Data management system of photoetching machine

Abstract

The invention relates to the technical field of data management, in particular to a data management system of a photoetching machine, which comprises a data acquisition part used for recording acquired data based on a data recording component suitable for industrial software so as to obtain a data record, the data record comprises mapping diagnosis context information and / or nested diagnosis context information corresponding to the predetermined operation; the data processing part is used for receiving the encrypted data records from the data acquisition part and carrying out data processing, and the data processing comprises filtering of the encrypted data records and pre-estimation of subsystem risks of the photoetching machine; and the data storage part is used for storing data related to the data acquisition part and the data processing part. Therefore, by adopting a mapping diagnosis context information / nested diagnosis context information mechanism, an encrypted data record filtering mechanism and a subsystem risk estimation mechanism, the industrial production efficiency and stability of the photoetching machine can be improved, and the software development cost of the photoetching machine is reduced.

Description

Technical Field

[0001] This application relates to the technical field of data management, and more specifically, to a data management system for a lithography machine. Background Art

[0002] Industrial manufacturing is centered around data, which is crucial for the stable operation, fault troubleshooting, and performance optimization of equipment. The software data management platform built based on the ELK (FileBeat, Elasticsearch, Logstash, Kibana) technology stack mainly consists of four key parts: data collection, data processing, data storage, and data visualization.

[0003] In the data collection part, Filebeat continuously monitors the data files generated locally by the lithography machine software through a file monitoring mechanism. Once new content is written to the data files, Filebeat will capture these changes and collect the new data. Then, Filebeat will send the collected data to the subsequent Logstash server according to the pre-configured transmission rules.

[0004] In the data processing part, Logstash is deployed on an independent server node and is connected to the lithography machine device running the Filebeat agent through the network. The Logstash server is responsible for receiving the data transmitted from multiple Filebeat agents through the port and processing the data through controls.

[0005] In the data storage part, the output plugin of Logstash sends the processed data to Elasticsearch. Elasticsearch stores the data according to the index, searches in the corresponding index according to the query conditions, and returns the matching results. Logstash can also send the data to a traditional database, such as mysql for storage.

[0006] In the data visualization part, Kibana is an open-source platform integrated with Elasticsearch, which has visualization functions, provides various types of charts, supports customization and dashboards. Kibana can query data interactively, preview, save and share in real time, analyze logs, and perform monitoring and alarming.

[0007] However, this software data management platform based on ELK also has the following problems.

[0008] First of all, in terms of data security, the security of data cannot be guaranteed during the transmission from Filebeat to Logstash. For example Figure 1As shown, due to the lack of an adequate encryption mechanism in the software data association platform, data is transmitted in plain text over the network, and malicious actors in the network can easily intercept, read, and tamper with the data. However, if Filebeat encrypts the data before sending it, Logstash needs to decrypt the data before processing it, which greatly affects the efficiency of data processing. Here, Figure 1 FIG. shows a schematic diagram of the data management process of an existing ELK-based software data management platform.

[0009] In addition, in terms of risk prediction, it is difficult to quickly and accurately predict key potential risk information. This is because traditional subsystem risk assessment methods usually rely on rules and experience, making it difficult to comprehensively and accurately assess. Moreover, expert judgment and experience reference are somewhat subjective, and different experts may have different judgments on the risk assessment results of the same subsystem due to differences in personal experience and cognition. As a result, the accuracy and reliability of risk assessment may be affected.

[0010] Furthermore, in terms of data analysis, it is difficult to quickly and accurately analyze data. This is because the relevance of data information is poor, the records are chaotic under multi-threaded and concurrent operations, and the efficiency of fault tracking is low. In particular, since the operation of lithography software involves multiple complex execution paths and subtasks, the confusion of information between threads, the isolation and dispersion of information, and the lack of context identification for data all increase the difficulty of problem tracking.

[0011] Therefore, there is a need to provide an improved data management system for lithography machines. SUMMARY OF THE INVENTION

[0012] Embodiments of the present application provide a data management system for a lithography machine, which can improve the industrial production efficiency and stability of the lithography machine and reduce the software development cost of the lithography machine by adopting a mapping diagnosis context information / nested diagnosis context information mechanism, an encrypted data record filtering mechanism, and a subsystem risk prediction mechanism.

[0013] According to one aspect of the present application, there is provided a data management system for a lithography machine, including: a data acquisition part for recording the collected data based on a data record component suitable for industrial software to obtain a data record, where the data record includes mapping diagnosis context information and / or nested diagnosis context information corresponding to a predetermined operation; a data processing part for receiving the encrypted data record from the data acquisition part and performing data processing, where the data processing includes filtering of the encrypted data record and prediction of the subsystem risk of the lithography machine; and a data storage part for storing data related to the data acquisition part and the data processing part.

[0014] In the data management system of the above-mentioned lithography machine, the data acquisition part sets at least one of the output time of the data record, the level of the data record, the storage path of the data record, the detailed content of the data record, and the rolling strategy of the data record file.

[0015] In the data management system of the above-mentioned lithography machine, the mapping diagnostic context information stores the task information of the predetermined operation in the form of key-value pairs, and the nested diagnostic context information stores the process and hierarchical structure information of the predetermined operation in the form of a stack.

[0016] In the data management system of the above-mentioned lithography machine, when the data acquisition part obtains the data record corresponding to the predetermined operation, it adds the mapping diagnostic context information of the current task thread of the operation to the data record, and stores the hierarchical path of the operation in the nested diagnostic context information stack.

[0017] In the data management system of the above-mentioned lithography machine, the data record includes information for identifying different task threads of the lithography machine operation and / or information for identifying different subsystems of the lithography machine.

[0018] In the data management system of the above-mentioned lithography machine, the encrypted data record is a data record encrypted by combining the AES encryption algorithm and the homomorphic encryption algorithm.

[0019] In the data management system of the above-mentioned lithography machine, the encrypted data record is a data record encrypted by the data acquisition part using the AES encryption algorithm, and the data processing part directly performs computational analysis on the encrypted data record using the homomorphic encryption algorithm.

[0020] In the data management system of the above-mentioned lithography machine, the filtering of the encrypted data record includes: verifying whether the encrypted data record meets the valid data standard; parsing the encrypted data record in response to the encrypted data record meeting the valid data standard; and filtering the encrypted data record in response to the encrypted data record not meeting the valid data standard.

[0021] In the data management system of the above-mentioned lithography machine, the estimation of the subsystem risk of the lithography machine includes evaluating the subsystem risk of the lithography machine with a pre-trained model, and evaluating the subsystem risk of the lithography machine with a pre-trained model includes: collecting data records of each component of the lithography machine software and related subsystems; labeling the data records with different risk levels, and creating a training data set according to a predetermined ratio of different risk levels; training a classifier with the training data set; and extracting data features from the newly collected subsystem data of the data acquisition part, vectorizing them, and inputting them into the classifier to predict the risk level of the newly collected subsystem data.

[0022] In the data management system of the above-mentioned lithography machine, it further includes: a data visualization part for visualizing the data stored in the data storage part.

[0023] The data management system of the lithography machine provided by the embodiments of the present application can improve the industrial production efficiency and stability of the lithography machine and reduce the software development cost of the lithography machine by adopting a mapping diagnostic context information / nested diagnostic context information mechanism, an encrypted data record filtering mechanism, and a subsystem risk prediction mechanism. Brief Description of the Drawings

[0024] By reading the following detailed description of the preferred specific embodiments, various other advantages and benefits of the present application will become clear to those of ordinary skill in the art. The accompanying drawings in the specification are only for the purpose of showing the preferred embodiments and are not considered to be a limitation of the present application. Obviously, the drawings described below are only some embodiments of the present application, and those of ordinary skill in the art can obtain other drawings based on these drawings without creative efforts. Moreover, throughout the drawings, the same reference numerals are used to represent the same components.

[0025] Figure 1 The schematic diagram of the data management process of the existing ELK-based software data management platform is shown.

[0026] Figure 2 The schematic block diagram of the data management system of the lithography machine according to the embodiments of the present application is shown.

[0027] Figure 3 The schematic diagram of the data management system of the lithography machine according to the embodiments of the present application implemented based on the ELK technology stack is shown. Detailed Description of the Embodiments

[0028] Next, exemplary embodiments according to the present application will be described in detail with reference to the accompanying drawings. Obviously, the described embodiments are only a part of the embodiments of the present application, rather than all the embodiments of the present application. It should be understood that the present application is not limited by the exemplary embodiments described here.

[0029] Figure 2 The schematic block diagram of the data management system of the lithography machine according to the embodiments of the present application is shown.

[0030] As Figure 2 shown, the data management system of the lithography machine according to the embodiments of the present application includes the following parts.

[0031] The data acquisition part. For example, in the case where the data management system of the lithography machine according to the embodiment of the present application is implemented by a software data management platform built based on the ELK (FileBeat, Elasticsearch, Logstash, Kibana) technology stack, the data acquisition part is FileBeat. In the embodiment of the present application, the data acquisition part records the collected data based on a data recording component suitable for industrial software to obtain a data record. For example, the data recording component suitable for industrial software can be a logging component such as Log4cplus, Log4j, Logback, etc., so as to obtain a data record in the form of a log. And, in the embodiment of the present application, the data acquisition part can set the output time of the data record, the level of the data record (such as Trace, DEBUG, INFO, WARN, ERROR), the storage path of the data record (local file or remote server), the detailed content of the data record, and the rolling policy of the data record file, etc., to ensure that the data record can be effectively recorded and managed.

[0032] In addition, in the embodiment of the present application, the data acquisition part can apply the MappedDiagnostic Context (MDC) and the Nested Diagnostic Context (NDC) to the data management system of the lithography machine to manage the data of the lithography machine industrial software, so as to realize the effective association and organization of the data record. Here, the MDC and NDC context information to be recorded is determined according to the industrial business process and operation characteristics of the lithography machine.

[0033] Specifically, in the embodiment of the present application, the MDC information is stored in the form of key-value pairs, and the stored data can be operation task information, including task number, operation type, device ID, device status, etc. Among them, when starting a new operation task, the MDC context information is initialized. And according to the specific situation of the operation task, the relevant key-value pair information is stored in the MDC. For example, for a lithography exposure task, information such as task number, exposure parameters, and device status is stored in the MDC. In this way, when obtaining a data record according to the collected data, the context information is automatically obtained from the MDC of the current thread and added to the entry of the data record. In addition, after the operation task ends, the context information stored in the MDC can be cleared, and all the key-value pairs previously stored in the MDC are deleted to release memory resources and avoid interference with subsequent operations.

[0034] In addition, the NDC information is stored in the form of a stack to construct nested context information, reflecting the hierarchical relationship and process sequence of operations. Specifically, according to the process and hierarchical structure of the operation, the hierarchical path of the current operation is pushed onto the NDC stack. When entering the next sub-step of the operation, the new hierarchical information is pushed onto the NDC stack; when a sub-step is completed, the corresponding information is popped from the NDC stack. Similarly, when clearing the NDC context information, all the hierarchical information in the NDC stack is popped to ensure that the NDC stack is empty and ready for the next operation task.

[0035] Moreover, in the embodiments of the present application, the data record may include information for identifying threads, so as to distinguish data records of different threads in a multi-threaded scenario. Furthermore, in the embodiments of the present application, the data record may include information for identifying subsystems.

[0036] In addition, in the embodiments of the present application, the data acquisition part may collect different user data in real time for permission management and fault troubleshooting. Additionally, the data acquisition system may operate in a multi-system environment, such as being compatible with systems like Windows and Linux, thereby ensuring the integrity of data records. Moreover, when the data management system of the lithography machine according to the embodiments of the present application is used for multi-machine collaborative work, the data acquisition part can be independently deployed on each machine to collect and transmit data in real time to achieve comprehensive monitoring.

[0037] Therefore, in the data management system of the lithography machine according to the embodiments of the present application, the data acquisition part records the collected data based on a data record component suitable for industrial software to obtain a data record. And the data acquisition part further sets at least one of the output time of the data record, the level of the data record, the storage path of the data record, the detailed content of the data record, and the rolling policy of the data record file.

[0038] In addition, the data acquisition part further establishes mapped diagnostic context information and / or nested diagnostic context information, wherein the mapped diagnostic context information stores the task information of the operation in the form of key-value pairs, and the nested diagnostic context information stores the process and hierarchical structure information of the operation in the form of a stack.

[0039] Moreover, when the data acquisition part obtains the data record corresponding to a predetermined operation, it adds the mapped diagnostic context information of the current task thread of the operation to the data record and stores the hierarchical path of the operation in the nested diagnostic context information stack.

[0040] In addition, the data record includes information for identifying different task threads of the lithography machine operation and / or information for identifying different subsystems of the lithography machine.

[0041] In this way, by adopting the MDC and NDC mechanisms, the data management system of the lithography machine according to the embodiments of the present application can automatically obtain and incorporate the MDC and NDC context information of the current operation task thread when recording the industrial software data of the lithography machine. Moreover, the data recording supports multiple data recording levels (such as trace, debug, information, warning, error), and is distinguished according to different operation stages and importance. When a failure occurs, relevant data records can be quickly screened out from a large amount of data records based on specific key values in the MDC or nested information in the NDC. Then, by performing correlation analysis on the screened data records and using the MDC and NDC to construct the complete execution path of the operation, the link and cause of the failure can be accurately located, and a detailed failure report including the failure occurrence time, operation steps, relevant context information, etc. can be generated. In addition, in a multi-threaded scenario, the data records of different threads can also be accurately distinguished to ensure that the operation process of each thread can be clearly recorded and traced through the MDC and NDC context information. Through the reasonable application of the MDC and NDC, the confusion of logs during multi-threaded operations is avoided, the timing consistency of multiple tests is ensured, and fine-grained verification can be performed for a specific thread to verify the status or data of the specified thread, thereby improving the data log management and fault troubleshooting efficiency of the industrial lithography machine software in a complex multi-threaded environment and enhancing the stability and reliability of the software system.

[0042] Moreover, the data management system of the lithography machine according to the embodiments of the present application can automatically identify and record the information of each subsystem involved in the software operation without manual recording. That is, when collecting and recording data, this subsystem information will be recorded together with other context information, providing a more comprehensive and accurate basis for subsequent fault troubleshooting and system analysis, and greatly improving the data recording efficiency and information integrity.

[0043] The data processing part is used to receive the encrypted data records from the data collection part and perform processing. For example, when the data management system of the lithography machine according to the embodiments of the present application is implemented by a software data management platform based on the ELK (FileBeat, Elasticsearch, Logstash, Kibana) technology stack, the data processing part is Logstash. That is, in the embodiments of the present application, different from the traditional software data management platform, Filebeat transmits data to Logstash in plain text, thus having security risks. To strengthen the security of data transmission, the data is transmitted in an encrypted form. And preferably, in order to further meet the requirement of directly processing encrypted data, in the embodiments of the present application, a combination of the AES encryption algorithm and the homomorphic encryption algorithm is used.

[0044] Among them, the AES (Advanced Encryption Standard) encryption algorithm, as a commonly used efficient symmetric encryption algorithm, can encrypt data quickly and ensure the basic security of transmission. Moreover, the homomorphic encryption algorithm allows specific calculations to be performed in the ciphertext state without decryption. By combining the AES encryption algorithm with the homomorphic encryption algorithm, their respective advantages can be brought into play to improve data security and processing efficiency.

[0045] Specifically, when transmitting data from the data acquisition part to the data processing part, first use the AES encryption algorithm to encrypt the data collected by the data acquisition part, that is, the plaintext data, to prevent sensitive information from being intercepted during transmission, and then transmit the encrypted data from the data acquisition part to the data processing part. Then, based on the homomorphic encryption characteristics, the data processing part directly calculates and analyzes the encrypted data using the homomorphic encryption algorithm, and extracts key information such as the data, its source identifier, and data type label according to preset rules.

[0046] Therefore, in the data management system of the lithography machine according to the embodiments of the present application, the data processing part is used to receive the encrypted data record from the data acquisition part and process it. And the encrypted data record is a data record encrypted by combining the AES encryption algorithm with the homomorphic encryption algorithm.

[0047] In addition, the encrypted data record is a data record encrypted by the data acquisition part using the AES encryption algorithm, and the data processing part directly calculates and analyzes the encrypted data record using the homomorphic encryption algorithm.

[0048] Moreover, after the data processing part extracts the data record, it can further verify whether the data record meets the valid data standard. Specifically, if the data record meets the valid data standard, it parses the encrypted data record and stores it in the data storage part for subsequent in-depth analysis, while the data record that does not meet the standard is directly filtered. Specifically, the data processing part can use a predetermined parsing tool, such as grok, to filter out the time, thread, level, subsystem, MDC information, NDC information, and log details in the data record. And as mentioned above, through the homomorphic encryption algorithm, direct filtering of encrypted data can be achieved, which not only ensures data security, saves the decryption time of the data processing part, and ensures that only valid data is decrypted and stored, but also can improve the density of valid data in the data storage part. At the same time, since a large amount of invalid data is filtered, the storage pressure on the data storage part is reduced, and the system operation efficiency and stability are improved.

[0049] Therefore, in the data management system of the lithography machine according to the embodiment of the present application, the processing of the data processing part includes verifying whether the encrypted data record meets the valid data standard. And, in response to the encrypted data record meeting the valid data standard, parsing the encrypted data record, and in response to the encrypted data record not meeting the valid data standard, filtering the encrypted data record.

[0050] Furthermore, as mentioned above, in the complex system architecture of the lithography machine, the stable operation of each subsystem is crucial, and the traditional risk assessment method is difficult to fully and accurately assess potential risks due to its reliance on rules and experience. Therefore, in the data management system of the lithography machine according to the embodiment of the present application, a pre-trained model is used to achieve a more accurate and intelligent subsystem risk assessment.

[0051] Specifically, the data processing part can collect data records of various components and related subsystems of the lithography machine software, such as logs, and security experts mark the data records as different risk levels, such as high, medium and low risk levels, according to the operation and potential threats, and then create a training data set according to the predetermined proportions of different risk levels, such as 70%, 15% and 15% of the high, medium and low risk levels. Subsequently, the data processing part extracts the features of the data records through feature engineering, selects a suitable classifier, such as the support vector machine (SVM) kernel function and multi-classification strategy, and uses grid search and other methods to tune the parameters. Then, take out the test set and input the model prediction, for example, use Python and related libraries to calculate the evaluation index, adjust according to the result analysis, and finally train a qualified classification model. In this way, the data processing part can extract data features and vectorize the newly collected data of the data collection part, input the model to predict the risk level, and store the results and visualization. In this way, the data management system of the lithography machine can take corresponding early warning and response measures according to the risk level to ensure the stable operation of the lithography machine software.

[0052] That is, by adopting the risk prediction mechanism of the pre-trained model, the data processing part can be used to efficiently collect multi-source data records, mark different risk levels, and store training data sets and test data sets during the data preparation stage. Then, the data processing unit performs pre-processing such as data cleaning on the data records, and extracts features using feature engineering methods such as word bag models. At this time, key features can also be further screened out to reduce the dimension. During model training, a suitable classifier is selected for multi-classification problems with different risk levels, such as SVM kernel functions and strategies, and the stored data is read. The adjustment parameters can also be monitored through data visualization. In addition, the evaluation indicators of the model can be stored and visualized for analysis, so as to optimize the model in a targeted manner. In this way, the risk prediction mechanism can process new data records in real time, predict the risk level, and accurately warn and respond according to different levels to ensure the stable and safe operation of the lithography machine software.

[0053] Therefore, in the data management system of a lithography machine according to an embodiment of the present application, the processing of the data processing part includes evaluating the subsystem risks of the lithography machine with a pre-trained model, and evaluating the subsystem risks of the lithography machine with a pre-trained model includes: collecting data records of each component of the lithography machine software and related subsystems; labeling the data records with different risk levels, and creating a training data set according to a predetermined ratio of different risk levels; training a classifier with the training data set; extracting data features from the newly collected subsystem data of the data acquisition part, vectorizing them, and inputting them into the classifier to predict the risk level of the newly collected subsystem data.

[0054] A data storage part, which is used to store data related to the data acquisition part and the data processing part. For example, in the case where the data management system of a lithography machine according to an embodiment of the present application is implemented by a software data management platform based on the ELK (FileBeat, Elasticsearch, Logstash, Kibana) technology stack, the data storage part is Elasticsearch.

[0055] Specifically, the data storage part is used to store data related to the data acquisition part and the data processing part, for example, including the data set, test set, time, level, subsystem, detailed information, etc. of the data records collected and recorded by the data acquisition part for the training of the pre-trained model as described above. The data storage part can store data according to an index, search in the corresponding index according to a query condition, and return the matching results. Here, the data storage part according to an embodiment of the present application can include Elasticsearch based on the ELK (FileBeat, Elasticsearch, Logstash, Kibana) technology stack and a traditional database, that is, the data processing part can also send data to the traditional database for storage.

[0056] Furthermore, the data management system of a lithography machine according to an embodiment of the present application may further include: a data visualization part, which is used to visualize the data stored in the data storage part.

[0057] For example, when the data management system of the lithography machine according to the embodiments of the present application is implemented by a software data management platform built based on the ELK (FileBeat, Elasticsearch, Logstash, Kibana) technology stack, the data visualization part is Kibana. Through the data visualization part, the data stored in the data storage part can be intuitively browsed. For example, through the data visualization part, users can create their own spaces and customize and display data in real time with various charts. At the same time, users can perform correlation analysis on the selected data records. For example, through the context information constructed by MDC and NDC, the execution path and sequence of operations can be sorted out, and the specific links and possible causes of faults can be found. For example, according to the hierarchical information in NDC, it can be determined which sub-process the fault occurs in, and combined with the device status information in MDC, the possible factors causing the fault can be analyzed. Thus, the data management system of the lithography machine according to the embodiments of the present application ensures the timeliness and sequentiality of problem diagnosis and reduces the development cost.

[0058] Moreover, the data storage part can store data based on the time of data records, and the data visualization part can display the operation sequence in a visual manner, facilitating users to comprehensively master the operation process.

[0059] In summary, in terms of problem diagnosis, the data management system of the lithography machine according to the embodiments of the present application can clearly record the hierarchy and sequence of lithography machine operations through MDC and NDC. Among them, NDC presents the step sequence in a stack structure, and MDC supplements key parameters and identifiers. In a multi-threaded concurrent scenario, MDC differentiates threads and combines with NDC to record operation steps, enabling sorting of data records, assisting in analyzing the timing relationship between threads, and quickly troubleshooting faults caused by incorrect operation sequences. The data processing part collects lithography machine software logs in real time, combines with MDC and NDC information, and quickly transmits them to the data storage part to establish an index, enabling instant query of new logs. In addition, based on MDC and NDC information, real-time rules can be set to monitor the running state, and an alarm is immediately triggered in case of an anomaly. The data visualization part can display the results to help operation and maintenance personnel quickly locate and solve problems, reducing the fault handling time. Moreover, the data processing part can automatically identify the system code of the lithography machine business through filtering, without manually recording in the code. After these information are incorporated into the data records, they are stored in the data storage part, facilitating screening and analysis of data records by subsystem and understanding the running status of each subsystem.

[0060] In addition, in terms of security guarantee, the data management system of the lithography machine according to the embodiments of the present application greatly improves the security and reliability of the data management system of the lithography machine in the data transmission link compared with the traditional unencrypted transmission method. Moreover, compared with the ordinary encryption method, the data management system of the lithography machine according to the embodiments of the present application achieves a direct filtering effect on the ciphertext data, which not only ensures the security of the data, but also saves a large amount of time cost for decryption in the data processing part. It also ensures that only truly valid data will be decrypted and stored in the data storage part, improving the density of the valid data stored in the data storage part. At the same time, since a large amount of invalid data is filtered out, the amount of data that needs to be processed and stored in the data storage part is reduced, thereby significantly reducing the storage pressure of the data storage part and improving the operation efficiency and stability of the entire system.

[0061] In addition, in terms of risk prediction, the data management system of the lithography machine according to the embodiments of the present application combines the data records of the lithography machine software and uses a classifier such as a support vector machine to predict the security risk level of the information in the data records. Thus, the data records can be effectively cleaned. Moreover, by combining methods such as the bag-of-words model to extract the key features of the data records, the feature dimension can be reduced and the model training efficiency can be improved. That is, the data management system of the lithography machine according to the embodiments of the present application can realize real-time processing of new data records and prediction of the risk level. By combining the data visualization part to display the results, early warning and response measures can be accurately taken according to different risk levels, comprehensively ensuring the stable and safe operation of the lithography machine software.

[0062] Those skilled in the art can understand that the data management system of the lithography machine according to the embodiments of the present application can adopt an information interaction framework based on the ELK (FileBeat, Elasticsearch, Logstash, Kibana) technology stack. For example, as Figure 3 shown, but other solutions can also be adopted. Moreover, in the field of data management of lithography machines, due to the different functional characteristics between the ELK technology stack and other solutions, such as Kafka, unique advantages can be demonstrated in many aspects. Here, Figure 3 FIG. shows a schematic diagram of the data management system of the lithography machine according to the embodiments of the present application implemented based on the ELK technology stack.

[0063] In terms of data processing and analysis capabilities, the ELK technology stack can achieve real-time full-text indexing and complex queries. Specifically, during the operation of a lithography machine, a large amount of complex data such as equipment status, process parameters, and various monitoring data during the exposure process will be generated. Elasticsearch can quickly index this data, and engineers can use its powerful query language (such as Query DSL) to perform complex queries at nearly real-time speeds. For example, when analyzing the pattern quality data of specific batches of lithography products under different exposure times and power combinations, relevant data can be quickly located and filtered out. Kafka is mainly a distributed stream processing platform that focuses on the efficient transmission and buffering of data. In terms of real-time indexing and complex query analysis of data, it is far less professional and powerful than Elasticsearch and is difficult to directly meet the needs of in-depth analysis of lithography process data.

[0064] Moreover, the ELK technology stack can achieve data integration and correlation analysis. Specifically, Logstash can collect data from various data sources of the lithography machine, such as sensors, control system logs, and operation records of different subsystems, and standardize and enrich the data according to specific rules, and then transmit the processed data to Elasticsearch. This enables lithography data from different devices and of different types to be integrated together for correlation analysis. For example, correlating the laser energy data of the lithography machine's exposure system with the motion accuracy data of the workbench to analyze their combined impact on the accuracy of the final lithography pattern. Although Kafka can efficiently transmit these data from different sources, it lacks the ability to deeply integrate and preprocess data to support complex correlation analysis.

[0065] In addition, in terms of data visualization and interactivity, the ELK technology stack can achieve rich and intuitive visual displays. Specifically, Kibana, as the visualization tool in the ELK technology stack, can present the lithography data stored in Elasticsearch in a rich, diverse, intuitive, and easy-to-understand manner. Real-time monitoring charts can be created to show the changing trends of the operating status parameters of the key components of the lithography machine (such as light source, objective lens, workbench, etc.) over time; it can also present the differences in various quality indicators of different batches of lithography products in the form of comparison charts. This kind of visual display allows engineers to quickly grasp the key information in the lithography process, discover problems and potential trends in a timely manner without spending a lot of time interpreting complex data reports. In contrast, Kafka itself does not provide powerful data visualization functions. To achieve similar visual analysis of lithography data, additional visualization tools need to be integrated, and it is difficult to compare with Kibana in the ELK technology stack in terms of the convenience of data docking and visualization customization.

[0066] Moreover, the ELK technology stack enables flexible interactive exploration. Specifically, when engineers use Kibana, they can flexibly conduct interactive exploration of the displayed data. For example, through operations such as zooming, filtering, and switching different dimensions, they can deeply analyze the details of the data of interest. For instance, when viewing the visualization chart of lithography pattern defect data, they can quickly locate information such as the batches and process conditions where specific types of defects occur more frequently through interactive operations, so as to targetedly optimize the process or debug the equipment. In the application scenarios of Kafka, such flexible data interactive exploration functions are usually not directly provided, which is not conducive to engineers quickly obtaining valuable information from a large amount of lithography data.

[0067] In addition, in terms of system architecture and scalability, the ELK technology stack can achieve the integrity and coordination of the overall architecture. Specifically, the ELK technology stack forms a complete ecosystem from data collection, processing, storage to visualization. Each component collaborates closely. FileBeat and Logstash are responsible for data collection and processing, Elasticsearch for efficient storage and retrieval, and Kibana for data visualization. This architecture with clear division of labor and collaborative work is very suitable for the data management and analysis scenarios of complex equipment such as lithography machines. In practical applications, each component can be horizontally or vertically scaled independently according to the amount of lithography data and processing requirements to adapt to lithography production environments of different scales and complexities. Although Kafka performs well in data transmission, it is only one link in the entire data processing process. To build a complete system suitable for lithography data processing, a large amount of additional work is required to integrate other data processing, storage, and visualization tools, and the complexity of system construction and maintenance is relatively high.

[0068] Moreover, the ELK technology stack can handle the scalability of lithography technology development. Specifically, as lithography technology continues to develop towards higher precision and higher efficiency, the requirements for the scalability of the equipment data processing system are also increasing day by day. The Elasticsearch cluster in the ELK technology stack supports horizontal scaling and can easily handle the growth of data volume by adding nodes. Both storage capacity and processing power can achieve linear scaling. For example, when a new and more advanced lithography machine is introduced into a lithography factory and the data volume increases significantly, the ELK technology stack can conveniently improve the overall performance by adding data nodes to ensure the smoothness and efficiency of data processing. Although the expansion of Kafka in the data transmission link is also relatively flexible, in the entire data processing chain, when facing more complex data processing requirements (such as more complex data analysis algorithms, fusion processing of more types of data, etc.) accompanied by the development of lithography technology, its limitations in scalability will gradually emerge, and it is difficult to adapt to the data processing challenges brought by the continuous development of lithography technology as comprehensively and flexibly as the ELK technology stack.

[0069] The basic principles of the present application have been described in conjunction with specific embodiments. However, it should be noted that the advantages, benefits, effects, etc. mentioned in the present application are only examples and not limitations. It cannot be considered that these advantages, benefits, effects, etc. are essential for each embodiment of the present application. Additionally, the specific details disclosed above are only for illustrative and easy-to-understand purposes, not limitations. These details do not limit the present application to necessarily adopting the above specific details for implementation.

[0070] The block diagrams of the devices, apparatuses, equipment, and systems involved in the present application are only illustrative examples and do not intend to require or imply that they must be connected, arranged, and configured in the manner shown in the block diagrams. As those skilled in the art will recognize, these devices, apparatuses, equipment, and systems can be connected, arranged, and configured in any manner. Words such as "including", "comprising", "having", etc. are open-ended terms, meaning "including but not limited to", and can be used interchangeably with each other. The word "or" and "and" used herein refer to the word "and / or" and can be used interchangeably with it, unless the context clearly indicates otherwise. The word "such as" used herein refers to the phrase "such as but not limited to" and can be used interchangeably with it.

[0071] It should also be noted that in the devices, equipment, and methods of the present application, each component or each step can be decomposed and / or recombined. These decompositions and / or recombinations should be regarded as equivalent solutions of the present application.

[0072] The above description of the disclosed aspects is provided to enable any person skilled in the art to make or use the present application. Various modifications to these aspects are very obvious to those skilled in the art, and the general principles defined herein can be applied to other aspects without departing from the scope of the present application. Therefore, the present application is not intended to be limited to the aspects shown herein, but rather to the broadest scope consistent with the principles and novel features disclosed herein.

[0073] The above description has been given for purposes of illustration and description. Additionally, this description does not intend to limit the embodiments of the present application to the forms disclosed herein. Although multiple example aspects and embodiments have been discussed above, those skilled in the art will recognize certain variations, modifications, changes, additions, and sub-combinations thereof.

Claims

1. A data management system for a lithography machine, wherein: include: a data collection part, for recording the collected data based on a data recording component suitable for industrial software to obtain a data record, wherein the data record includes mapped diagnostic context information and / or nested diagnostic context information corresponding to a predetermined operation; a data processing part, used for receiving the encrypted data record from the data acquisition part and performing data processing, wherein the data processing includes filtering the encrypted data record and estimating the subsystem risk of the lithography machine; and The data storage part is used to store data related to the data acquisition part and the data processing part.

2. The data management system for a lithography machine according to claim 1, wherein: The data collection part sets at least one of the output time of the data record, the level of the data record, the storage path of the data record, the detailed content of the data record and the rolling strategy of the data record file.

3. The data management system for a lithography machine according to claim 1, wherein: The mapped diagnostic context information stores the task information of the predetermined operation in the form of key-value pairs, and the nested diagnostic context information stores the process and hierarchical structure information of the predetermined operation in the form of a stack.

4. The data management system for a lithography machine according to claim 3, wherein: When obtaining the data record corresponding to the predetermined operation, the data collection part adds the mapping diagnostic context information of the current task thread of the operation to the data record, and stores the hierarchical path of the operation in the nested diagnostic context information stack.

5. The data management system for a lithography machine according to claim 1, wherein: The data record includes information for identifying different task threads of the lithography machine operation and / or information for identifying different subsystems of the lithography machine.

6. The data management system for a lithography machine according to claim 1, wherein: The encrypted data record is a data record encrypted using a combination of an AES encryption algorithm and a homomorphic encryption algorithm.

7. The data management system for a lithography machine according to claim 6, wherein: The encrypted data record is a data record encrypted by the data acquisition part using an AES encryption algorithm, and the data processing part directly performs calculations and analysis on the encrypted data record using a homomorphic encryption algorithm.

8. The data management system for a lithography machine according to claim 7, wherein: The filtering of the encrypted data records includes: Verifying whether the encrypted data record meets valid data standards; Responsive to the encrypted data record meeting valid data criteria, parsing the encrypted data record; and In response to the encrypted data record not meeting valid data criteria, the encrypted data record is filtered.

9. The data management system for a lithography machine according to claim 1, wherein: The estimation of the subsystem risk of the lithography machine includes evaluating the subsystem risk of the lithography machine using a pre-trained model, and evaluating the subsystem risk of the lithography machine using the pre-trained model includes: Collect data records of various components and related subsystems of the lithography machine software; labeling the data records as different risk levels and creating a training data set in predetermined proportions according to the different risk levels; training a classifier using the training data set; and Data features are extracted from the subsystem data newly collected by the data collection part and input into the classifier after being quantified to predict the risk level of the newly collected subsystem data.

10. The data management system for a lithography machine according to claim 9, wherein: Further including: The data visualization part is used to visualize the data stored in the data storage part.

Images