Method, apparatus and system for providing manufacturing and sales strategy derivation solution using artificial intelligence model based on automatic collection of global trend information and raw material information

Abstract

A method, apparatus, and system for providing a solution for deriving manufacturing and sales strategies for manufactured goods using an AI model based on the automatic collection of international trend information and raw material information is disclosed. In a method, apparatus, and system for providing a solution for deriving manufacturing and sales strategies for manufactured goods using an AI model based on the automatic collection of international trend information and raw material information according to an embodiment of the present invention, the apparatus for providing a solution for deriving manufacturing and sales strategies for manufactured goods using an AI model based on the automatic collection of international trend information and raw material information comprises: a memory; a communication unit;The system includes a processor connected to the communication unit and memory and configured to be linked with a manufacturing system including manufacturing equipment, an assembly system, or an autonomous mobile robot (AMR). The processor automatically collects international news information, industrial news information, economic news information, paper information, open source information, artificial intelligence model information, platform API information, sales data information, advertising data information, raw material price information, supply chain information, logistics information, exchange rate information, manufacturing process information, information related to autonomous mobile robots (AMR), camera recognition information, and manufacturing-related technology trend information from multiple external servers or databases in real-time or periodically. It classifies the collected data by category, removes duplicate data, extracts and summarizes data corresponding to the execution of the manufacturing process, and converts it into training data for manufacturing strategy analysis. It trains a first model based on the training data. Through the first model, it analyzes the sales volume of manufactured products, advertising efficiency, inventory levels, manufacturing costs, fluctuations in raw material prices, changes in market trends, and changes in international affairs to infer manufacturing strategies, sales strategies, advertising strategies, inventory strategies, and raw material procurement strategies. Based on the inferred strategy information, it determines the timing of manufacturing, the timing of raw material purchase, and the execution of the advertising budget. It generates timing, production volume control strategies, sales channel strategies, and manufacturing process optimization strategies, provides the generated strategy information through a display, and provides article information, paper information, artificial intelligence technology information, and market analysis information in a summarized or in-depth analysis form, automatically collects at least one newly disclosed data among newly disclosed artificial intelligence models, computer vision models, manufacturing algorithms, robot control algorithms, and open source technology information, and reflects the automatically collected newly disclosed data in the first model to retrain the first model or update the inference logic or inference parameters.

Description

Technology Field

[0001] The present invention relates to the field of artificial intelligence (AI)-based manufacturing and sales strategy analysis technology, and more specifically, to a method, apparatus, and system for providing a solution for deriving manufacturing and sales strategies for manufactured products using an AI model based on automatic collection of international trend information and raw material information, which automatically collects and analyzes international news, papers, market data, raw material price information, platform API data, manufacturing equipment data, logistics data, and technology trend data, and based thereon, automatically derives production strategies, sales strategies, inventory strategies, raw material supply strategies, advertising execution strategies, and manufacturing process optimization strategies for manufactured products using an AI model, and further enables manufacturing automation by linking with an autonomous mobile robot (AMR), a smart factory system, a 3D printing-based manufacturing system, and a CAD design system.

[0002] In particular, the present invention is not limited to a simple data analysis system or a general artificial intelligence recommendation system, but relates to an industrial AI decision-making system that automatically collects globally generated news articles, industry reports, research paper data, open source AI model information, raw material price information, manufacturing process data, sales platform API data, and market response data from multiple data sources in real-time or periodically, and then reflects this in an AI-based learning model to integrally infer manufacturing strategies and sales strategies.

[0003] Furthermore, the present invention can be configured to be applicable to the entire manufacturing industry and belongs to the field of AI-based strategic analysis platform technology that is universally applicable to clothing, underwear, household goods, electronic components, industrial equipment, medical devices, robot components, plastic injection molded products, metal processed products, food products, and other manufacturing-based industrial sectors.

[0004] Furthermore, the present invention belongs to the field of AI-based industrial decision-making automation technology that analyzes in real time fluctuations in raw material prices, changes in international supply chains, wars and changes in international affairs, changes in exchange rates, changes in consumption trends, changes in advertising efficiency, and changes in market demand occurring during the product manufacturing process, and reflects these findings in manufacturing and sales strategies.

[0005] In addition, the present invention can be used in conjunction with natural language-based manufacturing command processing technology, and can also be included in the field of AI-based manufacturing automation technology in which, when a user inputs a natural language command in the form of voice or text, an AI model analyzes it to automatically generate CAD design data, 3D model data, STL file data, and manufacturing process data, and enables the actual manufacturing process to be performed by linking it with a 3D printer, CNC equipment, or autonomous manufacturing equipment.

[0006] Furthermore, the present invention also belongs to the technical field of a Self-Evolving Industrial AI System, in which an AI model continuously and automatically learns the latest papers, the latest AI technologies, the latest open-source models, and the latest industrial technology information, and continuously advances manufacturing strategies and manufacturing functions based thereon. Background Technology

[0008] Recently, the manufacturing industry is facing a situation where it is difficult to maintain market competitiveness with existing static manufacturing and sales strategies alone, due to the increasing complexity of global supply chains, rising volatility in raw material prices, instability in international affairs, rapid changes in consumption trends, and the diversification of online sales platforms.

[0009] In particular, manufacturing companies must make decisions by comprehensively considering strategies such as raw material procurement, inventory management, production volume control, advertising budget, sales channel, and logistics before product production; however, they have traditionally relied mostly on manual data investigation and analysis.

[0010] For example, manufacturing companies formulate manufacturing and sales strategies by individually checking and analyzing data from Bloomberg, industry news, economic news, academic paper databases, market reports, international raw material price data, logistics information, and sales platform data; however, this approach faces the problem of difficulty in processing large volumes of data in real time and requires significant time and manpower for data analysis.

[0011] Furthermore, existing manufacturing strategy analysis systems generally rely on specific internal corporate data or data from a single ERP system, which has limitations in that they fail to comprehensively reflect external factors such as changes in international affairs, shifts in global technology trends, advancements in the latest AI technology, and fluctuations in raw material prices.

[0012] In particular, as the pace of AI technology development has recently accelerated, new AI models, new open-source technologies, new computer vision algorithms, new robot control algorithms, and new manufacturing automation technologies are continuously being released; however, existing manufacturing systems have not been equipped with a structure to automatically collect and incorporate these latest technologies.

[0013] In addition, most existing AI-based analysis systems remain limited to simple data summarization or simple recommendation functions, and have not advanced to the level of executing manufacturing strategies by directly linking with actual manufacturing processes, autonomous mobile robots (AMRs), smart factory equipment, CAD design systems, and 3D printing systems.

[0014] For example, existing systems often provided only sales volume forecasts or performed simple advertising recommendations, and integrated AI-based manufacturing strategy platforms that link the timing of raw material procurement, production, manufacturing process changes, component design modifications, and manufacturing equipment control required in the actual manufacturing process were provided only to a limited extent.

[0015] In addition, existing CAD design systems and 3D modeling systems had the problem that users had to design structures using CAD programs themselves, and general users without specialized CAD design knowledge faced significant difficulties in designing parts or generating STL files for manufacturing.

[0016] In particular, existing CAD systems had a problem in that they lacked the ability to automatically generate manufacturing part structures based solely on natural language command input, or to have AI automatically optimize and generate structures suitable for manufacturing purposes.

[0017] Furthermore, existing manufacturing AI systems lacked a structure for continuously incorporating the latest technologies, which presented a limitation in that it was difficult to automatically learn and reflect new papers, manufacturing algorithms, AI models, and industrial technologies even when they emerged.

[0018] Therefore, there is a continuously increasing need for an integrated AI manufacturing strategy platform that can automatically collect international news, papers, raw material information, sales platform information, and the latest AI technology information, and have an AI model continuously learn from them to automatically derive manufacturing strategies, sales strategies, raw material strategies, and manufacturing automation strategies, and furthermore, be linked with CAD design, 3D printing, and AMR systems. The problem to be solved

[0020] The present invention aims to solve the aforementioned problems by providing a method, apparatus, and system for deriving manufacturing and sales strategies for manufactured products using an AI model based on automatic collection of international trend information and raw material information, which automatically collects international news information, industry report information, paper information, raw material price information, platform API information, sales data, advertising data, manufacturing data, and technology trend data from multiple external data sources and analyzes them based on an AI model to automatically derive production strategies, sales strategies, advertising strategies, raw material procurement strategies, and inventory strategies for manufactured products.

[0021] The objective of the present invention is to provide a method, apparatus, and system for deriving product manufacturing and sales strategies utilizing an AI model based on the automatic collection of international trend information and raw material information, which enables manufacturing companies to establish more rapid and efficient manufacturing strategies by having the AI ​​model automatically analyze real-time changing international affairs, global supply chains, exchange rates, raw material prices, consumption trends, and the latest technology trends.

[0022] The objective of the present invention is to provide a method, apparatus, and system for deriving manufacturing and sales strategies for manufactured goods utilizing an AI model based on the automatic collection of international trend information and raw material information, which includes an autonomous evolutionary manufacturing AI platform capable of continuously enhancing the functionality of a manufacturing system by automatically collecting the latest AI papers, the latest open source technologies, the latest computer vision technologies, and the latest robot technologies, and reflecting them in an AI model.

[0023] The objective of the present invention is to provide a method, apparatus, and system for deriving manufacturing and sales strategies for manufactured goods utilizing an AI model based on the automatic collection of international trend information and raw material information, which integrates sales platform APIs, advertising platform APIs, and logistics data APIs to comprehensively analyze changes in sales volume, changes in advertising efficiency, and market response data, and enables the AI ​​model to automatically infer optimal sales strategies and advertising execution strategies.

[0024] The objective of the present invention is to provide a method, apparatus, and system for deriving manufacturing and sales strategies for manufactured goods utilizing an AI model based on the automatic collection of international trend information and raw material information, which analyzes international raw material price information and supply chain data to predict the possibility of future raw material price increases, supply shortages, and market volatility, and automatically derives raw material pre-purchase strategies and production strategies accordingly.

[0025] The objective of the present invention is to provide a method, apparatus, and system for deriving manufacturing and sales strategies for manufactured products utilizing an AI model based on automatic collection of international trend information and raw material information, which includes an AI-based manufacturing design system capable of automatically generating CAD structural data, 3D model data, and STL data based on a user's natural language command or voice command, and performing manufacturing automation by linking them with a 3D printing system or manufacturing equipment.

[0026] The objective of the present invention is to provide a method, apparatus, and system for deriving manufacturing and sales strategies for manufactured products using an AI model based on the automatic collection of international trend information and raw material information, which enables the manufacturing strategy results generated by the AI ​​model to be integrated with autonomous mobile robots (AMRs), smart factory systems, manufacturing equipment control systems, and production management systems to be reflected in actual manufacturing processes.

[0027] The objective of the present invention is to provide a method, apparatus, and system for deriving manufacturing and sales strategies for manufactured goods utilizing an AI model based on the automatic collection of international trend information and raw material information, thereby implementing an expandable system applicable to various manufacturing industries by providing a universal AI-based manufacturing strategy platform capable of responding to manufacturing and sales environments that differ by industrial sector. means of solving the problem

[0029] In a method, apparatus, and system for providing a solution for deriving manufacturing and sales strategies for manufactured products using an AI model based on automatic collection of international trend information and raw material information according to an embodiment of the present invention, the apparatus for providing a solution for deriving manufacturing and sales strategies for manufactured products using an AI model based on automatic collection of international trend information and raw material information comprises: a memory; a communication unit; The system includes a processor connected to the communication unit and memory and configured to be linked with a manufacturing system including manufacturing equipment, an assembly system, or an autonomous mobile robot (AMR). The processor automatically collects international news information, industrial news information, economic news information, paper information, open source information, artificial intelligence model information, platform API information, sales data information, advertising data information, raw material price information, supply chain information, logistics information, exchange rate information, manufacturing process information, information related to autonomous mobile robots (AMR), camera recognition information, and manufacturing-related technology trend information from multiple external servers or databases in real-time or periodically. It classifies the collected data by category, removes duplicate data, extracts and summarizes data corresponding to the execution of the manufacturing process, and converts it into training data for manufacturing strategy analysis. It trains a first model based on the training data. Through the first model, it analyzes the sales volume of manufactured products, advertising efficiency, inventory levels, manufacturing costs, fluctuations in raw material prices, changes in market trends, and changes in international affairs to infer manufacturing strategies, sales strategies, advertising strategies, inventory strategies, and raw material procurement strategies. Based on the inferred strategy information, it determines the timing of manufacturing, the timing of raw material purchase, and the execution of the advertising budget. Generates timing, production volume control strategies, sales channel strategies, and manufacturing process optimization strategies; provides the generated strategy information through a display; provides article information, paper information, AI technology information, and market analysis information in the form of summaries or in-depth analyses; and newly disclosed AI models, computer vision models, manufacturing algorithms,Automatically collects at least one new publicly available data among robot control algorithms and open source technical information, and reflects the automatically collected new publicly available data in the first model to retrain the first model or update inference logic or inference parameters, wherein the first model is an artificial intelligence model trained to output a manufacturing strategy, a sales strategy, an advertising strategy, an inventory strategy, and a raw material procurement strategy based on sales volume data, advertising efficiency data, inventory data, manufacturing cost data, raw material price fluctuation data, market trend change data, and international situation change data extracted from the training data for manufacturing strategy analysis, using the training data for manufacturing strategy analysis as input data, and wherein the first model is configured to infer strategy information for manufacturing timing, raw material purchasing timing, advertising execution timing, production volume control, sales channel selection, and manufacturing process optimization, wherein the manufacturing strategy includes strategy information for determining the production timing, production method, manufacturing process execution method, or manufacturing priority of a manufactured product, the sales strategy includes strategy information for determining the sales channel, advertising execution method, target market, or sales timing of a manufactured product, and the raw material procurement strategy includes the raw material purchasing timing, raw material purchasing quantity, whether raw materials are pre-purchased, or The manufacturing process optimization strategy includes strategic information for determining the method of supplying raw materials, and the manufacturing process optimization strategy includes strategic information for adjusting the method of operating manufacturing equipment, the sequence of manufacturing processes, the method of allocating manufacturing resources, or the method of performing manufacturing automation. The manufacturing system may be configured to update the output distribution, judgment criterion weight, prediction sensitivity, or calculation characteristics of the strategy priority calculation criteria corresponding to the manufacturing strategy analysis results by retraining the first model or the artificial intelligence inference model or updating the inference logic or inference parameters by the processor in accordance with the reflection of new manufacturing technology, a new artificial intelligence model, or new training data.

[0030] The processor according to one embodiment of the present invention receives a text or voice-based natural language manufacturing command from a user, analyzes manufacturing conditions, dimensional information, structural information, part specification information, bolt specification information, nut specification information, plate structure information, and manufacturing purpose information included in the natural language manufacturing command, automatically generates CAD data, three-dimensional (3D) model data, STL file data, or manufacturing structural data based on the analysis result of the natural language manufacturing command, generates 3D printer output data based on the generated CAD data or STL file data to control the output of a manufacturing module or part, generates manufacturing module data applicable to manufacturing equipment, assembly systems, or autonomous mobile robot (AMR) systems based on the generated part data, automatically collects new 3D algorithms, new CAD generation algorithms, new artificial intelligence generation models, or new manufacturing technology information, and updates output distributions, judgment criteria, or inference result characteristics corresponding to CAD generation results or manufacturing automation control results by relearning or updating inference parameters for the output characteristics of CAD generation algorithms and manufacturing automation control algorithms, and automatically generates a manufacturing structure, plate, bolt structure, nut structure, or machine part structure according to a natural language command input by a user, and Output, and the natural language manufacturing command includes command information for instructing a manufacturing purpose or manufacturing condition based on voice data or text data entered by a user, the CAD data is design data including shape information, dimension information, assembly information, or output information of a manufacturing structure, the STL file data includes three-dimensional shape data for performing three-dimensional shape output or 3D printer output, and the manufacturing module data may include structural data, control data, or output data usable in manufacturing equipment, assembly systems, or autonomous mobile robot systems.

[0031] The processor according to one embodiment of the present invention collects object recognition data from a plurality of cameras, sensors, or image input devices, performs object recognition, part recognition, manufacturing environment recognition, obstacle recognition, or work environment analysis based on image data which is object recognition data collected from the plurality of cameras or image input devices, automatically collects object recognition models, computer vision models, artificial intelligence inference models, and manufacturing-related source code provided from an external artificial intelligence platform, a research institution server, an open source repository, or a technology provider server, automatically trains a second model suitable for the manufacturing environment based on the collected artificial intelligence models and source code, generates an artificial intelligence inference model or inference algorithm capable of responding to new parts, new technologies, and new manufacturing conditions, generates a movement path, work method, manufacturing process execution method, or manufacturing equipment control strategy of an autonomous mobile robot (AMR) based on the generated artificial intelligence inference results, converts user voice commands into manufacturing control commands using a voice inference model or translation model capable of responding to multiple national languages, and when new data including new manufacturing technology, new artificial intelligence model, new manufacturing part information, or new manufacturing environment information is collected, automatically learns and reflects the new data to update the inference logic, control algorithm, or model parameters of the manufacturing system, and the second model is the plurality of An artificial intelligence model trained to output object recognition results, part recognition results, manufacturing environment recognition results, obstacle recognition results, or work environment analysis results within a manufacturing environment, using training data generated based on object recognition data and image data collected from a camera, sensor, or image input device as input data; and configured to generate environment recognition information including location information, shape information, and state information of objects within the manufacturing environment, and to update recognition results and inference results according to new parts, new manufacturing conditions, or changes in the manufacturing environment.The object recognition data includes at least one of object location information, part shape information, obstacle information, or work target information within a manufacturing environment, and the artificial intelligence inference model or inference algorithm includes control logic or control algorithm for changing a manufacturing control method or a manufacturing process execution method in response to changes in the manufacturing environment, addition of new parts, changes in new manufacturing conditions, or application of new technology, and the manufacturing equipment control strategy includes strategy information for controlling the operation sequence, movement method, work priority, or manufacturing process execution conditions of the manufacturing equipment, and the updating of the inference logic or control algorithm of the manufacturing system may include a process of updating the characteristics of the inference result of the artificial intelligence model or the control output value generated corresponding to each performance by retraining model parameters, feature vectors, or inference logic corresponding to manufacturing accuracy, manufacturing automation level, object recognition performance, or manufacturing strategy analysis performance in accordance with the reflection of new manufacturing technology or a new artificial intelligence model, or by updating inference parameters.

[0032] The processor according to one embodiment of the present invention configures a digital twin-based manufacturing simulation environment based on real-time manufacturing product status information, manufacturing environment information, manufacturing equipment operation information, and manufacturing process execution information; generates a first policy scenario using a third model trained to generate a first policy scenario for cases where a manufacturing control method or a manufacturing process execution method is changed in response to changes in the manufacturing environment, addition of new parts, changes in new manufacturing conditions, or application of new technology in the digital twin-based manufacturing simulation environment; generates a second policy scenario using a fourth model trained to generate a second policy scenario for cases where the manufacturing control method or a manufacturing process execution method is maintained in the digital twin-based manufacturing simulation environment; calculates a first policy profit / loss value including estimated costs due to manufacturing discontinuation, estimated costs due to process change, and estimated revenue due to process change based on the first policy scenario; calculates a second policy profit / loss value including estimated costs due to manufacturing maintenance based on the second policy scenario; calculates an optimal policy transition point for changing the manufacturing control method or the manufacturing process execution method by comparing the first policy profit / loss value and the second policy profit / loss value; and calculates the optimal policy transition point calculated after a pre-set reference point. In this case, a third policy scenario is generated through the third model by resimulating the first policy scenario by reflecting real-time manufacturing status information at the time of reaching the reference point, and a third policy profit and loss value including manufacturing stoppage costs, process re-change costs, and process reapplication revenue is recalculated based on the third policy scenario, and the optimal policy transition point is recalculated by reflecting the third policy profit and loss value.Based on the finally calculated optimal policy transition point, control is applied to manufacturing equipment or an autonomous mobile robot (AMR) by controlling the application of a manufacturing control method or a manufacturing process execution method corresponding to the first policy scenario or the third policy scenario; the third model is a policy decision artificial intelligence model trained to take policy scenario learning data, including real-time manufacturing product status information, manufacturing environment information, manufacturing equipment operation information, manufacturing process execution information, manufacturing cost information, production schedule information, supply chain information, and manufacturing downtime loss information, as input data, and output a first policy scenario including manufacturing downtime costs, process change costs, process change revenue, and production efficiency change information when changing the manufacturing control method or manufacturing process execution method in response to changes in the manufacturing environment or new manufacturing conditions, and is configured to dynamically generate a manufacturing strategy according to changes in the manufacturing status; and the fourth model is a policy scenario learning data, including real-time manufacturing product status information, manufacturing environment information, manufacturing equipment operation information, manufacturing process execution information, manufacturing cost information, production schedule information, supply chain information, and manufacturing maintenance loss information, as input data, and outputs production delay costs, manufacturing inefficiency costs, and supply when maintaining the manufacturing control method or manufacturing process execution method even in response to changes in the manufacturing environment or new manufacturing conditions. It is a policy evaluation AI model trained to output a second policy scenario including delay costs and maintenance strategy information, and can be configured to estimate the economic impact and operational risks when maintaining existing manufacturing control methods.

[0033] The processor according to one embodiment of the present invention identifies a worker through voice recognition and generates a worker ID to distinguish the worker; calculates a reliability score for each of manufacturing conditions, dimensional information, structural information, part specification information, bolt specification information, nut specification information, plate structure information, and manufacturing purpose information based on manufacturing history data including natural language manufacturing command data, defective product occurrence history, and profit and loss data accumulated and stored for the identified first worker; if the reliability score calculated for at least one item among the manufacturing conditions, dimensional information, structural information, part specification information, bolt specification information, nut specification information, plate structure information, and manufacturing purpose information is less than a preset threshold, the item is identified as caution information; and regarding the caution information, based on the first worker's natural language manufacturing command input tendency and manufacturing history data, generates a filtering algorithm including an input normalization rule and a semantic correction rule to normalize the natural language input value for the caution information and correct semantic misunderstanding or input deviation in conjunction with the reliability score calculation result; simulates the result of reinterpreting the manufacturing history data based on natural language by applying the filtering algorithm; and as a result of the simulation, the reliability score If improvement is not made beyond a threshold, a plurality of second workers whose reliability scores regarding the aforementioned caution information are within a certain rank are identified, and a fifth model trained to generate correction data for the first worker's caution information using the manufacturing history data of the plurality of second workers and the first worker's non-caution information as input is used to generate a caution information correction result reflecting the correction data, and when a new natural language manufacturing command of the first worker is input, the caution information is automatically corrected through the fifth model to generate a manufacturing command interpretation result, and the defect occurrence history includes manufacturing conditions, dimensional information, structural information, part specification information, bolt specification information, nut specification information,The plate structure information and manufacturing purpose information include input values ​​that cause defective products and correction history of said input values, and the reliability score is a value calculated based on repeatability, error frequency, correction history, and profit / loss impact extracted from the manufacturing history data, and the filtering algorithm is configured to normalize or correct the semantics of the natural language input values ​​for the attention information, and is dynamically updated by reflecting the weight distribution of repeatability, error frequency, and correction history used in calculating the reliability score, and the fifth model is an artificial intelligence model trained to learn manufacturing deviations and error patterns between workers by taking as input data time-series manufacturing data including the manufacturing history data of the first worker, manufacturing history data of a plurality of second workers, natural language manufacturing command data, and defect occurrence history, and to output corrected manufacturing conditions, dimensional information, structure information, part specification information, bolt specification information, nut specification information, plate structure information, and manufacturing purpose information for the attention information, and may be configured to include an error correction function to adaptively correct manufacturing command interpretation results according to worker characteristics and to correct natural language manufacturing command interpretation errors due to input deviations by worker.

[0034] A device according to one embodiment of the present invention may be controlled by a computer program stored on a medium to execute any one of the operation methods by the processor of the system described above, in combination with hardware. Effects of the invention

[0036] According to the present invention, by automatically collecting and analyzing international news information, paper information, raw material information, sales platform information, advertising information, logistics information, and the latest technology information, manufacturing companies can effectively respond quickly to market changes.

[0037] According to the present invention, an AI model analyzes global market changes, changes in international affairs, and fluctuations in raw material prices in real time to automatically derive manufacturing and sales strategies, thereby improving the speed of decision-making and increasing operational efficiency.

[0038] According to the present invention, by integrating and analyzing sales data, advertising data, and market response data, it is possible to optimize advertising budgets, sales channels, and inventory, thereby having the effect of improving the profitability of manufacturing companies.

[0039] According to the present invention, by continuously and automatically learning the latest AI papers, the latest open-source models, and the latest manufacturing technologies, it is possible to implement an autonomous evolutionary AI manufacturing platform in which the manufacturing system autonomously evolves its functions.

[0040] According to the present invention, CAD data and STL data can be automatically generated based on natural language commands, so even users lacking expertise in CAD design can easily design and manufacture parts for production.

[0041] According to the present invention, the manufacturing process can be automated by directly linking AI-based design data with 3D printers and manufacturing equipment, thereby enabling a reduction in manufacturing time and production costs.

[0042] According to the present invention, since manufacturing strategies can be reflected in the actual manufacturing process by linking with an autonomous mobile robot (AMR), smart factory equipment, and a manufacturing control system, there is an effect of improving the level of manufacturing automation and increasing the efficiency of smart factory implementation.

[0043] According to the present invention, by providing a preemptive raw material securing strategy in response to changes in the international supply chain and fluctuations in raw material prices, it is possible to reduce supply chain risks and improve manufacturing stability.

[0044] According to the present invention, by implementing an AI-based industrial operation platform that comprehensively analyzes manufacturing, logistics, sales, advertising, raw materials, technology trends, and manufacturing automation, it is possible to effectively support the digital transformation (DX) and AI-based smart manufacturing innovation of manufacturing companies. Brief explanation of the drawing

[0046] FIG. 1 is a block diagram schematically illustrating the basic configuration of a device according to one embodiment. FIGS. 2 to 4 are basic operation flowcharts of a method according to one embodiment. Specific details for implementing the invention

[0047] 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.

[0048] 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.

[0049] 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.

[0050] When it is stated that a component is "connected" to another component, it should be understood that it may be directly connected to or coupled with that other component, or that there may be other components in between.

[0051] 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.

[0052] 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.

[0053] 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.

[0054] 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.

[0055] Artificial Intelligence (AI) systems are computer systems that implement human-level intelligence; unlike existing rule-based smart systems, they are systems in which machines learn and make decisions autonomously. As AI systems improve in recognition accuracy and gain a more accurate understanding of user preferences with continued use, existing rule-based smart systems are gradually being replaced by deep learning-based AI systems.

[0056] Artificial intelligence technology consists of machine learning and component technologies utilizing machine learning. Machine learning is an algorithmic technology that autonomously classifies and learns the characteristics of input data, while component technologies are technologies that mimic the cognitive and judgmental functions of the human brain by utilizing machine learning algorithms such as deep learning, and are comprised of technological fields such as linguistic understanding, visual understanding, reasoning / prediction, knowledge representation, and motion control.

[0057] The various fields where artificial intelligence technology is applied are as follows. Linguistic understanding refers to technologies that recognize, apply, and process human language and text, including natural language processing, machine translation, dialogue systems, question answering, and speech recognition / synthesis. Visual understanding refers to technologies that perceive and process objects like human vision, including object recognition, object tracking, image search, person recognition, scene understanding, spatial understanding, and image enhancement. Inference and prediction refers to technologies that logically reason and predict by judging information, including knowledge / probability-based inference, optimization prediction, preference-based planning, and recommendation. Knowledge representation refers to technologies that automatically process human experiential information into knowledge data, including knowledge construction (data generation / classification) and knowledge management (data utilization). Motion control refers to technologies that control the autonomous driving of vehicles and the movement of robots, including motion control (navigation, collision, driving) and manipulation control (behavior control).

[0058] Generally, to apply machine learning algorithms to real-world situations, training is performed using a trial-and-error method due to the inherent characteristics of the fundamental methodologies. In particular, deep learning requires hundreds of thousands of iterations. Since it is impossible to execute this in a real physical external environment, training is instead performed through simulations that virtually recreate the actual physical environment on a computer.

[0059] 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.

[0060] Machine learning can refer to the process of training neural network models using experience in processing data. Through machine learning, computer software can improve 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.

[0061] 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.

[0062] 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.

[0063] 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.

[0064] The present invention relates to the field of artificial intelligence (AI)-based manufacturing and sales strategy analysis technology, and more specifically, to a method, apparatus, and system for providing a solution for deriving manufacturing and sales strategies for manufactured products using an AI model based on automatic collection of international trend information and raw material information, which automatically collects and analyzes international news, papers, market data, raw material price information, platform API data, manufacturing equipment data, logistics data, and technology trend data, and based thereon, automatically derives production strategies, sales strategies, inventory strategies, raw material supply strategies, advertising execution strategies, and manufacturing process optimization strategies for manufactured products using an AI model, and further enables manufacturing automation by linking with an autonomous mobile robot (AMR), a smart factory system, a 3D printing-based manufacturing system, and a CAD design system.

[0065] In particular, the present invention is not limited to a simple data analysis system or a general artificial intelligence recommendation system, but relates to an industrial AI decision-making system that automatically collects globally generated news articles, industry reports, research paper data, open source AI model information, raw material price information, manufacturing process data, sales platform API data, and market response data from multiple data sources in real-time or periodically, and then reflects this in an AI-based learning model to integrally infer manufacturing strategies and sales strategies.

[0066] Furthermore, the present invention can be configured to be applicable to the entire manufacturing industry and belongs to the field of AI-based strategic analysis platform technology that is universally applicable to clothing, underwear, household goods, electronic components, industrial equipment, medical devices, robot components, plastic injection molded products, metal processed products, food products, and other manufacturing-based industrial sectors.

[0067] Furthermore, the present invention belongs to the field of AI-based industrial decision-making automation technology that analyzes in real time fluctuations in raw material prices, changes in international supply chains, wars and changes in international affairs, changes in exchange rates, changes in consumption trends, changes in advertising efficiency, and changes in market demand occurring during the product manufacturing process, and reflects these findings in manufacturing and sales strategies.

[0068] In addition, the present invention can be used in conjunction with natural language-based manufacturing command processing technology, and can also be included in the field of AI-based manufacturing automation technology in which, when a user inputs a natural language command in the form of voice or text, an AI model analyzes it to automatically generate CAD design data, 3D model data, STL file data, and manufacturing process data, and enables the actual manufacturing process to be performed by linking it with a 3D printer, CNC equipment, or autonomous manufacturing equipment.

[0069] Furthermore, the present invention also belongs to the technical field of a Self-Evolving Industrial AI System, in which an AI model continuously and automatically learns the latest papers, latest AI technologies, latest open-source models, and the latest industrial technology information, and continuously advances manufacturing strategies and manufacturing functions based thereon.

[0070] Recently, the manufacturing industry is facing a situation where it is difficult to maintain market competitiveness with existing static manufacturing and sales strategies alone, due to the increasing complexity of global supply chains, rising volatility in raw material prices, instability in international affairs, rapid changes in consumption trends, and the diversification of online sales platforms.

[0071] In particular, manufacturing companies must make decisions by comprehensively considering strategies such as raw material procurement, inventory management, production volume control, advertising budget, sales channel, and logistics before product production; however, they have traditionally relied mostly on manual data investigation and analysis.

[0072] For example, manufacturing companies formulate manufacturing and sales strategies by individually checking and analyzing data from Bloomberg, industry news, economic news, academic paper databases, market reports, international raw material price data, logistics information, and sales platform data; however, this approach faces the problem of difficulty in processing large volumes of data in real time and requires significant time and manpower for data analysis.

[0073] Furthermore, existing manufacturing strategy analysis systems generally rely on specific internal corporate data or data from a single ERP system, which has limitations in that they fail to comprehensively reflect external factors such as changes in international affairs, shifts in global technology trends, advancements in the latest AI technology, and fluctuations in raw material prices.

[0074] In particular, as the pace of AI technology development has recently accelerated, new AI models, new open-source technologies, new computer vision algorithms, new robot control algorithms, and new manufacturing automation technologies are continuously being released; however, existing manufacturing systems have not been equipped with a structure to automatically collect and incorporate these latest technologies.

[0075] In addition, most existing AI-based analysis systems remain limited to simple data summarization or simple recommendation functions, and have not advanced to the level of executing manufacturing strategies by directly linking with actual manufacturing processes, autonomous mobile robots (AMRs), smart factory equipment, CAD design systems, and 3D printing systems.

[0076] For example, existing systems often provided only sales volume forecasts or performed simple advertising recommendations, and integrated AI-based manufacturing strategy platforms that link the timing of raw material procurement, production, manufacturing process changes, component design modifications, and manufacturing equipment control required in the actual manufacturing process were provided only to a limited extent.

[0077] In addition, existing CAD design systems and 3D modeling systems had the problem that users had to design structures directly using CAD programs, and general users without specialized CAD design knowledge faced significant difficulties in designing parts or generating STL files for manufacturing.

[0078] In particular, existing CAD systems had a problem in that they lacked the ability to automatically generate manufacturing part structures based solely on natural language command input, or to have AI automatically optimize and generate structures suitable for manufacturing purposes.

[0079] Furthermore, existing manufacturing AI systems lacked a structure for continuously incorporating the latest technologies, resulting in limitations that made it difficult to automatically learn and reflect new papers, manufacturing algorithms, AI models, and industrial technologies even when they emerged.

[0080] Therefore, there is a continuously increasing need for an integrated AI manufacturing strategy platform that can automatically collect international news, papers, raw material information, sales platform information, and the latest AI technology information, and have an AI model continuously learn from them to automatically derive manufacturing strategies, sales strategies, raw material strategies, and manufacturing automation strategies, and furthermore, be linked with CAD design, 3D printing, and AMR systems.

[0081] The present invention aims to solve the aforementioned problems by providing a method, apparatus, and system for deriving manufacturing and sales strategies for manufactured products using an AI model based on automatic collection of international trend information and raw material information, which automatically collects international news information, industry report information, paper information, raw material price information, platform API information, sales data, advertising data, manufacturing data, and technology trend data from multiple external data sources and analyzes them based on an AI model to automatically derive production strategies, sales strategies, advertising strategies, raw material procurement strategies, and inventory strategies for manufactured products.

[0082] The objective of the present invention is to provide a method, apparatus, and system for deriving product manufacturing and sales strategies utilizing an AI model based on the automatic collection of international trend information and raw material information, which enables manufacturing companies to establish more rapid and efficient manufacturing strategies by having the AI ​​model automatically analyze real-time changing international affairs, global supply chains, exchange rates, raw material prices, consumption trends, and the latest technology trends.

[0083] The objective of the present invention is to provide a method, apparatus, and system for deriving manufacturing and sales strategies for manufactured goods utilizing an AI model based on the automatic collection of international trend information and raw material information, which includes an autonomous evolutionary manufacturing AI platform capable of continuously enhancing the functionality of a manufacturing system by automatically collecting the latest AI papers, the latest open source technologies, the latest computer vision technologies, and the latest robot technologies, and reflecting them in an AI model.

[0084] The objective of the present invention is to provide a method, apparatus, and system for deriving manufacturing and sales strategies for manufactured goods utilizing an AI model based on the automatic collection of international trend information and raw material information, which integrates sales platform APIs, advertising platform APIs, and logistics data APIs to comprehensively analyze changes in sales volume, changes in advertising efficiency, and market response data, and enables the AI ​​model to automatically infer optimal sales strategies and advertising execution strategies.

[0085] The objective of the present invention is to provide a method, apparatus, and system for deriving manufacturing and sales strategies for manufactured goods utilizing an AI model based on the automatic collection of international trend information and raw material information, which analyzes international raw material price information and supply chain data to predict the possibility of future raw material price increases, supply shortages, and market volatility, and automatically derives raw material pre-purchase strategies and production strategies accordingly.

[0086] The objective of the present invention is to provide a method, apparatus, and system for deriving manufacturing and sales strategies for manufactured products utilizing an AI model based on automatic collection of international trend information and raw material information, which includes an AI-based manufacturing design system capable of automatically generating CAD structural data, 3D model data, and STL data based on a user's natural language command or voice command, and performing manufacturing automation by linking them with a 3D printing system or manufacturing equipment.

[0087] The objective of the present invention is to provide a method, apparatus, and system for deriving manufacturing and sales strategies for manufactured products using an AI model based on the automatic collection of international trend information and raw material information, which enables the manufacturing strategy results generated by the AI ​​model to be integrated with autonomous mobile robots (AMRs), smart factory systems, manufacturing equipment control systems, and production management systems to be reflected in actual manufacturing processes.

[0088] The objective of the present invention is to provide a method, apparatus, and system for deriving manufacturing and sales strategies for manufactured goods utilizing an AI model based on the automatic collection of international trend information and raw material information, thereby implementing an expandable system applicable to various manufacturing industries by providing a general-purpose AI-based manufacturing strategy platform capable of responding to manufacturing and sales environments that differ by industrial sector.

[0089] In a method, apparatus, and system for providing a solution for deriving manufacturing and sales strategies for manufactured products using an AI model based on automatic collection of international trend information and raw material information according to an embodiment of the present invention, the apparatus for providing a solution for deriving manufacturing and sales strategies for manufactured products using an AI model based on automatic collection of international trend information and raw material information comprises: a memory; a communication unit; The system includes a processor connected to the communication unit and memory and configured to be linked with a manufacturing system including manufacturing equipment, an assembly system, or an autonomous mobile robot (AMR). The processor automatically collects international news information, industrial news information, economic news information, paper information, open source information, artificial intelligence model information, platform API information, sales data information, advertising data information, raw material price information, supply chain information, logistics information, exchange rate information, manufacturing process information, information related to autonomous mobile robots (AMR), camera recognition information, and manufacturing-related technology trend information from multiple external servers or databases in real-time or periodically. It classifies the collected data by category, removes duplicate data, extracts and summarizes data corresponding to the execution of the manufacturing process, and converts it into training data for manufacturing strategy analysis. It trains a first model based on the training data. Through the first model, it analyzes the sales volume of manufactured products, advertising efficiency, inventory levels, manufacturing costs, fluctuations in raw material prices, changes in market trends, and changes in international affairs to infer manufacturing strategies, sales strategies, advertising strategies, inventory strategies, and raw material procurement strategies. Based on the inferred strategy information, it determines the timing of manufacturing, the timing of raw material purchase, and the execution of the advertising budget. Generates timing, production volume control strategies, sales channel strategies, and manufacturing process optimization strategies; provides the generated strategy information through a display; provides article information, paper information, AI technology information, and market analysis information in the form of summaries or in-depth analyses; and newly disclosed AI models, computer vision models, manufacturing algorithms,Automatically collects at least one new publicly available data among robot control algorithms and open source technical information, and reflects the automatically collected new publicly available data in the first model to retrain the first model or update inference logic or inference parameters, wherein the first model is an artificial intelligence model trained to output a manufacturing strategy, a sales strategy, an advertising strategy, an inventory strategy, and a raw material procurement strategy based on sales volume data, advertising efficiency data, inventory data, manufacturing cost data, raw material price fluctuation data, market trend change data, and international situation change data extracted from the training data for manufacturing strategy analysis, using the training data for manufacturing strategy analysis as input data, and wherein the first model is configured to infer strategy information for manufacturing timing, raw material purchasing timing, advertising execution timing, production volume control, sales channel selection, and manufacturing process optimization, wherein the manufacturing strategy includes strategy information for determining the production timing, production method, manufacturing process execution method, or manufacturing priority of a manufactured product, the sales strategy includes strategy information for determining the sales channel, advertising execution method, target market, or sales timing of a manufactured product, and the raw material procurement strategy includes the raw material purchasing timing, raw material purchasing quantity, whether raw materials are pre-purchased, or The manufacturing process optimization strategy includes strategic information for determining the method of supplying raw materials, and the manufacturing process optimization strategy includes strategic information for adjusting the method of operating manufacturing equipment, the sequence of manufacturing processes, the method of allocating manufacturing resources, or the method of performing manufacturing automation. The manufacturing system may be configured to update the output distribution, judgment criterion weight, prediction sensitivity, or calculation characteristics of the strategy priority calculation criteria corresponding to the manufacturing strategy analysis results by retraining the first model or the artificial intelligence inference model or updating the inference logic or inference parameters by the processor in accordance with the reflection of new manufacturing technology, a new artificial intelligence model, or new training data.

[0090] The processor according to one embodiment of the present invention receives a text or voice-based natural language manufacturing command from a user, analyzes manufacturing conditions, dimensional information, structural information, part specification information, bolt specification information, nut specification information, plate structure information, and manufacturing purpose information included in the natural language manufacturing command, automatically generates CAD data, three-dimensional (3D) model data, STL file data, or manufacturing structural data based on the analysis result of the natural language manufacturing command, generates 3D printer output data based on the generated CAD data or STL file data to control the output of a manufacturing module or part, generates manufacturing module data applicable to manufacturing equipment, assembly systems, or autonomous mobile robot (AMR) systems based on the generated part data, automatically collects new 3D algorithms, new CAD generation algorithms, new artificial intelligence generation models, or new manufacturing technology information, and updates output distributions, judgment criteria, or inference result characteristics corresponding to CAD generation results or manufacturing automation control results by relearning or updating inference parameters for the output characteristics of CAD generation algorithms and manufacturing automation control algorithms, and automatically generates a manufacturing structure, plate, bolt structure, nut structure, or machine part structure according to a natural language command input by a user, and Output, and the natural language manufacturing command includes command information for instructing a manufacturing purpose or manufacturing condition based on voice data or text data entered by a user, the CAD data is design data including shape information, dimension information, assembly information, or output information of a manufacturing structure, the STL file data includes three-dimensional shape data for performing three-dimensional shape output or 3D printer output, and the manufacturing module data may include structural data, control data, or output data usable in manufacturing equipment, assembly systems, or autonomous mobile robot systems.

[0091] The processor according to one embodiment of the present invention collects object recognition data from a plurality of cameras, sensors, or image input devices, performs object recognition, part recognition, manufacturing environment recognition, obstacle recognition, or work environment analysis based on image data which is object recognition data collected from the plurality of cameras or image input devices, automatically collects object recognition models, computer vision models, artificial intelligence inference models, and manufacturing-related source code provided from an external artificial intelligence platform, a research institution server, an open source repository, or a technology provider server, automatically trains a second model suitable for the manufacturing environment based on the collected artificial intelligence models and source code, generates an artificial intelligence inference model or inference algorithm capable of responding to new parts, new technologies, and new manufacturing conditions, generates a movement path, work method, manufacturing process execution method, or manufacturing equipment control strategy of an autonomous mobile robot (AMR) based on the generated artificial intelligence inference results, converts user voice commands into manufacturing control commands using a voice inference model or translation model capable of responding to multiple national languages, and when new data including new manufacturing technology, new artificial intelligence model, new manufacturing part information, or new manufacturing environment information is collected, automatically learns and reflects the new data to update the inference logic, control algorithm, or model parameters of the manufacturing system, and the second model is the plurality of An artificial intelligence model trained to output object recognition results, part recognition results, manufacturing environment recognition results, obstacle recognition results, or work environment analysis results within a manufacturing environment, using training data generated based on object recognition data and image data collected from a camera, sensor, or image input device as input data; and configured to generate environment recognition information including location information, shape information, and state information of objects within the manufacturing environment, and to update recognition results and inference results according to new parts, new manufacturing conditions, or changes in the manufacturing environment.The object recognition data includes at least one of object location information, part shape information, obstacle information, or work target information within a manufacturing environment, and the artificial intelligence inference model or inference algorithm includes control logic or control algorithm for changing a manufacturing control method or a manufacturing process execution method in response to changes in the manufacturing environment, addition of new parts, changes in new manufacturing conditions, or application of new technology, and the manufacturing equipment control strategy includes strategy information for controlling the operation sequence, movement method, work priority, or manufacturing process execution conditions of the manufacturing equipment, and the updating of the inference logic or control algorithm of the manufacturing system may include a process of updating the characteristics of the inference result of the artificial intelligence model or the control output value generated corresponding to each performance by retraining model parameters, feature vectors, or inference logic corresponding to manufacturing accuracy, manufacturing automation level, object recognition performance, or manufacturing strategy analysis performance in accordance with the reflection of new manufacturing technology or a new artificial intelligence model, or by updating inference parameters.

[0092] The processor according to one embodiment of the present invention configures a digital twin-based manufacturing simulation environment based on real-time manufacturing product status information, manufacturing environment information, manufacturing equipment operation information, and manufacturing process execution information; generates a first policy scenario using a third model trained to generate a first policy scenario for cases where a manufacturing control method or a manufacturing process execution method is changed in response to changes in the manufacturing environment, addition of new parts, changes in new manufacturing conditions, or application of new technology in the digital twin-based manufacturing simulation environment; generates a second policy scenario using a fourth model trained to generate a second policy scenario for cases where the manufacturing control method or a manufacturing process execution method is maintained in the digital twin-based manufacturing simulation environment; calculates a first policy profit / loss value including estimated costs due to manufacturing discontinuation, estimated costs due to process change, and estimated revenue due to process change based on the first policy scenario; calculates a second policy profit / loss value including estimated costs due to manufacturing maintenance based on the second policy scenario; calculates an optimal policy transition point for changing the manufacturing control method or the manufacturing process execution method by comparing the first policy profit / loss value and the second policy profit / loss value; and calculates the optimal policy transition point calculated after a pre-set reference point. In this case, a third policy scenario is generated through the third model by resimulating the first policy scenario by reflecting real-time manufacturing status information at the time of reaching the reference point, and a third policy profit and loss value including manufacturing stoppage costs, process re-change costs, and process reapplication revenue is recalculated based on the third policy scenario, and the optimal policy transition point is recalculated by reflecting the third policy profit and loss value.Based on the finally calculated optimal policy transition point, control is applied to manufacturing equipment or an autonomous mobile robot (AMR) by controlling the application of a manufacturing control method or a manufacturing process execution method corresponding to the first policy scenario or the third policy scenario; the third model is a policy decision artificial intelligence model trained to take policy scenario learning data, including real-time manufacturing product status information, manufacturing environment information, manufacturing equipment operation information, manufacturing process execution information, manufacturing cost information, production schedule information, supply chain information, and manufacturing downtime loss information, as input data, and output a first policy scenario including manufacturing downtime costs, process change costs, process change profits, and production efficiency change information when changing the manufacturing control method or manufacturing process execution method in response to changes in the manufacturing environment or new manufacturing conditions, and is configured to dynamically generate a manufacturing strategy according to changes in the manufacturing status; and the fourth model is a policy scenario learning data, including real-time manufacturing product status information, manufacturing environment information, manufacturing equipment operation information, manufacturing process execution information, manufacturing cost information, production schedule information, supply chain information, and manufacturing maintenance loss information, as input data, and outputs production delay costs, manufacturing inefficiency costs, and supply when maintaining the manufacturing control method or manufacturing process execution method even in response to changes in the manufacturing environment or new manufacturing conditions. It is a policy evaluation AI model trained to output a second policy scenario including delay costs and maintenance strategy information, and can be configured to estimate the economic impact and operational risks when maintaining existing manufacturing control methods.

[0093] The processor according to one embodiment of the present invention identifies a worker through voice recognition and generates a worker ID to distinguish the worker; calculates a reliability score for each of manufacturing conditions, dimensional information, structural information, part specification information, bolt specification information, nut specification information, plate structure information, and manufacturing purpose information based on manufacturing history data including natural language manufacturing command data, defective product occurrence history, and profit and loss data accumulated and stored for the identified first worker; if the reliability score calculated for at least one item among the manufacturing conditions, dimensional information, structural information, part specification information, bolt specification information, nut specification information, plate structure information, and manufacturing purpose information is less than a preset threshold, the item is identified as caution information; and regarding the caution information, based on the first worker's natural language manufacturing command input tendency and manufacturing history data, generates a filtering algorithm including an input normalization rule and a semantic correction rule to normalize the natural language input value for the caution information and correct semantic misunderstanding or input deviation in conjunction with the reliability score calculation result; simulates the result of reinterpreting the manufacturing history data based on natural language by applying the filtering algorithm; and as a result of the simulation, the reliability score If improvement is not made beyond a threshold, a plurality of second workers whose reliability scores regarding the aforementioned caution information are within a certain rank are identified, and a fifth model trained to generate correction data for the first worker's caution information using the manufacturing history data of the plurality of second workers and the first worker's non-caution information as input is used to generate a caution information correction result reflecting the correction data, and when a new natural language manufacturing command of the first worker is input, the caution information is automatically corrected through the fifth model to generate a manufacturing command interpretation result, and the defect occurrence history includes manufacturing conditions, dimensional information, structural information, part specification information, bolt specification information, nut specification information,Among plate structure information and manufacturing purpose information, the input value that causes defective products and the correction history of said input value are included; the reliability score is a value calculated based on repeatability, error frequency, correction history, and profit / loss impact extracted from the manufacturing history data; the filtering algorithm is configured to normalize or correct the semantics of the natural language input value for the attention information and is dynamically updated by reflecting the weight distribution of repeatability, error frequency, and correction history used in calculating the reliability score; the fifth model is an artificial intelligence model trained to learn manufacturing deviations and error patterns between workers by taking as input data time-series manufacturing data including the manufacturing history data of the first worker, manufacturing history data of a plurality of second workers, natural language manufacturing command data, and defect occurrence history, and to output corrected manufacturing conditions, dimensional information, structure information, part specification information, bolt specification information, nut specification information, plate structure information, and manufacturing purpose information for the attention information; and may be configured to include an error correction function to adaptively correct manufacturing command interpretation results according to worker characteristics and to correct natural language manufacturing command interpretation errors due to input deviations by worker.

[0094] A device according to one embodiment of the present invention may be controlled by a computer program stored on a medium to execute any one of the operation methods by the processor of the system described above, in combination with hardware.

[0095] According to the present invention, by automatically collecting and analyzing international news information, paper information, raw material information, sales platform information, advertising information, logistics information, and the latest technology information, manufacturing companies can effectively respond quickly to market changes.

[0096] According to the present invention, an AI model analyzes global market changes, changes in international affairs, and fluctuations in raw material prices in real time to automatically derive manufacturing and sales strategies, thereby improving the speed of decision-making and increasing operational efficiency.

[0097] According to the present invention, by integrating and analyzing sales data, advertising data, and market response data, it is possible to optimize advertising budgets, sales channels, and inventory, thereby having the effect of improving the profitability of manufacturing companies.

[0098] According to the present invention, by continuously and automatically learning the latest AI papers, the latest open-source models, and the latest manufacturing technologies, it is possible to implement an autonomous evolutionary AI manufacturing platform in which the manufacturing system autonomously evolves its functions.

[0099] According to the present invention, CAD data and STL data can be automatically generated based on natural language commands, so even users lacking expertise in CAD design can easily design and manufacture parts for production.

[0100] According to the present invention, the manufacturing process can be automated by directly linking AI-based design data with 3D printers and manufacturing equipment, thereby enabling a reduction in manufacturing time and production costs.

[0101] According to the present invention, since manufacturing strategies can be reflected in the actual manufacturing process by linking with an autonomous mobile robot (AMR), smart factory equipment, and a manufacturing control system, there is an effect of improving the level of manufacturing automation and increasing the efficiency of smart factory implementation.

[0102] According to the present invention, by providing a preemptive raw material securing strategy in response to changes in the international supply chain and fluctuations in raw material prices, it is possible to reduce supply chain risks and improve manufacturing stability.

[0103] According to the present invention, by implementing an AI-based industrial operation platform that comprehensively analyzes manufacturing, logistics, sales, advertising, raw materials, technology trends, and manufacturing automation, it is possible to effectively support the digital transformation (DX) and AI-based smart manufacturing innovation of manufacturing companies.

[0104] FIG. 1 is a block diagram schematically illustrating the basic configuration of a device according to one embodiment.

[0105] Referring to FIG. 1, a system for providing a solution for deriving manufacturing and sales strategies for manufactured goods using an AI model based on automatic collection of international trend information and raw material information according to one embodiment of the present invention may include a device (100) comprising an electronic device (100) including a processor (110) and a memory (120). The electronic device (100) according to one embodiment may be a server or a terminal. According to one embodiment, the processor (110) may be configured to perform operations or data processing regarding the control and / or communication of each component of the electronic device (100), and may be composed of one or more processors (110).

[0106] For example, the processor (110) may be configured to be coupled with a manufacturing system including manufacturing equipment, an assembly system, or an autonomous mobile robot (AMR).

[0107] The memory (120) can store information related to the method of providing a solution for deriving manufacturing and sales strategies for manufactured goods using an AI model based on automatic collection of international trend information and raw material information described above, or can store a program in which the method described above is implemented. The memory (120) may be a volatile memory (120) or a non-volatile memory (120).

[0108] According to one embodiment, the processor (110) can execute a program and control a device. The code of the program executed by the processor (110) can be stored in memory (120). Operations of the processor (110) can be performed by loading instructions stored in memory (120). The electronic device (100) 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 exchange data.

[0109] According to one embodiment, there are no limitations on the computation and data processing functions that the processor (110) can implement on the electronic device (100), but below, the process processing functions of a system that performs the method of providing a solution for deriving manufacturing and sales strategies for manufactured goods using an AI model based on automatic collection of international trend information and raw material information according to the present invention will be described.

[0110] Additionally, the electronic device (100) according to one embodiment may further include a communication unit (130).

[0111] Although not shown, the electronic device (100) is connected to a communication unit (130) and can transmit and receive signals and / or data to and from other electronic devices, other (multiple external) servers, other systems (e.g., manufacturing systems), and databases through the communication unit (130).

[0112] The communication unit (130) can be provided with a conductive line or a wireless communication chip, etc.

[0113] Hereinafter, the operation of the processor (110) can be described as the operation of the electronic device (100).

[0114] FIGS. 2 to 4 are basic operation flowcharts of a method according to one embodiment.

[0115] As illustrated in FIG. 2, the processor (110) can automatically collect data to be collected from a plurality of external servers or databases in real time or periodically (S210).

[0116] For example, the data to be collected may include at least one of international news information, industry news information, economic news information, paper information, open source information, artificial intelligence model information, platform API information, sales data information, advertising data information, raw material price information, supply chain information, logistics information, exchange rate information, manufacturing process information, information related to autonomous mobile robots (AMR), camera recognition information, and manufacturing-related technology trend information.

[0117] The processor (110) can classify the collected data by category, remove duplicate data, and extract and summarize data corresponding to the execution of the manufacturing process to convert it into training data for manufacturing strategy analysis (S220).

[0118] The processor (110) can train the first model (S230) based on the training data.

[0119] For example, the first model may be an artificial intelligence model trained to take training data for manufacturing strategy analysis as input data and output manufacturing strategy, sales strategy, advertising strategy, inventory strategy, and raw material securing strategy based on sales volume data, advertising efficiency data, inventory data, manufacturing cost data, raw material price fluctuation data, market trend change data, and international situation change data extracted from the training data for manufacturing strategy analysis.

[0120] In addition, the first model can be configured to infer first strategy information.

[0121] For example, the first strategic information may include a manufacturing strategy, sales strategy, advertising strategy, inventory strategy, and raw material securing strategy for the timing of manufacturing, the timing of raw material purchasing, the timing of advertising execution, production volume control, selection of sales channels, and optimization of the manufacturing process.

[0122] The manufacturing strategy may include strategic information for determining the production timing, production method, method of performing the manufacturing process, or manufacturing priority of the manufactured product.

[0123] The sales strategy may include strategic information for determining the sales channels, advertising execution methods, target markets, or sales timing of the manufactured product.

[0124] A raw material procurement strategy may include strategic information for determining the timing of raw material purchases, the quantity of raw materials to be purchased, whether to pre-purchase raw materials, or the method of supplying and sourcing raw materials.

[0125] A manufacturing process optimization strategy may include strategic information for coordinating the operation of manufacturing equipment, the sequence of manufacturing processes, the allocation of manufacturing resources, or the execution of manufacturing automation.

[0126] For example, the first model may refer to a multivariate prediction and decision support artificial intelligence model configured to take training data for manufacturing strategy analysis as input, convert sales-related indicators, advertising-related indicators, inventory-related indicators, manufacturing cost-related indicators, raw material price fluctuation indicators, market trend indicators, and international situation indicators extracted from the training data into feature vectors for training, and output production strategies, sales strategies, advertising strategies, inventory strategies, and raw material securing strategies for manufactured products simultaneously or selectively.

[0127] In addition, the first model can be configured to generate complex manufacturing strategies by learning the interrelationships between input data (e.g., correlation between raw material price fluctuations and production timing, correlation between advertising efficiency and sales channels, etc.).

[0128] Meanwhile, referring to FIGS. 2 and 3, after performing step S210, the processor (110) can input the data to be collected into the first model (S310).

[0129] When the processor (110) obtains output from the first model (S320), it can analyze the sales volume of the manufactured product, advertising efficiency, inventory volume, manufacturing cost, fluctuations in raw material prices, changes in market trends and changes in international affairs to infer the first strategy information described above (S330).

[0130] The processor (110) can generate second strategy information (S340) based on first strategy information.

[0131] For example, the second strategic information may include the timing of manufacturing, the timing of raw material purchasing, the timing of advertising budget execution, production volume control strategies, sales channel strategies, and manufacturing process optimization strategies.

[0132] The processor (110) can provide (S350) second strategy information to a worker and / or manufacturing system using the electronic device (100).

[0133] For example, the second strategy information may be provided through a display. The display may be provided on the operator's electronic device and / or on the electronic device (100), but is not limited thereto.

[0134] Additionally, in performing step S350, the processor (110) may provide article information, paper information, artificial intelligence technology information, and market analysis information to the worker and / or manufacturing system in the form of a summary or in-depth analysis.

[0135] Meanwhile, referring to FIG. 4, the processor (110) can automatically collect (S410) at least one of the newly disclosed artificial intelligence models, computer vision models, manufacturing algorithms, robot control algorithms, and open source technology information.

[0136] Accordingly, with reference to FIGS. 2 and FIGS. 4, the processor (110) may retrain the first model (S230) by reflecting newly collected public data into the first model, and / or, although not illustrated, the processor (110) may update the inference logic or inference parameters of the first model.

[0137] In the present invention, “retraining” refers to the process of re-optimizing the weights or parameters of an artificial intelligence model using newly collected data or an updated data set, and “updating inference parameters” may refer to the process of adjusting the inference results by updating thresholds, weight coefficients, judgment criteria, or output correction values ​​applied during the inference stage without retraining the entire model.

[0138] Meanwhile, the manufacturing system may be configured to update the output distribution, judgment criterion weight, prediction sensitivity, or strategy priority calculation criterion output characteristics corresponding to the manufacturing strategy analysis result by retraining the first model or artificial intelligence inference model by the processor (110) or updating the inference logic or inference parameters according to the reflection of new manufacturing technology, new artificial intelligence model, or new training data.

[0139] For example, in the present invention, the term “output characteristics of manufacturing strategy analysis results” may refer to structural and numerical characteristics of a model output including at least one of the output distribution, judgment criteria weight, prediction sensitivity, decision threshold, and strategy priority calculation method for manufacturing strategy, sales strategy, advertising strategy, inventory strategy, and raw material securing strategy output by the first model or artificial intelligence inference function.

[0140] In addition, updating output characteristics may mean that the reference value, judgment boundary, or output probability distribution of the strategy output is changed as the parameters or inference logic of the first model are readjusted in accordance with the reflection of new data.

[0141] Meanwhile, although not illustrated, a processor (110) according to one embodiment of the present invention may receive text or voice-based natural language manufacturing commands from a user.

[0142] For example, a natural language manufacturing command may include command information for instructing a manufacturing purpose or manufacturing conditions based on voice data or text data entered by a user.

[0143] For example, receiving natural language manufacturing commands may be performed through an input device (e.g., microphone, mouse, touch panel, keyboard, etc.) separately provided in the electronic device (100) and / or through the operator's electronic device, but is not limited thereto.

[0144] The processor (110) can analyze manufacturing conditions, dimensional information, structural information, part specification information, bolt specification information, nut specification information, plate structure information and manufacturing purpose information included in natural language manufacturing commands.

[0145] For example, the processor (110) may be configured to analyze context information, intent information, and work purpose information included in natural language manufacturing commands based on natural language processing to extract design parameters corresponding to manufacturing purposes and manufacturing conditions.

[0146] The processor (110) can automatically generate CAD data, three-dimensional (3D) model data, STL file data, or manufacturing structure data based on the analysis results of natural language manufacturing commands.

[0147] For example, CAD data may be design data including shape information, dimensional information, assembly information, or output information of a structure for manufacturing.

[0148] STL file data may include three-dimensional shape data for performing three-dimensional shape output or 3D printer output.

[0149] The processor (110) can generate 3D printer output data based on generated CAD data or STL file data and control the manufacturing system to enable the output of a manufacturing module or part.

[0150] The processor (110) can generate manufacturing module data applicable to manufacturing equipment, assembly systems, or autonomous mobile robot (AMR) systems included in a manufacturing system based on the generated part data.

[0151] For example, manufacturing module data may include structural data, control data, or output data available for use in manufacturing equipment, assembly systems, or autonomous mobile robot systems.

[0152] The processor (110) can automatically generate and output manufacturing structures, plates, bolt structures, nut structures, or machine part structures according to natural language commands entered by the user.

[0153] Meanwhile, the processor (110) can automatically collect new 3D algorithms, new CAD generation algorithms, new artificial intelligence generation, as well as model or new manufacturing technology information.

[0154] The processor (110) can update output distribution, judgment criteria, or inference result characteristics corresponding to CAD generation results or manufacturing automation control results based on new 3D algorithms, new CAD generation algorithms, new artificial intelligence generation, as well as models or new manufacturing technology information.

[0155] For example, updating the output distribution, judgment criteria, or inference result characteristics can be performed by the processor (110) in a manner that relearns or updates the inference parameters of the output characteristics of the CAD generation algorithm and the output characteristics of the manufacturing automation control algorithm.

[0156] Furthermore, a processor (110) according to one embodiment of the present invention can distinguish workers by identifying workers through voice recognition and generating a worker ID when receiving a natural language manufacturing command.

[0157] The processor (110) can obtain manufacturing history data including natural language manufacturing command data, defective product occurrence history, and profit and loss data accumulated and stored for the identified first worker.

[0158] For example, the defect occurrence history may include input values ​​that cause defects among manufacturing conditions, dimensional information, structural information, part specification information, bolt specification information, nut specification information, plate structure information, and manufacturing purpose information, as well as the modification history of said input values.

[0159] For example, profit and loss data may include manufacturing-related financial data including at least one of production costs, raw material costs, logistics costs, advertising costs, manufacturing stoppage costs, rework costs, sales revenue, supply delay losses, and process change costs associated with the production of manufactured goods.

[0160] The processor (110) can calculate a reliability score for each of the manufacturing conditions, dimensional information, structural information, part specification information, bolt specification information, nut specification information, plate structure information, and manufacturing purpose information based on manufacturing history data.

[0161] For example, the reliability score may be a value calculated based on repeatability, error frequency, correction history, and profit and loss impact extracted from manufacturing history data.

[0162] The processor (110) can identify an item as caution information if the reliability score calculated for at least one item among manufacturing conditions, dimensional information, structural information, part specification information, bolt specification information, nut specification information, plate structure information and manufacturing purpose information is less than a preset threshold.

[0163] When attention information is identified, the processor (110) can generate a filtering algorithm based on the first worker's natural language manufacturing command input tendency and manufacturing history data regarding the attention information.

[0164] For example, the filtering algorithm may be linked to the result of calculating a confidence score and may include input normalization rules and semantic correction rules to normalize natural language input values ​​regarding attention information and correct semantic misunderstandings or input deviations.

[0165] Specifically, the filtering algorithm is configured to normalize or semantize natural language input values ​​for attention information and can be dynamically updated by reflecting the weight distribution of repeatability, error frequency, and correction history used in calculating the confidence score.

[0166] In addition, the filtering algorithm in the present invention is not limited to a simple natural language normalization function, but can function as a preprocessing and feature refinement module for evaluating the reliability of manufacturing conditions, dimensional information, structural information, part specification information, bolt specification information, nut specification information, plate structure information, and manufacturing purpose information based on the interpretation results of the first worker's natural language manufacturing command.

[0167] That is, the filtering algorithm can be configured to analyze semantic consistency between expressions extracted from natural language manufacturing commands and manufacturing history data, remove or correct distortion factors in feature values ​​used to calculate reliability scores for each manufacturing information item, and provide the correction result as the input for calculating reliability scores.

[0168] Accordingly, the filtering algorithm operates as an intermediate processing step to refine data quality between the natural language interpretation stage and the confidence score calculation stage, and can play a role in improving the accuracy of the confidence score by reflecting expression variations by operator and error patterns based on manufacturing history.

[0169] The filtering algorithm can be configured to analyze the semantic similarity of input values ​​in response to deviations in natural language expressions by worker, and to correct input feature values ​​based on semantic consistency with previously stored manufacturing history data.

[0170] The filtering algorithm is a component that performs preprocessing and semantic refinement functions to normalize the expression form, term usage method, and input deviation of natural language manufacturing commands input by a worker, and to ensure consistency in semantic interpretation, prior to the final manufacturing command correction performed by the fifth model described below. The fifth model functions as a post-correction model that generates the final manufacturing command interpretation result by learning manufacturing deviations and error patterns among workers based on the input data refined through the filtering algorithm and manufacturing history data for each worker.

[0171] Dynamic updating in the filtering algorithm does not mean a change in a fixed rule-based structure, but rather means that the input normalization rule and semantic correction rule are readjusted to reflect the weight distribution as the weight distribution changes based on repeatability, error frequency, and correction history derived from manufacturing history data; this can be implemented in a form that is adaptively updated based on the learning results or statistical analysis results of the fifth model.

[0172] Accordingly, the processor (110) can simulate the result of reinterpreting manufacturing history data based on natural language by applying a filtering algorithm.

[0173] As a result of the simulation, if the reliability score does not improve above a threshold, the processor (110) can identify multiple second workers whose reliability score is within a certain rank for the attention information.

[0174] The processor (110) can generate a correction result for attention information that reflects the correction data through a fifth model that is trained to generate correction data for attention information of the first worker by taking manufacturing history data of multiple second workers and non-attention information of the first worker as inputs.

[0175] For example, non-cautionary information may include manufacturing conditions, dimensional information, structural information, part specification information, bolt specification information, nut specification information, plate structure information, or manufacturing purpose information in which the reliability score is calculated to be above a preset threshold.

[0176] For example, the fifth model may be an artificial intelligence model trained to take time-series manufacturing data per worker, including manufacturing history data of a first worker, manufacturing history data of a plurality of second workers, natural language manufacturing command data, and defect occurrence history, as input data, learn manufacturing deviations and error patterns between workers, and output corrected manufacturing conditions, dimensional information, structural information, part specification information, bolt specification information, nut specification information, plate structure information, and manufacturing purpose information regarding attention information.

[0177] The fifth model may be configured to include an error correction function to adaptively correct manufacturing command interpretation results according to worker characteristics and to correct natural language manufacturing command interpretation errors due to input deviations per worker.

[0178] Afterward, when a new natural language manufacturing command from the first worker is input, the processor (110) can automatically correct attention information through the fifth model to generate a manufacturing command interpretation result.

[0179] Meanwhile, although not illustrated, a processor (110) according to one embodiment of the present invention is connected to a plurality of cameras, sensors, or image input devices through a communication unit (130) and can collect object recognition data from these devices.

[0180] For example, object recognition data may include at least one of object location information, part shape information, obstacle information, or work target information within a manufacturing environment.

[0181] Specifically, the processor (110) can perform object recognition, part recognition, manufacturing environment recognition, obstacle recognition, or work environment analysis based on image data. For example, the image data may be object recognition data collected from a camera or image input device.

[0182] The processor (110) can automatically collect object recognition models, computer vision models, artificial intelligence inference models, and manufacturing-related source code provided from external artificial intelligence platforms, research institution servers, open source repositories, or technology provision servers.

[0183] The processor (110) can automatically train a second model suitable for a manufacturing environment based on the collected artificial intelligence model and source code, and can generate an artificial intelligence inference model or inference algorithm capable of responding to new parts, new technologies and new manufacturing conditions.

[0184] For example, the second model may be an artificial intelligence model trained to output object recognition results, part recognition results, manufacturing environment recognition results, obstacle recognition results, or work environment analysis results within a manufacturing environment, using training data generated based on object recognition data and image data collected from a plurality of cameras, sensors, or image input devices as input data.

[0185] The second model may be configured to generate environment recognition information including location information, shape information, and state information of objects within a manufacturing environment, and to update recognition results and inference results according to new parts, new manufacturing conditions, or changes in the manufacturing environment.

[0186] For example, the second model may be an artificial intelligence model that recognizes the location, shape, and state of objects within a manufacturing environment based on object recognition data and image data collected through an image input device or sensor, and thereby converts the state of manufacturing equipment, autonomous mobile robots (AMRs), and the manufacturing process execution environment into quantified environmental recognition information.

[0187] The output of this second model can be transmitted to the first model or the manufacturing equipment control unit and utilized to generate manufacturing process execution paths, work sequences, and equipment control strategies.

[0188] For example, environmental perception information generated in the second model can be converted into a feature vector including manufacturing process risk, workability status, or the possibility of obstacle occurrence, and used as input data for manufacturing strategy inference in the first model.

[0189] For example, an artificial intelligence inference model or inference algorithm may include control logic or control algorithms for changing the manufacturing control method or the method of performing the manufacturing process in response to changes in the manufacturing environment, the addition of new parts, changes in new manufacturing conditions, or the application of new technology.

[0190] The processor (110) can generate a movement path, work method, manufacturing process execution method, or manufacturing equipment control strategy for an autonomous mobile robot (AMR) based on the artificial intelligence inference results generated through the second model.

[0191] For example, a manufacturing equipment control strategy may include strategic information for controlling the operation sequence, movement method, work priority, or manufacturing process execution conditions of the manufacturing equipment.

[0192] For example, the processor (110) can convert a user voice command into a manufacturing control command using a voice inference model or a translation model capable of responding to multiple national languages.

[0193] The speech inference model and / or translation model may be stored in memory (120) or replaced with a model from an external translation server, but are not limited thereto.

[0194] Meanwhile, when new data including new manufacturing technology, new artificial intelligence model, new manufacturing part information, or new manufacturing environment information is collected, the processor (110) can automatically learn and reflect the new data to update the inference logic, control algorithm, or model parameters of the manufacturing system.

[0195] For example, updating the inference logic or control algorithm of a manufacturing system may include a process of updating the characteristics of the inference results or control output values ​​of an artificial intelligence model generated corresponding to each performance by retraining model parameters, feature vectors, or inference logic corresponding to manufacturing accuracy, manufacturing automation level, object recognition performance, or manufacturing strategy analysis performance, or by updating inference parameters, in accordance with the reflection of new manufacturing technology or a new artificial intelligence model.

[0196] In this invention, “improvement in manufacturing accuracy, level of manufacturing automation, object recognition performance, and manufacturing strategy analysis performance” does not refer to consequential effects, but rather to results resulting from changes in the model’s internal structure and output function caused by the expansion of training data due to the reflection of new data, the reconstruction of feature vectors, the readjustment of model parameters, and the updating of inference logic.

[0197] The manufacturing system may be composed of an integrated artificial intelligence-based manufacturing decision-making system including a first model and a second model, manufacturing environment recognition information generated in the second model may be used as an input variable for the first model, and manufacturing strategy information generated in the first model may be converted into motion control information for a manufacturing equipment control unit or an autonomous mobile robot (AMR) and executed.

[0198] Meanwhile, although not illustrated, a processor (110) according to one embodiment of the present invention can configure a digital twin-based manufacturing simulation environment based on real-time manufacturing product status information, manufacturing environment information, manufacturing equipment operation information and manufacturing process execution information.

[0199] The processor (110) can generate a first policy scenario through a previously learned third model in response to changes in the manufacturing environment, addition of new parts, changes in new manufacturing conditions, or application of new technology in a digital twin-based manufacturing simulation environment.

[0200] For example, the first policy scenario may be a scenario for changing the manufacturing control method or the method of performing the manufacturing process.

[0201] In the present invention, a policy scenario may refer to a simulation-based decision data set for predicting the results of applying multiple manufacturing policies in response to changes in the manufacturing environment, production schedule, supply chain, manufacturing equipment status, or manufacturing conditions in a digital twin-based manufacturing simulation environment.

[0202] For example, the third model may be a policy decision artificial intelligence model trained to take policy scenario learning data, including real-time manufacturing product status information, manufacturing environment information, manufacturing equipment operation information, manufacturing process execution information, manufacturing cost information, production schedule information, supply chain information, and manufacturing downtime loss information, as input data, and to output a first policy scenario including manufacturing downtime costs, process change costs, process change revenue, and production efficiency change information when changing the manufacturing control method or manufacturing process execution method in response to changes in the manufacturing environment or new manufacturing conditions.

[0203] The third model can be configured to dynamically generate manufacturing strategies based on changes in manufacturing conditions.

[0204] Additionally, the processor (110) can use a previously learned fourth model to generate a second policy scenario for maintaining a manufacturing control method or a manufacturing process execution method in a digital twin-based manufacturing simulation environment.

[0205] For example, the fourth model may be a policy evaluation artificial intelligence model trained to take policy scenario learning data, including real-time manufacturing product status information, manufacturing environment information, manufacturing equipment operation information, manufacturing process execution information, manufacturing cost information, production schedule information, supply chain information, and manufacturing maintenance loss information, as input data, and to output a second policy scenario including production delay costs, manufacturing inefficiency costs, supply delay costs, and maintenance strategy information when maintaining a manufacturing control method or a manufacturing process execution method even in the event of a change in the manufacturing environment or a change in new manufacturing conditions.

[0206] The fourth model can be configured to estimate the economic impact and operational risks when maintaining existing manufacturing control methods.

[0207] The processor (110) can calculate a first policy profit and loss value including an estimated cost due to manufacturing discontinuation, an estimated cost due to process change, and an estimated profit due to process change based on a first policy scenario.

[0208] The processor (110) can calculate a second policy profit and loss value including the expected cost of manufacturing maintenance based on a second policy scenario.

[0209] Accordingly, the processor (110) can calculate the optimal policy switching time for changing the manufacturing control method or the manufacturing process execution method by comparing the first policy profit / loss value and the second policy profit / loss value.

[0210] At this time, if the optimal policy transition time is calculated to be after a preset reference time, the processor (110) can generate a third policy scenario through a third model by resimulating the first policy scenario by reflecting real-time manufacturing status information at the time of reaching the reference time.

[0211] The processor (110) can recalculate the third policy profit and loss value, including manufacturing discontinuation costs, process re-change costs, and process re-application revenue, based on the third policy scenario.

[0212] The processor (110) can recalculate the optimal policy transition time by reflecting the third policy profit and loss value.

[0213] The processor (110) can control the application of a manufacturing control method or a manufacturing process execution method corresponding to a first policy scenario or a third policy scenario to the manufacturing equipment or an autonomous mobile robot (AMR) based on the final calculated optimal policy switching time.

[0214] The first policy scenario, the second policy scenario, and the third policy scenario may each include policy execution results corresponding to changes in manufacturing strategy, maintenance of manufacturing strategy, and reapplication of strategy based on resimulation, and may be configured to include expected cost changes, production efficiency changes, and supply chain risk changes over the time axis.

[0215] In the present invention, the first policy profit / loss value, the second policy profit / loss value, and the third policy profit / loss value are not merely arithmetic calculation values, but may refer to probability-based prediction values ​​or model-based estimates calculated based on the inference results of the third model and the fourth model. Here, the expected profit / loss is an expected value calculated based on the output probability distribution of the third model and the fourth model, and is a statistical prediction value that includes uncertainty.

[0216] In other words, the profit and loss value is a value generated by predicting the results of changes in the manufacturing environment, process conditions, and production status using statistical or AI-based inference methods, and can be composed of an estimated indicator for decision support rather than a fixed value of actual costs or revenues.

[0217] For example, the profit and loss value may be calculated in the form of a composite cost function or compensation function that includes at least one of manufacturing stoppage losses, changes in production efficiency, supply delay losses, raw material price fluctuation costs, manufacturing equipment operating costs, and expected sales revenue.

[0218] In addition, in the present invention, the optimal policy switching point is not determined by a simple threshold comparison result, but can be understood as a reinforcement learning-based policy selection result in which a policy selection is performed by comparing the output results of the third model and the fourth model in a digital twin-based manufacturing simulation environment.

[0219] In other words, the optimal policy transition point is the point in time derived from the process of selecting the optimal policy based on a compensation function (e.g., profit and loss value) according to the manufacturing environment state, and is the result of a decision made to adaptively respond to a dynamic manufacturing environment.

[0220] For example, in a digital twin-based manufacturing simulation environment, manufacturing state information is utilized as state information, whether the manufacturing control method or the manufacturing process execution method is changed is utilized as action information, and profit and loss values ​​are utilized as a reward function to perform policy optimization.

[0221] The optimal policy transition point is defined as the point in time when the difference between the expected profit / loss or expected cost is maximized based on the comparison result between the first policy profit / loss value and the second policy profit / loss value, and is characterized by being determined by sequentially evaluating profit / loss estimates that change along the time axis in a digital twin-based manufacturing simulation environment.

[0222] In addition, the digital twin-based manufacturing simulation environment in the present invention refers to a virtual manufacturing environment that is continuously updated by reflecting the status information of the actual manufacturing site in real time, and the reflection of real-time manufacturing status information can be performed through state synchronization between the digital twin environment and the actual manufacturing system.

[0223] Based on this synchronization process, the operating status of manufacturing equipment, the progress status of processes, and the resource allocation status are reflected in the virtual environment, and accordingly, scenario re-simulation and policy re-evaluation can be performed in real time.

[0224] A device according to one embodiment of the present invention may be controlled by a computer program stored in a medium to execute any one of the operation methods by the processor (110) of the system described above, in combination with hardware.

[0225] 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.

[0226] 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.

[0227] 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.

[0228] 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.

[0229] Therefore, other implementations, other embodiments, and equivalents to the claims also fall within the scope of the claims set forth below.

Claims

Claim 1 In a device for providing a solution for deriving manufacturing and sales strategies for manufactured goods using an AI model based on automatic collection of international trend information and raw material information, the device comprises: memory; a communication unit; The system includes a processor that is connected to the communication unit and memory and configured to be linked with a manufacturing system including manufacturing equipment, an assembly system, or an autonomous mobile robot (AMR). The processor automatically collects international news information, industrial news information, economic news information, paper information, open source information, artificial intelligence model information, platform API information, sales data information, advertising data information, raw material price information, supply chain information, logistics information, exchange rate information, manufacturing process information, information related to autonomous mobile robots (AMR), camera recognition information, and manufacturing-related technology trend information from a plurality of external servers or databases in real-time or periodically. It classifies the collected data by category, removes duplicate data, extracts and summarizes data corresponding to the execution of the manufacturing process, and converts it into training data for manufacturing strategy analysis. It trains a first model based on the training data. Through the first model, it analyzes the sales volume of manufactured products, advertising efficiency, inventory volume, manufacturing cost, fluctuations in raw material prices, changes in market trends, and changes in international affairs to infer manufacturing strategies, sales strategies, advertising strategies, inventory strategies, and raw material procurement strategies. Based on the inferred strategy information, it determines the timing of manufacturing, the timing of raw material purchase, and the execution of the advertising budget. Generating timing, production volume control strategies, sales channel strategies, and manufacturing process optimization strategies; providing the generated strategy information through a display and providing article information, paper information, artificial intelligence technology information, and market analysis information in a summarized or in-depth analysis form; automatically collecting at least one newly disclosed data among newly disclosed artificial intelligence models, computer vision models, manufacturing algorithms, robot control algorithms, and open source technology information; and reflecting the automatically collected newly disclosed data in the first model.The first model is retrained or the inference logic or inference parameters are updated. The first model is an artificial intelligence model trained to output a manufacturing strategy, a sales strategy, an advertising strategy, an inventory strategy, and a raw material procurement strategy based on sales volume data, advertising efficiency data, inventory data, manufacturing cost data, raw material price fluctuation data, market trend change data, and international situation change data extracted from the training data for manufacturing strategy analysis, using the training data for manufacturing strategy analysis as input data. The model is configured to infer strategic information for manufacturing timing, raw material purchasing timing, advertising execution timing, production volume control, sales channel selection, and manufacturing process optimization. The manufacturing strategy includes strategic information for determining the production timing, production method, manufacturing process execution method, or manufacturing priority of a manufactured product. The sales strategy includes strategic information for determining the sales channel, advertising execution method, target market, or sales timing of a manufactured product. The raw material procurement strategy includes strategic information for determining the raw material purchasing timing, raw material purchase quantity, whether to pre-purchase raw materials, or raw material supply method. The manufacturing process optimization strategy includes manufacturing equipment operation method, manufacturing process sequence, and manufacturing resources. A device for providing a solution for deriving manufacturing and sales strategies for manufactured goods utilizing an AI model based on automatic collection of international trend information and raw material information, characterized in that it includes strategic information for adjusting a distribution method or a method for performing manufacturing automation, and the manufacturing system is configured to update the output characteristics of an output distribution, judgment criterion weight, prediction sensitivity, or strategy priority calculation criterion corresponding to the manufacturing strategy analysis result by retraining the first model or the artificial intelligence inference model by the processor or updating the inference logic or inference parameters according to the reflection of new manufacturing technology, a new artificial intelligence model, or new training data. Claim 2 In claim 1, the processor receives a text or voice-based natural language manufacturing command from a user, analyzes manufacturing conditions, dimensional information, structural information, part specification information, bolt specification information, nut specification information, plate structure information, and manufacturing purpose information included in the natural language manufacturing command, automatically generates CAD data, three-dimensional (3D) model data, STL file data, or manufacturing structural data based on the analysis results of the natural language manufacturing command, generates 3D printer output data based on the generated CAD data or STL file data to control the output of a manufacturing module or part, generates manufacturing module data applicable to manufacturing equipment, assembly systems, or autonomous mobile robot (AMR) systems based on the generated part data, automatically collects new 3D algorithms, new CAD generation algorithms, new artificial intelligence generation models, or new manufacturing technology information, and updates output distributions, judgment criteria, or inference result characteristics corresponding to CAD generation results or manufacturing automation control results by relearning or updating inference parameters for the output characteristics of CAD generation algorithms and manufacturing automation control algorithms, and automatically generates a manufacturing structure, plate, bolt structure, nut structure, or machine part structure according to a natural language command entered by a user. Generating and outputting, wherein the natural language manufacturing command includes command information for instructing a manufacturing purpose or manufacturing condition based on voice data or text data input by a user, wherein the CAD data is design data including shape information, dimensional information, assembly information, or output information of a manufacturing structure, wherein the STL file data includes three-dimensional shape data for performing three-dimensional shape output or 3D printer output, and wherein the manufacturing module data is structural data usable in manufacturing equipment, assembly systems, or autonomous mobile robot systems,Device for providing a solution for deriving manufacturing and sales strategies for manufactured goods utilizing an AI model based on automatic collection of international trend information and raw material information, characterized by including control data or output data. Claim 3 In claim 2, the processor collects object recognition data from a plurality of cameras, sensors, or image input devices, performs object recognition, part recognition, manufacturing environment recognition, obstacle recognition, or work environment analysis based on image data which is object recognition data collected from the plurality of cameras or the image input devices, automatically collects object recognition models, computer vision models, artificial intelligence inference models, and manufacturing-related source code provided from an external artificial intelligence platform, a research institution server, an open source repository, or a technology provider server, automatically trains a second model suitable for the manufacturing environment based on the collected artificial intelligence models and source code, generates an artificial intelligence inference model or inference algorithm capable of responding to new parts, new technologies, and new manufacturing conditions, generates a movement path, work method, manufacturing process execution method, or manufacturing equipment control strategy of an autonomous mobile robot (AMR) based on the generated artificial intelligence inference results, converts user voice commands into manufacturing control commands using a voice inference model or translation model capable of responding to multiple national languages, and when new data including new manufacturing technology, new artificial intelligence model, new manufacturing part information, or new manufacturing environment information is collected, automatically learns and reflects the new data to update the inference logic, control algorithm, or model parameters of the manufacturing system, and The second model is an artificial intelligence model trained to output object recognition results, part recognition results, manufacturing environment recognition results, obstacle recognition results, or work environment analysis results within a manufacturing environment, using training data generated based on object recognition data and image data collected from the plurality of cameras, sensors, or image input devices as input data; it is configured to generate environment recognition information including location information, shape information, and state information of objects within the manufacturing environment, and to update recognition results and inference results according to new parts, new manufacturing conditions, or changes in the manufacturing environment, and the object recognition data isA device for providing a solution for deriving manufacturing and sales strategies for manufactured goods utilizing an AI model based on automatic collection of international trend information and raw material information, characterized by comprising at least one of object location information, part shape information, obstacle information, or work target information within a manufacturing environment; wherein the artificial intelligence inference model or inference algorithm comprises control logic or control algorithm for changing a manufacturing control method or a manufacturing process execution method in response to changes in the manufacturing environment, addition of new parts, changes in new manufacturing conditions, or application of new technology; wherein the manufacturing equipment control strategy comprises strategy information for controlling the operation sequence, movement method, work priority, or manufacturing process execution conditions of the manufacturing equipment; and wherein the updating of the inference logic or control algorithm of the manufacturing system comprises a process of updating the characteristics of the inference result or control output value of the artificial intelligence model generated corresponding to each performance by retraining model parameters, feature vectors, or inference logic corresponding to manufacturing accuracy, manufacturing automation level, object recognition performance, or manufacturing strategy analysis performance in accordance with the reflection of new manufacturing technology or a new artificial intelligence model.

Images