System and method for optimizing goods production and logistics management using artificial intelligence

An AI-driven logistics management system optimizes production, packaging, and shipping processes by predicting delays and defects, automating scheduling, and precisely calculating packaging, thereby enhancing delivery efficiency and reducing costs.

KR103043128B1Active Publication Date: 2026-11-06CONCON CO LTD
View PDF 1 Cites 0 Cited by

Patent Information

Application Number
KR1020250177237
Authority / Receiving Office
KR · KR
Patent Type
Patents
Current Assignee / Owner
Filing Date
2025-11-20
Publication Date
2026-11-06
Estimated Expiration
2045-11-20

AI Technical Summary

Technical Problem

Existing merchandise production systems face challenges in integrated management due to separate operation of production, inspection, packaging, and shipping processes, leading to manual rescheduling when issues occur, delivery delays due to unpredictable lead times and defect rates, and increased transportation costs and damage rates from inaccurate packaging calculations.

Method used

A logistics management optimization system using artificial intelligence (AI) that automates the process from customer consultation to manufacturing and logistics, predicts lead times and defect rates, and optimizes schedules and packaging specifications to minimize delays and costs.

Benefits of technology

The system increases on-time delivery rates, reduces processing time, minimizes bottlenecks, and lowers logistics costs by dynamically adjusting to production variables and accurately calculating packaging needs.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure 112025130317017-PAT00004_ABST
    Figure 112025130317017-PAT00004_ABST
Patent Text Reader

Abstract

According to the present disclosure, a goods production and logistics optimization system using artificial intelligence (AI) may include a memory for storing instructions, a processor, a consultation unit, a rendering unit, and a scheduling unit. When the instructions are executed by the processor, the system can be controlled to generate a work order by receiving order information and design files required for the production of goods from a customer through a consultation unit, render the actual design and packaging state of the goods based on the work order or order information through a rendering unit, analyze order conditions including the quantity, delivery date, and price of the goods using an artificial intelligence (AI) model based on the work order through a scheduling unit, select the optimal company by considering at least one of the estimated lead time, defect rate, or equipment availability of the manufacturing company, packaging company, and transportation company, and determine at least one of the production, packaging, or shipping schedules.
Need to check novelty before this filing date? Find Prior Art

Description

Technology Field

[0001] The embodiments of the present disclosure relate to the fields of artificial intelligence (AI), goods production platforms, and supply chain management (SCM).

[0002] More specifically, the invention relates to a logistics management optimization system and method that receives customer order information, performs consultation and design rendering through an artificial intelligence model, optimizes the entire process by comprehensively analyzing the schedules, costs, and defect rates of multiple partner companies responsible for production, packaging, and transportation, and dynamically readjusts the schedule in response to variables occurring during the production process. Background Technology

[0004] With the recent increase in demand for personalized merchandise, platforms connecting multiple manufacturers, packaging companies, and logistics providers are emerging. However, existing merchandise production systems often struggled with integrated management because the production, inspection, packaging, and shipping processes were frequently operated separately. In particular, when unexpected issues such as defects or delays occurred during the production process, there was the inconvenience of having to manually re-plan and coordinate the entire schedule.

[0005] In addition, there was a problem where delivery delays or rework occurred frequently due to the lack of functionality to accurately predict and reflect in plans the varying lead times and defect rates of each supplier in advance.

[0006] Meanwhile, during the packaging stage, there were limitations in that the specifications and volume of the product could not be precisely calculated in advance, leading to the use of unnecessarily large packaging materials or improper cushioning materials, which increased transportation costs and a higher rate of damage during delivery. Prior art literature

[0008] Korean Registered Patent No. 10-2578727 (September 14, 2023) The problem to be solved

[0009] Conventional AI-based product manufacturing and logistics management optimization systems and methods had several drawbacks: the manufacturing, packaging, and logistics processes were operated in a fragmented manner, requiring manual readjustment of the entire schedule in the event of a problem; delivery delays occurred due to the inability to predict lead times and defect rates for each supplier; and transportation costs and damage rates increased due to the inaccuracy of packaging volumes.

[0010] The embodiments of the present disclosure aim to solve various problems, including those mentioned above, and to provide a system and method for optimizing product manufacturing and logistics management using artificial intelligence (AI), which automates and dynamically optimizes the entire process from customer consultation to manufacturing and logistics using artificial intelligence.

[0011] Specifically, the present invention aims to provide a system and method for performing AI consultation based on customer order information and design files, and for automatically generating a list of necessary work steps and materials to derive a work order. In addition, the present invention aims to predict transportation costs and damage rates in advance by providing a rendering function that visualizes not only the physical design of the goods but also packaging specifications, volume, and cushioning material placement based on the generated work order.

[0012] The present invention aims to comprehensively optimize the entire production, packaging, and delivery plan by using an artificial intelligence model to comprehensively predict order conditions, including the quantity, delivery date, and price of goods, as well as the expected lead time, defect rate, equipment availability, and delivery risk of multiple partner companies.

[0013] However, these tasks are exemplary and do not limit the scope of the present disclosure. means of solving the problem

[0015] According to one aspect of the present disclosure, a system for optimizing goods production and logistics using artificial intelligence (AI) comprises a memory for storing instructions, a processor, a consultation unit, a rendering unit, and a scheduling unit. When the instructions are executed by the processor, the system receives order information and design files required for the production of goods from a customer through the consultation unit and generates a work order; renders the actual design and packaging state of the goods based on the work order or order information through the rendering unit; analyzes order conditions including the quantity, delivery date, and price of the goods using an artificial intelligence (AI) model based on the work order through the scheduling unit; selects an optimal company by considering at least one of the expected lead time, defect rate, or equipment availability of the production company, packaging company, and transportation company; and determines at least one of the production, packaging, or shipping schedule.

[0016] According to the present embodiment, the consultation unit receives basic order information including at least one of an IP image file, a type of goods, a production quantity, a desired deadline, or a packaging design, and when an inquiry regarding a production style or feel is received from a customer, analyzes the customer's inquiry using an artificial intelligence (AI) model, suggests an option corresponding to the inquiry by referring to a production dictionary database including at least one of sewing fabric, embroidery, gloss, size, or process options, and can generate a work order based on the order information and the suggested option.

[0017] According to the present embodiment, the consultation unit receives an inquiry from a customer regarding whether sample production is possible or whether production is possible for quantities less than a preset minimum order quantity, refers to an FAQ database through an artificial intelligence (AI) model, and in response to the inquiry, can automatically provide a predefined answer suggesting that sampling be done by ordering the product in the minimum order unit, or, if the quantity exceeds a standard quantity requiring a separate quote, that separate verification regarding sample production availability and costs is required.

[0018] According to the present embodiment, the rendering unit automatically maps an Intellectual Property (IP) design to a 3D physical shape of a good based on design and specification information included in a work order and renders it, calculates at least one element among packaging specifications, packaging volume, cushioning material placement, or label location required to reduce the damage rate of the good and optimize transportation costs using an artificial intelligence (AI) model through simulation, generates a rendering image that visualizes the packaging state of the good based on the calculated result, and transmits the calculated packaging volume and specification information to the scheduling unit so that the scheduling unit can use it for selecting a transportation company and calculating costs.

[0019] According to the present embodiment, the rendering unit may first provide a first rendering image by combining an IP image file input through the consultation unit and information about the form of the goods before the work order is generated, and after the work order is finally confirmed, generate and provide a second rendering image including the actual design and packaging state of the goods based on the detailed specifications specified in the work order.

[0020] According to the present embodiment, when a delay or defect occurs in at least one of the production, packaging, or shipping processes of goods, the scheduling unit can recalculate the entire schedule in real time, including at least one of the production, packaging, or shipping schedules, by reflecting the delay or defect information in an artificial intelligence (AI) model, and transmit at least one of the work instruction, packaging instruction, or shipping instruction reflecting the adjusted schedule to a terminal of at least one of the manufacturers, packaging companies, or transportation companies affected by the schedule change.

[0021] According to the present embodiment, the scheduling unit analyzes data on the three-dimensional (3D) size, material, and production quantity of individual goods specified in the work order using an artificial intelligence (AI) model, calculates in advance packaging specifications including optimal packaging box dimensions, required volume of cushioning material, and label placement to minimize the damage rate during transportation of goods and optimize transportation costs based on the analysis results, compares with a database of company-specific features including at least one feature among the ability to process large orders, specialization in small orders, or expertise regarding items with a high risk of damage, selects the optimal packaging company and transportation company that meet the calculated packaging specifications, and controls the dispatch by transmitting a work order including the packaging specifications to the terminal of the selected company.

[0022] delete

[0023] According to the present embodiment, when the consultation unit receives order information and design files from a customer, it analyzes at least one element among the design complexity, number of colors, or implementation difficulty of the design file through an artificial intelligence (AI) model, and infers a production risk including an expected defect rate or expected delivery delay probability for the order information using an artificial intelligence (AI) model based on past defect rates, time history, and past order history data that had specifications similar to the design requirement specifications stored in a company-specific feature database, and if the inferred production risk exceeds a preset threshold, it may notify the customer of the risk or display alternatives including design modification and material change before generating a work order.

[0024] According to the present embodiment, the artificial intelligence (AI) model determines the suitability for production in advance by analyzing technical specifications including at least one of the resolution, color profile, or design complexity of the design file based on the input of the design file through the consultation unit, and determines the customer's priority regarding delivery date, cost, and quality based on the delivery date, price, or quantity included in the order conditions through the scheduling unit, selects the optimal company by applying differential weights to the estimated lead time, defect rate, and cost of the manufacturing company, packaging company, and transportation company according to the determined priority, detects the possibility of expected delay by receiving real-time process progress status data from the terminal of the selected company and comparing it with the optimized schedule, and if the possibility of delay is detected, controls can readjust the schedule of the subsequent process or the company to perform the subsequent process in order to comply with the final delivery date of the goods.

[0025] Other aspects, features, and advantages other than those described above will become clear from the following specific details, claims, and drawings for implementing the invention.

[0026] In addition, these general and specific aspects may be implemented using a system, method, computer program, or any combination of a system, method, or computer program. Effects of the invention

[0028] According to the exemplary embodiment of the present disclosure as described above, a system and method for optimizing product manufacturing and logistics management using artificial intelligence (AI) can be implemented, which automates and dynamically optimizes the entire process from customer consultation to manufacturing and logistics. Of course, the scope of the present disclosure is not limited by such effects.

[0029] According to the product manufacturing and logistics management optimization system and method using artificial intelligence (AI) according to the present invention, the process from AI chatbot consultation to work order generation and pre-verification through rendering can be automated to increase customer response efficiency and convenience.

[0030] Furthermore, the present invention analyzes and predicts the capabilities (e.g., lead time, defect rate, availability) of multiple manufacturing, packaging, and transportation companies in real time using an artificial intelligence model to combine optimal partners and optimize the overall schedule, thereby having the effect of increasing on-time delivery rates and reducing overall processing time compared to existing methods that relied on manual planning.

[0031] The present invention can minimize bottlenecks and reduce the rework rate by immediately detecting unexpected variables, such as delays or defects occurring during the manufacturing process, and dynamically readjusting the overall plan.

[0032] The present invention can reduce unnecessary transportation costs and minimize overall logistics costs and errors by precisely calculating and rendering packaging volume and specifications in advance, and by reducing label errors and damage rates during delivery. Brief explanation of the drawing

[0034] FIG. 1 is a drawing for explaining a system utilizing an artificial intelligence model according to an exemplary embodiment of the present disclosure. FIG. 2 is a diagram illustrating the learning of a neural network according to an exemplary embodiment of the present disclosure. FIG. 3 is a block diagram showing the configuration of a system according to an exemplary embodiment of the present disclosure. FIG. 4 is a flowchart of a method for optimizing goods production and logistics using artificial intelligence (AI) according to an exemplary embodiment of the present disclosure. FIG. 5 shows a flowchart of a method for generating a work instruction sheet for a consultation department according to an exemplary embodiment of the present disclosure. FIG. 6 shows a flowchart of an AI model weight adjustment method of a scheduling unit according to an exemplary embodiment of the present disclosure. FIG. 7 illustrates a schematic configuration diagram of a goods production and logistics linkage system using an artificial intelligence model according to an exemplary embodiment of the present disclosure. Specific details for implementing the invention

[0035] The present disclosure is capable of various modifications and may have various embodiments; specific embodiments are illustrated in the drawings and described in detail in the detailed description. The effects and features of the present disclosure and the methods for achieving them will become clear by referring to the embodiments described below in detail together with the drawings. However, the present disclosure is not limited to the embodiments disclosed below and may be implemented in various forms.

[0036] In the following embodiments, terms such as first, second, etc. are used not in a limiting sense, but for the purpose of distinguishing one component from another component.

[0037] In the following examples, singular expressions include plural expressions unless the context clearly indicates otherwise.

[0038] In the following embodiments, terms such as "include" or "have" mean that the features or components described in the specification are present, and do not preclude the possibility that one or more other features or components may be added.

[0039] In the following embodiments, when a part such as a layer, region, or component is described as being on or above another part, it includes not only cases where it is directly on top of another part, but also cases where another region, component, etc. is interposed in between.

[0040] In the drawings, the size of components may be exaggerated or reduced for convenience of explanation. For example, the size and thickness of each component shown in the drawings are depicted arbitrarily for convenience of explanation, and therefore the present disclosure is not necessarily limited to what is depicted.

[0041] Where an embodiment can be implemented differently, a specific sequence of operations may be performed differently from the order described. For example, two steps described consecutively may be performed substantially simultaneously or proceed in the reverse order of the description.

[0042] In this specification, “A and / or B” indicates the case where it is A, B, or A and B. And, “at least one of A and B indicates the case where it is A, B, or A and B.”

[0043] In the following embodiments, when layers, regions, components, etc. are described as being connected, this includes cases where the layers, regions, components are directly connected, or / or cases where other layers, regions, components are interposed between the layers, regions, components to be indirectly connected. For example, when layers, regions, components, etc. are described as being electrically connected in this specification, it indicates cases where the layers, regions, components, etc. are directly electrically connected, and / or cases where other layers, regions, components, etc. are interposed between them to be indirectly electrically connected.

[0044] The x-axis, y-axis, and z-axis are not limited to the three axes of an orthogonal coordinate system but can be interpreted in a broader sense that includes them. For example, the x-axis, y-axis, and z-axis may be orthogonal to each other, but they may also refer to different directions that are not orthogonal to each other.

[0045] The advantages and features of the present disclosure and the methods for achieving them will become clear by referring to the embodiments described below in detail together with the accompanying drawings. However, the present disclosure is not limited to the embodiments disclosed below but may be implemented in various different forms. These embodiments are provided merely to ensure that the disclosure is complete and to fully inform those skilled in the art of the scope of the present disclosure, and the present disclosure is defined only by the scope of the claims.

[0046] The terms used in this disclosure are for describing the embodiments and are not intended to limit this disclosure. In this disclosure, the singular form may include the plural form unless specifically stated otherwise in the text. The terms “comprises” and / or “comprising” used in this disclosure do not exclude the presence or addition of one or more other components in addition to the components mentioned. Throughout the disclosure, the same reference numerals refer to the same components, and “and / or” may include each of the mentioned components and all combinations of one or more. Although terms such as “first,” “second,” etc., are used to describe various components, these components are not limited by these terms. These terms are used merely to distinguish one component from another. Accordingly, the first component mentioned below may be the second component within the technical scope of this disclosure.

[0047] The word "exemplary" is used in this disclosure to mean "used as an example or illustration." Any embodiment described as "exemplary" in this disclosure should not be interpreted as being preferred or having an advantage over other embodiments.

[0048] Embodiments of the present disclosure may be described in terms of functions or blocks performing functions. A block, which may be referred to as a part or module, etc., of the present disclosure may be physically implemented by analog or digital circuits such as logic gates, integrated circuits, microprocessors, microcontrollers, memory, passive electronic components, active electronic components, optical components, hardwired circuits, etc., and may optionally be driven by firmware and software. Additionally, the term “part” as used in the disclosure refers to a hardware element such as software, FPGA, or ASIC, and the “part” may perform certain roles. However, the meaning of “part” is not limited to software or hardware. The “part” may be configured to reside in an addressable storage medium or may be configured to run one or more processors. Accordingly, as an example, a "part" may include elements such as software elements, object-oriented software elements, class elements, and task elements, as well as processes, functions, attributes, procedures, subroutines, segments of program code, drivers, firmware, microcode, circuits, data, databases, data structures, tables, arrays, and variables. The functionality provided within the elements and "parts" may be combined into a smaller number of elements and "parts" or further separated into additional elements and "parts."

[0049] Embodiments of the present disclosure may be implemented using at least one software program running on at least one hardware device and may perform network management functions to control elements.

[0050] Spatially relative terms such as "below," "beneath," "lower," "above," and "upper" may be used to facilitate the description of the relationship between one component and other components as illustrated in the drawings. Spatially relative terms should be understood as encompassing different orientations of components during use or operation, in addition to the orientations depicted in the drawings. For example, if a component depicted in a drawing is inverted, a component described as "below" or "beneath" of another component may be placed "above" of that component. Therefore, the exemplary term "below" may encompass both the lower and upper directions. Components may also be oriented in other directions, and accordingly, spatially relative terms may be interpreted according to the orientation.

[0051] Unless otherwise defined, all terms used in this disclosure (including technical and scientific terms) may be used in a meaning commonly understood by those skilled in the art to which this disclosure pertains. Additionally, terms defined in commonly used dictionaries are not to be interpreted ideally or excessively unless explicitly and specifically defined otherwise.

[0052] Hereinafter, embodiments of the present disclosure will be described in detail with reference to the attached drawings. When describing with reference to the drawings, identical or corresponding components are given the same reference numerals, and redundant descriptions thereof will be omitted.

[0054] FIG. 1 is a drawing for explaining a system utilizing an artificial intelligence model according to an exemplary embodiment of the present disclosure.

[0055] As illustrated in FIG. 1, a system (100) utilizing an artificial intelligence model may include a plurality of user terminals (e.g., a first user terminal (110-1), a second user terminal (110-2)), a server (120), and a database (130).

[0056] According to one embodiment, the database (130) is configured separately from the server (120), but it is not necessarily configured this way, and the database (130) may be located inside the server (120). For example, the server (120) may include various artificial intelligences for performing machine learning algorithms. According to another embodiment, a plurality of user terminals (110-1, 110-2), the server (120), and the database (130) may be connected so as to communicate with each other through a network (105).

[0057] According to one embodiment, a network (105) may enable wireless or wired communication between a plurality of user terminals (e.g., a first user terminal (110-1), a second user terminal (110-2)), a server (120), and a database (130). For example, the network (105) may support wireless communication in at least one of LTE (long-term evolution), LTE-A (LTE Advanced), CDMA (code division multiple access), WCDMA (wideband CDMA), WiBro (Wireless BroadBand), WiFi (wireless fidelity), Bluetooth, and NFC (near field communication).

[0058] According to one embodiment, the database (130) has the ability to store various data. The data stored in the database (130) is data acquired, processed, or used by a plurality of user terminals (e.g., a first user terminal (110-1), a second user terminal (110-2)) and at least one component of the server (120), and may include software (e.g., a program). The database (130) may include volatile and / or non-volatile memory.

[0060] FIG. 2 is a diagram illustrating the learning of a neural network according to one embodiment.

[0061] As shown in FIG. 2, the machine learning model may train a neural network (210) to process review responses received from multiple user terminals (e.g., the first user terminal (110-1) and the second user terminal (110-2) of FIG. 1) item by item. Additionally, the machine learning model may train a neural network (210) to extract user stay records based on user movement path information. According to one embodiment, the machine learning model may be a separate entity from the server (120), but this is not limited to this.

[0062] According to one embodiment, the neural network (210) includes an input layer (220) and an output layer (230), where training samples are input and training outputs are generated. The network is trained based on the difference between the training outputs and the labels. The labels are defined based on items related to review responses and may also be defined based on user dwell records according to movement path information. The neural network (210) is connected in groups of multiple nodes and may be defined by weights and activation functions between the nodes.

[0063] According to one embodiment, a machine learning model can train a neural network (210) using a Gradient Descent (GD) technique or a Stochastic Gradient Descent (SGD) technique. The device can utilize a loss function designed by the output and label of the neural network. The machine learning model can calculate a training error using a predefined loss function. The loss function can predefine the label, output, and parameter as input variables, and these parameters can be set as weights within the neural network (210).

[0064] According to one embodiment, a machine learning model can find weights that affect training error using a backpropagation technique. Here, weights represent relationships between nodes within a neural network (210). The machine learning model can use an SGD technique utilizing labels and outputs to optimize the weights found by the backpropagation technique.

[0065] According to one embodiment, a machine learning model extracts a first object from a review response, obtains a first label corresponding to this object, and applies it to a first neural network to generate a first training output corresponding to the first object. Subsequently, the first neural network can be trained based on the first training output and the first label. The machine learning model extracts a second object from movement path information, obtains a second label which is a user dwell record corresponding to this object, and applies it to a second neural network to generate a second training output corresponding to the second object. Subsequently, the second neural network can be trained based on the second training output and the second label.

[0066] According to one embodiment, a machine learning model can generate a first training feature vector based on the constituent features, location features, and pattern features of a review response, and various methods may be used for feature extraction. Additionally, the machine learning model can generate a second training feature vector based on the constituent features, length features, and pattern features of movement path information, and various methods may be used to extract features.

[0067] According to one embodiment, a machine learning model can obtain a training output by applying a first training feature vector to a neural network (210). Subsequently, the machine learning model can train a review item extraction algorithm of the neural network (210) based on the training output and the first label. The machine learning model can train the review item extraction algorithm by calculating a training error corresponding to the training output and optimizing the connection relationships between nodes within the neural network (210) to minimize this error. The server (120) can extract items from review responses by utilizing the first neural network that has completed training. The machine learning model can obtain a training output by applying a second training feature vector to the neural network (210).

[0069] FIG. 3 is a block diagram showing the configuration of a system according to an exemplary embodiment of the present disclosure.

[0070] A system (100) according to one embodiment may include a processor (310) and a memory (320), and some of the illustrated components may be omitted or substituted. A system (100) according to one embodiment may include a server or a terminal.

[0071] According to one embodiment, the processor (310) may be composed of one or more processors, configured to perform operations or data processing regarding the control and / or communication of each component of the system (100). The memory (320) may store information related to the method described above or store a program in which the method described above is implemented. The memory (320) may be volatile memory or non-volatile memory. The memory (320) may store various file data, and the stored file data may be updated according to the operation of the processor (310).

[0072] According to one embodiment, the processor (310) can execute a program and control the system (100). The code of the program executed by the processor (310) can be stored in memory (320). Operations of the processor (310) can be performed by loading instructions stored in memory (320). The system (100) can be connected to an external device (e.g., a personal computer or a network) through a communication interface and exchange data.

[0074] FIG. 4 is a flowchart of a method for optimizing goods production and logistics using artificial intelligence (AI) according to an exemplary embodiment of the present disclosure.

[0075] Although the process steps, method steps, algorithms, etc., in the flowchart of FIG. 4 are described in a sequential order, such processes, methods, and algorithms may be configured to operate in any suitable order. The steps of the processes, methods, and algorithms described in various embodiments of the present invention do not need to be performed in the order described in the present invention.

[0076] In operation 410, the consultation unit (330) can generate a work order by receiving order information and design files required for the production of goods from a customer. The order information may include at least one of an IP image file, a type of goods, a production quantity, or a desired deadline. The consultation unit (330) can receive this information through an artificial intelligence model and generate a standardized work order for subsequent processes based on it.

[0077] In operation 420, the rendering unit (340) can render the physical design and packaging state of the goods. The rendering unit (340) can automatically map the IP design to the 3D physical shape of the goods based on the generated work order or input order information. Additionally, the rendering unit (340) can generate a rendering image by visualizing the packaging state, including the packaging volume, cushioning material placement, label position, etc.

[0078] In operation 430, the scheduling unit (350) can analyze order conditions (e.g., quantity, delivery date, price) using an artificial intelligence (AI) model. The scheduling unit (350) can prepare an optimized process by extracting the quantity of goods specified in the work order, the delivery date and price conditions requested by the customer as variables.

[0079] In operation 440, the scheduling unit (350) can select an optimal supplier by considering supplier factors (e.g., lead time, defect rate). The scheduling unit (350) can refer to a database of multiple manufacturers, packaging companies, and transportation companies. The database may include at least one factor among the supplier's expected lead time, historical defect rate, equipment availability, or shipping risk. The AI ​​model can select an optimal combination of suppliers by matching the analyzed order conditions with the factors mentioned in the database.

[0080] In operation 450, the scheduling unit (350) can automatically optimize and dispatch production, packaging, and shipping schedules. The scheduling unit (350) can optimize the entire supply chain schedule so that the production completion time and the packaging start time, and the packaging completion time and the shipping start time are organically linked, going beyond simply selecting individual companies. According to the optimized schedule, the system (100) can automatically transmit work instructions or dispatch instructions to the terminals of the selected companies.

[0081] According to one embodiment, the system (100) receives order information and design files required for the production of goods from a customer through a consultation unit (330) to generate a work order, renders the actual design and packaging state of the goods based on the work order or order information through a rendering unit (340), analyzes order conditions including the quantity, delivery date, and price of the goods using an artificial intelligence (AI) model based on the work order through a scheduling unit (350), selects an optimal company by considering at least one of the expected lead time, defect rate, or equipment availability of the manufacturing company, packaging company, and transportation company, and determines at least one of the production, packaging, or shipping schedules.

[0082] According to one embodiment, the consultation unit (330) may refer to an interface that automates or assists in interaction with a customer. The consultation unit (330) may receive order information required for the production of goods from a customer through an AI chatbot, an input form on a webpage, or an application interface. The design file may refer to an image file (e.g., ai, psd, jpg, png) containing the IP that the customer wants to produce. The work order may refer to a document created by the consultation unit (330) in a standardized data format based on the order information and the design file, so that the subsequent processes, the rendering unit (340) and the scheduling unit (350), can refer to it. This may have the effect of preventing omissions or errors in information that may occur during the manual consultation process and ensuring that all processes operate based on the same data.

[0083] According to one embodiment, the rendering unit (340) can automatically combine the specifications (e.g., type of goods, size, material) specified in the work order with the design file. The rendering unit (340) can generate a rendering image in which the customer's design file is applied to the three-dimensional physical shape of the goods using 3D modeling technology. In particular, the packaging state rendering may include operations that go beyond simply showing an image of the finished goods to visualizing the optimal packaging box size, the placement of cushioning material, or the attachment location of the shipping label that matches the quantity and volume of the goods.

[0084] According to one embodiment, the scheduling unit (350) can function as a core optimization engine of the system (100). The artificial intelligence (AI) model of the scheduling unit (350) can analyze the quantity of goods included in the work order, the delivery date requested by the customer, and price conditions, etc., to identify the constraints and goals (e.g., cost minimization, delivery priority) of the order.

[0085] According to one embodiment, the system (100) can find a combination of companies that best satisfy the order conditions from a pool of multiple pre-registered manufacturers, packaging companies, and transport companies. An artificial intelligence (AI) model may consider complex factors including each company's estimated lead time, defect rate based on historical data, and real-time facility availability (e.g., current workload). For example, in the case of an order with a very tight delivery time, the AI ​​model may be trained to prioritize the selection of a highly reliable company with a short lead time and a low historical defect rate, even if the cost is somewhat higher. Such company-related data (e.g., estimated lead time, defect rate) is merely an example and is not limited thereto, and various additional factors including the company's location (=logistics distance) or area of ​​expertise may be considered.

[0086] According to one embodiment, the scheduling unit (350) can establish an integrated schedule for the entire process (e.g., manufacturing, packaging, shipping) so that selected companies can perform tasks sequentially. The scheduling unit (350) can assign a task start schedule for the packaging company to match the expected completion time of the manufacturing company's task. Additionally, the schedule can be optimized to minimize bottlenecks between processes, such as dispatching a vehicle for the transportation company to match the expected completion time of the packaging. Once optimization is complete, the system (100) can automatically transmit work instruction details (e.g., manufacturing drawing, quantity, shipping address) corresponding to each company.

[0088] FIG. 5 shows a flowchart of a method for generating a work instruction sheet for a consultation department according to an exemplary embodiment of the present disclosure.

[0089] Although the process steps, method steps, algorithms, etc., in the flowchart of FIG. 5 are described in a sequential order, such processes, methods, and algorithms may be configured to operate in any suitable order. The steps of the processes, methods, and algorithms described in various embodiments of the present invention do not need to be performed in the order described in the present invention.

[0090] In operation 510, the consultation unit (330) can receive basic order information (e.g., IP, type of goods, quantity, etc.). The customer can input basic information required for production, such as an IP image file, desired type of goods (e.g., pouch, keyring), production quantity, desired deadline, and packaging design, through an AI chatbot interface.

[0091] In operation 520, when the consultation department (330) receives an inquiry regarding production style or feel, the AI ​​can analyze the content of the inquiry. For example, a customer may inquire about abstract styles or feelings (e.g., a doll with a cozy feel or a sticker with a hip style) rather than specific specifications. In this case, an artificial intelligence (AI) model (e.g., a large-scale language model) can analyze the customer's intent of the inquiry through natural language processing technology.

[0092] In operation 530, the consultation unit (330) may suggest process options (e.g., fabric, embroidery, etc.) by referring to a production dictionary database. An artificial intelligence (AI) model may search the production dictionary database for technical options corresponding to the analyzed inquiry. The production dictionary database may include at least one of the following information: types of sewing fabric, embroidery techniques, gloss treatment, characteristics by size, or unit prices by process.

[0093] In operation 540, the consultation unit (330) can generate a work order based on order information and proposed options. When the customer selects and confirms the proposed options, the consultation unit (330) can integrate the basic order information received in operation 510 and the confirmed process options. Based on the final production specifications thus integrated, a standardized work order to be used in subsequent rendering and scheduling processes can be generated.

[0094] According to one embodiment, the consultation unit (330) receives basic order information including at least one of an IP image file, a type of goods, a production quantity, a desired deadline, or a packaging design, and when an inquiry regarding a production style or feel is received from a customer, it analyzes the customer's inquiry using an artificial intelligence (AI) model, suggests an option corresponding to the inquiry by referring to a production dictionary database including at least one of a sewing fabric, embroidery, gloss, size, or process option, and can generate a work order based on the order information and the suggested option.

[0096] According to one embodiment, the consultation unit (330) receives an inquiry from a customer regarding whether sample production is possible or whether production is possible for less than a preset minimum order quantity, refers to an FAQ database through an artificial intelligence (AI) model, and in response to the inquiry, may automatically provide a predefined answer suggesting that the customer order and sample the product in the minimum order unit, or, if the product is larger than a standard quantity requiring a separate quote, that the customer needs to check separately regarding the possibility of sample production and the cost.

[0097] According to one embodiment, the consultation department (330) can handle policy inquiries such as sample production or production of less than the minimum order quantity. Such inquiries may frequently occur when a customer wants to check the quality of a product before placing a formal order or wants to order a small quantity.

[0098] According to one embodiment, the artificial intelligence (AI) model may be a question-response model that identifies the intent of a customer's natural language inquiry (e.g., "Is one sample available?") or a rule-based engine that detects specific keywords (e.g., sample, 1) and triggers defined logic. The FAQ database may refer to a knowledge base in which policies and answers for exceptional inquiries are predefined and stored. The content of the customer's inquiry is merely an example and is not limited thereto, and may vary depending on the settings.

[0099] According to one embodiment, the proposal to sample by ordering in the minimum order unit of the product may be based on the premise that the system (100) checks the minimum order quantity (MOQ) information of the product. If the minimum order quantity of the product is set to 1, the artificial intelligence (AI) model may interpret the customer's sample request as an order of 1 and guide the customer to proceed with an order of 1 on the website without a separate inquiry. This can have the effect of inducing the customer to obtain a sample through a standard order without a separate process for sample production.

[0100] According to one embodiment, cases where the quantity exceeds a standard quantity requiring a separate quote may occur when the minimum order quantity of the product is set to a relatively large value, such as 100 units. Sample production of such products may require a formal quote as it incurs separate mold costs or preparation costs. Therefore, instead of providing a definitive schedule or cost via an automated response, the AI ​​model may indicate that separate verification (e.g., confirmation by a person in charge, supplier quote) is required regarding the feasibility and cost of sample production. This prevents the provision of inaccurate information to customers and can effectively link complex inquiries to manual consultation. This standard quantity (e.g., 100 units) is merely an example and is not limited thereto; it may vary depending on the type of product or the complexity of the manufacturing process.

[0102] According to one embodiment, the rendering unit (340) automatically maps the Intellectual Property (IP) design to the 3D physical shape of the goods based on design and specification information included in the work order and renders it, and calculates at least one of the packaging specifications, packaging volume, cushioning material placement, or label location required to reduce the damage rate of the goods and optimize transportation costs through simulation using an artificial intelligence (AI) model, generates a rendering image that visualizes the packaging state of the goods from the calculated result, and transmits the calculated packaging volume and specification information to the scheduling unit (350) so that the scheduling unit (350) can use it for selecting a transportation company and calculating costs.

[0103] According to one embodiment, the rendering unit (340) can render by automatically mapping an IP (Intellectual Property) design onto a 3D physical shape of a good based on design and specification information included in a work order. The IP design may refer to 2D or 3D design data including a character image, logo, or text provided by a customer. The 3D physical shape of the good may refer to standard 3D model data of the good to be produced (e.g., pouch, mug). The rendering unit (340) can automatically apply and place the IP design on the surface of the 3D physical shape model using texture mapping or similar graphic technology.

[0104] According to one embodiment, an artificial intelligence (AI) model can be trained with the goal of reducing the damage rate of goods and optimizing transportation costs. The AI ​​model can receive input of the specifications and quantity of goods from a work order and simulate the process of packaging the goods. The AI ​​model can calculate the total volume (or volume) of the goods, estimate the amount of cushioning material required by considering the risk of damage based on the material of the goods (e.g., glass), and determine the minimum standard box specifications capable of accommodating all of them. Additionally, the AI ​​model can calculate the location where a shipping label can be attached on the packaged shape.

[0105] According to one embodiment, the rendering unit (340) can generate a packaging state rendering image showing how the goods are arranged inside the box based on elements such as the calculated packaging specifications and cushioning material arrangement. This serves to show the final delivery state to the customer in advance and can be used as a visual work guide for the packaging company.

[0106] According to one embodiment, the rendering unit (340) can transmit the calculated packaging volume and specification information to the scheduling unit (350). The scheduling unit (350) can use the calculated packaging volume and specification information to accurately calculate the cost required for actual transportation in advance. For example, the scheduling unit (350) can select the most cost-effective means of transportation among courier, quick service, or cargo transportation based on the packaging volume and weight, and instruct the transportation company to dispatch the vehicle. This can have the effect of preventing additional transportation costs that may occur due to inaccurate packaging information.

[0108] According to one embodiment, the rendering unit (340) may first provide a first rendering image by combining an IP image file input through the consultation unit (330) and information about the form of the goods before the work order is generated, and after the work order is finally confirmed, may generate and provide a second rendering image including the actual design and packaging state of the goods based on the detailed specifications specified in the work order.

[0109] According to one embodiment, the rendering unit (340) may first provide a first rendering image. This is intended to quickly provide a rough draft to the customer before a work order is generated, that is, before detailed production specifications are finalized. The first rendering image may be generated by combining an IP image file input through the consultation unit (330) with information on the form of goods (e.g., pouch, mug). The first rendering image may be provided by visualizing a combination of a rough shape and design that does not reflect detailed specifications (e.g., exact color code, material, location of auxiliary materials).

[0110] According to one embodiment, the rendering unit (340) may generate and provide a secondary rendering image after the work order is finally confirmed. The work order may specify detailed specifications of the goods (e.g., w 130 x h 97.5 x d 30 (mm), velboa fabric, color 7499c). The rendering unit (340) may precisely render the actual design of the goods by reflecting all of these detailed specifications. In addition, the secondary rendering image may be generated to include not only the actual design of the goods but also the final packaging state based on the calculated packaging specifications and volume. The detailed specifications of the goods are merely examples and are not limited thereto, and may vary depending on the settings.

[0111] According to one embodiment, the rendering unit (340) can enable the customer to receive rapid visual feedback (=1st rendering) during the initial consultation stage. In addition, after all specifications are finalized, by accurately verifying the final product and packaging state (=2nd rendering), it is possible to prevent manufacturing errors and increase customer satisfaction.

[0112] According to one embodiment, the rendering unit (340) can generate a first rendering image and then provide a user interface (UI) through the consultation unit (330) that allows the customer to directly manipulate the image. The customer can adjust the size, position, or angle of the IP image file on the first rendering image in real time. This customer manipulation information can be combined with goods form information by the consultation unit (330) and automatically converted into detailed specification data of a work order (e.g., IP print coordinates, size). This allows the customer to intuitively determine the desired design layout without the help of an expert, thereby shortening the time required to generate the work order and reducing design errors.

[0113] According to one embodiment, at the time when a first rendering image is generated, the rendering unit (340) can provide real-time feedback on production risk by linking with the artificial intelligence (AI) model of the consultation unit (330). For example, if a customer reduces the size of an IP design on the first rendering image to below a specific threshold (e.g., 2mm) or places it too close to the edge of the goods (e.g., seam), the artificial intelligence (AI) model can detect this as a factor that causes a sharp increase in production difficulty or a high defect rate. The rendering unit (340) can detect the factor with a high defect rate and display a corresponding alternative (e.g., "This placement has a high printing defect rate. Would you like to move it inward by 1mm?") on the first rendering interface. This can have the effect of correcting potential defect factors in advance, that is, from the consultation stage with the customer, before the work order is generated.

[0114] According to one embodiment, a secondary rendering image may be generated after the optimization of the scheduling unit (350) is completed. When the scheduling unit (350) determines the optimal packaging box specifications, the required volume of cushioning material, and the label placement, the rendering unit (340) may receive the determined data. Based on the determined data, the rendering unit (340) may generate a three-dimensional packaging guide as part of the secondary rendering image, which shows not only the actual design of the goods but also how the goods and cushioning material are arranged inside the selected packaging box.

[0116] According to one embodiment, if a delay or defect occurs in at least one of the production, packaging, or shipping processes of goods, the scheduling unit (350) can recalculate the entire schedule in real time, including at least one of the production, packaging, or shipping schedules, by reflecting the delay or defect information in an artificial intelligence (AI) model, and transmit at least one of the work instruction, packaging instruction, or shipping instruction reflecting the adjusted schedule to at least one of the production company, packaging company, or shipping company affected by the schedule change.

[0117] According to one embodiment, the scheduling unit (350) may detect in real time when a delay or defect occurs in at least one of the production, packaging, or shipping processes of the goods, or receive such information from a cooperating company. For example, this may include a production delay caused by a breakdown of equipment at the manufacturer, or the occurrence of defective products during the quality inspection (QC) process.

[0118] According to one embodiment, an artificial intelligence (AI) model can receive a variable to reflect delay or defect information and update the estimated completion time of the current process. Based on the updated estimated completion time, the AI ​​model can recalculate the previously optimized overall schedule in real time.

[0119] According to one embodiment, an artificial intelligence (AI) model can analyze the cascading effect of a delay in a specific process on all subsequent processes. The artificial intelligence (AI) model can re-optimize an overall schedule, including at least one of a production, packaging, or shipping schedule, under changed constraints.

[0120] According to one embodiment, the scheduling unit (350) can identify the companies affected by the recalculated schedule. For example, the work start date of a packaging company may be changed due to production delays, or the pick-up time of a transporter may be changed. The scheduling unit (350) can automatically transmit new instructions reflecting the adjusted schedule (e.g., changed work instructions, adjusted packaging instructions, dispatch instructions for changed dispatch times) to the terminal of at least one of the identified manufacturing company, packaging company, or transporter. This can have the effect of minimizing bottlenecks and preventing overall delivery delays by automating the process of manually readjusting subsequent schedules and making contact when a problem occurs.

[0122] According to one embodiment, the scheduling unit (350) analyzes data on the three-dimensional (3D) size, material, and production quantity of individual goods specified in the work order using an artificial intelligence (AI) model, calculates in advance packaging specifications including optimal packaging box dimensions, required volume of cushioning material, and label placement to minimize the damage rate during transportation of goods and optimize transportation costs based on the analysis results, compares with a database of company-specific features including at least one feature among the ability to process large orders, specialization in small orders, or expertise regarding items with a high risk of damage, selects the optimal packaging company and transportation company that meet the calculated packaging specifications, and controls the dispatch by transmitting a work order including the packaging specifications to the terminal of the selected company.

[0123] According to one embodiment, the scheduling unit (350) can analyze data specified in the work order using an artificial intelligence (AI) model. The data may include the three-dimensional (3D) size (e.g., width, height, length), material (e.g., ceramic, PVC, fabric) and total production quantity of individual goods.

[0124] According to one embodiment, an artificial intelligence (AI) model can calculate the total volume (or volume) and weight of the goods based on data specified in a work order. Additionally, by considering the risk of damage according to the material (e.g., ceramics), it can estimate the required volume (or volume) of cushioning material to minimize the damage rate during transportation. The scheduling unit (350) can calculate the smallest standard packaging box size capable of accommodating the goods by summing the total volume of the goods and the required volume of cushioning material. This process is intended to optimize transportation costs and can have the effect of reducing volume costs incurred by using unnecessarily large boxes.

[0125] According to one embodiment, the scheduling unit (350) may compare the calculated packaging specifications with a database of company-specific features. The database of company-specific features may include information on the specialized capabilities of each packaging company and transport company. For example, the ability to process large orders may be used to identify companies equipped with automated facilities capable of quickly packaging thousands of goods. The ability to specialize in small orders may be used to identify companies capable of processing small orders at a reasonable cost. Expertise in items with a high risk of breakage may refer to a feature that identifies specialized companies possessing special handling and cushioning packaging techniques, such as for goods made of ceramic or glass materials. These features are merely examples and are not limited thereto, and additional features such as the location or cost structure of the company may be added.

[0126] According to one embodiment, the scheduling unit (350) can select the optimal packaging company and transportation company that meet the calculated packaging specifications (e.g., have expertise in handling goods with a high risk of damage) through the comparison. Once the selection is complete, the scheduling unit (350) can complete the dispatch by automatically sending a work order to the company that includes the calculated packaging specifications (e.g., box size to be used, type of cushioning material required).

[0128] FIG. 6 shows a flowchart of an AI model weight adjustment method of a scheduling unit according to an exemplary embodiment of the present disclosure.

[0129] Although process steps, method steps, algorithms, etc. are described in a sequential order in the flowchart of FIG. 6, such processes, methods, and algorithms may be configured to operate in any suitable order. The steps of the processes, methods, and algorithms described in various embodiments of the present invention do not need to be performed in the order described in the present invention.

[0130] In operation 610, the scheduling unit (350) can receive performance data (e.g., defect rate, time taken, etc.) after the completion of the process. After the production, packaging, or transportation of the goods is completed, the scheduling unit (350) can receive performance data including the actual defect rate, the actual time taken for the work, and the actual cost incurred from the company that performed the process.

[0131] In operation 620, the scheduling unit (350) can calculate a prediction error by comparing the predicted value with the actual performance data using an AI model. The artificial intelligence (AI) model can compare the estimated time required, estimated defect rate, and estimated cost values ​​predicted during the vendor selection stage of operation 440 with the actual performance data received in operation 610. Through the comparison of data, the artificial intelligence (AI) model can calculate a prediction error indicating how accurate or how far off the prediction was.

[0132] In operation 630, the scheduling unit (350) can dynamically adjust the AI ​​model and company weights based on the prediction error. Based on the calculated prediction error (operation 620), the prediction model of the artificial intelligence (AI) model itself can be corrected. Alternatively, the weight parameters of the corresponding company (e.g., reliability, speed, quality score) stored in the company-specific feature database can be adjusted. For example, the scheduling unit (350) can increase the speed weight of a company whose delivery time is consistently faster than the prediction.

[0133] In operation 640, the scheduling unit (350) can improve prediction accuracy by applying adjusted weights for the next order. When the system (100) selects a supplier for a new order in the future, it can apply the dynamically adjusted weight parameters in operation 630. By applying the adjusted weights, the estimated time required and defect rate of each supplier can be predicted more accurately.

[0134] According to one embodiment, the scheduling unit (350) receives performance data including the actual defect rate, actual time taken, and actual cost incurred after the production, packaging, or transportation of goods is completed from a terminal of a selected manufacturing company, packaging company, or transportation company, and calculates a prediction error by using an artificial intelligence (AI) model to compare the received performance data with the predicted time taken, predicted defect rate, and predicted cost predicted in the selection step of claim 1, corrects the prediction model of the artificial intelligence (AI) model based on the calculated prediction error, or adjusts a weighting parameter including at least one of the reliability, speed, or quality score of the corresponding company stored in the company-specific feature database, and when selecting a company for the next order, applies the adjusted weighting parameter to predict the expected time taken and defect rate of each company.

[0136] According to one embodiment, when the scheduling unit (350) selects a company for the next order, it can predict the expected time required and defect rate of each company by applying the adjusted weighting parameter. At this time, the artificial intelligence (AI) model may not apply the adjusted weighting parameter uniformly to all next orders, but may apply it differentially according to the characteristics of the next order.

[0137] According to one embodiment, an artificial intelligence (AI) model can first identify customer priorities (e.g., delivery date first, cost first, quality first) as characteristics of the next order. For example, if the next order is determined to have 'delivery date' as the highest priority, the AI ​​model can select a vendor by assigning the highest evaluation weight to a speed score adjusted based on actual time taken and a reliability score reflecting adherence to the predicted delivery date, among the weight parameters stored in the vendor feature database. Conversely, if the next order is determined to have 'quality' as the highest priority, the AI ​​model can control the selection of a vendor by assigning the highest evaluation weight to a quality score adjusted based on the actual defect rate.

[0138] According to one embodiment, an artificial intelligence (AI) model can store and apply weight parameters in a more granular manner by linking them to the detailed conditions of an order. For example, a manufacturer A may have a history of consistently achieving low prediction errors and high reliability scores for small orders (e.g., fewer than 100 units) and simple designs (e.g., stickers), but for large orders (e.g., more than 1000 units) or complex designs (e.g., embroidered pouches), the actual time required may be longer than predicted or the actual defect rate may be high.

[0139] According to one embodiment, when adjusting weight parameters, the reliability score of Manufacturer A may be stored and updated as a conditional weight, such as a small / simple order or a large / complex order, rather than as a single value. Accordingly, the artificial intelligence (AI) model may select Manufacturer A as the top candidate if the characteristics of the next order are analyzed as a small / simple order. On the other hand, if the next order is analyzed as a large / complex order, the selection ranking of Manufacturer A may be lowered or excluded. The application of these conditional weight parameters can have the effect of improving prediction accuracy and scheduling success rate compared to the method using simple average scores by accurately reflecting the actual strengths and weaknesses of each company in the context of the order. The quantity criteria of these orders (e.g., 100, 1000) or the classification of designs (e.g., simple, complex) are merely examples and are not limited thereto, and may vary depending on the training data of the system (100).

[0140] According to one embodiment, when an artificial intelligence (AI) model adjusts weight parameters of a company-specific feature database, it can classify order characteristics into multiple categories and store and update conditional weight parameters separated for each category.

[0141] According to one embodiment, the category classifying the characteristics of the order may be defined by combining at least one of the quantity of the order (e.g., large quantity, heavy, small quantity), the type of goods (e.g., printed material, plush toy, acrylic keyring), the complexity of the design (e.g., simple 2D printing, complex multi-color embroidery), or the characteristics of the material (e.g., glass with a high risk of breakage, standard PVC). For example, the weight parameter of 'A manufacturer' may store the reliability score for the 'sticker / large quantity / simple' category and the reliability score for the 'plush toy / small quantity / complex' category as different values. These order classification criteria are merely examples and are not limited thereto, and may be subdivided or changed according to the learning data or operational strategy of the system (100).

[0142] According to one embodiment, when selecting a vendor for a next order, an artificial intelligence (AI) model may first analyze and identify the characteristics of the next order as one of the defined categories. The AI ​​model may extract only the conditional weight parameters corresponding to the identified categories from each vendor's database and use them as criteria for vendor selection.

[0143] According to one embodiment, an artificial intelligence model can accurately reflect the actual strengths and weaknesses of each company in accordance with the context of the order by applying conditional weight parameters. For example, Manufacturer A may have very high reliability in producing 'stickers' but a high defect rate in producing 'plush toys'. Therefore, if the ordered item is 'stickers', the AI ​​model can select Manufacturer A as the top candidate. On the other hand, if the next order is for 'plush toys', the AI ​​model can control the selection ranking to lower or exclude the company because the conditional weight for the 'plush toy' category is low, even if Manufacturer A has a high overall average score.

[0145] According to one embodiment, when the consultation unit (330) receives order information and design files from a customer, it analyzes at least one element of the design complexity, number of colors, or difficulty of implementation of the design file through an artificial intelligence (AI) model, and infers a production risk including an expected defect rate or expected delivery delay probability for the order information using an artificial intelligence (AI) model based on past defect rates, time history, and past order history data that had specifications similar to the design requirement specifications stored in a company-specific feature database, and if the inferred production risk exceeds a preset threshold, it may notify the customer of the risk or display alternatives including design modification and material change before generating a work order.

[0146] According to one embodiment, when the consultation unit (330) receives order information and a design file from a customer, it can verify in advance the suitability of the design file for production through an artificial intelligence (AI) model.

[0147] According to one embodiment, design complexity may mean a value measured by quantifying at least one of, for example, the number of vector nodes included in the design, the number of layers, or, in the case of an embroidery design, the total stitch count.

[0148] According to one embodiment, an artificial intelligence (AI) model can analyze the analyzed design requirement specifications in conjunction with the history of past defect rates and required times stored in a vendor-specific feature database. Additionally, the AI ​​model can refer to historical order data that had design requirement specifications similar to the current order.

[0149] According to one embodiment, an artificial intelligence (AI) model can infer the production risk of current order information based on such historical data. The production risk can be inferred as a quantitative figure, such as an expected defect rate (%) or an expected delivery delay probability (%). For example, if there is a history that past orders with similar complexity and number of colors showed high defect rates or delivery delays at a specific supplier, the AI ​​model can infer that the production risk of the current order is relatively higher than that of previous orders.

[0150] According to one embodiment, the system (100) can determine whether the inferred manufacturing risk exceeds a preset threshold. If it exceeds the threshold, the system (100) may notify the customer of the risk before generating a work order. Alternatively, the system (100) may display to the customer alternatives that can reduce the risk (e.g., suggestions to modify parts of the design, suggestions to change to a different material that is easier to manufacture).

[0152] According to one embodiment, the artificial intelligence (AI) model determines in advance the suitability for production by analyzing technical specifications including at least one of the resolution, color profile, or design complexity of the design file based on the input of the design file through the consultation unit (330), determines the customer's priority regarding the delivery date, cost, and quality based on the delivery date, price, or quantity included in the order conditions through the scheduling unit (350), selects the optimal company by applying differential weights to the estimated time required, defect rate, and cost of the manufacturing company, packaging company, and transportation company according to the determined priority, detects the possibility of expected delay by receiving real-time process progress status data from the terminal of the selected company and comparing it with the optimized schedule, and if the possibility of delay is detected, controls can readjust the schedule of the subsequent process or the company to perform the subsequent process in order to comply with the final delivery date of the goods.

[0153] According to one embodiment, an artificial intelligence (AI) model can perform control functions across the consultation unit (330) and the scheduling unit (350). The artificial intelligence (AI) model can automatically analyze the technical specifications of a design file when the design file is input through the consultation unit (330). The technical specifications may include design complexity, which refers to whether the minimum resolution required for printing or embroidery (e.g., 300 dpi) is met, whether the color profile required by the production process (e.g., CMYK, RGB) is complied with, or whether an excessively complex design (e.g., thin lines that cannot be implemented) is included. Through this analysis, the artificial intelligence (AI) model can determine in advance whether the design file is suitable for proceeding with the subsequent production process. This can have the effect of preventing process interruptions or defects caused by unsuitable files in advance.

[0154] According to one embodiment, an artificial intelligence (AI) model can determine customer priorities by analyzing order conditions through a scheduling unit (350). For example, orders with very short delivery dates may be determined with the highest priority for delivery date, and orders with very large production quantities where unit price negotiation is important may be determined with the highest priority for cost. Customer priorities (e.g., delivery date, cost, quality) are merely examples and are not limited thereto, and may vary depending on the settings.

[0155] According to one embodiment, an artificial intelligence (AI) model may apply differential weights to evaluation factors (e.g., estimated lead time, defect rate, cost) when selecting a vendor based on the determined priority. For example, if the delivery date is determined to be the top priority, the vendor may be evaluated by assigning the highest weight to the estimated lead time factor and lowering the weight to the cost factor.

[0156] According to one embodiment, an artificial intelligence (AI) model can select an optimal company based on evaluation scores to which differential weights are applied. Once the selection is complete, the scheduling unit (350) can receive real-time process progress status data from the terminal of the selected company. The real-time process progress status data may include at least one of a signal indicating the completion of a process step (e.g., fabric preparation, cutting, sewing) of the manufacturing company or location information (GPS) of the transportation company.

[0157] According to one embodiment, an artificial intelligence (AI) model can continuously compare received real-time process progress data with an existing optimized schedule. Through this comparison, if it is predicted that the current process is delayed longer than expected and will affect the final delivery date of the goods, this can be detected as a potential delay.

[0158] According to one embodiment, if a possibility of delay is detected, the artificial intelligence (AI) model can automatically readjust the schedule of subsequent processes to ensure compliance with the final delivery date of the goods. For example, if it is detected that production will be delayed by two hours, the start time of work for the packaging company can be delayed by two hours, and the dispatch time of the transportation company can be readjusted in conjunction. Alternatively, if the AI ​​model determines that the delay is severe and that the existing company cannot meet the delivery deadline, it can reassign the company to perform the subsequent processes to another currently available company.

[0160] FIG. 7 illustrates a schematic configuration diagram of a goods production and logistics linkage system using an artificial intelligence model according to an exemplary embodiment of the present disclosure.

[0161] In FIG. 7, the system (e.g., the system (100) of FIG. 3) receives order information from a customer (710), and the artificial intelligence model (720) analyzes this to connect and coordinate the optimal partner companies (730, 732, 734).

[0162] According to one embodiment, the customer (710) may refer to a subject who commissions the production of goods (e.g., dolls, keyrings, mugs). The customer (710) may input initial data, including a design file of the goods to be produced, the required quantity, a desired delivery date, and an affordable price (or budget), into an artificial intelligence model (720) through a user interface (e.g., a webpage, a mobile application) of the system (100).

[0163] According to one embodiment, the artificial intelligence model (720) can identify the customer's needs by analyzing the initial data received from the customer (710). For example, the artificial intelligence model (720) can analyze the complexity of the design file (e.g., number of colors, presence or absence of embroidery data) and evaluate the urgency and production difficulty of the order by combining quantity and delivery date information.

[0164] According to one embodiment, the artificial intelligence model (720) can select an optimal combination of suppliers from a pool of suppliers registered in the system and coordinate the schedule based on the analyzed customer needs and order information. The supplier pool may include a manufacturing company (730), a packaging company (732), and a transportation company (734). The components of the supplier pool are merely examples and are not limited thereto, and may vary depending on the settings.

[0165] According to one embodiment, the manufacturer (730) may refer to a factory or workshop that actually produces goods. The artificial intelligence model (720) selects the optimal manufacturer (730) by comprehensively considering the current equipment availability, production unit cost, and past defect rate data of the manufacturer (e.g., a specialized acrylic manufacturer) that is most suitable for the type of goods to be produced (e.g., acrylic keyring) and the quality required by the customer, and can transmit a work order (e.g., production drawing, quantity, delivery date) to the terminal of the manufacturer (730).

[0166] According to one embodiment, the packaging company (732) may refer to a company that inspects finished goods and performs individual packaging (e.g., OPP plastic bags, backing paper) or combined packaging for shipping (e.g., shipping boxes containing cushioning material). The artificial intelligence model (720) may select a packaging company (732) that possesses packaging expertise for the item (e.g., production of custom cushioning material) by considering the characteristics (e.g., glass cups with a high risk of breakage) and quantity of the goods.

[0167] According to one embodiment, the carrier (734) may refer to a logistics company (e.g., courier, quick service, freight transport) that delivers the packaged goods to a final destination (e.g., customer's address, third logistics warehouse). The artificial intelligence model (720) can deliver dispatch instructions by optimizing both the intermediate transport from the manufacturer (730) to the packaging company (732) and the final transport schedule from the packaging company (732) to the customer (710) or the final destination.

[0168] According to one embodiment, the artificial intelligence model (720) can manage the entire process so that these companies (730, 732, 734) operate in conjunction rather than individually. For example, the artificial intelligence model (720) can monitor the expected completion time of production of the manufacturing company (730) in real time, instruct the transport company (734) to dispatch a pickup vehicle at that time, and simultaneously transmit relevant information in advance so that the packaging company (732) can prepare personnel and materials to start the next task immediately.

[0169] The distinction between manufacturing, packaging, and transportation companies is merely one example of functional classification and is not limited thereto; it is also possible for a single company to perform two or more functions (e.g., manufacturing and packaging) simultaneously.

[0170] According to one embodiment, the scheduling unit (350) can establish a split production plan for bulk orders or urgent orders using an artificial intelligence (AI) model. For example, if a customer orders 10,000 goods with an urgent delivery date of within 3 days, the artificial intelligence (AI) model may determine that meeting the delivery deadline is impossible with only the facility availability of a single manufacturer (730). In this case, the artificial intelligence (AI) model may automatically split the order into multiple small orders (e.g., 5,000, 3,000, 2,000) and simultaneously assign multiple manufacturers (e.g., Factory A, Factory B, Factory C) capable of processing each split order, taking into account real-time facility availability and geographical location. The scheduling unit (350) may establish a schedule to collect the goods produced from each manufacturer into a single packaging company (732).

[0171] According to one embodiment, the scheduling unit (350) can automatically batch and process multiple individual orders using an artificial intelligence (AI) model. The artificial intelligence (AI) model can analyze multiple customer orders received by the system (100) in real time. If different customers request that the same or similar goods (e.g., stickers of the same specifications) be produced by the same manufacturer, the artificial intelligence (AI) model can automatically merge those orders into a single bulk order. Based on the merged order, the scheduling unit (350) can negotiate unit prices with the manufacturer or control the production process to increase production efficiency by batch processing. Additionally, logistics costs can be reduced by optimizing the shipping schedule to bundle multiple completed packages destined for the same destination.

[0172] According to one embodiment, the scheduling unit (350) can use an artificial intelligence (AI) model to analyze in real time the delivery date entered by the customer and the availability of manufacturing facilities internally owned by the system (100), and can present a price to the customer. For example, if the customer requests a specific delivery date (e.g., November 10), the artificial intelligence (AI) model can calculate a first cost by calculating the optimal schedule to comply with the delivery date. At the same time, the artificial intelligence (AI) model can calculate a second delivery date (e.g., November 15) that is more flexible based on the schedule of manufacturing companies that are currently idle or available for operation during non-peak hours. If the customer accepts the second delivery date, the scheduling unit (350) can distribute the overall production load of the system (100) and increase operational efficiency, so it can also propose a discounted second cost to the customer.

[0173] According to one embodiment, the system (100) may include a material management function that centrally procures and manages common auxiliary materials (e.g., fabric of a specific color, cotton, standard-sized carabiners) required for the production of goods. When the consultation unit (330) generates a work order, an artificial intelligence (AI) model can automatically calculate the required amount of fabric and auxiliary materials needed for the work order. The scheduling unit (350) can select an optimal manufacturer and, at the same time, automatically generate a logistics order to send the calculated required amount of materials from the central material warehouse of the system (100) to the manufacturer. This can fundamentally prevent quality inconsistencies and cost increases caused by individual manufacturers procuring different auxiliary materials. Additionally, the manufacturer can shorten the overall lead time by allowing it to focus solely on pure processing (e.g., sewing, printing) tasks.

[0174] According to one embodiment, the consultation unit (330) can propose active alternatives to the customer by linking the customer's order information with the real-time resource status of the scheduling unit (350). For example, if a customer orders 100 pouches of design A, the artificial intelligence (AI) model can determine through the scheduling unit (350) that the equipment availability of the pouch manufacturer is low and the risk of delivery delay is high. In this case, the artificial intelligence (AI) model can automatically search for items among other types of goods suitable for design A (e.g., keyrings, mugs) that have high current equipment availability and favorable production unit costs. The consultation unit (330) can present an alternative to the customer that can simultaneously increase the operational efficiency of the system (100) and customer satisfaction (e.g., "The pouches you requested will take N days due to high current orders. If you produce keyrings with the same design, they can be produced in M ​​days at a lower cost.").

[0175] According to one embodiment, the rendering unit (340) can generate reference images for marketing using an artificial intelligence (AI) generative model, going beyond simple 3D modeling. When the rendering unit (340) completes the 3D physical design of the goods, the artificial intelligence (AI) model can generate multiple lifestyle rendering images by naturally synthesizing the 3D model into an actual usage environment (e.g., the goods placed on a desk, attached to a bag, worn by a model).

[0177] Up to this point, only the system (100) has been described, but the present disclosure is not limited thereto. For example, the method of operating such a system is also considered to be within the scope of the present disclosure.

[0178] Although the present disclosure has been described with reference to the embodiments illustrated in the drawings, this is merely illustrative, and those skilled in the art will understand that various modifications and equivalent alternative embodiments are possible therefrom. Accordingly, the true technical scope of protection of the present disclosure should be determined by the technical spirit of the appended claims.

Claims

Claim 1 A system for optimizing goods production and logistics using artificial intelligence (AI), comprising: a memory for storing instructions; a processor; a consultation unit; a rendering unit; and a scheduling unit; wherein, when the instructions are executed by the processor, the system receives order information and design files required for goods production from a customer through the consultation unit to generate a work order; renders the actual design and packaging state of the goods based on the work order or the order information through the rendering unit; analyzes order conditions including the quantity, delivery date, and price of the goods using an artificial intelligence (AI) model based on the work order through the scheduling unit; selects an optimal company by considering at least one of the estimated lead time, defect rate, or equipment availability of the manufacturing company, packaging company, and transportation company; and controls the determination of at least one of the production, packaging, or shipping schedule. Claim 2 A system according to claim 1, wherein the consultation unit receives basic order information including at least one of an IP (Intellectual Property) image file, type of goods, production quantity, desired deadline, or packaging design, and when an inquiry regarding production style or feel is received from a customer, analyzes the customer's inquiry using an artificial intelligence (AI) model, suggests an option corresponding to the inquiry by referring to a production dictionary database including at least one of sewing fabric, embroidery, gloss, size, or process options, and generates the work order based on the order information and the suggested option. Claim 3 In claim 1, the consultation department receives an inquiry from a customer regarding whether sample production is possible or whether production is possible for a quantity less than a preset minimum order quantity, refers to an FAQ database through an artificial intelligence (AI) model, and in response to the inquiry suggests ordering and sampling the product in the minimum order unit, or if the product is larger than a standard quantity requiring a separate quote, provides guidance that separate verification of the sample production possibility and cost is required. Claim 4 A system according to claim 1, wherein the rendering unit automatically maps an Intellectual Property (IP) design onto a 3D physical shape of the goods based on design and specification information included in the work order and renders it, calculates at least one element among packaging specifications, packaging volume, cushioning material placement, or label location required to reduce the damage rate of the goods and optimize transportation costs through simulation using an artificial intelligence (AI) model, generates a rendering image that visualizes the packaging state of the goods from the calculated result, and transmits the calculated packaging volume and specification information to the scheduling unit so that the scheduling unit uses it for selecting a transportation company and calculating costs. Claim 5 In claim 1, the rendering unit first provides a first rendering image by combining an IP image file input through the consultation unit and information regarding the form of the goods before the work order is generated, and after the work order is finally confirmed, a system that generates and provides a second rendering image including the actual design and packaging state of the goods based on the detailed specifications specified in the work order. Claim 6 In claim 1, the scheduling unit is a system in which, when a delay or defect occurs in at least one of the production, packaging, or shipping processes of the goods, the artificial intelligence (AI) model recalculates the entire schedule including at least one of the production, packaging, or shipping schedules in real time by reflecting the delay or defect information, and transmits at least one of a work instruction, a packaging instruction, or a shipping instruction reflecting the adjusted schedule to a terminal of at least one of the production company, packaging company, or transportation company affected by the schedule change. Claim 7 In claim 1, the scheduling unit analyzes data on the three-dimensional (3D) size, material, and production quantity of individual goods specified in the work order using the artificial intelligence (AI) model, calculates in advance packaging specifications including optimal packaging box dimensions, required volume of cushioning material, and label placement to minimize the damage rate during transportation of the goods and optimize transportation costs based on the analysis results, compares with a database of company-specific features including at least one of the ability to process large orders, specialization in small orders, or expertise regarding items at risk of damage, selects the optimal packaging company and transportation company that meet the calculated packaging specifications, and controls the dispatch by transmitting a work order including the packaging specifications to the terminal of the selected company. Claim 8 delete Claim 9 In claim 1, the consulting unit, when receiving order information and a design file from a customer, analyzes at least one of the design complexity, number of colors, or difficulty of implementation of the design file through the artificial intelligence (AI) model, and infers a production risk including an expected defect rate or expected delivery delay probability for the order information using the artificial intelligence (AI) model based on past defect rates, time history, and past order history data that had specifications similar to the design requirement specifications stored in a company-specific feature database, and if the inferred production risk exceeds a preset threshold, notifies the customer of the risk or displays alternatives including design modification and material change before generating a work order. Claim 10 In claim 1, the artificial intelligence (AI) model determines in advance the suitability for production by analyzing technical specifications including at least one of the resolution, color profile, or design complexity of the design file based on the input of the design file through the consultation unit, and determines the customer's priority regarding delivery date, cost, and quality based on the delivery date, price, or quantity included in the order conditions through the scheduling unit, selects the optimal company by applying differential weights to the estimated time required, defect rate, and cost of the manufacturing company, packaging company, and transportation company according to the determined priority, detects the expected possibility of delay by receiving real-time process progress status data from the terminal of the selected company and comparing it with the optimized schedule, and if the possibility of delay is detected, controls the schedule of the subsequent process or the company to perform the subsequent process to be readjusted in order to comply with the final delivery date of the goods.

Citation Information

Patent Citations

  • Method for tracking management and processing of logistics based on artificial intelligence

    KR102578727B1