Methods, devices, and systems for automatically indexing and retrieving manufacturing data metadata

Abstract

An apparatus according to one embodiment collects manufacturing process history data including manufacturing data, and based on the manufacturing process history data, extracts equipment data, material data, process condition data, and manufacturing environment data, which are metadata constituting each manufacturing data, to generate metadata combinations, generates combination scores for each metadata combination based on the manufacturing process history data, selects manufacturing data corresponding to a metadata combination with a combination score higher than a target score as optimal manufacturing data, automatically generates an index for searching manufacturing data based on the optimal manufacturing data, receives specific metadata constituting search conditions from a user's terminal, extracts specific optimal manufacturing data corresponding to a metadata combination matching the specific metadata through the automatically generated index, and provides the specific optimal manufacturing data to the user's terminal.

Description

Technology Field

[0001] The following embodiments relate to a technology that generates a metadata combination from a manufacturing process history, scores it, indexes manufacturing data exceeding a standard as optimal manufacturing data, and then automatically searches for and provides manufacturing data that matches metadata conditions entered by a user. Background Technology

[0002] In the manufacturing process, manufacturing data is generated by combining various elements such as equipment information, material information, process condition information, and manufacturing environment information, and this data is utilized as key information for quality control, process analysis, and improvement of production efficiency.

[0003] However, in actual industrial settings, manufacturing data is often stored individually by process and equipment, which limits the ability to collectively identify the relationships between various metadata constituting the data or to analyze them at the combination level.

[0004] Furthermore, because manufacturing process history accumulates in a vast and complex form, it is not easy to quickly search for manufacturing data that satisfies specific conditions or to select and utilize only high-quality manufacturing data.

[0005] Existing data search methods often rely on simple keyword-based searches or single-condition filtering, which limits their ability to efficiently search manufacturing data involving a complex combination of equipment, materials, and process conditions.

[0006] As a result, systematic indexing based on the entire combination of metadata included in the manufacturing data cannot be achieved, and the functionality to automatically provide suitable manufacturing data based on specific metadata is also insufficient.

[0007] In particular, due to the lack of capabilities to quantitatively compare quality deviations between manufacturing data or to evaluate performance by data combination to select superior combinations, data-driven decision support and process optimization are not being smoothly implemented in manufacturing sites.

[0008] Therefore, there is a growing need for technology that automatically generates metadata combinations from manufacturing process history, scores the quality of each combination to derive optimal manufacturing data according to standardized criteria, and supports the efficient retrieval of manufacturing data that meets user requirements. Prior art literature

[0009] Korean Registered Patent No. 10-2722862 (Published Oct. 29, 2024) Korean Registered Patent No. 10-2127785 (Published Jun. 29, 2020) Korean Registered Patent No. 10-2791759 (Published Apr. 3, 2025) Korean Registered Patent No. 10-1527994 (Published Jun. 10, 2015) The problem to be solved

[0010] The embodiments aim to search for and provide manufacturing data that matches specific metadata conditions entered by a user by generating a metadata combination including information on equipment, materials, process conditions, and manufacturing environment from a manufacturing process history and automatically configuring the said metadata combination into an index.

[0011] The embodiments calculate a multi-stage combination score based on the ratio of good products, cumulative usage patterns, economic efficiency, and energy efficiency of metadata combinations, and thereby select manufacturing data having performance above a standard as optimal manufacturing data.

[0012] The embodiments aim to derive specific optimal manufacturing data that optimally matches user conditions by determining metadata combinations that completely or partially match specific metadata entered by a user and comparing the corresponding combination scores.

[0013] The objectives of the present invention are not limited to those mentioned above, and other unmentioned objectives will be clearly understood from the description below. means of solving the problem

[0014] A method for automatic indexing and search optimization of manufacturing data metadata comprises: a step of collecting manufacturing process history data containing manufacturing data; a step of generating metadata combinations by extracting equipment data, material data, process condition data, and manufacturing environment data, which are metadata constituting each manufacturing data, based on the manufacturing process history data; a step of generating a combination score for each of the metadata combinations based on the manufacturing process history data; a step of selecting manufacturing data corresponding to a metadata combination in which the combination score is higher than a preset target score as optimal manufacturing data; a step of automatically generating an index for searching manufacturing data based on the optimal manufacturing data; a step of receiving specific metadata constituting search conditions from a user's terminal; a step of extracting specific optimal manufacturing data that matches the specific metadata through the automatically generated index; and a step of providing the specific optimal manufacturing data to the user's terminal.

[0015] The step of generating a combination score for each of the metadata combinations based on the manufacturing process history data comprises: calculating the ratio of good products of the manufacturing result for each of the metadata combinations, assigning a combination score of 0 for metadata combinations where the ratio of good products is less than a preset reference ratio, and selecting metadata combinations that are greater than or equal to the reference ratio as candidate combinations; for each of the candidate combinations, generating a recency weight through the time difference between the last time of use and the current time, and generating a first combination score by applying the recency weight to the cumulative number of uses; for each of the candidate combinations, calculating the time and cost required to produce one manufacturing result, and generating a second combination score by evaluating economic efficiency based on the time and cost; for each of the candidate combinations, generating a third combination score based on the energy efficiency used to produce one manufacturing result; and generating a final combination score for the candidate combinations based on the first combination score, the second combination score, and the third combination score.

[0016] The step of extracting specific optimal manufacturing data that matches the specific metadata through the automatically generated index; comprises: checking whether there exists complete matching optimal manufacturing data that includes all of the specific metadata through the automatically generated index; if it is confirmed that the complete matching optimal manufacturing data exists, extracting the complete matching optimal manufacturing data with the highest combination score as the specific optimal manufacturing data; if it is confirmed that the complete matching optimal manufacturing data does not exist, checking whether there exists partial matching optimal manufacturing data that includes some of the specific metadata; if it is confirmed that the partial matching optimal manufacturing data exists, generating a similarity weight based on the similarity between the partial matching optimal manufacturing data and the specific metadata; generating a cost weight based on the change cost required to change the specific metadata not included in the partial matching optimal manufacturing data; generating a corrected combination score by applying the similarity weight and the cost weight to the combination score of the partial matching optimal manufacturing data; extracting the partial matching optimal manufacturing data with the highest corrected combination score as the specific optimal manufacturing data; and if it is confirmed that the partial matching optimal manufacturing data does not exist, determining that there is no specific optimal manufacturing data that matches the specific metadata, and replacing the optimal manufacturing data with the highest combination score as the optimal manufacturing data. Includes the operation of extracting data.

[0017] The step of providing the specific optimal manufacturing data to the user's terminal includes: classifying each metadata included in the specific optimal manufacturing data into maintenance data, change data, and irrelevant data based on whether it matches the specific metadata; displaying the maintenance data in a preset first color to provide the specific optimal manufacturing data to the user's terminal; displaying the change data in a preset second color to provide the specific optimal manufacturing data to the user's terminal; and displaying the irrelevant data in a preset third color to provide the specific optimal manufacturing data to the user's terminal.

[0018] The step of providing the specific optimal manufacturing data to the user's terminal further includes, based on the fact that the change data includes at least one of equipment data or material data, the operation of searching for a plurality of external sales sites for the equipment or material corresponding to the change data, the operation of identifying a sales site available for purchase at the lowest price among the external sales sites as a recommended sales site, and the operation of providing a sales link to the recommended sales site along with the specific optimal manufacturing data to the user's terminal.

[0019] A method for automatic indexing and search optimization of manufacturing data metadata further comprises: a step of setting metadata used to perform the most recent manufacturing process as specific metadata based on manufacturing process history data received from the user's terminal when at least one of the scheduled time for equipment replacement, the time for material depletion, the time for new equipment installation, the time for the start of a manufacturing process, or the time for a change in the manufacturing environment is detected to have arrived; a step of querying the automatically generated index based on the specific metadata to extract specific optimal manufacturing data based on candidate time points that matches the specific metadata; and a step of providing the specific optimal manufacturing data based on candidate time points to the user's terminal in advance.

[0020] A device according to one embodiment may be combined with hardware and controlled by a computer program stored on a medium to execute the method of any one of the methods described above. Effects of the invention

[0021] The embodiments generate a metadata combination including information on equipment, materials, process conditions, and manufacturing environment from a manufacturing process history, and automatically configure the said metadata combination into an index, thereby enabling the search and provision of manufacturing data that matches specific metadata conditions entered by a user.

[0022] The embodiments calculate a multi-stage combination score based on the ratio of good products, cumulative usage patterns, economic efficiency, and energy efficiency of the metadata combination, and thereby select manufacturing data having performance above a standard as optimal manufacturing data.

[0023] The embodiments can derive specific optimal manufacturing data that optimally matches user conditions by determining metadata combinations that completely or partially match specific metadata entered by a user and comparing the corresponding combination scores.

[0024] Meanwhile, the effects according to the embodiments are not limited to those mentioned above, and other unmentioned effects will be clearly understood by those skilled in the art from the description below. Brief explanation of the drawing

[0025] FIG. 1 is a schematic diagram showing the configuration of a system according to one embodiment. FIG. 2 is a flowchart illustrating the process of automatically indexing manufacturing data metadata and optimizing search according to one embodiment. FIG. 3 is a flowchart illustrating the process of generating a combination score according to one embodiment. FIG. 4 is a flowchart illustrating the process of extracting specific optimal manufacturing data according to one embodiment. FIG. 5 is a flowchart illustrating the process of providing specific optimal manufacturing data to a user's terminal according to one embodiment. FIG. 6 is a flowchart illustrating the process of providing a sales link together with the process of providing modified data according to one embodiment. FIG. 7 is a flowchart illustrating a process for automatically extracting specific optimal manufacturing data and providing it to a user at a time when specific optimal manufacturing data is needed, according to one embodiment. FIG. 8 is an example diagram of the configuration of a device according to one embodiment. Specific details for implementing the invention

[0026] Hereinafter, embodiments are described in detail with reference to the attached drawings. However, various modifications may be made to the embodiments, and thus the scope of the patent application is not limited or restricted by these embodiments. It should be understood that all modifications, equivalents, and substitutions to the embodiments are included within the scope of the rights.

[0027] Specific structural or functional descriptions of the embodiments are disclosed for illustrative purposes only and may be modified and implemented in various forms. Accordingly, the embodiments are not limited to the specific disclosed forms, and the scope of this specification includes modifications, equivalents, or substitutions that fall within the technical concept.

[0028] Terms such as "first" or "second" may be used to describe various components, but these terms should be interpreted solely for the purpose of distinguishing one component from another. For example, the first component may be named the second component, and similarly, the second component may be named the first component.

[0029] When it is stated that a component is "connected" to another component, it should be understood that it may be directly connected to or joined to that other component, or that there may be other components in between.

[0030] The terms used in the embodiments are for illustrative purposes only and should not be interpreted as intended to be limiting. Singular expressions include plural expressions unless the context clearly indicates otherwise. In this specification, terms such as "comprising" or "having" are intended to indicate the existence of the features, numbers, steps, actions, components, parts, or combinations thereof described in the specification, and should be understood as not precluding the existence or addition of one or more other features, numbers, steps, actions, components, parts, or combinations thereof.

[0031] Unless otherwise defined, all terms used herein, including technical or scientific terms, have the same meaning as generally understood by those skilled in the art to which the embodiments pertain. Terms such as those defined in commonly used dictionaries should be interpreted as having a meaning consistent with their meaning in the context of the relevant technology, and should not be interpreted in an ideal or overly formal sense unless explicitly defined in this application.

[0032] In addition, when describing with reference to the attached drawings, identical components are assigned the same reference numeral regardless of drawing symbols, and redundant descriptions thereof are omitted. In describing the embodiments, if it is determined that a detailed description of related prior art could unnecessarily obscure the essence of the embodiments, such detailed description is omitted.

[0033] The embodiments can be implemented in various forms of products such as personal computers, laptop computers, tablet computers, smartphones, televisions, smart home appliances, intelligent automobiles, kiosks, and wearable devices.

[0034] FIG. 1 is a schematic diagram showing the configuration of a system according to one embodiment.

[0035] Referring to FIG. 1, a system according to one embodiment may include a user terminal (100) and a device (200) capable of communicating with each other through a communication network.

[0036] First, the communication network can be configured regardless of the mode of communication, such as wired or wireless, and can be implemented in various forms to enable communication between servers and between servers and terminals.

[0037] The user's terminal (100) is a terminal used by a user who manages the smooth execution of the manufacturing process and plans equipment, materials, process conditions, and manufacturing environment, and can be implemented as a mobile phone, desktop PC, laptop PC, tablet PC, smartphone, etc., but is not limited thereto, and can also be implemented as various types of communication devices that can be connected to an external server. For example, as shown in FIG. 1, the user's terminal (100) may be a smartphone, and may be adopted differently depending on the embodiment. At this time, the user may generally be a manufacturing process manager or manufacturing operation manager who operates the manufacturing process, but is not limited thereto.

[0038] The user's terminal (100) may be configured to perform all or part of the computational functions, storage / reference functions, input / output functions, and control functions that a conventional computer has. The user's terminal (100) may be configured to communicate with the device (200) via wired or wireless means.

[0039] The user's terminal (100) may be connected to a website operated by a person or organization providing a service using the device (200), or may have an application developed and distributed by a person or organization providing a service using the device (200) installed. The user's terminal (100) may be linked with the device (200) through the website or application.

[0040] In FIG. 1 and the description below, for convenience of explanation, only one user terminal (100) is illustrated and described, but the number of terminals can vary depending on the embodiment. As long as the processing capacity of the device (200) allows, there is no particular limit to the number of terminals.

[0041] The device (200) may be a private server owned by a person or organization providing a service using the device (200), a cloud server, or a peer-to-peer (P2P) set of distributed nodes. The device (200) may be configured to perform all or part of the computational functions, storage / reference functions, input / output functions, and control functions that a conventional computer possesses. The device (200) may be configured to communicate with the user's terminal (100) via wired or wireless means.

[0042] Additionally, the device (200) can communicate wirelessly or via wired connection with websites including SNS platforms such as blogs, cafes, Instagram, Facebook, Twitter, and YouTube, and web pages including articles, and the device (200) can access the websites to obtain information.

[0043] Additionally, the device (200) collects manufacturing process history data including manufacturing data, and based on the manufacturing process history data, extracts equipment data, material data, process condition data, and manufacturing environment data, which are metadata constituting each manufacturing data, to generate metadata combinations, and based on the manufacturing process history data, generates combination scores for each metadata combination, selects manufacturing data corresponding to a metadata combination with a combination score higher than a target score as optimal manufacturing data, automatically generates an index for searching manufacturing data based on the optimal manufacturing data, receives specific metadata constituting search conditions from the user's terminal (100), extracts specific optimal manufacturing data corresponding to a metadata combination that matches the specific metadata through the automatically generated index, and can provide the specific optimal manufacturing data to the user's terminal (100).

[0044] Meanwhile, singular expressions described in the claims and the description of the invention may be understood to include plurals.

[0045] In the present invention, Artificial Intelligence (AI) refers to a technology that imitates human learning ability, reasoning ability, and perceptual ability, and implements them on a computer, and may include concepts such as machine learning and symbolic logic. Machine Learning (ML) is an algorithmic technology that classifies or learns the characteristics of input data on its own. AI technology can analyze input data as a machine learning algorithm, learn from the results of the analysis, and make judgments or predictions based on the results of the learning. Furthermore, technologies that mimic the functions of the human brain, such as cognition and judgment, by utilizing machine learning algorithms can also be understood as falling within the category of AI. For example, technological fields such as linguistic understanding, visual understanding, reasoning / prediction, knowledge representation, and motion control may be included.

[0046] Machine learning can refer to the process of training neural network models using experience in processing data. It implies that through machine learning, computer software improves its own data processing capabilities. A neural network model is constructed by modeling the correlations between data, and these correlations can be expressed by multiple parameters. A neural network model extracts and analyzes features from given data to derive correlations between them; machine learning can be defined as the process of optimizing the model's parameters by repeating this process. For example, a neural network model can learn the mapping (correlation) between inputs and outputs for data given as input-output pairs. Alternatively, even when only input data is provided, a neural network model can derive regularities between the given data and learn those relationships.

[0047] An artificial intelligence learning model or neural network model can be designed to implement the structure of the human brain on a computer and may include multiple network nodes that have weights and simulate neurons of a human neural network. The multiple network nodes may have interconnected relationships by simulating the synaptic activity of neurons, where neurons exchange signals through synapses. In an artificial intelligence learning model, multiple network nodes may be located in layers of different depths and exchange data according to convolutional connections. The artificial intelligence learning model may be, for example, an Artificial Neural Network (ANN) or a Convolutional Neural Network (CNN). As an embodiment, the artificial intelligence learning model may be machine learned according to methods such as supervised learning, unsupervised learning, and reinforcement learning. Machine learning algorithms for performing machine learning may include Decision Tree, Bayesian Network, Support Vector Machine, Artificial Neural Network, Ada-boost, Perceptron, Genetic Programming, and Clustering.

[0048] Among these, CNNs are a type of multilayer perceptron designed to use minimal preprocessing. CNNs consist of one or more convolutional layers and standard artificial neural network layers stacked on top, additionally utilizing weights and pooling layers. Thanks to this structure, CNNs can fully utilize two-dimensional input data. Compared to other deep learning architectures, CNNs demonstrate good performance in both image and audio fields. CNNs can also be trained using standard backpropagation. CNNs have the advantage of being easier to train than other feedforward artificial neural network techniques and using a small number of parameters.

[0049] Convolutional networks are neural networks comprising sets of nodes with bounded parameters. Many computer vision tasks have been significantly improved, driven by the increased size of available training data and the availability of computational power, combined with algorithmic advancements such as discriminative linear units and dropout training. In the case of massive datasets, such as those available for many tasks today, outfitting is not critical, and increasing the network size improves test accuracy. Optimal utilization of computing resources becomes a limiting factor. To address this, distributed, scalable implementations of deep neural networks can be employed.

[0050] FIG. 2 is a flowchart illustrating the process of automatically indexing manufacturing data metadata and optimizing search according to one embodiment.

[0051] Referring to FIG. 2, first in step S201, the device (200) can collect manufacturing process history data including manufacturing data.

[0052] Specifically, the device (200) can collect manufacturing process history data including manufacturing data generated during the execution of a manufacturing process through a user's terminal (100) that communicates with the device (200) via wired or wireless means, or through an external manufacturing system (e.g., MES, ERP, facility control system, SCADA, PLC-based automation equipment, sensor network system, quality inspection equipment, environmental monitoring device, etc.).

[0053] Here, manufacturing process history data refers to a set of data recorded corresponding to the time and work unit at which a specific manufacturing process was actually performed, and each manufacturing process history data may be composed of manufacturing data units that include equipment data, material data, process condition data, and manufacturing environment data constituting the manufacturing process.

[0054] Equipment data refers to data regarding equipment used in the process of performing a manufacturing process. For example, equipment data may include, but is not limited to, items such as equipment ID, equipment name, equipment type (e.g., cutting machine, molding machine, injection molding machine, etc.), process steps performed by the equipment (machining, molding, assembly, inspection, etc.), start and end times of equipment operation, operating time and cumulative operating time of the equipment in the manufacturing process, date of the last inspection of the equipment, maintenance history, failure history, frequency or number of uses of the equipment, and equipment status data (e.g., sensor data measuring temperature, vibration, current values, etc.).

[0055] Material data refers to data regarding raw materials and auxiliary materials input into a manufacturing process. For example, material data may include, but is not limited to, items such as material name, material code, barcode / QR code ID, batch number, production lot number, supplier information (supplier name, supplier code), input quantity, input time, input sequence, mixing or blending ratio, usage stage and method, material specifications, grade, and specification information.

[0056] Process condition data refers to operating condition values ​​set in the equipment when a manufacturing process is performed. For example, process condition data may include, but is not limited to, items such as temperature, pressure, rotational speed (RPM), current, voltage, torque value, machining speed, machining time, cutting depth, forming pressure, humidity, heat treatment temperature, cooling time, drying time, inter-equipment interoperability conditions, and control parameter values.

[0057] Manufacturing environment data refers to external environmental data of the workplace where the manufacturing process is performed. For example, manufacturing environment data may include, but is not limited to, items such as workplace temperature, humidity, air quality, particulate matter concentration, carbon dioxide concentration, illuminance (brightness of lighting), noise level, workplace ventilation status, and outside temperature.

[0058] Meanwhile, the manufacturing process history data may further include quality inspection data for manufactured products produced based on the manufacturing data, and for example, the quality inspection data may include, but is not limited to, items such as inspection item names (dimensional inspection, appearance inspection, strength inspection, etc.), inspection standard values ​​and tolerance ranges, actual measured values, good / defective product judgment values, inspection time and inspection personnel information, inspection equipment ID, and inspection equipment calibration history.

[0059] The above manufacturing process history data can be directly input from the user's terminal (100) or automatically generated by an external manufacturing system and transmitted to the device (200). The device (200) stores the received manufacturing process history data in a database (DB) provided in the device (200) and can subsequently be used for metadata combination generation, combination score calculation, index generation, search processing, etc.

[0060] Meanwhile, the items constituting the above manufacturing process history data are exemplary, and some items may be added, specific items omitted, or the structure of the items changed depending on the type of manufacturing system, process environment, and user requirements.

[0061] This embodiment should be understood to include various variations and extended forms.

[0062] In step S202, the device (200) can generate a metadata combination by extracting equipment data, material data, process condition data, and manufacturing environment information, which are metadata constituting each manufacturing data, based on manufacturing process history data.

[0063] Specifically, the device (200) analyzes a plurality of manufacturing data included in the manufacturing process history data collected in step S201 and parses the internal structure of the manufacturing process history data to identify field values ​​mapped to each manufacturing data, thereby separating and extracting items corresponding to equipment data, material data, process condition data, and manufacturing environment data.

[0064] For example, the device (200) can identify items related to equipment data (e.g., equipment ID, equipment name, equipment type, process step, etc.), items related to material data (e.g., material name, material code, batch number, supplier information, etc.), items related to process condition data (e.g., process temperature, process pressure, rotational speed (RPM), processing time, etc.), and items related to manufacturing environment data (e.g., workplace temperature, humidity, illuminance, etc.) within the manufacturing process history data and separate them into metadata elements.

[0065] In addition, the device (200) can perform normalization to convert the same item into a unified structure, taking into account that manufacturing process history data may use different data formats or item names depending on the external manufacturing system (MES, ERP, sensor network, etc.).

[0066] For example, if the same equipment identification item is recorded with different names such as “machine_id”, “equip_id”, or “equipment ID”, the device (200) can unify them into the same item using an internal mapping table or a keyword dictionary. This process can be applied in the same way to material data, process condition data, and manufacturing environment data.

[0067] If there are missing values ​​in the manufacturing process history data, the device (200) may correct the item to a default value according to a predefined missing value processing rule, replace it with a representative value (such as an average or median) used during a recent period of the same manufacturing process, or exclude the missing item and configure the metadata combination. This data cleaning process is intended to ensure the reliability of the metadata combination.

[0068] Meanwhile, the device (200) may include only the core items that constitute the metadata combination among the extracted equipment data, material data, process condition data, and manufacturing environment data, taking into account the characteristics of the manufacturing process and the representativeness of the data. For example, in the case of equipment data, only items that directly influence the determination of the combination conditions of the manufacturing process, such as “Equipment ID,” “Equipment Name,” “Equipment Type,” and “Process Step,” may be included, and non-essential items that vary by process, such as the start time of equipment operation, the end time, inspection history, and sensor values, may be excluded from the metadata combination. This can be applied equally to material data, process condition data, and manufacturing environment data, and the metadata combination may be composed by including only the core items that are essential for composing the combination in each data group.

[0069] For example, the device (200) can generate a metadata combination by grouping key items extracted from one manufacturing data, such as {“Equipment ID A”, “Material Code B”, “Temperature 180°C and Pressure 20 bar”, “Workplace Temperature 28°C”}. The constituent items of the generated metadata combination can be expanded or contracted according to the type of manufacturing process, equipment structure, product quality requirements, and user settings, and, for example, if quality inspection data of the manufacturing result influences the metadata combination selection process, the quality inspection item may be included in the metadata combination.

[0070] Additionally, each metadata element constituting the metadata combination may be expressed as a single value, but depending on the embodiment, it may be expressed as a range or interval of values ​​instead of a single value.

[0071] For example, the workplace temperature of the manufacturing environment data may be recorded as a specific value such as “28℃,” but according to the embodiment, it may also be expressed as a specific range or an allowable range such as “25℃ to 30℃.” These range values ​​may be set considering the variability of the manufacturing process, and the device (200) may treat the range-type values ​​as well as components of the metadata combination. This method of expressing ranges may be applied equally to process condition data (e.g., process temperature 170℃ to 190℃, pressure 18 bar to 22 bar), material data (e.g., mixing ratio 40% to 45%), equipment data (e.g., equipment operating speed 1,500 RPM to 1,700 RPM), etc., and the device (200) may utilize both single values ​​and range values ​​as valid metadata in the metadata combination generation and subsequent combination score calculation and search processes.

[0072] In addition, if multiple manufacturing data are recorded in the same manufacturing process history data, the device (200) can independently generate and store metadata combinations for each manufacturing data.

[0073] In step S203, the device (200) can generate a combination score for each metadata combination based on manufacturing process history data.

[0074] Specifically, for each of the multiple metadata combinations generated in step S202, the device (200) can calculate multiple evaluation items capable of evaluating the performance of the metadata combination, such as the quality level of the manufacturing result, frequency of use, efficiency, and energy consumption, and can combine the evaluation items to generate a single combination score.

[0075] At this time, the combination score is an indicator to comprehensively reflect the performance and reliability when the metadata combination is used in an actual manufacturing process, and the specific process for generating the combination score will be explained with reference to Fig. 3.

[0076] In step S204, the device (200) can select manufacturing data corresponding to a metadata combination in which the combination score is higher than the target score as optimal manufacturing data.

[0077] Here, the target score is a preset value and may vary depending on the example.

[0078] Specifically, the device (200) can check the combination score of each metadata combination generated in step S203 and the process of FIG. 3 and compare it with a preset target score.

[0079] Additionally, the device (200) can classify metadata combinations in which the combination score is determined to be greater than or equal to the target score as “recommended suitable metadata combinations.”

[0080] For example, if the combination scores calculated in step S203 are 0.82 for the first metadata combination, 0.65 for the second metadata combination, and 0.91 for the third metadata combination, and the target score is 0.80, the device (200) can select the first and third metadata combinations as recommended suitable metadata combinations.

[0081] Additionally, the device (200) can finally designate the manufacturing data corresponding to the recommended suitable metadata combination selected in this way (i.e., a combination of equipment data, material data, process condition data, and manufacturing environment data included in the recommended suitable metadata combination) as the “optimal manufacturing data.”

[0082] In step S205, the device (200) can automatically generate an index for searching manufacturing data based on optimal manufacturing data.

[0083] Specifically, the device (200) can analyze the optimal manufacturing data selected in step S204 and perform a process of converting the metadata elements (equipment data, material data, process condition data, manufacturing environment data, etc.) included in the optimal manufacturing data into an index structure so that they can be used as search keys.

[0084] First, the device (200) can identify the value of each metadata element included in the optimal manufacturing data and the attribute of the element, and convert it into an indexable key-value form or a multi-dimensional index form based on this.

[0085] For example, each item such as equipment ID, equipment type, process step / material name, batch number, supplier information / process temperature, process pressure, rotational speed (RPM) / workplace temperature, humidity, illuminance, etc., can be generated as an independent search key or as a combination search key consisting of multiple items.

[0086] The device (200) can use various data structures, such as hash tables, B-Tree, R-Tree, KD-Tree, inverted indexes, and graph-based indexes, which are commonly used as index structures for identifying optimal manufacturing data, and each data structure can be selected according to the scale of manufacturing data, search speed requirements, etc.

[0087] Additionally, the device (200) can generate multiple indexes based on each metadata item so that the same optimal manufacturing data can be retrieved by different metadata items.

[0088] For example, search efficiency can be maximized by creating separate indexes based on equipment, materials, process conditions, and manufacturing environments.

[0089] Meanwhile, the device (200) may check for duplicate metadata during the process of creating an index, merge duplicate metadata into the same index item, or select a representative value according to a specified priority rule to improve the storage efficiency of the index.

[0090] Additionally, if additional optimal manufacturing data is generated or existing optimal manufacturing data is updated, the device (200) may be configured to automatically update (refresh or re-index) the corresponding index.

[0091] And the device (200) stores the generated index in a database provided in the device (200), and subsequently, when a user inputs specific metadata in step S206, enables the device to quickly search for manufacturing data that matches the specific metadata based on the generated index.

[0092] In this way, at step S205, the device (200) can automatically build an index that corresponds to various search conditions from optimal manufacturing data, thereby providing a fast and accurate search processing environment for various metadata-based search requests input from the user's terminal (100).

[0093] In step S206, the device (200) can receive specific metadata constituting search conditions from the user's terminal (100).

[0094] Specifically, the device (200) can receive one or more specific metadata indicating conditions of manufacturing data that the user wants to search for from the user's terminal (100).

[0095] Here, specific metadata may be one or more of equipment data, material data, process condition data, or manufacturing environment data that constitute the metadata combination.

[0096] Additionally, specific metadata may consist of a single item or multiple items, and if multiple items are provided, they may be transmitted including logical conditions such as “AND conditions” or “OR conditions.”

[0097] For example, the device (200) may receive {“Equipment ID A”} as specific metadata from the user’s terminal (100), the device (200) may receive {“Equipment ID A” AND “Material Code B”} as specific metadata from the user’s terminal (100), and the device (200) may receive {“Process Pressure 20 bar” OR “Process Pressure 22 bar”} as specific metadata from the user’s terminal (100).

[0098] Additionally, the device (200) may determine whether the specific metadata transmitted as a search condition is in the form of a range (e.g., temperature 150~180℃), a set (e.g., material grade A or B), or a pattern (e.g., equipment ID starting with 'MFG-***'), and may automatically set a search comparison operator suitable for each condition.

[0099] In step S207, the device (200) can extract specific optimal manufacturing data that matches specific metadata through an automatically generated index.

[0100] Specifically, the device (200) can use the search condition (specific metadata) received in step S206 as a search key to search the index (equipment standard index, material standard index, process condition standard index, manufacturing environment standard index, etc.) generated in step S205, and identify optimal manufacturing data that matches the specific metadata.

[0101] First, the device (200) can determine whether there is “full match” optimal manufacturing data for specific metadata.

[0102] At this time, if there is complete matching optimal manufacturing data, the device (200) can extract specific optimal manufacturing data among them.

[0103] Meanwhile, if there is no complete match optimal manufacturing data, the device (200) can additionally determine whether there is “partial match” optimal manufacturing data containing a part of specific metadata.

[0104] At this time, the device (200) can extract specific optimal manufacturing data among the partial matching metadata combinations if such combinations exist.

[0105] Meanwhile, if both the fully matched optimal manufacturing data and the partially matched optimal manufacturing data do not exist, the device (200) may determine that there is no specific optimal manufacturing data that matches specific metadata, and in this case, the device (200) may extract the optimal manufacturing data with the highest combination score among all optimal manufacturing data selected in step S204 as “alternative optimal manufacturing data.”

[0106] Additionally, in this process, the device (200) may indicate to the user's terminal (100) that there is no optimal manufacturing data matching specific metadata in the process of providing specific optimal manufacturing data to the user's terminal (100) to be performed later, in order to guide the user to the situation.

[0107] At this time, the process of extracting specific optimal manufacturing data will be explained with reference to Fig. 4.

[0108] In step S208, the device (200) can provide specific optimal manufacturing data to the user's terminal (100).

[0109] Specifically, the device (200) can transmit specific optimal manufacturing data extracted through step S207 to the user's terminal (100), and to do so, the specific optimal manufacturing data can be converted into a structured data format in the form of JSON, XML, or key-value and transmitted to the user's terminal (100) through a communication network.

[0110] Additionally, the device (200) may display specific optimal manufacturing data in a table format, list format, or detailed view format, depending on the screen structure or UI type of the user's terminal, and may clearly inform the user whether the data is a complete matching result, a partial matching result, or alternative optimal manufacturing data.

[0111] Furthermore, the device (200) can calculate additional information such as the expected production time, manufacturing cost, energy consumption, and quality risk when applying specific optimal manufacturing data to an actual manufacturing process based on evaluation factors such as the ratio of good products, economic efficiency, energy efficiency, and cumulative usage patterns considered in the process of calculating combination scores to select optimal manufacturing data, and provide this information to the user's terminal (100) together with the specific optimal manufacturing data.

[0112] This allows users to check the expected benefits and risk factors in advance when applying specific manufacturing data to the actual process.

[0113] Additionally, when partial matching or alternative optimal manufacturing data is provided, the device (200) can analyze items among specific metadata that are inconsistent with the user request, changes required to apply the items to the actual process (e.g., equipment change, material replacement, process condition reset, etc.), and estimated costs associated with the changes and display them together on the user's terminal (100).

[0114] This additional information can help users immediately determine whether the manufacturing data is suitable for application to the actual process.

[0115] As a result, the device (200) can simultaneously improve the efficiency of manufacturing data utilization and the accuracy of decision-making in the manufacturing process by automatically generating, evaluating, and indexing metadata combinations from the manufacturing process history and quickly and accurately searching for and providing manufacturing data that optimally matches the conditions entered by the user.

[0116] FIG. 3 is a flowchart illustrating the process of generating a combination score according to one embodiment.

[0117] Referring to FIG. 3, first, in step S301, the device (200) calculates the ratio of good products of the manufactured result for each metadata combination, assigns a combination score of 0 to metadata combinations where the ratio of good products is less than the reference ratio, and selects metadata combinations where the ratio is greater than or equal to the reference ratio as candidate combinations.

[0118] Here, the standard ratio is a preset value (e.g., 80%, 85%, 90%, etc.) according to the user environment or quality policy, and may vary by example depending on the manufacturing quality level.

[0119] Specifically, the device (200) can verify quality inspection data of the manufacturing result that matches each metadata combination generated in step S202 through the manufacturing process history data.

[0120] Additionally, the device (200) can calculate the total number of manufactured products and the number of manufactured products judged to be good (e.g., OK, PASS, etc.) when the manufacturing process is performed according to the combination of metadata through quality inspection data.

[0121] Afterward, the device (200) can calculate the ratio of good products by dividing the number of good products manufactured by the total number of manufactured products, for example, in the manner of good product ratio = (number of good products / total number of manufactured products).

[0122] Additionally, the device (200) can compare the calculated good product ratio with a preset standard ratio (e.g., 80%, 90%, etc.), and if the device (200) determines that the good product ratio is lower than the standard ratio, it determines that the metadata combination does not have sufficient quality reliability during the manufacturing process, sets the combination score to 0, and terminates the score calculation immediately without performing the additional calculation process of the combination score in steps S302 to S305 that are performed thereafter.

[0123] Conversely, if the device (200) confirms that the ratio of good products is greater than or equal to the reference ratio, it designates the corresponding metadata combination as a valid candidate combination for calculating the combination score, performs subsequent steps (S302~S305), and can calculate the final combination score by reflecting additional score elements.

[0124] In step S302, the device (200) can generate a recency weight for each candidate combination through the time difference between the last time of use and the current time, and generate a first combination score by applying the recency weight to the cumulative number of uses.

[0125] Specifically, the device (200) can identify manufacturing data that matches each candidate combination among the manufacturing process history data collected in step S201, and through the identified manufacturing data and the manufacturing process history data containing the manufacturing data, the most recent time among the times when the candidate combination was last used, i.e., the time of process execution (e.g., the time of process start or the time of process end), can be set as “t_last”.

[0126] Additionally, the device (200) can set the current time to “t_now” through a system internal clock or an external time synchronization module and calculate the time difference Δt = t_now - t_last.

[0127] At this time, the unit of the time difference Δt (day, hour, minute, etc.) can be set in various ways depending on the manufacturing system settings or the operating cycle of the manufacturing line, and the device (200) can normalize the time difference based on this setting information and use it.

[0128] Additionally, the device (200) can generate a recency weight in a form inversely proportional to the calculated time difference Δt, and, for example, the recency weight can be calculated in the form of an exponential function such as e^(-λ × Δt).

[0129] Here, λ is a weighting coefficient used to adjust the degree of recency reflection, and it can be pre-set according to the data update cycle of the manufacturing process, quality standards, user requirements, etc. As the recency weighting decreases as the value of Δt increases, metadata combinations that have not been used for a long time will have lower scores.

[0130] Additionally, the device (200) can detect exceptional situations where the Δt value becomes abnormally large or negative (e.g., time error in a manufacturing system, incorrect process record, etc.), and in such cases, the reliability of the data can be ensured by applying predefined exception handling rules, such as correcting the Δt value to 0 or replacing it with the average value of a previous normal record.

[0131] Subsequently, the number of manufacturing data matching the corresponding candidate combination among the manufacturing process history data collected in step S201 of the device (200) can be set as the cumulative usage count (f), which represents the total number of times the combination was used in past manufacturing processes.

[0132] The device (200) can calculate a first combination score by combining the calculated recency weight and the cumulative number of uses, and at this time, the device (200) can calculate the first combination score through recency weight x cumulative number of uses.

[0133] Therefore, the higher the recency weight (i.e., the more recently the combination is used) or the greater the cumulative usage count f, the higher the metadata combination will have a first combination score, while conversely, combinations with a low recency weight or a small cumulative usage count may yield a relatively low first combination score.

[0134] In step S303, the device (200) can calculate the time and cost required to produce one manufacturing result for each candidate combination, and evaluate economic efficiency based on the time and cost to generate a second combination score.

[0135] Specifically, the device (200) can identify manufacturing data that matches each candidate combination among the manufacturing process history data collected in step S201, and can calculate the manufacturing time (T) required for a unit of manufacturing product by checking the process start time and process end time included in the identified manufacturing data.

[0136] For example, the device (200) can calculate the manufacturing time by subtracting the process start time from the process end time, and if there are multiple manufacturing results corresponding to the candidate combination, the manufacturing time of each manufacturing result can be averaged to set a representative manufacturing time.

[0137] In addition, if the manufacturing process consists of multiple stages (e.g., preparation stage, processing stage, post-processing stage, etc.), the device (200) may accumulate the time required for all stages according to manufacturing system settings or user-defined rules, or select only specific stages to reflect in the calculation of manufacturing time required.

[0138] Additionally, the device (200) can calculate the manufacturing cost (C) per unit of manufacturing result based on the equipment data and material data included in the manufacturing data that matches the candidate combination.

[0139] For example, the device (200) can calculate the equipment operating time and the equipment usage cost per unit time to calculate the equipment usage cost included in the candidate combination, and can calculate the material cost by linking the material input amount and material unit price information to calculate the material cost.

[0140] In this case, if the unit price of materials or the unit cost of equipment is not listed in the manufacturing process history data, the device (200) can perform wired or wireless communication with an external server or sales site to query the latest unit price information, or use existing average unit price information stored in an internal database (DB) as a substitute value.

[0141] The device (200) can calculate economic efficiency (E) using manufacturing time (T) and manufacturing cost (C) values, and can calculate economic efficiency based on a formula such as E = 1 / (α·T + β·C).

[0142] Here, α and β are predefined weight values ​​considering the manufacturing environment, equipment operating cost structure, production scale, user settings, etc., and can be set so that the value of economic efficiency (E) increases as the manufacturing time (T) or manufacturing cost (C) decreases.

[0143] In addition, the device (200) can ensure the reliability of the economic efficiency calculation by determining that the manufacturing time (T) or manufacturing cost (C) is abnormally high (e.g., equipment malfunction, exceptional process delay, data error, etc.) is an outlier and applying outlier processing rules such as moving average-based correction, median substitution, and IQR (Interquartile Range)-based outlier removal techniques.

[0144] Accordingly, the device (200) can set the calculated economic efficiency (E) as a second combination score, and determine that the higher the economic efficiency of the metadata combination, the better the manufacturing performance in terms of time and cost.

[0145] In step S304, the device (200) can generate a third combination score for each candidate combination based on the energy efficiency used to produce one manufacturing result.

[0146] Specifically, the device (200) can identify manufacturing data that matches each candidate combination among the manufacturing process history data collected in step S201, and can verify energy usage-related data recorded in the manufacturing process history data including the identified manufacturing data.

[0147] To this end, the manufacturing process history data may further include energy usage-related data indicating the type and amount of energy consumed during the execution of the manufacturing process. This energy usage-related data may include electricity consumption (kWh), gas consumption (Nm^3), steam consumption, compressed air consumption, heat consumption (kcal or MJ), etc., and such data can be automatically collected through power meters, energy monitoring devices, IoT-based energy sensors, or MES-linked equipment attached to the manufacturing equipment.

[0148] The device (200) can calculate the total energy consumption (E_total) consumed when producing a manufactured product through manufacturing process history data matched with candidate combinations.

[0149] At this time, the device (200) can calculate the total energy amount by applying an energy unit conversion factor (e.g., gas 1 Nm^3 → kWh conversion) to the power consumption and gas consumption recorded in the energy usage data to integrate different energy units into a single unit such as kWh. For example, the total energy consumption can be calculated in the manner of E_total = power consumption (kWh) + (gas consumption × gas conversion factor).

[0150] In addition, if there are multiple manufacturing results corresponding to the candidate combination, the device (200) can collect all the total energy amounts of each manufacturing result and calculate the average energy consumption (E_avg) per manufacturing result.

[0151] In this process, the device (200) may evaluate energy efficiency (E_eff) by additionally reflecting items such as the average current value during equipment operation, equipment operation rate, process speed, and step-by-step energy usage pattern through equipment data included in the manufacturing process history data.

[0152] At this time, the device (200) can calculate energy efficiency (E_eff) in a manner such as 1 / E_avg or E_eff = (production amount / total energy consumption), and various models may be applied depending on the manufacturing process environment, equipment structure or user settings.

[0153] Generally, the lower E_avg is—that is, the less energy is consumed per manufactured product—the higher the energy efficiency (E_eff) is calculated.

[0154] Meanwhile, if the energy usage recorded during the manufacturing process is abnormally high or shows rapid fluctuations (e.g., equipment malfunction, energy measurement sensor error, environmental factors, etc.), the device (200) may determine such values ​​as outliers and apply predefined outlier processing rules such as moving average-based correction, median-based substitution, and upper and lower percentile (IQR) filtering, thereby ensuring the reliability of the energy efficiency (E_eff).

[0155] Accordingly, the device (200) can set the energy efficiency (E_eff) calculated as above as a third combination score, and the higher the energy efficiency value, the more the corresponding metadata combination can be evaluated as having superior manufacturing performance in terms of energy usage.

[0156] In step S305, the device (200) can generate a final combination score for candidate combinations based on a first combination score, a second combination score, and a third combination score.

[0157] Specifically, for each metadata combination selected as a candidate combination in step S301, the device (200) can calculate a final combination score by applying weights to the three combination scores suitable for the purpose of evaluating metadata combinations and summing them using a weighted average method, based on the first combination score calculated in step S302 (e.g., a score based on recency and cumulative usage patterns), the second combination score calculated in step S303 (e.g., an economic efficiency score based on time and cost), and the third combination score calculated in step S304 (e.g., an energy efficiency score).

[0158] At this time, the device (200) can generate a final combination score for a candidate combination through a first combination score with a first weight applied + a second combination score with a second weight applied + a third combination score with a third weight applied.

[0159] Here, the first to third weights may be pre-set values ​​considering the priority policy of the manufacturing system, user requirements, manufacturing process characteristics (e.g., energy-intensive processes, processes with high material variability, etc.), and may vary depending on the embodiment.

[0160] As a result, the device (200) can automatically determine the excellence of manufacturing data and efficiently and accurately select optimal manufacturing data by comprehensively evaluating quality reliability, usage patterns, economic efficiency, and energy efficiency for each metadata combination generated from manufacturing process history data and calculating an objective and quantified final combination score.

[0161] FIG. 4 is a flowchart illustrating the process of extracting specific optimal manufacturing data according to one embodiment.

[0162] Referring to FIG. 4, first, in step S401, the device (200) can check whether there is complete matching optimal manufacturing data.

[0163] Specifically, the device (200) can search for a metadata combination containing specific metadata (e.g., specific equipment ID, specific material code, specific process condition value, specific manufacturing environment value, etc.) received from the user's terminal (100) in step S206 by referring to the index generated in step S205 (i.e., an index for searching manufacturing data).

[0164] At this time, the index for searching manufacturing data may be composed of a data structure in which a combination of metadata (equipment data, material data, process condition data, manufacturing environment data) corresponding to each optimal manufacturing data is registered in the form of a key or hash, and the device (200) can search for an index entry containing specific metadata in the index.

[0165] First, the device (200) can retrieve an index key corresponding to each specific metadata to extract a list of all candidate optimal manufacturing data (metadata combinations) containing specific metadata, and then determine whether there is a complete match by checking whether the extracted candidates include all of the specific metadata entered by the user.

[0166] For example, when a user inputs specific metadata such as {Equipment ID=A, Material Code=B}, the device (200) searches for combinations containing Equipment ID=A and combinations containing Material Code=B among the metadata combinations stored in the index, and checks whether there exists an intersection combination satisfying both conditions to determine whether there is a complete match.

[0167] Additionally, if the device (200) includes manufacturing process history data or optimal manufacturing data in a non-standardized format or range (e.g., range specification values ​​such as temperature 170–180°C, humidity 40–50%), it may apply a rule for determining inclusion based on whether specific metadata is a single value or a range value. For example, if specific metadata is a temperature of 175°C, the device (200) may determine whether a complete match is made by determining whether the temperature range (170–180°C) recorded in the metadata combination includes the specific metadata value.

[0168] Meanwhile, if the format of a specific metadata differs from the storage format of the index (e.g., differences in expression such as “Facility A”, “Facility No. A”, “EQUIP-A”, etc.), the device (200) can check whether there is a complete match after converting it into the same item by applying a pre-stored mapping table or normalization rule.

[0169] This allows metadata recorded in different expression styles to be matched according to a unified standard.

[0170] The device (200) can determine whether there exists one or more fully matched optimal manufacturing data containing all specific metadata according to the above process.

[0171] If it is confirmed that there is a fully matched optimal manufacturing data in step S401, in step 402, the device (200) can extract the fully matched optimal manufacturing data with the highest combination score as specific optimal manufacturing data.

[0172] Specifically, if the device (200) determines in step S401 that there exists one or more fully matched optimal manufacturing data that includes all specific metadata received from the user's terminal (100), the device (200) may consider the fully matched optimal manufacturing data as a valid candidate that exactly matches the search conditions requested by the user.

[0173] Accordingly, the device (200) can check the combination score calculated through step S203 and the procedure described in FIG. 3 for each of the fully matched optimal manufacturing data, and extract the fully matched optimal manufacturing data with the highest combination score among them as specific optimal manufacturing data.

[0174] At this time, if there are two or more fully matched optimal manufacturing data having the highest combination score because the combination scores are identical, the device (200) can select a single specific optimal manufacturing data by applying a preset priority judgment criterion.

[0175] Here, the priority determination criteria may consist of usage frequency, time of recent use, manufacturing time, manufacturing cost, quality inspection history, etc., and these may be defined differently depending on system settings or user-specified requirements.

[0176] Additionally, according to an embodiment, if there are multiple fully matched optimal manufacturing data having the highest combination score, the device (200) may extract each of them as specific optimal manufacturing data and provide them to the user in the form of multiple candidates.

[0177] In this case, the device (200) can display multiple candidate data in the form of a list, table, or priority sorting and provide them to the user's terminal (100) so that the user can compare and select.

[0178] If it is confirmed in step S401 that there is no fully matched optimal manufacturing data, in step S403, the device (200) can check whether there is partially matched optimal manufacturing data.

[0179] Specifically, if the device (200) determines in step S401 that there is no fully matched optimal manufacturing data containing all specific metadata received from the user's terminal (100), the device (200) can detect whether there is optimal manufacturing data containing one or more of the specific metadata by searching the index automatically generated in step S205 based on the search conditions configured by the specific metadata received in step S206.

[0180] At this time, the device (200) can determine whether there is candidate data containing at least one of the specific metadata by separating each specific metadata into individual keys and querying the index item of the optimal manufacturing data corresponding to each metadata key.

[0181] The device (200) can classify data as a “partial match” rather than a “complete match” even if there is optimal manufacturing data corresponding to some of the metadata when there are multiple specific metadata, and can determine whether there is a partial match by comparing whether the metadata included in each optimal manufacturing data matches the specific metadata provided by the user.

[0182] At this time, the device (200) can determine whether partial matching is possible by (1) configuring a specific metadata set {M1, M2, M3, …} for partial matching determination, (2) checking a metadata set {C1, C2, C3, …} included in the optimal manufacturing data, and (3) checking whether the number of intersections of the two sets |{M1, M2, …} ∩ {C1, C2, …}| is one or more.

[0183] Additionally, the device (200) may additionally determine optimal manufacturing data having values ​​included within the range as partial matching targets when values ​​corresponding to some of the specific metadata are defined as a range (e.g., temperature 25~30℃, humidity 40~60%, etc.).

[0184] Through this process, the device (200) can determine whether there is one or more partial matching optimal manufacturing data containing some of the specific metadata.

[0185] If it is confirmed in step S403 that partial matching optimal manufacturing data exists, in step S404, the device (200) generates similar weights and cost weights for the partial matching optimal manufacturing data, generates a correction combination score accordingly, and can extract the partial matching optimal manufacturing data with the highest correction combination score as specific optimal manufacturing data.

[0186] Specifically, if the device (200) determines in step S403 that there is one or more partial matching optimal manufacturing data containing some of the specific metadata received from the user's terminal (100), the device (200) may consider the partial matching optimal manufacturing data as a valid candidate that partially matches the search conditions requested by the user.

[0187] Accordingly, the device (200) can calculate a similarity weight (W_sim) by evaluating the degree of agreement between the two data by comparing specific metadata received from the user's terminal (100) with metadata included in the partial matching optimal manufacturing data.

[0188] For example, if there are multiple specific metadata, the device (200) may define a similarity weight by dividing the number of specific metadata included in the partial matching optimal manufacturing data by the total number of specific metadata, and the similarity weight value may be set to increase as the inclusion ratio increases.

[0189] This method is intended to give higher weight to partial matching optimal manufacturing data that includes metadata closer to the search criteria requested by the user.

[0190] Additionally, the device (200) can generate cost weights (W_cost) by considering the change costs required to reflect specific metadata not included in the partial matching optimal manufacturing data into the actual manufacturing process.

[0191] For example, if equipment data included in specific metadata is not included in the partial matching optimal manufacturing data, the device (200) can determine the equipment change costs required to introduce or replace the equipment (equipment included in the partial matching optimal manufacturing data) through manufacturing process history data or an ERP system, and in the case of material data, can calculate the material change costs based on the cost of purchasing new materials (materials included in the partial matching optimal manufacturing data) or the cost of disposing of existing materials.

[0192] On the other hand, for items that can be applied by adjusting only the set value without requiring equipment changes, such as process condition data or manufacturing environment data, the device (200) can determine that the change cost is almost zero or zero and reflect this in the cost weighting.

[0193] At this time, the device (200) may apply a cost weighting formula based on an inverse relationship such that the cost weighting decreases as the change cost increases, and the cost data may be obtained from manufacturing process history data, an equipment management system, or a pre-stored cost table.

[0194] Subsequently, the device (200) can generate a corrected combination score by applying similar weights and cost weights to the combination score (i.e., base score) for the partial matching optimal manufacturing data generated through step S203 and the process of FIG. 3.

[0195] For example, the device (200) can calculate a corrected combination score through combination score X similar weight X cost weight.

[0196] This is intended to select more realistic optimal candidate data by considering not only the basic combination score but also the suitability of the user's search criteria and the burden of change costs.

[0197] Additionally, the device (200) can compare each of the partial matching optimal manufacturing data for which a correction combination score is calculated, and finally select the data with the highest correction combination score as the specific optimal manufacturing data.

[0198] In this case as well, if there are two or more partial matching optimal manufacturing data having the highest combination score with the same correction combination score, the device (200) can select a single specific optimal manufacturing data by applying a preset priority judgment criterion.

[0199] Here, the priority determination criteria may consist of usage frequency, time of recent use, manufacturing lead time, manufacturing cost, quality inspection history, combination score, similarity weight, and the magnitude of cost weight, and these may be defined differently depending on system settings or user-specified requirements.

[0200] Additionally, according to an embodiment, if there are multiple partial matching optimal manufacturing data having the highest correction combination score, the device (200) may extract each of them as specific optimal manufacturing data and provide them to the user in the form of multiple candidates.

[0201] In this case, the device (200) can display multiple candidate data in the form of a list, table, or priority sorting and provide them to the user's terminal (100) so that the user can compare and select.

[0202] If it is confirmed in step S403 that there is no partial matching optimal manufacturing data, in step S405, the device (200) determines that there is no specific optimal manufacturing data and can extract the optimal manufacturing data with the highest combination score as replacement optimal manufacturing data.

[0203] Specifically, if the device (200) confirms in step S401 that there is no fully matched optimal manufacturing data, and subsequently in step S403 that there is no partially matched optimal manufacturing data containing some of the specific metadata, the device (200) may determine that there is no manufacturing data that directly corresponds to the conditions requested by the user within the currently generated index (manufacturing data search index).

[0204] In this case, considering that alternative recommended data for performing the manufacturing process is still needed, the device (200) may compare the combination scores of all optimal manufacturing data calculated in step S203 and the process of FIG. 3, and then select the optimal manufacturing data with the highest combination score as the alternative optimal manufacturing data. This method is intended to maintain the search function and practicality of the system by recommending manufacturing data with the best manufacturing performance among the entire manufacturing process history, even if there is no result that completely matches the user search conditions.

[0205] Additionally, if there are multiple optimal manufacturing data with the highest combination score, the device (200) may apply predefined priority judgment criteria (e.g., recent usage frequency, manufacturing time, manufacturing cost, quality inspection history, energy efficiency, etc.) to finally select a single alternative optimal manufacturing data. According to an embodiment, the device (200) may be configured to provide all of the multiple alternative optimal manufacturing data having the same score to the user. In this case, the device (200) may display the multiple candidate data on the user terminal (100) in a format such as a list, table, or priority sorting screen, so that the user can directly select one.

[0206] Meanwhile, if the device (200) fails to find specific optimal manufacturing data directly corresponding to specific metadata, it may be configured to output a notification message to the user’s terminal (100) stating, “Since there is no manufacturing data matching the search conditions, alternative optimal manufacturing data is provided,” along with the process of providing specific optimal manufacturing data to the user in step S205. This notification message is intended to clearly inform the user that the result is an alternative recommendation result rather than a complete or partial match result.

[0207] As a result, the device (200) determines in stages whether there is a match (complete match / partial match) with specific metadata entered by the user, selects the optimal candidate by reflecting the change cost and similarity if necessary, and provides alternative manufacturing data with the highest performance when there are no matching items at all, thereby consistently providing stable and reliable manufacturing data recommendation results in various search situations.

[0208] FIG. 5 is a flowchart illustrating the process of providing specific optimal manufacturing data to a user's terminal according to one embodiment.

[0209] Referring to FIG. 5, first, in step S501, the device (200) can classify each metadata included in specific optimal manufacturing data into maintenance data, change data, and irrelevant data based on whether it matches specific metadata.

[0210] Specifically, when specific optimal manufacturing data is determined in step S207, the device (200) can extract all metadata elements included in the optimal manufacturing data and check whether each element is identical to an item included in the specific metadata.

[0211] To this end, the device (200) can normalize specific metadata and metadata within specific optimal manufacturing data into the same comparison format (e.g., key-value structure, JSON object, predefined metadata standard format) and then compare whether the two data are identical.

[0212] The device (200) can determine that the metadata is maintenance data if the items included in the specific metadata and the metadata of the specific optimal manufacturing data are completely identical (e.g., values ​​such as the same equipment ID, same material code, same set temperature, same environmental conditions match). Maintenance data may refer to items that already satisfy the conditions requested by the user and do not require a change in the manufacturing process. For example, if the specific metadata is “Equipment ID = A101” and the equipment data of the specific optimal manufacturing data is also “A101”, the device (200) can classify this as maintenance data.

[0213] On the other hand, the device (200) may determine that the metadata is changed data if it is included in specific metadata but has a value different from the metadata included in specific optimal manufacturing data. Change data may refer to items that require setting changes, material changes, or equipment replacement when applied to the manufacturing process. For example, if the specific metadata is “Equipment ID = A101” but the equipment data of the specific optimal manufacturing data is “A205”, the device (200) may classify this as changed data. In the case of process conditions, if values ​​such as designated temperature or pressure are different, it may be determined as changed data.

[0214] Additionally, the device (200) may determine metadata that is included in specific optimal manufacturing data but not in specific metadata as irrelevant data. Irrelevant data is not related to user request conditions and corresponds to items whose change status is not directly related to satisfying specific metadata. For example, if specific metadata includes only equipment data and material data, process condition data or environmental data within specific optimal manufacturing data may be classified as irrelevant data.

[0215] The device (200) can utilize pre-stored metadata type-specific matching rules (e.g., equipment information is compared based on equipment ID, equipment type, and process step, material information is compared based on material code and batch number, process conditions are compared based on set values, etc.) to classify such maintenance data, change data, and irrelevant data, and if the expression method between data is different or the units are different (e.g., temperature unit °C ↔ °F, process pressure bar ↔ MPa), the units can be unified through an internal system conversion module and then compared.

[0216] Additionally, the device (200) may perform an exception handling procedure in which, if certain metadata or some metadata within certain optimal manufacturing data is missing, the item is classified as irrelevant data or the comparison criterion is replaced based on a predefined default value or recent value.

[0217] In step S502, the device (200) can provide specific optimal manufacturing data to the user's terminal (100) by displaying the maintenance data in a first color.

[0218] Here, the first color may vary according to the embodiment as a preset color value. For example, the first color may be blue, but is not limited thereto.

[0219] Specifically, the device (200) determines that an item matching a specific metadata received from a user's terminal (100) among the metadata elements constituting specific optimal manufacturing data is retention data, and can render such retention data by applying a first color (Color1) so that it can be visually distinguished.

[0220] The first color may vary depending on system settings, user preferences, manufacturing environment, or UI style, and may be set to, for example, blue (RGB value: 0, 122, 255) or HEX code (#007AFF), but is not limited thereto.

[0221] Additionally, the device (200) may map each metadata element constituting specific optimal manufacturing data to a UI element (e.g., text label, tag-type display, table cell, list item, etc.) to perform color display of maintenance data, and apply a first color style only to the item determined to be maintenance data.

[0222] The color style can be composed of various visual attributes such as background color, text color, border color, and highlight color, and the device (200) can select and apply appropriate attributes according to the display environment.

[0223] The device (200) may also be configured to additionally display an icon, tag, or text label (e.g., “Maintain”) meaning “Maintain” along with the application of a first color so that, in the process of displaying specific optimal manufacturing data, the user can intuitively recognize that it is maintenance data.

[0224] Additionally, if there are multiple maintenance data or multiple categories of color displays are applied within the same manufacturing data, the device (200) may apply a combination of auxiliary visual effects, such as transparency control, shading, and highlighting, to prevent visual interference between colors.

[0225] In this way, the device (200) displays maintenance data in a first color and provides it to the user's terminal (100), thereby improving visual readability so that the user can identify at a glance the items among the specific optimal manufacturing data that already match the requirements.

[0226] In step S503, the device (200) can provide specific optimal manufacturing data to the user's terminal (100) by displaying the change data in a second color.

[0227] Here, the second color may vary according to the embodiment as a preset color value. For example, the second color may be red, but is not limited thereto.

[0228] Specifically, the device (200) determines that an item inconsistent with specific metadata received from the user's terminal (100) among the metadata elements constituting specific optimal manufacturing data is changed data, and can render such changed data by applying a second color (Color2) so that it can be visually distinguished.

[0229] The second color may be defined as a preset color value in the system, which may be set as a color to intuitively indicate that a change is needed. For example, the second color may be red (RGB value: 255, 59, 48 or HEX code #FF3B30), but may be defined as other colors such as orange, pink, purple, etc. depending on the embodiment, but is not limited thereto.

[0230] After identifying the changed data, the device (200) may map the corresponding metadata element to a UI element (e.g., a table cell, a list item, a tag, a balloon icon, etc.) and apply a second color to the UI element. The color application method may include one or more of various visual effects such as text color change, background color change, border highlighting, and icon highlighting, and the device (200) may select and display an appropriate color application method according to the user interface style or terminal performance.

[0231] In addition, the device (200) may apply additional auxiliary visual display elements, such as a text label indicating “change needed” (e.g., “Change”), a warning icon, an exclamation mark, and an accent border, in addition to the second color display, in order to clearly recognize to the user that the data is changed. This display method allows for the immediate identification of items requiring actual modification in the manufacturing process, thereby enabling the rapid identification of whether a change has been made.

[0232] Meanwhile, when displayed together with other color-separated items such as maintenance data and irrelevant data, the device (200) can perform additional visual correction functions, such as adjusting contrast, changing transparency, and highlighting borders, so that the second color is sufficiently distinguished within the screen.

[0233] In addition, if there are multiple change data, the device (200) can improve user readability by performing additional functions such as sorting by change data group, displaying emphasis priority, and displaying the amount of change.

[0234] In this way, the device (200) displays the change data in a second color and provides it to the user terminal (100), thereby clearly distinguishing items requiring actual change measures among specific optimal manufacturing data and supporting the user in intuitively understanding the measures required when applying the process.

[0235] In step S504, the device (200) can provide specific optimal manufacturing data to the user's terminal (100) by displaying irrelevant data in a third color.

[0236] Here, the third color may vary according to the embodiment as a preset color value. For example, the third color may be black, but is not limited thereto.

[0237] Specifically, the device (200) determines that items among the metadata constituting specific optimal manufacturing data that are not related to specific metadata received from the user's terminal (100) are irrelevant data, and applies a third color (Color3) to such irrelevant data so that it can be visually distinguished from maintenance data and change data.

[0238] A third color may be defined as a preset color value in the system, and a neutral color may be used to indicate that the item is not directly related to changes in the manufacturing process. For example, the third color may be black (RGB value: 0, 0, 0 or HEX code #000000), but is not limited thereto, and may include shades of gray, dark blue, or other custom colors.

[0239] The device (200) maps irrelevant data to UI display elements (e.g., table cells, list items, text labels, tag-type display elements, etc.) and can apply a third color style to the UI elements. The method of applying the style can be implemented in various forms, such as changing text color, processing background color, changing border color, and adjusting transparency, and the device (200) can apply display rules in a way that maintains visual contrast so as not to be confused with the colors of the maintained data and the changed data.

[0240] In addition, the device (200) may apply additional auxiliary visual display elements, such as a text label meaning “irrelevant” (e.g., “-”), a dotted border, and a secondary tone, in addition to the third color display, to clearly inform the user that the data is irrelevant. Such auxiliary elements can effectively convey that the irrelevant data is not an item that necessarily needs to be changed or maintained during the manufacturing process selection process.

[0241] When multiple metadata is displayed in different colors, the device (200) may perform additional UI correction operations, such as adjusting transparency, adjusting color contrast, and improving tag alignment methods, to minimize visual interference between colors. Additionally, when multiple irrelevant data exist, the device (200) may group and display the irrelevant data or organize it using a folding / unfolding method to improve user readability.

[0242] In this way, the device (200) displays irrelevant data in a third color and provides it to the user terminal (100), thereby clearly distinguishing and displaying items among specific optimal manufacturing data that are not directly related to the user's specific metadata, and improving the efficiency and understanding of the manufacturing data review process.

[0243] As a result, the device (200) automatically classifies each metadata included in specific optimal manufacturing data into data to be maintained, changed, or irrelevant based on whether it matches the user's search conditions, and provides them visually distinguished based on color, thereby allowing the user to intuitively identify items suitable for requirements, items that need to be changed, and irrelevant items, which can greatly improve the understanding and utilization efficiency of specific optimal manufacturing data.

[0244] FIG. 6 is a flowchart illustrating the process of providing a sales link together with the process of providing modified data according to one embodiment.

[0245] Referring to FIG. 6, first, in step S601, the device (200) can search for a plurality of external sales sites for equipment or materials corresponding to the change data based on the fact that the change data includes at least one of equipment data or material data.

[0246] Specifically, the device (200) checks the results classified into maintenance data, change data, and irrelevant data through step S501, depending on whether each metadata included in specific optimal manufacturing data matches specific metadata received from the user's terminal (100), and can determine whether at least one of equipment data or material data is included in the change data.

[0247] That is, the device (200) can determine whether a change or purchase of equipment or materials is necessary during the process of applying the specific optimal manufacturing data to the actual manufacturing process by checking whether the metadata items included in the change data are equipment-related items such as equipment ID, equipment name, equipment model name, equipment specification code, manufacturer information, etc., or material-related items such as material code, material name, material specification, material grade, batch information, etc.

[0248] If the device (200) determines that at least one of equipment data or material data is included in the change data, it can automatically generate search keywords for performing a product search request to an external sales site based on the equipment or material items included in the change data.

[0249] For example, the device (200) may form a search keyword string by extracting at least one of the equipment ID, equipment name, equipment model name, specification code, and manufacturer name included in the equipment data within the change data, and in the case of material data, may form a search keyword including the material code, material name, specification information, grade, batch number, supplier information, etc.

[0250] Additionally, the device (200) can improve search accuracy by applying predefined text normalization rules, such as removing unnecessary words during the keyword generation process or converting into a form combining model name and specification information (e.g., “model name + specification”).

[0251] Subsequently, the device (200) can communicate wirelessly or via wired connection with multiple pre-registered external sales sites (e.g., specialized sales platforms for industrial equipment, online malls for factory automation parts, shopping malls for raw material suppliers, industrial material categories of global open markets, etc.) to perform a search request for equipment or materials. At this time, considering that the availability of APIs and API methods may differ for each external sales site, the device (200) can automatically select a request format according to the communication standards of each site, such as REST API, SOAP API, GraphQL API, and HTTPS-based HTTP GET / POST requests.

[0252] In addition, for sites where an API is not provided, the device (200) can collect product information by scraping or crawling through HTML DOM-based web page structure analysis, CSS Selector extraction, XPath analysis, etc.

[0253] The device (200) can parse response data received from an external sales site and convert and store product name, model name, specification information, option information, compatibility information, seller name, price information (unit price, discount amount, whether tax is included, etc.), inventory information (stock quantity, availability for immediate purchase, expected arrival date, etc.), delivery information (delivery fee, delivery period), detailed description, etc. into structured data in the form of JSON, XML, or Key-Value. At this time, considering that field names may differ by site, the device (200) can utilize a pre-stored field mapping table to integrate and map different names, such as “price”, “unit_price”, “salePrice”, and “price”, into the same price item.

[0254] Furthermore, the device (200) can identify whether the same item is the same based on a pre-established compatibility dictionary or product code mapping table, taking into account that the same equipment or the same material may be sold at multiple external sales sites with different product codes. For example, if similar model names such as “ABC-1000”, “ABC1000”, or “ABC Series 1000” are found at different sites, it can be determined that they are the same item through string similarity analysis (e.g., Levenshtein distance, Cosine Similarity, N-gram analysis, etc.), and if the specification information (voltage, dimensions, capacity, etc.) is partially different, it can be classified as an optional variant within a user-specified allowable range.

[0255] Additionally, when the same item is sold with different specifications or options at multiple external sales sites, the device (200) can refine the search results by reflecting predefined filtering criteria (e.g., same model name priority, same specification priority, recent update information priority, user setting priority) or search options received from the user's terminal (100) (lowest price priority, stock availability priority, etc.).

[0256] In this way, the device (200) can automatically perform product search and information collection procedures targeting multiple external sales sites so as to secure purchaseable information necessary to reflect the equipment or materials included in the change data into the actual manufacturing process.

[0257] Meanwhile, if the device (200) determines that at least one of the equipment data or material data is not included in the change data, it may not perform the process of FIG. 6 (steps S601 to S603).

[0258] In step S602, the device (200) can identify a sales site available for purchase at the lowest price among external sales sites as a recommended sales site.

[0259] Specifically, the device (200) can calculate the final purchase cost (TotalCost) of each product candidate by extracting items such as price information, option costs, shipping costs, and whether tax (VAT) is included from the response data of an external sales site that has been parsed and normalized in step S601.

[0260] At this time, the device (200) calculates the total cost including option price, shipping fee, tax, etc., along with the basic price, taking into account that the unit price display method may differ depending on the sales site, and in cases where the units are different (e.g., unit price per item vs. bundle unit price), it can convert to a comparable standard by performing unit matching using a pre-stored unit conversion table or configuration information.

[0261] The device (200) can compare the converted final purchase cost for each external sales site, identify the sales site with the lowest cost among them as the lowest price sales site, and set it as the recommended sales site.

[0262] Additionally, if there are multiple sales sites with the same final purchase cost, the device (200) may select a single recommended sales site or provide multiple recommended sales sites to the user in parallel by applying predefined priority criteria (e.g., delivery time, seller reliability, stock quantity, time of last update, user preferred site, etc.).

[0263] Meanwhile, if the collected price data is determined to be an abnormal value (e.g., 0 won, negative number, abnormally high value, etc.), the device (200) may apply a data verification rule to exclude the data from comparison, and if there is a response delay or communication failure at an external sales site, it may apply a re-request policy or perform price comparison only on sites that provided a normal response.

[0264] Accordingly, the device (200) can reliably recommend a sales site where equipment or materials corresponding to the change data can be purchased at the lowest cost through a price comparison procedure targeting multiple external sales sites.

[0265] In step S603, the device (200) can provide a sales link to a recommended sales site along with specific optimal manufacturing data to the user's terminal (100).

[0266] Specifically, the device (200) can obtain a sales link corresponding to a finally selected recommended sales site among sales sites where equipment or materials included in the change data can be purchased through steps S601 and S602, and this sales link may be in the form of an HTML-based web URL, a product detail page URL, a purchase API endpoint, a cart link, or an app deep link provided by an external sales site.

[0267] Additionally, the device (200) can convert the secured sales link into an internal standard format (JSON, XML, Key-Value structure, etc.) and normalize it into a form that can be transmitted to the user's terminal (100).

[0268] Additionally, the device (200) can generate a display packet to be transmitted to a user's terminal (100) by combining specific optimal manufacturing data and a sales link.

[0269] At this time, the display packet may be configured to include metadata (maintenance data, change data, irrelevant data) included in specific optimal manufacturing data, color display information of the data (steps S502~S504), detailed information such as combination scores or correction combination scores, along with the name of a recommended sales site, seller name, price information, shipping information, and a purchase link.

[0270] Afterward, the device (200) can transmit the generated display packet to the user's terminal (100), and the user terminal can simultaneously display specific optimal manufacturing data and a purchase link in accordance with the user interface (UI) based on the display packet.

[0271] For example, a “Buy” button or a “Check Lowest Price” button can be displayed next to a change data item on the user terminal screen, and when the button is clicked, a link to the recommended sales site is executed to move to an external page or proceed with the purchase process.

[0272] Accordingly, the device (200) can support decision-making for improving or changing the manufacturing process by providing a sales link to a recommended sales site along with specific optimal manufacturing data so that the user can easily purchase equipment or materials corresponding to the change data.

[0273] As a result, the device (200) can automatically search for equipment or materials that need to be changed in specific optimal manufacturing data, identify the lowest price sales site, and provide the corresponding purchase link, thereby enabling the user to secure the equipment and materials required for manufacturing process application at the most efficient cost and with the minimum procedure.

[0274] FIG. 7 is a flowchart illustrating a process for automatically extracting specific optimal manufacturing data and providing it to a user at a time when specific optimal manufacturing data is needed, according to one embodiment.

[0275] Referring to FIG. 7, first, in step S701, the device (200) can detect that at least one of the following has arrived: the time when equipment replacement is scheduled, the time when materials are exhausted, the time when new equipment is installed, the time when the manufacturing process starts, or the time when the manufacturing environment changes.

[0276] Specifically, the device (200) receives data regarding equipment status information, material inventory information, equipment configuration information, manufacturing process schedule information, and manufacturing environment information through an external manufacturing system (e.g., MES (Manufacturing Execution System), ERP (Enterprise Resource Planning), SCM (Supply Chain Management), equipment control device, equipment diagnostic system, sensor network system, etc.) that communicates with the user's terminal (100) or the device (200) via wired or wireless communication, and can determine whether the time point has arrived by determining whether the received data satisfies a pre-set reference value or event occurrence condition.

[0277] First, in order to detect the scheduled time for equipment replacement, the device (200) may receive the cumulative operating time of the equipment, expected lifespan information, replacement cycle of major components, self-diagnosis results of the equipment, and output values ​​of a failure prediction model (e.g., an AI-based risk prediction model) through at least one of a user's terminal (100) or an external manufacturing system. The device (200) may detect that the scheduled time for equipment replacement has arrived when conditions satisfying the equipment replacement criteria are confirmed, such as when the cumulative operating time is greater than or equal to the maximum allowable time per equipment, when the remaining lifespan prediction value is below a reference threshold, or when the failure risk prediction value exceeds a critical risk level.

[0278] Additionally, to detect when materials are depleted, the device (200) may receive information regarding the remaining amount of materials, the material consumption rate, the material replenishment schedule, and the estimated time of depletion from a material management system or a user's terminal (100). The device (200) may determine that the time of material depletion has arrived if the remaining amount of materials is less than or equal to a reference inventory amount, or if the estimated time of depletion is within a specific threshold time (e.g., 1 hour or 30 minutes) of the current time.

[0279] Furthermore, to detect the time of new equipment installation, the device (200) may receive new equipment registration information (e.g., equipment ID, installation location, installation time), equipment configuration change logs, or network connection events (Logon Events) of the new equipment from an external equipment management system. If the new equipment is registered as an asset within the device (200) system or is recognized on the network, the device (200) may detect that the time of new equipment installation has arrived.

[0280] Meanwhile, to detect the start time of the manufacturing process, the device (200) may receive a process start command (Manual Start) or an automatic start event generated by a process scheduler from a user's terminal (100) or a manufacturing system, and may also determine that the start time of the manufacturing process has arrived when it receives an operation start signal (e.g., current application signal, motor rotation start signal, etc.) from an equipment sensor.

[0281] Additionally, to detect a time when the manufacturing environment changes, the device (200) may receive changes in manufacturing environment information, such as workplace temperature, humidity, air quality (e.g., PM2.5 and PM10 particulate matter concentration), illuminance, vibration, and noise, from an external system based on an environment sensor. The device (200) may detect that a time when the manufacturing environment changes has arrived if at least one of the environment values ​​deviates from a predefined standard range or changes by more than a specific percentage (e.g., ±5%) compared to the existing value.

[0282] In this way, the device (200) can reliably detect that at least one time has arrived among the scheduled time for equipment replacement, the time for material depletion, the time for new equipment installation, the time for the start of the manufacturing process, or the time for the change in the manufacturing environment by comprehensively determining the equipment status, material inventory, whether new equipment is installed, whether the manufacturing process is running, and whether there is a change in the surrounding manufacturing environment.

[0283] In step S702, the device (200) can set the metadata used to perform the most recent manufacturing process to specific metadata based on the manufacturing process history data received from the user's terminal (100).

[0284] Specifically, the device (200) can extract manufacturing process history data as reference manufacturing process history data from among the manufacturing process history data collected in step S201, including (i) manufacturing process history data directly collected through the user's terminal (100) and (ii) manufacturing process history data collected from an external control system (e.g., equipment control device, process control system, material management system, etc.) in which the user has administrator privileges.

[0285] Additionally, the device (200) can identify the manufacturing process performed closest to the current time by analyzing time and identification information, such as the time of process execution (e.g., start time and end time), process step number, and process identifier (Process ID), included in the reference manufacturing process history data.

[0286] Subsequently, the device (200) can acquire manufacturing data corresponding to the most recent manufacturing process identified above, and extract metadata including at least one of equipment data, material data, process condition data, and manufacturing environment data included in the manufacturing data and set it as specific metadata.

[0287] Meanwhile, the device (200) may operate to exclude some metadata during the specific metadata setting process according to the time-based event type detected in step S701.

[0288] For example, if it is determined that the time for equipment replacement or the time for new equipment installation has arrived, the device (200) may configure specific metadata by excluding equipment data from the manufacturing data corresponding to the most recent manufacturing process. If it is determined that the time for material depletion has arrived, specific metadata may be set by excluding material data from the most recent manufacturing process data, and if the time for the start of the manufacturing process has arrived, specific metadata may be checked by excluding process condition data. Additionally, if the time for a change in the manufacturing environment has arrived, specific metadata may be set by excluding manufacturing environment data from the corresponding manufacturing process data.

[0289] By setting specific metadata that reflects the time-based event situation in this way, the device (200) can operate so that the index-based query and optimal manufacturing data derivation process performed in subsequent steps is more aligned with the current manufacturing environment and the user's work situation.

[0290] In step S703, the device (200) can query an index automatically generated based on specific metadata to extract specific optimal manufacturing data based on candidate time points that match the specific metadata.

[0291] Specifically, the device (200) can use at least one of the equipment data, material data, process condition data, and manufacturing environment data included in the specific metadata set in step S702 as a search key to search for optimal manufacturing data that matches the search key.

[0292] At this time, the device (200) can select a matching target for a plurality of optimal manufacturing data stored in the index in the same manner as step S207 and the process described in FIG. 4, based on whether specific metadata is included, the degree of matching between data, and the similarity of the metadata combination.

[0293] Consequently, the device (200) can extract the most suitable candidate time-based specific optimal manufacturing data from an index automatically generated based on specific metadata set in step S702.

[0294] In step S704, the device (200) can provide specific optimal manufacturing data based on candidate time points in advance to the user's terminal (100).

[0295] Specifically, the device (200) can configure specific optimal manufacturing data based on candidate time points derived in step S703 into a data packet and transmit it to the user's terminal (100).

[0296] At this time, when the device (200) determines that at least one of the following times has arrived—such as the time scheduled for equipment replacement, the time when materials are depleted, the time when new equipment is installed, the time when the manufacturing process starts, or the time when the manufacturing environment changes—it can provide specific optimal manufacturing data based on candidate times in real time using the information of the time as a trigger, thereby enabling the user to quickly check appropriate manufacturing conditions before starting the manufacturing process or when equipment and materials need to be replaced.

[0297] Additionally, the device (200) can guide the user's terminal (100) to perform necessary preparatory work in advance before performing the actual manufacturing process by providing information on items that need to be changed to apply the data (e.g., whether equipment change is necessary, whether material replacement is necessary, whether process conditions need to be reset, whether manufacturing environment adjustment is necessary, etc.) for specific optimal manufacturing data based on candidate time points provided in advance.

[0298] In this way, the device (200) can improve the responsiveness and efficiency of the manufacturing process and perform a pre-guidance function to support user judgment when a time-based event occurs.

[0299] As a result, the device (200) can increase the responsiveness and efficiency of the entire manufacturing process by providing the user with specific optimal manufacturing data based on candidate time points in advance, thereby enabling the user to identify optimal manufacturing conditions and prepare necessary change operations before the time of performing the manufacturing process.

[0300] FIG. 8 is an example diagram of the configuration of a device according to one embodiment.

[0301] A device (200) according to one embodiment includes a processor (210) and a memory (220). The processor (210) may include at least one device described with reference to FIGS. 1 through 7 or may perform at least one method described with reference to FIGS. 1 through 7. A person or organization using the device (200) may provide services related to some or all of the methods described with reference to FIGS. 1 through 7.

[0302] The memory (220) may store information related to the methods described above or store a program in which the methods described below are implemented. The memory (220) may be volatile memory or non-volatile memory.

[0303] The processor (210) can execute a program and control the device (200). The code of the program executed by the processor (210) can be stored in memory (220). The device (200) can be connected to an external device (e.g., a personal computer or a network) through an input / output device (not shown in the drawing) and can exchange data via wired or wireless communication. The processor (210) can be operatively connected to the components of the device (200). The processor (210) can load commands or data received from other components of the device (200) into memory (220), process the commands or data stored in memory (220), and store the resulting data.

[0304] Additionally, the device (200) may further include a communication circuit. The communication circuit may establish a communication channel with an external device (e.g., a user's terminal (100)) and transmit and receive various data with the external device. According to various embodiments, the communication circuit may include a cellular communication module and be configured to be connected to a cellular network (e.g., 3G, LTE, 5G, Wibro, or Wimax). According to various embodiments, the communication circuit may include a short-range communication module and transmit and receive data with the external device using short-range communication (e.g., Wi-Fi, Bluetooth, Bluetooth Low Energy (BLE), UWB), but is not limited thereto.

[0305] Additionally, the device (200) may be used to train an artificial neural network or to use a trained artificial neural network. The memory (220) may contain an artificial neural network that is being trained or has been trained. The processor (210) may train or execute an artificial neural network algorithm stored in the memory (220). The device (200) for training the artificial neural network and the device (200) for using the trained artificial neural network may be the same or separate.

[0306] The embodiments described above may be implemented as hardware components, software components, and / or combinations of hardware and software components. For example, the devices, methods, and components described in the embodiments may be implemented using one or more general-purpose or special-purpose computers, such as, for example, a processor, a controller, an arithmetic logic unit (ALU), a digital signal processor, a microcomputer, a field programmable gate array (FPGA), a programmable logic unit (PLU), a microprocessor, or any other device capable of executing and responding to instructions. The processing unit may execute an operating system (OS) and one or more software applications executed on said operating system. Additionally, the processing unit may access, store, manipulate, process, and generate data in response to the execution of the software. For ease of understanding, the processing unit may be described as being used as a single unit, but those skilled in the art will understand that the processing unit may include multiple processing elements and / or multiple types of processing elements. For example, the processing unit may include multiple processors or one processor and one controller. Additionally, other processing configurations, such as parallel processors, are also possible.

[0307] The method according to the embodiment may be implemented in the form of program instructions that can be executed through various computer means and recorded on a computer-readable medium. The computer-readable medium may include program instructions, data files, data structures, etc., either alone or in combination. The program instructions recorded on the medium may be those specifically designed and configured for the embodiment, or they may be those known and available to those skilled in the art of computer software. Examples of computer-readable recording media include magnetic media such as hard disks, floppy disks, and magnetic tapes; optical recording media such as CD-ROMs and DVDs; magneto-optical media such as floptical disks; and hardware devices specifically configured to store and execute program instructions, such as ROM, RAM, and flash memory. Examples of program instructions include machine code, such as that generated by a compiler, as well as high-level language code that can be executed by a computer using an interpreter, etc. The hardware devices described above may be configured to operate as one or more software modules to perform the operation of the embodiment, and vice versa.

[0308] Software may include computer programs, code, instructions, or a combination of one or more of these, and may configure a processing unit to operate as desired or command the processing unit independently or collectively. Software and / or data may be permanently or temporarily embodied in any type of machine, component, physical device, virtual equipment, computer storage medium or device, or transmitted signal wave so as to be interpreted by the processing unit or to provide instructions or data to the processing unit. Software may be distributed over networked computer systems and may be stored or executed in a distributed manner. Software and data may be stored on one or more computer-readable recording media.

[0309] Although the embodiments have been described above with reference to the limited drawings, those skilled in the art can apply various technical modifications and variations based on the above. For example, suitable results may be achieved even if the described techniques are performed in a different order than described, and / or if the components of the described system, structure, device, circuit, etc. are combined or assembled in a form different from described, or replaced or substituted by other components or equivalents.

[0310] Therefore, other implementations, other embodiments, and equivalents to the claims also fall within the scope of the claims set forth below.

Claims

Claim 1 A method for automatic indexing of manufacturing data metadata and search optimization performed by a device comprises: collecting manufacturing process history data including manufacturing data; generating metadata combinations by extracting equipment data, material data, process condition data, and manufacturing environment data, which are metadata constituting each manufacturing data, based on the manufacturing process history data; generating combination scores for each of the metadata combinations based on the manufacturing process history data; selecting manufacturing data corresponding to a metadata combination in which the combination score is higher than a preset target score as optimal manufacturing data; automatically generating an index for searching manufacturing data based on the optimal manufacturing data; receiving specific metadata constituting search conditions from a user terminal; and extracting specific optimal manufacturing data that matches the specific metadata through the automatically generated index. and the step of providing the specific optimal manufacturing data to the terminal of the user; and the step of extracting specific optimal manufacturing data that matches the specific metadata through the automatically generated index; comprises: an operation of checking whether there exists a complete matching optimal manufacturing data that includes all of the specific metadata through the automatically generated index; an operation of extracting the complete matching optimal manufacturing data with the highest combination score as the specific optimal manufacturing data if it is confirmed that the complete matching optimal manufacturing data exists; an operation of checking whether there exists a partial matching optimal manufacturing data that includes some of the specific metadata if it is confirmed that the complete matching optimal manufacturing data does not exist; and an operation of generating a similarity weight based on the similarity between the partial matching optimal manufacturing data and the specific metadata if it is confirmed that the partial matching optimal manufacturing data exists.The method includes generating cost weights based on the change costs required to change specific metadata not included in the partial matching optimal manufacturing data, generating a corrected combination score by applying the similar weights and the cost weights to the combination score of the partial matching optimal manufacturing data, extracting the partial matching optimal manufacturing data with the highest corrected combination score as specific optimal manufacturing data, and, if it is confirmed that the partial matching optimal manufacturing data does not exist, determining that there is no specific optimal manufacturing data matching the specific metadata, and extracting the optimal manufacturing data with the highest combination score as alternative optimal manufacturing data; and the step of providing the specific optimal manufacturing data to the user's terminal includes classifying each metadata included in the specific optimal manufacturing data into maintenance data, change data, and irrelevant data based on whether it matches the specific metadata, displaying the maintenance data in a preset first color to provide the specific optimal manufacturing data to the user's terminal, displaying the change data in a preset second color to provide the specific optimal manufacturing data to the user's terminal, and displaying the irrelevant data in a preset third color to provide the specific optimal manufacturing data to the user's terminal; and the step of providing the specific optimal manufacturing data to the user's terminal includes, among the equipment data or material data in the change data A method for automatic indexing of manufacturing data metadata and search optimization, further comprising, based on including at least one, the operation of searching for a plurality of external sales sites for equipment or materials corresponding to the change data, the operation of identifying a sales site among the external sales sites that can be purchased at the lowest price as a recommended sales site, and the operation of providing a sales link to the recommended sales site along with the specific optimal manufacturing data to a user's terminal. Claim 2 A method for automatic indexing and search optimization of manufacturing data metadata according to claim 1, wherein the step of generating a combination score for each of the metadata combinations based on the manufacturing process history data comprises: calculating the ratio of good products of manufacturing results for each of the metadata combinations, assigning a combination score of 0 for metadata combinations where the ratio of good products is less than a preset reference ratio, and selecting metadata combinations where the ratio is greater than or equal to the reference ratio as candidate combinations; for each of the candidate combinations, generating a recency weight through the time difference between the last time of use and the current time, and generating a first combination score by applying the recency weight to the cumulative number of uses; for each of the candidate combinations, calculating the time and cost required to produce one manufacturing result, and generating a second combination score by evaluating economic efficiency based on the time and cost; for each of the candidate combinations, generating a third combination score based on the energy efficiency used to produce one manufacturing result; and generating a final combination score for the candidate combinations based on the first combination score, the second combination score, and the third combination score. Claim 3 delete

Images