Neat set analysis method and device, electronic equipment, readable storage medium and chip

Abstract

The embodiment of the invention provides a suit analysis method and device, electronic equipment, a readable storage medium and a chip, and the method comprises the steps: receiving a natural language request for a target material; analyzing the natural language request to determine an analysis instruction related to the target material; querying and obtaining related data of the target material based on the analysis instruction; obtaining a first intelligent agent; determining an inventory analysis result corresponding to the target material according to the demand parameter and the inventory data; when it is judged that the subitem materials are in shortage according to the inventory analysis result, a second agent is obtained; acquiring at least one item of extended data related to the target material; the second intelligent agent determines at least one material shortage reason corresponding to the historical material shortage record according to the extended data; and according to the inventory analysis result and the material shortage reason, determining a neat set analysis result corresponding to the target material. Through the scheme of the invention, the time consumption of manual root cause checking by a planner is remarkably reduced, and the analysis efficiency of material set analysis is enhanced.

Description

Technical Field

[0001] This invention relates to the field of data processing technology, and more specifically, to a set of analysis methods, apparatus, electronic devices, readable storage media, and chips. Background Technology

[0002] Traditional material kitting analysis mainly relies on the fixed logic and reporting functions of Enterprise Resource Planning (ERP) systems. The typical process involves the system calculating material requirements based on the production plan and Bill of Materials (BOM), then matching this with supply data such as inventory and in-transit records to ultimately generate a static list of material shortages.

[0003] However, with the increasing complexity of supply chains and the growing demands for production agility, the inherent limitations of the model are becoming increasingly apparent. The depth of analysis is severely insufficient; traditional results can only answer what is missing and how much is missing. Root cause analysis of why materials are missing (such as supplier delivery delays, quality control delays, and logistical anomalies) relies heavily on planners manually checking across systems, which is inefficient and limited by personal experience. Summary of the Invention

[0004] The purpose of this invention is to provide a kitting analysis method, apparatus, electronic device, readable storage medium, and chip that can solve the problem of insufficient depth in kitting analysis.

[0005] In view of this, an embodiment of the first aspect of the present invention provides a homogeneous analysis method.

[0006] A second aspect of the present invention provides a kitting analysis apparatus.

[0007] An embodiment of the third aspect of the present invention provides an electronic device.

[0008] An embodiment of the fourth aspect of the present invention provides a readable storage medium.

[0009] An embodiment of the fifth aspect of the present invention provides a chip.

[0010] To achieve the above objectives, an embodiment of the first aspect of the present invention provides a kitting analysis method, comprising: receiving a natural language request for a target material, the natural language request being used to query the kitting status of the target material, the target material including multiple sub-items; parsing the natural language request to determine analysis instructions related to the target material; querying and obtaining relevant data of the target material based on the analysis instructions, the relevant data including at least the bill of materials, inventory data, procurement data, and production plan data of the multiple sub-items; acquiring a first intelligent agent, the first intelligent agent determining demand parameters based on the procurement data and production plan data; determining an inventory analysis result corresponding to the target material based on the demand parameters and inventory data; acquiring a second intelligent agent when it is determined that a sub-item is in short supply based on the inventory analysis result; acquiring at least one piece of extended data related to the target material; acquiring at least one type of historical shortage record; the second intelligent agent determining at least one shortage reason corresponding to the historical shortage record based on the extended data; and determining a kitting analysis result corresponding to the target material based on the inventory analysis result and the shortage reason.

[0011] The kitting analysis method provided by this invention constructs an intelligent kitting analysis process with natural language interaction as the entry point and intelligent agent collaborative analysis as the engine.

[0012] The kitting analysis method first receives the user's natural language query about the target material and then uses a large language model to parse out the structured analysis instructions.

[0013] Subsequently, based on the analysis instructions, the system automatically queries the relevant data such as the bill of materials, inventory, procurement and production plan of the target material, and calls the first intelligent agent (inventory analysis intelligent agent) to calculate the material demand parameters. Then, by comparing with the inventory data, it generates inventory analysis results containing specific shortage details.

[0014] When a material shortage is detected, a second intelligent agent (material shortage analysis agent) is activated. By correlating extended data with historical material shortage records, the agent can intelligently diagnose the cause of the material shortage.

[0015] Ultimately, by combining the results of inventory analysis with in-depth analysis of the reasons for material shortages, a comprehensive and in-depth analysis is generated that not only identifies what is missing and how much is missing, but also explains why it is missing. This enables an intelligent upgrade of the understanding of the status of production materials, moving from superficial inquiry to root cause insight.

[0016] Understandably, traditional kitting analysis typically only provides a static list of shortages. This invention, however, introduces a large language model and a multi-agent collaborative mechanism to achieve intelligent and in-depth analysis. When the Enterprise Resource Planning (ERP) system identifies a material shortage, it automatically activates the shortage analysis agent. By correlating extended data (such as supplier delivery status, quality inspection results, and logistics information) and historical shortage records, the agent intelligently diagnoses the root cause of the shortage. This significantly reduces the time planners spend manually investigating root causes, decreases over-reliance on personal experience, and enhances the perception capability and overall analysis efficiency of kitting analysis.

[0017] In some technical solutions, optionally, the first intelligent agent is connected to a resource planning data interface. After the first intelligent agent determines the demand parameters based on the procurement data and production plan data, the solution further includes: obtaining at least one sub-item material status change parameter through the resource planning data interface; determining the quantity of each sub-item material based on the status change parameter; generating a secondary analysis instruction for the target material when the quantity of any sub-item material decreases; updating the relevant data of the target material based on the secondary analysis instruction; and re-determining the kitting analysis result based on the updated relevant data.

[0018] In this solution, the first intelligent agent responsible for core computing is connected to the resource planning data interface of the production execution layer, enabling it to continuously monitor and acquire parameters reflecting real-time consumption and status changes of materials (such as production requisition, accidental scrapping, quality inspection lock-up, etc.).

[0019] Once a decrease in the available quantity of any sub-item material is detected, the system will automatically and event-drivenly generate a secondary analysis command, triggering a data update and a complete kitting analysis process reassessment.

[0020] In some technical solutions, optionally, determining the kitting analysis result corresponding to the target material based on inventory analysis results and material shortage reasons includes: acquiring supply data related to multiple sub-items; determining whether at least one sub-item has corresponding in-transit supply information based on the supply data; if in-transit supply information exists, acquiring a third intelligent agent connected to the logistics data interface; determining the transportation status of at least one sub-item in the in-transit supply information based on the third intelligent agent; determining the in-transit transportation parameters of at least one sub-item based on the logistics data interface; determining the material in-transit data based on the transportation status and in-transit transportation parameters; and determining the kitting analysis result corresponding to the target material based on inventory analysis results, material shortage reasons, and material in-transit data.

[0021] In this solution, materials in transit are identified by querying supply data. A third-party intelligent agent connected to an external logistics data interface is then invoked to obtain the precise transportation status and detailed transportation parameters of the materials in transit in real time, thereby generating structured material transit data. The third-party intelligent agent integrates the inventory analysis results representing the current situation, the root causes of material shortages, and the material transit data predicting the future to generate a complete set of analysis results.

[0022] In some technical solutions, optionally, the inventory analysis results corresponding to the target material are determined based on demand parameters and inventory data, including: inputting the bill of materials, inventory data, procurement data, and production plan data of the target material into the first intelligent agent; determining the required quantity of each sub-item material according to the arrangement order of multiple sub-items in the bill of materials; obtaining the available inventory quantity of each sub-item material in the inventory data; determining the in-transit quantity of each sub-item material based on the procurement data and in-transit transportation parameters; comparing the required quantity of each sub-item material with the available inventory quantity and the in-transit quantity, and determining the comparison result; determining the shortage quantity of each sub-item material based on the comparison result; summarizing the shortage quantity of each sub-item material, and generating an inventory analysis result including a detailed shortage list and an overall completeness judgment.

[0023] This solution starts by aggregating multi-source data such as bill of materials, inventory, procurement, and production plans. It first expands upon the bill of materials structure and calculates the required quantity of each sub-item material to meet the production plan.

[0024] Subsequently, by obtaining the current available inventory quantity in the warehouse and combining it with procurement data and real-time logistics information, the reliable quantity in transit that can arrive within the planning period is dynamically determined.

[0025] Subsequently, for each sub-item material, its required quantity is precisely compared with the sum of the available inventory quantity and the quantity in transit, thereby determining and calculating the precise shortage quantity of each material.

[0026] Ultimately, the system summarizes the shortage status of all sub-items and generates a structured inventory analysis result that includes a detailed list of specific shortage materials as well as an assessment of the overall availability of target materials. This transforms the complex availability assessment into an automated, quantitative, and logically rigorous calculation process.

[0027] In some technical solutions, optionally, after determining the shortage quantity of each sub-item material based on the comparison results, the solution further includes: determining the required quantity of each sub-item material based on the production plan data; determining the availability status of each sub-item material based on the shortage quantity and the required quantity; and determining the query status of each sub-item material based on the availability status, so as to lock the sub-item materials that meet the production requirements corresponding to the production plan data.

[0028] In this solution, after calculating the shortage quantity, the total demand quantity of each sub-item material is first determined based on the production plan data. Then, based on this demand quantity and the shortage quantity, the availability status of each material is determined. Subsequently, based on this availability status, the system assigns a technical query status to each material to control its allocation, and finally performs a system-level inventory lock operation on materials with a status of fully available.

[0029] In some technical solutions, optionally, after determining the kitting analysis results corresponding to the target material based on the inventory analysis results and the reasons for material shortage, the solution further includes: acquiring a fourth intelligent agent; marking at least one sub-item material determined to be in a kitting state in the inventory analysis results as a material to be monitored, and inputting it into the fourth intelligent agent; establishing a status monitoring system for the material to be monitored through the fourth intelligent agent; when a status change event is detected for the material to be monitored through the status monitoring system, acquiring status change parameters, wherein the status change event includes at least one of inventory being reserved and locked by other demands, scrapping during production, or quality inspection failure; determining whether the status change parameters cause the material to be monitored to become in a shortage state; if it is determined to become in a shortage state, triggering a new round of kitting analysis for the target material to re-determine the kitting analysis results.

[0030] In this scheme, after generating the initial kitting analysis results, a fourth intelligent agent (real-time monitoring agent) is introduced to proactively establish continuous status monitoring for materials that have been determined to be kitted.

[0031] Once the system detects negative events that reduce the availability of materials, such as inventory misappropriation, production scrapping, or quality inspection failures, it will immediately assess whether the event is sufficient to cause the materials to go from being in stock to being in short supply.

[0032] If the determination is correct, a completely new set of analysis processes based on the latest facts is automatically triggered, thus constructing a closed-loop feedback system that analyzes and monitors events before re-analyzing them. This transforms the entire solution from a one-off static snapshot tool into an intelligent guardian system that can dynamically evolve with the production process, proactively warn of risks, and self-update, fundamentally ensuring the continuous reliability of production plan analysis and the ability to respond instantly to sudden disturbances.

[0033] A second aspect of the present invention provides a kitting analysis device, comprising: a request receiving module for receiving a natural language request for a target material, the natural language request being used to query the kitting status of the target material, the target material including multiple sub-items; a request parsing module for parsing the natural language request to determine analysis instructions related to the target material; a data querying module for querying and obtaining relevant data of the target material based on the analysis instructions, the relevant data including at least the bill of materials, inventory data, purchase data, and production plan data of the multiple sub-items; and a demand determination module for acquiring a first intelligent agent, the first intelligent agent determining the purchase quantity... The system includes: a demand parameter determination module (based on production plan data); an inventory analysis module (based on demand parameters and inventory data to determine the inventory analysis results corresponding to the target material); a shortage acquisition module (based on inventory analysis results to determine when a sub-item material is in short supply and to acquire a second agent); a data acquisition module (based on data acquisition to acquire at least one piece of extended data related to the target material); a history record module (based on historical shortage records of at least one type); a cause analysis module (based on extended data to determine at least one cause of shortage corresponding to historical shortage records); and a kitting analysis module (based on inventory analysis results and shortage causes to determine the kitting analysis results corresponding to the target material).

[0034] An embodiment of the third aspect of this application provides an electronic device including a processor, a memory, and a program or instructions stored in the memory and executable on the processor, wherein the program or instructions, when executed by the processor, implement the steps of the kitting analysis method as described in the first aspect.

[0035] An embodiment of the fourth aspect of this application provides a readable storage medium on which a program or instructions are stored, which, when executed by a processor, implement the steps of the kitting analysis method as described in the first aspect.

[0036] An embodiment of the fifth aspect of this application provides a chip including a processor and a communication interface, the communication interface and the processor being coupled together, the processor being used to run a program or instructions to implement the steps of the kitting analysis method as described in the first aspect.

[0037] Additional aspects and advantages of the technical solutions of the present invention will become apparent in the following description or may be learned by practice of the invention. Attached Figure Description

[0038] Figure 1 One of the flowcharts of the kitting analysis method according to this application is shown;

[0039] Figure 2 A second flowchart illustrating the kitting analysis method according to this application is shown;

[0040] Figure 3A third flowchart illustrating the kitting analysis method according to this application is shown;

[0041] Figure 4 A fourth flowchart illustrating the kitting analysis method according to this application is shown;

[0042] Figure 5 Fifth of the flowcharts illustrating the kitting analysis method according to this application is shown;

[0043] Figure 6 A flowchart of the kitting analysis method according to this application is shown in diagram six;

[0044] Figure 7 A schematic block diagram of the structure of the kit analysis apparatus according to this application is shown;

[0045] Figure 8 A schematic block diagram of the structure of an electronic device according to this application is shown;

[0046] Figure 9 A diagram illustrating the call relationships between framework components according to this application is shown;

[0047] Figure 10 A flowchart illustrating the multi-agent homogenization analysis according to this application is shown.

[0048] Among them, 900: kitting analysis device; 902: request receiving module; 904: request parsing module; 906: data query module; 908: demand determination module; 910: inventory analysis module; 912: shortage acquisition module; 914: data acquisition module; 916: historical record module; 918: cause analysis module; 920: kitting analysis module; 1000: electronic device; 1109: memory; 1110: processor. Detailed Implementation

[0049] To better understand the above-described objectives, features, and advantages of the embodiments of the present invention, the embodiments of the present invention will be further described in detail below with reference to the accompanying drawings and specific embodiments. It should be noted that, unless otherwise specified, the embodiments and features described in these embodiments can be combined with each other.

[0050] Material kitting analysis is a core function of an enterprise ERP system, directly impacting the smooth execution of production plans and the successful delivery of orders. In the complex world of smart manufacturing, the diversification of material supply sources and the accelerated pace of production make accurate and efficient material kitting increasingly important and complex.

[0051] Currently, kitting analysis in the manufacturing industry is mainly based on the logic of traditional ERP systems. Its typical workflow is as follows: the system calculates the demand for products at each level of the bill of materials (BOM) based on the production plan and the bill of materials, then matches it with the current inventory and supply documents, and finally generates a shortage report for planners to refer to.

[0052] However, with the deepening of enterprise digital transformation, the traditional model of global procurement is increasingly revealing its inherent defects and limitations in the face of supplier delays, urgent order deliveries, and shortages of key components. It is no longer able to meet the high demands of modern intelligent manufacturing for agility, intelligence, and automation. The main shortcomings of existing technologies are reflected in the following aspects:

[0053] Poor system interaction and user experience: Traditional systems often rely on complex menu navigation and fixed-format reports, making operation cumbersome and information presentation unintuitive. Managers, sales personnel, and professionals outside of production planning find it difficult to directly and quickly obtain the required material availability status, let alone conduct flexible, conversational, in-depth queries.

[0054] The analysis results are not in-depth: Traditional kitting analysis can only tell "what materials are missing?" and "how much are missing?", but it lacks the ability to deeply analyze the crucial question of "why are they missing?". This is due to the severe data silos, such as: the delivery status of supply documents, logistics information of documents in transit, weather conditions for documents in transit, inventory status of other documents, quality inspection status of materials, and supplier delays under the procurement model. Planners need to spend a lot of time querying, verifying, and communicating, resulting in low analysis efficiency and a high reliance on personal experience.

[0055] Static analysis lacks real-time capability and predictability: Traditional kitting analysis often generates reports at a specific point in time by accessing the kitting analysis menu. This approach fails to provide continuous monitoring and dynamic early warnings. For example, if a material has a quality issue during product assembly in the workshop, a traditional system will not trigger a new kitting analysis. Introducing an intelligent agent allows for real-time monitoring of all relevant data changes. Once a rule is triggered, the agent informs the user of the material quality issue, analyzes whether there are alternative materials, and whether emergency allocation from other suppliers is possible. Alternatively, the user can decide whether to re-initiate the kitting analysis.

[0056] Many specific details are set forth in the following description in order to provide a full understanding of this application. However, embodiments of the invention may also be implemented in other ways different from those described herein. Therefore, the scope of protection of this application is not limited to the specific embodiments disclosed below.

[0057] The kit analysis method, apparatus, electronic device, readable storage medium, and chip provided in this application will be described in detail below with reference to specific embodiments and application scenarios.

[0058] like Figure 1 As shown, this embodiment provides a homogeneity analysis method, including:

[0059] Step S100: Receive a natural language request for the target material. The natural language request is used to query the completeness status of the target material, which includes multiple sub-items.

[0060] Step S102: Parse the natural language request to determine the analysis instructions related to the target material;

[0061] Step S104: Based on the analysis command, query and obtain relevant data of the target material. The relevant data includes at least the bill of materials, inventory data, procurement data and production plan data of multiple sub-items.

[0062] Step S106: Obtain the first intelligent agent, which determines the demand parameters based on the procurement data and production plan data;

[0063] Step S108: Determine the inventory analysis results corresponding to the target material based on the demand parameters and inventory data;

[0064] Step S110: When a shortage of a sub-item material is determined based on the inventory analysis results, a second intelligent agent is obtained;

[0065] Step S112: Obtain at least one piece of extended data related to the target material;

[0066] Step S114: Obtain at least one type of historical material shortage record;

[0067] Step S116: The second agent determines at least one cause of material shortage corresponding to the historical material shortage record based on the extended data;

[0068] Step S118: Determine the kitting analysis results corresponding to the target material based on the inventory analysis results and the reasons for material shortage.

[0069] The kitting analysis method provided by this invention constructs an intelligent kitting analysis process with natural language interaction as the entry point and intelligent agent collaborative analysis as the engine.

[0070] The kitting analysis method first receives the user's natural language query about the target material and then uses a large language model to parse out the structured analysis instructions.

[0071] Subsequently, based on the analysis instructions, the system automatically queries the relevant data such as the bill of materials, inventory, procurement and production plan of the target material, and calls the first intelligent agent (inventory analysis intelligent agent) to calculate the material demand parameters. Then, by comparing with the inventory data, it generates inventory analysis results containing specific shortage details.

[0072] When a material shortage is detected, a second intelligent agent (material shortage analysis agent) is activated. By correlating extended data with historical material shortage records, the agent can intelligently diagnose the cause of the material shortage.

[0073] Ultimately, by combining the results of inventory analysis with in-depth analysis of the reasons for material shortages, a comprehensive and in-depth analysis is generated that not only identifies what is missing and how much is missing, but also explains why it is missing. This enables an intelligent upgrade of the understanding of the status of production materials, moving from superficial inquiry to root cause insight.

[0074] Understandably, traditional kitting analysis typically only provides a static list of shortages. This invention, however, introduces a large language model and a multi-agent collaborative mechanism to achieve intelligent and in-depth analysis. When the Enterprise Resource Planning (ERP) system identifies a material shortage, it automatically activates the shortage analysis agent. By correlating extended data (such as supplier delivery status, quality inspection results, and logistics information) and historical shortage records, the agent intelligently diagnoses the root cause of the shortage. This significantly reduces the time planners spend manually investigating root causes, decreases over-reliance on personal experience, and enhances the perception capability and overall analysis efficiency of kitting analysis.

[0075] For example, Material Kit Analysis is an important function in manufacturing. It refers to the process of systematically checking and analyzing whether all the materials required to produce a product or fulfill an order are available. It ensures that all necessary materials are in place at the planned time, thereby guaranteeing smooth production or delivery.

[0076] Large Language Models (LLMs) are deep learning-based artificial intelligence models that, when trained on massive amounts of text data, are able to understand and generate human language. These models typically have billions or even hundreds of billions of parameters, giving them powerful language understanding and generation capabilities.

[0077] An agent is an important concept in the field of artificial intelligence, referring to an entity capable of perceiving its environment and acting upon it through actuators. An agent can make rational decisions based on perceived information to achieve its design goals. Agents can be software programs, robots, or other automated systems. In enterprise applications, an agent can be understood as a digital employee with specific professional capabilities.

[0078] Specifically, a final product or component used for assembly is the target material. The target material is composed of multiple levels of parts and raw materials, and these components are collectively referred to as sub-item materials.

[0079] The structural and quantitative relationships between sub-items are precisely defined by the bill of materials.

[0080] For example, a server as a target material includes multiple sub-materials such as a motherboard, a central processing unit, memory modules, and hard drives. The motherboard itself may also be a component composed of lower-level sub-materials such as chipsets, capacitors, and slots.

[0081] This method involves performing a precise, one-by-one completeness check on all sub-items associated with the target material.

[0082] Natural language requests refer to queries made directly by users, such as production planners, workshop supervisors, sales managers, or enterprise managers, using everyday spoken or written language through the system's conversational interface.

[0083] Such conversational interactive interfaces include, but are not limited to: web chat windows, chatbots or voice assistants integrated into office software.

[0084] Users' requests are directly targeted at specific products, semi-finished products, or production orders they are interested in—that is, the target materials. For example, a user might enter, "How many units of model A equipment can we deliver next Wednesday?" or "If we want to increase production of product B by 500 units, what parts are currently missing? Why are they missing?" This approach allows any business-related personnel to intuitively and conveniently initiate complex material availability queries without requiring specialized training or in-depth understanding of the system's internal logic.

[0085] Parsing natural language requests to determine analytical instructions related to the target material is coordinated by the central processing unit of the large language model.

[0086] Parsing refers to leveraging the powerful natural language understanding and contextual reasoning capabilities of large language models to perform deep semantic analysis on unstructured text input by users. Its goal is to identify the user's underlying query intent and transform ambiguous linguistic descriptions into structured, executable analysis instructions within the system.

[0087] The analysis command contains a series of explicit parameters and actions, such as: identifying the specific code or name of the target material, extracting the production quantity (e.g., 100 vehicles) and time range (e.g., next week) to be analyzed from the request, determining the analysis dimensions that the user cares about (whether it is simply checking for material shortages or root cause analysis), and generating a start command to invoke the subsequent intelligent agent workflow.

[0088] The system retrieves and integrates necessary information from multiple internal and external data sources based on analytical instructions. The relevant data retrieved constitutes the core dataset essential for kitting calculations and analysis. This data includes at least: a Bill of Materials (BOM) defining the product composition, specifying all sub-items and their quantities; inventory data reflecting current warehouse inventory (including available inventory and reserved inventory); procurement data reflecting future supply plans (such as purchase orders already placed, supplier-committed delivery dates and quantities); and source production planning data triggering material demand (such as master production schedules and work orders, specifying what needs to be produced, how much, and when).

[0089] The first intelligent agent specifically refers to the inventory analysis intelligent agent responsible for executing the core set of computational logic. The first intelligent agent is a software entity that encapsulates specialized algorithms for material requirements calculation, inventory matching, and shortage detection.

[0090] The demand parameter is a key intermediate calculation result, which refers to the net demand for each sub-item at a specified point in time after taking into account future supply commitments.

[0091] The first intelligent agent calculates the gross requirement of each sub-item material by expanding the bill of materials layer by layer based on the planned output and time of the target material in the production plan data.

[0092] Then, based on the confirmed quantity and time of delivery within the planned time zone in the procurement data, the gross demand is reduced to obtain the precise, time-segmented demand parameters for each material. The demand parameters include the required quantity and required time for each sub-item.

[0093] The first intelligent agent will calculate the demand parameters for each sub-item material and compare them with the current inventory data in real time and accurately.

[0094] The comparison logic checks whether available inventory is sufficient to meet net demand. Through this comparison, the system can accurately calculate which materials are in stock, which materials are in short supply, and the specific quantities in short supply. This report, containing a detailed shortage list and an overall stock availability conclusion, is the inventory analysis result. It answers the fundamental questions of what is missing and how much is missing.

[0095] When the calculations of the first agent indicate a material shortage, the system will not simply stop at outputting a shortage list, but will automatically trigger a deeper analysis process and obtain a second agent, which is the material shortage analysis agent, whose responsibility is to explore the root cause behind the shortage.

[0096] The second intelligent agent is a software entity that encapsulates material demand cause analysis algorithms, stores historical shortage records, and matches historical shortage records.

[0097] The second intelligent agent acquires at least one piece of extended data related to the target material through at least one data interface. This extended data comes from supplier management systems (such as supplier performance ratings and historical on-time delivery rates), quality management systems (such as material quality inspection history and defect rates), and even external data sources (such as weather information and port congestion news). Acquiring extended data aims to break down data silos between traditional systems and provide multi-dimensional evidence to support the analysis of the causes of material shortages.

[0098] Historical material shortage records refer to data archives of similar material shortage events that have occurred in the past, which may include information such as the cause of the shortage, the handling process, the time of resolution, and the final impact.

[0099] Historical material shortage records are structured and stored in a knowledge base for the second intelligent agent to compare, reference, and reason during cause analysis. This helps to quickly identify recurring problems or common cause patterns. The second intelligent agent accesses the knowledge base through a data interface to call up historical material shortage records and stores the current material shortage cause in the knowledge base to update it.

[0100] The activated second intelligent agent, the shortage analysis agent, uses specific extended data of the current shortage scenario to correlate, compare, and reason with similar cases in historical shortage records. Through this multi-source information fusion and intelligent reasoning, the system can diagnose the root cause of the current shortage.

[0101] Specifically, the second intelligent agent pre-sets various types of material shortage reasons, such as supplier delivery delays, substandard incoming materials, unexpected production losses, accidents during transportation, and sudden increases in planned demand. Each type of material shortage reason corresponds to a set of characteristic parameter patterns.

[0102] The intelligent system will convert the extended data of the current material shortage event into extended parameters and perform similarity matching with the parameter patterns of these preset material shortage cause types.

[0103] When the similarity between at least one extended parameter and the parameter pattern is greater than a preset similarity threshold, the second agent determines the reason for the material shortage of the sub-item as the corresponding historical reason for the material shortage recorded in the historical material shortage record.

[0104] Through matching, the second agent can not only determine the most likely cause of the material shortage, but also extract more specific details from the historical records with the highest matching degree. For example, the second agent might output: the root cause of this material shortage is a delayed order delivery from supplier A, whose average on-time delivery rate over the past three months is only 75%, or the shortage is due to an excessive rate of defective products found in incoming inspection of substitute material B, which is currently being addressed. This method greatly improves the accuracy and interpretability of the cause analysis.

[0105] For example, the reason for the material shortage is: the order delivery from the key supplier, Party A, is delayed by 5 days, and historical records show that the supplier has had three similar delays in the past three months. Meanwhile, the inventory of alternative material C is currently undergoing quality inspection, which is expected to be completed tomorrow afternoon.

[0106] By integrating the quantitative inventory analysis results generated by the first intelligent agent with the qualitative reasons for material shortages generated by the second intelligent agent, the resulting kitting analysis is a comprehensive, in-depth, and highly insightful report. It not only tells users what is lacking but also explains the reasons for the shortages, thus providing direct and powerful evidence for users' subsequent decisions and actions (such as urging suppliers, activating alternative materials, and adjusting production plans), greatly improving decision-making efficiency and accuracy.

[0107] In some embodiments, the system may optionally receive natural language requests for the target material, including requests that support multi-turn dialogue interaction. The large language model can combine historical dialogue context to coherently understand and interpret the user's subsequent follow-up questions, thereby completing the entire analysis process from initial query to in-depth discussion in a single session.

[0108] In some embodiments, the relevant data of the target material can be queried and obtained. The data source queried is not limited to the enterprise's internal ERP system, but also includes manufacturing execution systems, supplier portals, warehouse management systems and Internet of Things platforms, in order to obtain extended data such as real-time inventory levels, work-in-process status, supplier inventory (VMI) and production line consumption rate.

[0109] In some embodiments, optionally, at least one type of historical shortage record is obtained, including retrieving historical shortage cases related to the current shortage material, supplier, or purchaser from a knowledge base, and using a machine learning model to perform similarity matching on the cases to provide a reference for prioritizing the root cause reasoning of the second agent.

[0110] In some embodiments, optionally, after determining the kitting analysis results based on the inventory analysis results and the reasons for material shortages, the method further includes: integrating the kitting analysis results, the reasons for material shortages, and the information in transit using a large language model to generate a multimodal analysis report containing structured data, natural language summaries, and visualization charts (such as kitting status dashboards and fishbone diagrams of material shortage root causes), and automatically sending it to the relevant responsible persons through preset channels (such as email, enterprise chat tools, and mobile application push notifications).

[0111] In some embodiments, optionally, the first intelligent agent is connected to a resource planning data interface, such as... Figure 2 As shown, after acquiring the first intelligent agent, and the first intelligent agent determines the demand parameters based on the procurement data and production plan data, the process also includes:

[0112] Step S1070: Obtain at least one sub-item material's status change parameters through the resource planning data interface;

[0113] Step S1072: Determine the quantity of each sub-item material based on the status change parameters;

[0114] Step S1076: When the quantity of any sub-item material decreases, generate a secondary analysis instruction for the target material;

[0115] Step S1076: Update the relevant data of the target material based on the secondary analysis command;

[0116] Step S1078: Based on the updated relevant data, re-determine the kitting analysis results.

[0117] In this embodiment, by connecting the first intelligent agent responsible for core computing to the resource planning data interface of the production execution layer, it can continuously monitor and obtain parameters reflecting the real-time consumption and status changes of materials (such as production requisition, accidental scrapping, quality inspection lock-up, etc.).

[0118] Once a decrease in the available quantity of any sub-item material is detected, the system will automatically and event-drivenly generate a secondary analysis command, triggering a data update and a complete kitting analysis process reassessment.

[0119] Understandably, this constructs an automated closed-loop feedback mechanism from event occurrence to analysis update, thereby ensuring that the complete set of analysis results can keep up with the ever-changing actual situation on the production site. It overcomes the inherent defects of traditional methods in static analysis, lack of real-time performance and predictability from the root, and realizes continuous protection of the feasibility of production plans and risk warning.

[0120] Among them, the first intelligent agent is connected to the resource planning data interface, which establishes a real-time data perception channel for the first intelligent agent.

[0121] The resource planning data interface is a key system connection point that enables the first agent to proactively and in real time obtain dynamic information from a wider range of operational data sources within the enterprise.

[0122] This interface typically connects to manufacturing execution systems, shop floor control systems, or advanced planning and scheduling systems. These systems record real-time events and data at the production execution level, such as work order initiation and reporting, actual material requisition and consumption, scrap and rework during production, and material status locks due to quality checks. Through this connection, the first agent no longer relies solely on relatively static inventory and purchasing master data, but can access data streams reflecting real-time changes in material availability.

[0123] Status change parameters refer to specific event parameters captured from the production execution layer through the resource planning data interface that can cause an instantaneous change in the actual or planned available quantity of a certain sub-item material.

[0124] These specific event parameters are micro-level and event-driven. For example, a production work order may actually start and a specific quantity of materials may be taken from the warehouse; unplanned material scrapping may occur during the assembly process; a batch of materials may be deemed unqualified and isolated during the incoming quality inspection; or a new urgent order may lock up a portion of the inventory for dedicated use.

[0125] The first intelligent agent continuously listens for such events through this interface to obtain specific parameters related to the event, such as material code, change quantity, change type, and occurrence time.

[0126] Upon receiving the status change parameters, the first intelligent agent immediately recalculates or updates the latest available quantity of one or more affected sub-items based on the change type (e.g., requisition, scrap, lock) and the change quantity. This available quantity is dynamic and changes in real time due to every relevant event on the production floor.

[0127] Once the above calculations reveal a decrease in the available quantity of any sub-item material, whether due to requisition, scrapping, or other negative events, the first intelligent agent will automatically determine that this change may jeopardize the accuracy of the kitting analysis results previously calculated based on old data.

[0128] Therefore, the first intelligent agent automatically generates a secondary analysis instruction for the target material. This secondary analysis instruction is the same in nature and objective as the analysis instruction generated by the user's initial natural language request, but its trigger source is an internal system event, rather than external user input, and its purpose is to verify or update the initial analysis conclusions.

[0129] Upon receiving an internally triggered secondary analysis instruction, the system will first refresh the relevant data of the target material based on the latest status change event, with the most crucial step being updating the inventory data.

[0130] Using the updated relevant data, the first agent, and potentially the second agent, is driven to completely re-execute the kitting analysis process. This will produce a completely new kitting analysis result based on the latest facts. This mechanism enables the entire system to have proactive adaptability and real-time correction capabilities, allowing it to react immediately to unexpected disturbances in the production process and ensuring the continuous reliability of production planning analysis.

[0131] In some embodiments, optionally, at least one sub-item material status change parameter can be obtained through the resource planning data interface. The status change parameter can be further subdivided into different types, such as planned consumption event parameters (e.g., standard requisition for work orders), unplanned loss event parameters (e.g., production scrapping, quality destruction), and availability lock event parameters (e.g., reserved by higher priority orders).

[0132] The first intelligent agent configures different listening rules and processing priorities for different event types. For example, it immediately triggers reanalysis for sudden large-scale scrap events, while it performs cumulative batch processing on standard planned requisition events before triggering them.

[0133] In some embodiments, optionally, after determining the quantity of each sub-item material based on the state change parameters, the proportion of the reduction to the total demand or total inventory of the sub-item material is calculated and compared in advance with a preset threshold. A secondary analysis instruction is generated only when the reduction or reduction proportion exceeds the threshold, thereby avoiding frequent and unnecessary full-scale analysis triggered by negligible quantity changes and optimizing system computing resources.

[0134] In some embodiments, optionally, the entire process analysis is not always performed from scratch. Instead, the task coordination agent makes the judgment: if the state change only involves individual materials and does not affect overall production efficiency, the first agent can be instructed to perform a partial recalculation of the affected parts and quickly update the relevant parts of the analysis report; if the change has a significant impact, a complete re-analysis is initiated. This further improves the system's response efficiency.

[0135] In some embodiments, optionally, such as Figure 3 As shown, the kitting analysis results corresponding to the target material, determined based on inventory analysis results and the reasons for material shortages, include:

[0136] Step S1180: Obtain supply data related to multiple sub-item materials;

[0137] Step S1182: Determine whether there is corresponding in-transit supply information for at least one sub-item material based on the supply data;

[0138] Step S1184: If there is in-transit supply information, then obtain the third intelligent agent, and the third intelligent agent is connected to the logistics data interface;

[0139] Step S1186: Determine the transportation status of at least one sub-item material in the in-transit supply information based on the third intelligent agent;

[0140] Step S1188: Determine the in-transit transportation parameters of at least one sub-item material based on the logistics data interface;

[0141] Step S1190: Determine the material in transit data based on the transportation status and in-transit transportation parameters;

[0142] Step S1192: Determine the kitting analysis results corresponding to the target material based on the inventory analysis results, the reasons for material shortage, and the material in transit data.

[0143] In this embodiment, materials in transit are identified by querying supply data, and a third-party intelligent agent connected to an external logistics data interface is specifically invoked to obtain the precise transportation status and detailed transportation parameters of the materials in transit in real time, thereby generating structured material transit data. The third-party intelligent agent integrates the inventory analysis results representing the current situation, the root causes of material shortages, and the material transit data predicting the future to generate a complete set of analysis results.

[0144] Specifically, the system will proactively query supply data, which refers to all supply plans and vouchers that have been committed but have not yet been converted into actual usable inventory.

[0145] The core forms of supply data are purchase orders, transfer requests, and supplier delivery notes.

[0146] Supply data indicates the material requirements that a company has issued to external or internal entities, as well as the promised supply quantity and date. It is the fundamental basis for determining whether materials are in transit.

[0147] The system analyzes the aforementioned supply data, checking for records marked as shipped, in transit, transported, or with a planned delivery date at a future point in time but not yet in the warehouse. These records are identified as supply information in transit. The significance lies in accurately identifying sub-items from all future supply plans that have already left the supplier and are currently moving towards the system's warehouse. The dynamics of these sub-items directly impact short-term completeness.

[0148] When the system confirms the existence of in-transit supply information, it automatically activates a third-party intelligent agent (or in-transit tracking intelligent agent) specializing in logistics tracking. The third-party intelligent agent is connected to the logistics data interface, which means that the capabilities of the third-party intelligent agent are not limited to the enterprise's internal systems, but can interact with external logistics service providers, express delivery companies' tracking systems, or the enterprise's transportation management platform in real time through application programming interfaces to obtain information on the movement of goods in the physical world.

[0149] The third intelligent agent uses a logistics data interface to query and report the latest transportation status of each supply in transit, thereby transforming a static purchase order or delivery order into a dynamic logistics object whose location can be known.

[0150] Transport status is a key parameter describing the stage of a shipment in the logistics chain, and is usually represented by a standard logistics status code.

[0151] In addition to basic transportation status, the third intelligent agent will also acquire in-transit transportation parameters, which include more specific, predictive and actionable details, including but not limited to: the name and waybill number of the logistics company carrying the goods, the real-time geographical location of the goods (such as latitude and longitude, cities passed through), detailed transportation trajectory history, the latest scan time, and the estimated arrival time predicted based on the logistics company's algorithm.

[0152] For example, in-transit parameters can also provide early warnings of anomalies, such as delays caused by weather or traffic congestion.

[0153] The third-party intelligent agent cleans, correlates, and formats the acquired transportation status and various in-transit parameters to generate a unified, structured material in-transit data report. This report clearly indicates for each shortage or in-transit sub-item: which order it is on, its current location, who is transporting it, and when it is expected to arrive and become available inventory.

[0154] Understandably, by introducing dynamic tracking and intelligent analysis of materials in transit, traditional static kitting checks are upgraded to supply chain situational awareness with a timeline perspective and predictive capabilities. This breaks down information silos and seamlessly connects internal ERP data with external logistics dynamics. This transparency helps enhance collaboration efficiency between internal departments and with external suppliers and logistics providers, enabling rapid identification and response to breakpoints.

[0155] In some embodiments, optionally, when it is identified that the materials in transit will be severely delayed, the third intelligent agent can automatically activate the emergency response process, such as linking with the material shortage analysis intelligent agent to quickly find alternative materials or suppliers, or directly pushing the warning information to the procurement system to generate an expedited delivery task.

[0156] In some embodiments, the generated material transit data can be optionally converted into a visual timeline or map trajectory and integrated into a complete set of analysis reports. Users can click to interact and view detailed logistics node information, improving the intuitiveness of information acquisition.

[0157] In some embodiments, optionally, such as Figure 4 As shown, the inventory analysis results corresponding to the target material are determined based on demand parameters and inventory data, including:

[0158] Step S1080: Input the bill of materials, inventory data, procurement data and production plan data of the target material into the first intelligent agent;

[0159] Step S1082: Determine the required quantity of each sub-item material according to the order of the multiple sub-items in the bill of materials;

[0160] Step S1084: Obtain the available inventory quantity of each sub-item material in the inventory data;

[0161] Step S1086: Determine the quantity of each sub-item material in transit based on the procurement data and in-transit transportation parameters;

[0162] Step S1088: Compare the required quantity of each sub-item material with the available inventory quantity and the quantity in transit, and determine the comparison result;

[0163] Step S1090: Determine the shortage quantity of each sub-item material based on the comparison results;

[0164] Step S1092: Summarize the shortage quantity of each sub-item material and generate inventory analysis results including a detailed shortage list and an overall completeness judgment.

[0165] In this embodiment, starting with the aggregation of multi-source data such as bill of materials, inventory, procurement and production plans, the process first expands based on the bill of materials structure and calculates the required quantity of each sub-item material to meet the production plan.

[0166] Subsequently, by obtaining the current available inventory quantity in the warehouse and combining it with procurement data and real-time logistics information, the reliable quantity in transit that can arrive within the planning period is dynamically determined.

[0167] Subsequently, for each sub-item material, its required quantity is precisely compared with the sum of the available inventory quantity and the quantity in transit, thereby determining and calculating the precise shortage quantity of each material.

[0168] Ultimately, the system summarizes the shortage status of all sub-items and generates a structured inventory analysis result that includes a detailed list of specific shortage materials as well as an assessment of the overall availability of target materials. This transforms the complex availability assessment into an automated, quantitative, and logically rigorous calculation process.

[0169] Specifically, the first intelligent agent performs the initialization and data loading phases of the calculation, where all necessary raw business data is aggregated and input into the first intelligent agent. The bill of materials defines the parent-child relationship between the target material and all its sub-items, as well as the precise unit quantity.

[0170] Inventory data provides the current physical inventory of all sub-items in the warehouse; procurement data includes placed purchase orders and their promised delivery dates and quantities; production planning data specifies the exact quantity and time requirements of the target materials to be produced. Unifying the input of these multi-source data provides a unique and accurate data foundation for subsequent precise calculations.

[0171] The order in which multiple sub-items in a bill of materials usually refers to the hierarchical expansion order of the bill of materials or the process route order.

[0172] The first agent, following this sequence and combining it with the planned output of the target materials in the production plan data, calculates the required quantities layer by layer and item by item. For example, if 100 finished products A are to be produced, and the bill of materials specifies that each A requires 2 parts B, then the gross required quantity of part B is 200. This calculation traverses all levels of the bill of materials, down to the lowest level of raw materials, to obtain the total required quantity of each sub-item of material to complete the production plan.

[0173] Available inventory quantity refers to the physical quantity of sub-item materials in the current warehouse that are not reserved or occupied by other demands and can be freely used for the current production plan, as obtained from inventory data.

[0174] The number in transit is a dynamically calculated expected value.

[0175] Based on procurement data, the "in-transit quantity" is used to filter out the number of purchase orders whose planned delivery dates fall within the timeframe required for the current production plan. Furthermore, by combining in-transit transportation parameters, the system intelligently determines how many of these ordered materials will arrive on time and be available for production.

[0176] For example, if a purchase order for a certain material contains 1,000 units, but logistics tracking shows that only 600 units will arrive before the required production date, then the quantity in transit is counted as 600 units. This definition goes beyond the static logic of traditional ERP systems that rely solely on planned dates, introducing accurate forecasting based on real-time logistics.

[0177] For each sub-item material, the first agent performs the following formulaic comparison: the demand quantity is compared to the sum of the available inventory quantity and the quantity in transit. This comparison is performed item by item, synchronously. The comparison result is the calculation conclusion, which clearly indicates whether the current and upcoming total supply for the material is greater than or equal to, or less than, its demand quantity.

[0178] When the above comparison shows that the demand quantity is greater than the sum of the available inventory and the quantity in transit, the difference between the sum of the available inventory and the quantity in transit and the demand quantity is the shortage quantity of the sub-item. For example, if the demand is 200 units, the available inventory is 50 units, and the reliable quantity in transit is 100 units, then the shortage quantity = 200 - (50 + 100) = 50 units. If the supply is greater than or equal to the demand, the shortage quantity is recorded as 0. This calculation generates an accurate shortage quantification index for each material.

[0179] The first AI will summarize the calculation results of all sub-items and generate a detailed shortage list. The detailed shortage list is a structured list that clearly lists all materials with a shortage quantity greater than zero and their corresponding specific shortage quantities.

[0180] Simultaneously, the system will make an overall availability assessment based on the existence and severity of material shortages, such as complete availability, partial shortage but sufficient to meet the main plan, or severe shortage preventing operation. This report, combining specific details with overall conclusions, constitutes the complete inventory analysis results, providing planners with direct and quantifiable action guidelines.

[0181] In some embodiments, optionally, the required quantity of each sub-item is determined according to the arrangement order of multiple sub-items in the bill of materials. This arrangement order is not limited to a single-layer BOM unfolding order, but can also be a network dependency order based on the manufacturing process route. The first intelligent agent can dynamically calculate the optimal demand unfolding order and demand timing based on the material's supply and demand time, substitution relationships, and process constraints, providing a timing basis for accurate kitting judgment.

[0182] In some embodiments, the shortage quantity of each sub-item material can be determined based on the comparison results. When calculating the shortage, the quantity of unavailable inventory that has been locked by other higher priority work orders or customer orders in the system will also be considered and deducted simultaneously to ensure that the available inventory quantity calculated in this analysis is a net quantity that can be truly controlled by the current production plan and to avoid inventory allocation conflicts.

[0183] In some embodiments, optionally, if a shortage of a certain sub-item material is found during the comparison process, the first intelligent agent will automatically query a predefined knowledge base of material substitution relationships. If a certified substitute material exists, a rapid completeness check for the substitute material will be initiated in parallel, and the feasibility of substitution will be included as a key note in the final inventory analysis results, providing planners with an immediate solution.

[0184] In some embodiments, optionally, such as Figure 5 As shown, after determining the shortage quantity of each sub-item material based on the comparison results, the process also includes:

[0185] Step S10902: Determine the required quantity of materials corresponding to each sub-item based on the production plan data;

[0186] Step S10904: Determine the availability status of each sub-item material based on the shortage quantity and the required quantity;

[0187] Step S10906: Determine the query status of each sub-item material based on the completeness status, so as to lock the sub-item materials that meet the production requirements corresponding to the production plan data.

[0188] In this embodiment, after calculating the shortage quantity, the total demand quantity of each sub-item material is first determined based on the production plan data. Then, the availability status of each material is determined based on this demand quantity and the shortage quantity. Subsequently, based on this availability status, the system determines a technical query status for each material to control its allocation, and finally performs a system-level inventory lock operation on materials with a status of fully available.

[0189] Understandably, converting the analysis results into proactive reservations of physical inventory effectively prevents confirmed available resources from being occupied by other demands before execution, thus achieving a seamless connection from virtual analysis to physical support and ensuring the seriousness and feasibility of production plans.

[0190] Specifically, after the first intelligent agent completes the core material demand and supply comparison and accurately calculates the shortage quantity of each sub-item material, the system will not terminate the process, but must immediately enter a critical resource reservation and status management phase.

[0191] Although the first agent used production planning data to calculate demand in the previous steps, it is necessary to explicitly refer to this data again here to obtain the required quantity of each sub-item material to complete a specific production task. The required quantity is the benchmark denominator for measuring the shortage ratio and also the absolute standard for judging whether a material is sufficient.

[0192] The calculated shortage quantity of each material is compared with its demand quantity to assign a qualitative kitting status label to each sub-item. This status is usually not a simple yes or no, but may include multiple levels, such as: fully kitting (shortage quantity is 0, demand is 100% met), partially kitting (shortage quantity is greater than 0 but less than demand quantity, for example, demand is 1000 units, shortage is 200 units), and fully kitting (available and in-transit supply is 0, shortage quantity equals demand quantity).

[0193] The query status is a status field used in inventory management systems or material master data to control material availability. The system determines and sets the query status of each material based on the completeness status logic.

[0194] For example, for materials whose completeness is determined to be fully complete, their query status is set to available or planned to be reserved; for materials that are partially or completely in short supply, their query status is marked as insufficient supply or pending follow-up.

[0195] For all sub-items whose completeness is confirmed to meet the needs of the current production plan after calculation, the system will automatically perform a locking operation. The locking operation is a specific and irreversible system action, which means that in the inventory management record, a specific quantity of materials is marked as reserved by a specific production plan, thereby updating their status in the system to be unavailable for allocation or use by other work orders or demands.

[0196] This process assigns a unique label to confirmed resources, ensuring that the validity of the production plan analysis conclusions is not undermined by competition from other subsequent demands.

[0197] In some embodiments, the query status of each sub-item material can be determined according to the kitting status. The defined query status may include multiple refined statuses such as fully kitted, partially kitted, and pending release - which can be preempted by a better plan.

[0198] The system can be configured with rules to not only lock the full amount of materials that are fully complete, but also lock the quantity of materials that are already complete in the shortage. At the same time, it can associate the shortage with a pending replenishment status to finely manage each piece of inventory.

[0199] In some embodiments, optionally, sub-items of materials that meet production needs are locked, with the locking operation accompanied by a configurable lock validity period. The lock is automatically released after the timeout to prevent materials from being held ineffectively for an extended period due to plan changes or user forgetfulness, thus affecting the overall supply chain liquidity.

[0200] In some embodiments, optionally, the system performs a lock conflict check before the locking operation is executed. That is, it checks whether the target material has been reserved by other higher priority plans (such as confirmed customer orders). If a conflict exists, it is adjudicated according to preset priority rules, and a conflict report is generated to inform the user, rather than directly overwriting the original lock, thereby ensuring the supply of resources for the enterprise's core business.

[0201] In some embodiments, optionally, such as Figure 6 As shown, after determining the kitting analysis results corresponding to the target material based on the inventory analysis results and the reasons for material shortages, the following steps are also included:

[0202] Step S1202: Obtain the fourth agent;

[0203] Step S1204: Mark at least one sub-item material that is determined to be in a complete state in the inventory analysis results as a material to be monitored, and input it into the fourth intelligent agent;

[0204] Step S1206: Establish status monitoring of the materials to be monitored through the fourth intelligent agent;

[0205] Step S1208: When a status change event is detected by the status listener for the material to be monitored, the status change parameters are obtained. The status change event includes at least one of the following: inventory is reserved and locked by other demands, scrapped during production, or fails quality inspection.

[0206] Step S1210: Determine whether the status change parameter causes the monitored material to change to a shortage state;

[0207] Step S1212: If the situation is determined to be in a shortage state, a new round of kitting analysis for the target material is triggered to re-determine the kitting analysis results.

[0208] In this embodiment, after generating the initial kitting analysis results, a fourth intelligent agent (real-time monitoring intelligent agent) is introduced to actively establish continuous status monitoring for materials that have been determined to be kitted.

[0209] Once the system detects negative events that reduce the availability of materials, such as inventory misappropriation, production scrapping, or quality inspection failures, it will immediately assess whether the event is sufficient to cause the materials to go from being in stock to being in short supply.

[0210] If the determination is correct, a completely new set of analysis processes based on the latest facts is automatically triggered, thus constructing a closed-loop feedback system that analyzes and monitors events before re-analyzing them. This transforms the entire solution from a one-off static snapshot tool into an intelligent guardian system that can dynamically evolve with the production process, proactively warn of risks, and self-update, fundamentally ensuring the continuous reliability of production plan analysis and the ability to respond instantly to sudden disturbances.

[0211] The fourth agent is a specialized execution unit introduced to undertake continuous monitoring responsibilities. The system dynamically acquires, invokes, or instantiates the fourth agent, which corresponds to the real-time monitoring agent in the system architecture. Unlike the preceding agents responsible for calculation and analysis, the fourth agent is a resident, event-driven daemon process whose core capability is to perceive changes in the state of materials in real time.

[0212] From the inventory analysis results generated in this study, all sub-items that were determined to be in a complete state were selected and specially marked as materials to be monitored. These materials, which are considered safe at the current point in time, have been formally included in the key monitoring list of the fourth intelligent agent.

[0213] Subsequently, this list was input into the fourth agent, becoming the target set for its real-time monitoring task.

[0214] The fourth intelligent agent establishes a status listener in the system background based on the received list of materials to be monitored. This means that the fourth intelligent agent will subscribe to or poll the data streams or event buses of relevant systems within the enterprise, such as work order transaction logs of the manufacturing execution system, inventory change records of the warehouse management system, and inspection result notifications of the quality system, and specifically filter out event information related to the materials to be monitored.

[0215] Status change events refer to specific events that cause a negative change in the actual or available quantity of the monitored material.

[0216] Change event types include, but are not limited to: material inventory being reserved and locked by other demand (i.e., being occupied by new, higher priority orders), scrapping during production (accidental damage during assembly or processing), and quality inspection failure (incoming materials or in-process materials failing inspection).

[0217] Upon detecting such an event, the fourth agent will immediately obtain state change parameters. These parameters describe the event itself in detail, such as: which new requirement locked it, how many were locked, how many were scrapped at which workstation, and which batch of materials failed quality inspection for what reason.

[0218] The fourth intelligent agent, or the central coordination logic that works in conjunction with it, will perform real-time calculations and judgments based on the acquired state change parameters.

[0219] The core issue is to assess whether the reduction in the quantity of materials caused by this incident is sufficient to reverse the material shortage from a state of completeness to one of scarcity.

[0220] For example, if the original complete quantity of material A was 1000 units, and this event caused 50 units to be locked by new orders, then its available quantity becomes 950 units, which may still meet the original planned demand (assuming the demand is 900 units), and thus it will not become a shortage. However, if 200 units are locked, then the available quantity is only 800 units, which cannot meet the original demand of 900 units, and thus it is determined to become a shortage. This judgment is a key threshold check that determines whether a reanalysis should be initiated.

[0221] Once it is determined that a monitored material has changed to a shortage status due to a status change, a new round of kitting analysis for the target material will be automatically triggered. The entire process, from parsing, data querying, inventory calculation to root cause analysis, or an optimized incremental process, will be automatically re-initiated.

[0222] The revised kitting analysis, based on the latest data including the impact of this unexpected event, will generate an updated analysis report, promptly informing users of the changed situation and new risks. This gives the system strong adaptability and real-time immunity to production disruptions, ensuring that the information used for planning decisions is always synchronized with the production site, greatly improving the resilience of the supply chain and the reliability of production plans.

[0223] In some embodiments, optionally, the status change events are not limited to the types listed, but may also include any internal or external system events that cause a reduction in the quantity of available materials or a delay in delivery, such as purchase order cancellation or reduction, supplier notification of delivery delay, warehouse inventory shortage, or materials being transferred to other plants. The fourth agent listens to these diverse events by connecting to a wider range of data interfaces such as the purchasing system and the warehousing system.

[0224] In some embodiments, optionally, the determination of whether a state change parameter causes the monitored material to enter a shortage state can be configured to consider the material's safety stock threshold and the production plan's buffer time. For example, even if the material quantity can still meet the current net demand after a reduction, if it is already below the preset safety stock level, or has encroached on the time buffer reserved to cope with uncertainty, the fourth agent can still determine that it has entered a risk state and trigger an early warning notification, rather than necessarily triggering a complete reanalysis immediately.

[0225] In some embodiments, the state monitoring mechanism established by the fourth agent can optionally be a combination of an event-driven mode and a periodic snapshot comparison mode. In addition to monitoring real-time event streams, the fourth agent will also periodically acquire snapshots of the inventory, in-transit, and demand of key materials, and discover slow changes or quantity differences that have not been captured by the event system by comparing the differences between snapshots of adjacent periods, thereby achieving more comprehensive state monitoring coverage.

[0226] In some embodiments, optionally, when a status change event occurs but is determined not to have caused the material to become scarce, the fourth agent will still generate a status change log, recording the event content, time, quantity affected, and system judgment result. The status change log is saved in association with the original kitting analysis results, forming a complete material status change audit trail for subsequent problem backtracking or analysis model optimization.

[0227] In one specific embodiment, optionally, a multi-agent homogenization analysis device based on a large language model is provided, the framework component call relationship diagram of which is as follows: Figure 9 As shown:

[0228] At the application layer, users inquire about the availability status of materials via natural language interaction. The Large Language Model (LLM) layer receives the user's natural language, parses it, and generates task execution instructions. Then, the LLM calls the intelligent agent coordination layer. After receiving the LLM instructions, the intelligent agent is dispatched by the task coordination intelligent agent. The first step is to call the data query intelligent agent to query static data in the ERP system and return it to the task coordination intelligent agent. The second step is to call the inventory analysis intelligent agent to perform availability analysis and calculate the availability status of materials, and return it to the task coordination intelligent agent. The third step is to call the real-time monitoring intelligent agent to monitor the material data in the ERP system. Finally, the inventory analysis intelligent agent calls the shortage analysis intelligent agent and the in-transit tracking intelligent agent based on the analysis results (the data sources for this part are the ERP system, external application data sources, knowledge base, rules, etc.).

[0229] The application layer includes the user interface design (UI) interface, where users can input natural language to query the availability status of materials through a large language model.

[0230] Among them, the data query agent and the inventory analysis agent are respectively connected to the ERP system data source.

[0231] In one specific embodiment, optionally, a multi-agent homogenization analysis method based on a large language model is described in the following process: Figure 10 As shown, it includes:

[0232] Step S200: The user sends a natural language request to check the kitting status of a certain material;

[0233] Step S202: The large language model coordinates central reasoning, distribution, and aggregation;

[0234] Step S204: The task-coordinating intelligent agent begins its work;

[0235] Step S206: The inventory analysis agent performs kitting calculations and automatically locks the inventory;

[0236] Step S208: Determine if there is a material shortage;

[0237] If the result of step S208 is yes, then step S210 is executed: the material shortage analysis agent analyzes why there is a material shortage.

[0238] If the result of step S208 is negative, proceed to step S212: determine whether the supply document is in transit;

[0239] When the result of step S212 is yes, step S214 is executed: the in-transit tracking agent analyzes the in-transit status of materials and logistics information.

[0240] If the result of step S212 is negative, return to step S202;

[0241] After step S214, proceed to step S216: Output natural language executable insights;

[0242] Among them, after step S204, step S300 also needs to be executed: the real-time monitoring agent triggers the event driven by the real-time data source through the API event stream;

[0243] After step S204, step S400 must be executed: the data query agent queries the material BOM and inventory data, procurement data, and production plan data.

[0244] Specifically, users can engage in multi-round dialogues and in-depth discussions by inputting natural language queries through the system's interactive interface (e.g., chat box, voice prompts, etc.). For example: How many cars can we produce now? Or what materials are we lacking to produce 100 cars next week? Why are we short of wheels? When will we be able to assemble enough parts for 2000 cars?

[0245] Large Language Model (LLM) Coordination Center: LLM utilizes its powerful natural language understanding capabilities to analyze the user's deep intent and generate instructions, calling the corresponding agents to execute the instructions. At this point, the task coordination agent is first called to initiate the kit analysis process.

[0246] The task coordination agent receives instructions from the LLM and is responsible for advancing the specific workflow. According to the preset logical order, it first calls the data query agent, the inventory analysis agent, and the real-time data monitoring agent, and manages the data transfer between them.

[0247] The instructions parsed by LLM are passed to the data query agent to retrieve the key data required for complete set analysis, such as the bill of materials (BOM), inventory data, procurement data, and production plan data for material A, and the query results are fed back to the task coordination agent.

[0248] The inventory analysis agent receives the results from the data query agent (Bill of Materials, inventory data, procurement data, and production plan data). It expands the BOM, compares the material requirements with available inventory item by item, and accurately calculates which materials are readily available, which are in short supply, and the quantities of any shortages. Furthermore, the system can automatically lock in materials that meet the demand, reserving them for production. The analysis results are then fed back to the task coordination agent.

[0249] The real-time monitoring agent receives the results from the inventory analysis agent and monitors the inventory materials that meet the demand. It monitors whether the inventory of a certain material is reserved or locked by other demand documents, the scrapping status of materials during production, and the quality inspection status of materials. If a material is damaged during the assembly of a product, the monitoring results are fed back to the Artificial Intelligence (AI) big language model. After the AI ​​big language model informs the user, the user decides whether to restart the kitting analysis process or urgently transfer materials from other suppliers based on the model's results.

[0250] When a material shortage is detected in the system, the material shortage analysis agent is activated. It will further query the data and analyze the reasons for the shortage, such as: Did the supplier fail to deliver on time? Was it returned due to quality inspection failure? Did production losses exceed expectations? Or did the demand plan suddenly increase? The final analysis results are fed back to the AI ​​large language model.

[0251] The system determines whether supply data is in transit. When it detects in-transit documents, it activates an intelligent tracking agent. This agent tracks the status of purchase orders and delivery notes, and may integrate third-party logistics data to provide information such as the real-time location of materials, estimated arrival time, and transportation carrier. This transforms static in-transit data into dynamic and predictable material information, and feeds the data back into the AI ​​large-scale language model.

[0252] Output Natural Language Reports and Actionable Insights: All analysis results and data are fed back to the AI ​​Large Language Model (LLM). The AI ​​Large Language Model summarizes and integrates all information, transforming it into a well-structured and easy-to-understand natural language report. The natural language report includes: completeness status, missing items, reasons for missing items, and information on items in transit. At the same time, it also generates actionable insights, such as: recommending to use a certain material to replace the material when it is in short supply, or delaying the production plan until material A arrives, or contacting the supplier immediately to expedite delivery, etc.

[0253] like Figure 7As shown in the illustration, this application embodiment also provides a kitting analysis device 900, which includes: a request receiving module 902, used to receive a natural language request for a target material, the natural language request being used to query the kitting status of the target material, the target material including multiple sub-items; a request parsing module 904, used to parse the natural language request to determine the analysis instructions related to the target material; a data query module 906, used to query and obtain relevant data of the target material based on the analysis instructions, the relevant data including at least the bill of materials, inventory data, procurement data, and production plan data of the multiple sub-items; and a demand determination module 908, used to obtain a first intelligent agent, the first intelligent agent determining the demand based on the procurement data. The system includes: a production plan data module 910 for determining demand parameters; an inventory analysis module 912 for determining inventory analysis results corresponding to the target material based on demand parameters and inventory data; a shortage acquisition module 914 for acquiring a second agent when a sub-item material is in short supply based on inventory analysis results; a data acquisition module 914 for acquiring at least one piece of extended data related to the target material; a historical record module 916 for acquiring at least one type of historical shortage record; a cause analysis module 918 for the second agent to determine at least one cause of shortage corresponding to the historical shortage record based on the extended data; and a kitting analysis module 920 for determining kitting analysis results corresponding to the target material based on inventory analysis results and shortage causes.

[0254] like Figure 8 As shown, this application embodiment also provides an electronic device 1000, including a processor 1110, a memory 1109, and a program or instructions stored in the memory 1109 and executable on the processor 1110. When the program or instructions are executed by the processor 1110, they implement the various processes of the above-described embodiments of the kitting analysis method and achieve the same technical effect. To avoid repetition, they will not be described again here.

[0255] Optionally, the processor 1110 is configured to receive a natural language request for a target material, the natural language request being used to query the completeness status of the target material, the target material including multiple sub-item materials;

[0256] Optionally, the processor 1110 is also configured to parse natural language requests to determine analysis instructions related to the target material;

[0257] Optionally, the processor 1110 is also used to query and obtain relevant data of the target material based on analysis instructions. The relevant data includes at least the bill of materials, inventory data, procurement data and production plan data of multiple sub-items.

[0258] Optionally, the processor 1110 is also used to acquire a first intelligent agent, which determines the demand parameters based on the procurement data and production plan data;

[0259] Optionally, the processor 1110 is also configured to determine the inventory analysis results corresponding to the target material based on the demand parameters and inventory data;

[0260] Optionally, the processor 1110 is also used to obtain a second intelligent agent when it is determined that there is a shortage of sub-item materials based on the inventory analysis results;

[0261] Optionally, the processor 1110 is also configured to acquire at least one piece of extended data related to the target material;

[0262] Optionally, the processor 1110 is also configured to acquire at least one type of historical shortage record;

[0263] Optionally, the processor 1110 is further configured to enable the second agent to determine at least one cause of material shortage corresponding to a historical material shortage record based on extended data;

[0264] Optionally, the processor 1110 is also configured to determine the kitting analysis results corresponding to the target material based on the inventory analysis results and the reasons for the shortage.

[0265] The memory 1109 can be used to store software programs and various data. The memory 1109 may primarily include a first storage area for storing programs or instructions and a second storage area for storing data. The first storage area may store the operating system, application programs or instructions required for at least one function (such as sound playback, image playback, etc.). Furthermore, the memory 1109 may include volatile memory or non-volatile memory, or both. The non-volatile memory may be read-only memory (ROM), programmable read-only memory (PROM), erasable programmable read-only memory (EPROM), electrically erasable programmable read-only memory (EEPROM), or flash memory. Volatile memory can be random access memory (RAM), static random access memory (SRAM), dynamic random access memory (DRAM), synchronous dynamic random access memory (SDRAM), double data rate synchronous dynamic random access memory (DDRSDRAM), enhanced synchronous dynamic random access memory (ESDRAM), synchronous link dynamic random access memory (SLDRAM), and direct memory bus RAM (DRRAM). The memory 1109 in this embodiment includes, but is not limited to, these and any other suitable types of memory.

[0266] This application also provides a readable storage medium storing a program or instructions. When the program or instructions are executed by a processor, they implement the various processes of the above-described kitting analysis method embodiments and achieve the same technical effects. To avoid repetition, these will not be described again here. Furthermore, the readable storage medium improves the data storage capacity and data processing speed of the kitting analysis method in this application.

[0267] A computer-readable storage medium can be a tangible device that holds and stores instructions for use by an instruction execution device. A computer-readable storage medium can be an electronic storage device, a magnetic storage device, an optical storage device, an electromagnetic storage device, a semiconductor storage device, or any suitable combination of the foregoing, but is not limited thereto. A non-exhaustive list of more specific examples of computer-readable storage media includes: portable computer floppy disks, hard disks, random access memory (RAM), read-only memory (ROM), erasable and programmable read-only memory (EPROM or flash memory), static random access memory (SRAM), portable optical disc read-only memory (CD-ROM), digital universal disk (DVD), memory cards, floppy disks, encoding mechanical devices (e.g., punched cards or grooves with raised structures for recording instructions), and any suitable combination of the foregoing. The computer-readable storage medium used herein should not be construed as the transmission of signals themselves, such as radio waves or other freely propagating electromagnetic waves, electromagnetic waves propagating through waveguides or other transmission media, or electrical signals transmitted through wires.

[0268] The processor is the processor in the electronic device described in the above embodiments. The readable storage medium includes computer-readable storage media, such as computer read-only memory (ROM), random access memory (RAM), magnetic disk, or optical disk.

[0269] This application also provides a chip, which includes a processor and a communication interface. The communication interface and the processor are coupled. The processor is used to run programs or instructions to implement the various processes of the above-described kitting analysis method embodiments and achieve the same technical effects. To avoid repetition, it will not be described again here. Furthermore, the chip improves the data processing speed corresponding to the kitting analysis method in this application.

[0270] It should be understood that the chip mentioned in the embodiments of this application may also be referred to as a system-on-a-chip, system chip, chip system, or system-on-a-chip, etc.

[0271] In this invention, the terms "first," "second," and "third" are used for descriptive purposes only and should not be construed as indicating or implying relative importance; the term "multiple" refers to two or more unless otherwise explicitly defined. The terms "install," "connect," "link," and "fix" should be interpreted broadly. For example, "connect" can be a fixed connection, a detachable connection, or an integral connection; "link" can be a direct connection or an indirect connection through an intermediate medium. Those skilled in the art can understand the specific meaning of the above terms in this invention according to the specific circumstances.

[0272] In the description of this invention, it should be understood that the terms "upper," "lower," "left," "right," "front," "rear," etc., indicate the orientation or positional relationship based on the orientation or positional relationship shown in the accompanying drawings. They are only for the convenience of describing this invention and simplifying the description, and do not indicate or imply that the device or unit referred to must have a specific orientation or be constructed and operated in a specific orientation. Therefore, they should not be construed as limitations on this invention.

[0273] In the description of this specification, the terms "one embodiment," "some embodiments," "specific embodiment," etc., refer to specific features, structures, materials, or characteristics described in connection with an embodiment or example that are included in at least one embodiment or example of the present invention. In this specification, the illustrative expressions of the above terms do not necessarily refer to the same embodiment or example. Furthermore, the specific features, structures, materials, or characteristics described may be combined in any suitable manner in one or more embodiments or examples.

[0274] The above are merely preferred embodiments of the present invention and are not intended to limit the invention. Various modifications and variations can be made to the present invention by those skilled in the art. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the scope of protection of the present invention.

Claims

1. A homogeneity analysis method, characterized in that, include: Receive a natural language request for a target material, the natural language request being used to query the completeness status of the target material, the target material including multiple sub-items; Parse the natural language request to determine the analysis instructions related to the target material; Based on the analysis command, relevant data of the target material is queried and obtained. The relevant data includes at least the bill of materials, inventory data, procurement data and production plan data of multiple sub-items. A first intelligent agent is obtained, and the first intelligent agent determines the demand parameters based on the procurement data and the production plan data; Based on the demand parameters and the inventory data, determine the inventory analysis results corresponding to the target material; When a shortage of the sub-item material is determined based on the inventory analysis results, a second intelligent agent is obtained; Obtain at least one piece of extended data related to the target material; Retrieve at least one type of historical material shortage record; The second intelligent agent determines at least one cause of material shortage corresponding to the historical material shortage record based on the extended data; Based on the inventory analysis results and the reasons for material shortage, determine the kitting analysis results corresponding to the target material.

2. The fitting analysis method according to claim 1, characterized in that, The first intelligent agent is connected to the resource planning data interface. After the first intelligent agent is acquired and determines the demand parameters based on the procurement data and the production planning data, the method further includes: At least one status change parameter of the sub-item material is obtained through the resource planning data interface; The quantity of each sub-item material is determined based on the status change parameters; When the quantity of any of the aforementioned sub-item materials decreases, a secondary analysis instruction is generated for the target material; Based on the secondary analysis command, update the relevant data of the target material; Based on the updated relevant data, the results of the fitting analysis were re-determined.

3. The fitting analysis method according to claim 1, characterized in that, The step of determining the kitting analysis result corresponding to the target material based on the inventory analysis result and the reason for the material shortage includes: Obtain supply data related to the multiple sub-item materials; Based on the supply data, determine whether at least one sub-item material has corresponding in-transit supply information; If the in-transit supply information exists, a third intelligent agent is obtained, and the third intelligent agent is connected to the logistics data interface; The transportation status of at least one of the sub-items in the in-transit supply information is determined by the third intelligent agent. Determine at least one of the in-transit transportation parameters of the sub-item material based on the logistics data interface; The material in transit data is determined based on the transportation status and the in-transit transportation parameters. Based on the inventory analysis results, the reasons for material shortages, and the material in transit data, determine the kitting analysis results corresponding to the target material.

4. The fitting analysis method according to claim 3, characterized in that, The step of determining the inventory analysis result corresponding to the target material based on the demand parameters and the inventory data includes: Input the bill of materials, inventory data, procurement data, and production plan data of the target material into the first intelligent agent; The required quantity of each sub-item is determined according to the order of arrangement of the multiple sub-items in the bill of materials; Obtain the available inventory quantity of each of the sub-items in the inventory data; The quantity of each sub-item material in transit is determined based on the procurement data and the in-transit transportation parameters; The required quantity of each sub-item material is compared with the available inventory quantity and the quantity in transit to determine the comparison result; The shortage quantity of each of the sub-items is determined based on the comparison results; Summarize the shortage quantities of each of the aforementioned sub-items to generate an inventory analysis result that includes a detailed shortage list and an overall availability assessment.

5. The fitting analysis method according to claim 4, characterized in that, After determining the shortage quantity of each of the sub-items based on the comparison results, the method further includes: Determine the required quantity of each of the sub-items based on the production plan data; The availability status of each of the sub-items is determined based on the shortage quantity and the demand quantity; The query status of each sub-item material is determined based on the completeness status, so as to lock the sub-item material that meets the production requirements corresponding to the production plan data.

6. The fitting analysis method according to any one of claims 1 to 5, characterized in that, After determining the kitting analysis result corresponding to the target material based on the inventory analysis result and the reason for the material shortage, the method further includes: Acquiring the fourth intelligent agent; At least one of the sub-items identified as being in a complete set state in the inventory analysis results is marked as a material to be monitored and input into the fourth intelligent agent; The fourth intelligent agent establishes a status monitoring system for the material to be monitored. When a status change event is detected by the status listener for the monitored material, the status change parameters are obtained. The status change event includes at least one of the following: inventory is reserved and locked by other demands, scrapped during production, or fails quality inspection. Determine whether the status change parameter causes the monitored material to become in a shortage state; If the situation is determined to be in a shortage state, a new round of kitting analysis for the target material is triggered to redetermine the kitting analysis results.

7. A kitting analysis apparatus, characterized in that, include: The request receiving module is used to receive natural language requests for target materials. The natural language requests are used to query the matching status of the target materials, and the target materials include multiple sub-items. The request parsing module is used to parse the natural language request to determine the analysis instructions related to the target material; The data query module is used to query and obtain relevant data of the target material based on the analysis command. The relevant data includes at least the bill of materials, inventory data, procurement data and production plan data of multiple sub-items. The demand determination module is used to acquire a first intelligent agent, which determines the demand parameters based on the procurement data and the production plan data. The inventory analysis module is used to determine the inventory analysis results corresponding to the target material based on the demand parameters and the inventory data. The shortage detection module is used to detect a second intelligent agent when the shortage of the sub-item material is determined based on the inventory analysis results. The data acquisition module is used to acquire at least one piece of extended data related to the target material; The history module is used to retrieve at least one type of historical material shortage record; The cause analysis module is used by the second intelligent agent to determine at least one cause of material shortage corresponding to the historical material shortage record based on the extended data; The kitting analysis module is used to determine the kitting analysis result corresponding to the target material based on the inventory analysis result and the reason for the material shortage.

8. An electronic device, characterized in that, It includes a processor, a memory, and a program or instructions stored in the memory and executable on the processor, wherein the program or instructions, when executed by the processor, implement the steps of the matching analysis method as described in any one of claims 1 to 6.

9. A readable storage medium, characterized in that, The readable storage medium stores a program or instructions that, when executed by a processor, implement the steps of the kitting analysis method as described in any one of claims 1 to 6.

10. A chip, characterized in that, The chip includes a processor and a communication interface, the communication interface being coupled to the processor, the processor being used to run programs or instructions to implement the steps of the kitting analysis method as described in any one of claims 1 to 6.

Images