 , such as a computer system or a data center, to save image transmission. time and facilitates the integration and execution of the anomaly diagnosis method M1, but the present disclosure is not limited thereto. In some embodiments, the device that includes both the image preprocessing device 130 and the determination system 140 may further include a database 150 to store images captured by the photography device 120 .

In step S400, the determination system 140 will determine whether the processed image has features. In some embodiments, step S400 includes comparing the processed image and a comparison image, where the comparison image is a normal state image (step S410 ). For example, the comparison image may be the image captured by the photography device 120 in Figure 3 . After capturing the comparison image, the photography device 120 can pre-store the comparison image in the database 150 for subsequent access by the determination system 140 . In some other embodiments, the comparison images can also be images captured by other photography devices 120 and uploaded to the database 150 .

In step S500, the alarm module 160 will issue an alarm according to the determination result of the determination system 140. In some embodiments, when the determination system 140 determines that the processed image has characteristics, it determines that the image is abnormal, and issues an alarm through the alarm module 160 . Specifically, the classification of the judgment results can be preset inside the judgment system 140, and the judgment results can be classified accordingly according to the characteristics obtained after comparison. The determination system 140 will decide whether to notify the alarm module 160 to issue an alarm based on whether the comparison result has a corresponding classification. However, other suitable determination methods can also be used.

The following will provide a detailed description of the determination system 140 and the alarm module 160 for some embodiments to better understand the execution of step S500. FIG. 5 is a schematic diagram illustrating an abnormal state image of the processing space 110 according to an embodiment of the present disclosure. In some embodiments, the comparison image includes pipelines 112, 114, and 116, and the comparison image defines the liquid level in the pipelines 112, 114, and 116 in the image, and the characteristics include the image and the liquid level positions of the comparison image are different. , and when the liquid level is located inside the pipelines 112, 114, and 116, the alarm issued by the alarm module 160 includes the pipeline back suction detection result. Please refer to Figures 3 and 5. Specifically, in some embodiments, the liquid levels in pipelines 112, 114, and 116 will be aligned with the liquid outlets 112a, 114a, and 116a under normal conditions (as shown in Figure 3 shown). Therefore, the positions of the liquid outlets 112a, 114a, and 116a are defined as standard liquid level positions by comparing the images. However, in the abnormal state image in Figure 5, it can be found that the liquid level position 116b in the pipeline 116 is different from the position of the liquid outlet 116a. When the liquid level position 116b is located in the pipeline 116 and is not aligned with the standard liquid level position (ie, the liquid outlet 116a), the determination system 140 will determine that the abnormal state image belongs to the category of "pipeline back suction abnormality". Abnormal pipeline back suction may cause the liquid output controlled by the automation system to be different from the expected value and affect the standardization of the manufacturing process. The determination system 140 will transmit the determination result of the pipeline back suction abnormality to the alarm module 160, and the alarm module 160 will issue an alarm containing the pipeline back suction abnormality to notify personnel to perform maintenance according to the abnormality.

FIG. 6 is a schematic diagram illustrating another abnormal state image of the processing space 110 according to an embodiment of the present disclosure. In some embodiments, the comparison image includes pipelines 112, 114, and 116, and the feature includes bubbles 200 inside the pipelines 112, 114, and 116 in the processed image, and when there are bubbles 200 inside the pipelines 112, 114, and 116, The alarm issued by the alarm module 160 includes the pipeline splash detection result. Please refer to Figures 3 and 6. Specifically, pipelines 112, 114, and 116 should only be filled with liquid under normal conditions (as shown in Figure 3). However, in the abnormal state image in Figure 6, it can be found that there are multiple bubbles 200 in the pipeline 112. The accumulation of bubbles 200 will affect the liquid discharge state of the pipeline 112. The reason is that when the automation system controls the liquid discharge from the pipeline 112, the liquid will be squeezed by the bubbles 200 in the pipeline 112, causing the liquid pressure at the liquid outlet 112a to increase, further As a result, the liquid is ejected from the liquid outlet 112a in a splashing manner. The liquid sprayed from the liquid outlet 112a may splash out of the opening 118a, thereby affecting the amount of liquid entering the liquid tank 118. Therefore, when the determination system 140 determines that the image is an abnormal state image and finds that the bubble 200 is located in the pipeline 112, the determination system 140 will determine that the abnormal state image belongs to the category of "pipeline splash abnormality". The determination system 140 will transmit the determination result of the pipeline splash abnormality to the alarm module 160, and the alarm module 160 will issue an alarm containing the pipeline splash abnormality to notify personnel to perform maintenance according to the abnormality.

In some embodiments, the comparison image includes pipelines 112, 114, 116, the feature includes liquid flowing out of the pipeline 112, 114, 116 in the processed image, and when the liquid flows outside the pipeline 112, 114, 116, the alarm model Alarms issued by group 160 include pipeline drip detection results or pipeline splash detection results. Please refer to Figures 3, 5 and 6. Specifically, under normal conditions, when the automation system does not control the liquid outflow from pipelines 112, 114, and 116, the liquid level will pass through the balance between atmospheric pressure and liquid pressure. Maintain at the positions of liquid outlets 112a, 114a, and 116a (as shown in Figure 3). However, in the abnormal state images in Figures 5 and 6, it can be found that liquid leaks from the pipeline 112. Furthermore, abnormal conditions can be classified based on the trajectory of liquid leakage.

Please refer to Figure 5 first. In some embodiments, when the liquid travels along a linear trajectory, the alarm issued by the alarm module 160 includes the pipeline drip detection result. For example, when the pipeline is not controlled by the automation system and the liquid leaves the liquid outlet 112a linearly along the direction In other embodiments, if the pipeline 112 has other breaches other than the liquid outlet 112a, and the liquid will linearly leak out of the pipeline 112 along the direction X from the breach, the determination system 140 will also determine this abnormality. The status image belongs to the category of "pipeline leakage abnormality". This abnormal dripping state will cause the automation system to be unable to control the liquid output, thus affecting the standardization of the manufacturing process. The determination system 140 will transmit the determination result of the pipeline leakage abnormality to the alarm module 160, and the alarm module 160 will issue an alarm containing the pipeline leakage abnormality to notify personnel to perform repairs according to the abnormality.

Please refer to Figure 6 again. In some embodiments, when the liquid travels along a non-linear trajectory, the alarm issued by the alarm module 160 includes the pipeline splash detection result. For example, when the liquid is sprayed divergently toward the liquid tank 118 with the liquid outlet 112a as the center, the determination system 140 will determine that the abnormal state image belongs to the category of "pipeline splash abnormality". This divergent liquid spraying state may occur simultaneously with the aforementioned abnormal state of bubbles 200 in the pipeline 112 and affect the amount of liquid entering the liquid tank 118 . The determination system 140 will transmit the determination result of the pipeline splash abnormality to the alarm module 160, and the alarm module 160 will issue an alarm containing the pipeline splash abnormality to notify personnel to perform maintenance according to the abnormality.

FIG. 7 is a schematic diagram illustrating another abnormal state image of the processing space 110 according to an embodiment of the present disclosure. In some embodiments, the feature includes that the mechanism configuration of the processing space 110 in the image is different from the mechanism configuration of the processing space 110 in the comparison image, and the alarm issued by the alarm module 160 includes the mechanism abnormality detection result. Please refer to Figures 3 and 7. Specifically, the determination system 140 can also compare the mechanism configuration in the image. For example, by comparing the positions of pipelines 112, 114, and 116 in the normal state image in Figure 3 and the abnormal state image in Figure 7, it can be found that pipeline 114 is abnormally skewed in Figure 7. At this time, the determination system 140 will determine the abnormality. Status images belong to the category of "organizational abnormality". Furthermore, when the pipeline has holes, deflection, deformation, or displacement and other physical structural changes, it is a structural abnormality. The determination system 140 will transmit the determination result of the mechanism abnormality to the alarm module 160, and the alarm module 160 will issue an alarm containing the mechanism abnormality to notify personnel to perform maintenance according to the abnormality.

From the above detailed description of the specific embodiments of the present disclosure, it can be clearly seen that in the abnormality diagnosis method of the present disclosure, the status of the internal processing space of the machine can be remotely monitored and collected in real time through a photography device. In addition, with the assistance of the judgment system, abnormal conditions can be classified without relying on manpower, thereby improving the maintenance efficiency of personnel and reducing the labor cost of monitoring machines.

The foregoing description is merely illustrative and descriptive of exemplary embodiments of the present disclosure, and is not intended to be exhaustive or to limit the precise forms of the invention disclosed in the present disclosure. The above teachings may be modified or varied.

The embodiments were chosen and described in order to explain the contents of the present disclosure and their practical applications to thereby inspire others skilled in the art to utilize the disclosure and various embodiments and with various modifications as are suited to the particular use contemplated. Alternative embodiments will be apparent to those skilled in the art to which this disclosure belongs without departing from the spirit and scope of this disclosure. Therefore, the scope of the present disclosure is determined by the appended invention claims rather than by the foregoing specification and the exemplary embodiments described therein.

100:Machine 110: Processing space 112,114,116:Pipeline 112a, 114a, 116a: Liquid outlet 116b: Liquid level position 118:Liquid tank 118a:Open your mouth 120: Photography installation 130:Image preprocessing device 140:Judgment system 150:Database 160:Alarm module 200: Bubbles M1:Method S100, S200, S300, S400, S410, S500: steps X: direction

The drawings illustrate one or more embodiments of the disclosure and, together with the written description, serve to explain principles of the disclosure. Wherever possible, the same reference numbers will be used throughout the drawings to refer to similar or identical elements of the embodiments, where: Figure 1 is a flow chart illustrating an abnormality diagnosis method according to an embodiment of the present disclosure. Figure 2 is a schematic diagram illustrating a machine and a processing space according to an embodiment of the present disclosure. Figure 3 is a schematic diagram illustrating a normal state image of a processing space according to an embodiment of the present disclosure. FIG. 4 is a functional block diagram illustrating an abnormality diagnosis method according to an embodiment of the present disclosure. Figure 5 is a schematic diagram illustrating an abnormal state image of a processing space according to an embodiment of the present disclosure. Figure 6 is a schematic diagram illustrating another abnormal state image of the processing space according to an embodiment of the present disclosure. FIG. 7 is a schematic diagram illustrating another abnormal state image of the processing space according to an embodiment of the present disclosure.

Domestic storage information (please note in order of storage institution, date and number) without Overseas storage information (please note in order of storage country, institution, date, and number) without

M1:Method

S100, S200, S300, S400, S500: steps

Claims (9)

An abnormality diagnosis method is applied to a machine. The machine has a processing space. The processing space has a pipeline and a liquid tank. The pipeline has a liquid outlet facing downwardly toward the liquid tank and is configured to pass through the liquid tank. The liquid outlet inputs a liquid into the liquid tank. The abnormality diagnosis method includes: capturing an image of the processing space through a photography device of the machine; performing an image preprocessing on the image; and processing the processed space. The image is sent to a determination system outside the machine; the determination system determines whether the processed image has at least one feature, wherein the at least one feature includes a liquid in the pipeline in the processed image. There is a difference between the liquid level position and a position of the liquid outlet; and an alarm is issued through an alarm module according to a judgment result of the judgment system. The abnormality diagnosis method as claimed in claim 1, wherein the image preprocessing includes sharpening the image. The abnormality diagnosis method as claimed in claim 1, wherein when the determination system determines that the processed image has the at least one characteristic, it determines that the image is abnormal, and issues the alarm through the alarm module. An abnormality diagnosis method, applied to a machine, the machine There is a processing space, the processing space has a plurality of pipelines and a liquid tank, each of the pipelines has a liquid outlet facing downwardly to the liquid tank and is respectively configured to input a liquid into the liquid through the liquid outlet. In the tank, the abnormality diagnosis method includes: capturing an image of the processing space through a photography device of the machine; performing an image preprocessing on the image; and transmitting the processed image to an external device outside the machine. A determination system; using the determination system to determine whether the processed image has at least one characteristic, including comparing the processed image and a comparison image, wherein the comparison image is a normal state image, and the normal state image includes a plurality of pipelines, respectively corresponding to the pipelines in the processed image, and the at least one feature includes at least one of the pipelines in the processed image relative to the normal state image. The counterpart in the pipeline is skewed; and an alarm is issued through an alarm module according to a determination result of the determination system. The abnormality diagnosis method as described in claim 4, wherein the pipelines in the normal state image are each completely filled with a liquid, and each of the pipelines has a liquid outlet aligned with a liquid surface of the liquid. The anomaly diagnosis method as described in claim 4, wherein the comparison image includes at least one pipeline, and the feature includes the processed image There is at least one bubble inside the at least one pipeline, and when there is at least one bubble inside the at least one pipeline, the alarm issued by the alarm module includes a pipeline splash detection result. An abnormality diagnosis method is applied to a machine. The machine has a processing space. The processing space has a pipeline and a liquid tank. The pipeline has a liquid outlet facing downwardly toward the liquid tank and is configured to pass through the liquid tank. The liquid outlet inputs a liquid into the liquid tank. The abnormality diagnosis method includes: capturing an image of the processing space through a photography device of the machine; performing an image preprocessing on the image; and processing the processed space. The image is sent to a determination system outside the machine; the determination system determines whether the processed image has at least one feature, wherein the at least one feature includes the liquid in the pipeline in the processed image. It flows down from the outside of the pipeline through the liquid outlet; and an alarm is issued through an alarm module according to a judgment result of the judgment system. The abnormality diagnosis method as described in claim 7, wherein when the liquid travels along a linear trajectory, the alarm issued by the alarm module includes the pipeline drip detection result. The abnormality diagnosis method as described in claim 7, wherein when the liquid travels along a non-linear trajectory, the alarm package issued by the alarm module Contains the pipeline splash detection results.

Images