Digital intelligent management system and method for unmanned aerial vehicle research, development and manufacturing enterprises

Abstract

The invention relates to the technical field of digital intelligent management, in particular to a digital intelligent management system and method for an unmanned aerial vehicle research, development and manufacturing enterprise, and the system comprises a knowledge management module, a data embedding processing module, an LLM center model, an AI agent, an enterprise resource planning module, a user interaction feedback module, and a security scale module. The knowledge management module is used for providing an internal knowledge base, encrypting authority management and real-time updating; the data embedding processing module is used for vector database and multi-mode embedding; the system realizes operations such as data input, storage, learning, calling, updating and the like to realize an intelligent working system, staff in a company exchange information with zero delay, and by combining intelligent auxiliary analysis of data, misjudgment caused by redundant and repeated invalid delay and information mismatching is avoided, and the working efficiency is improved. According to the method, internal information circulation of a company is effectively improved, the information island problem of each department is solved, the data secrecy capability in an enterprise is improved, and the management risk early warning capability is enhanced.

Description

Technical Field

[0001] The present invention relates to the field of digital intelligent management technology, and specifically to a digital intelligent management system and method for unmanned aerial vehicle (UAV) research and development and manufacturing enterprises. Background Art

[0002] In company management, due to the difficulty of timely communication among multiple departments, difficulty in personnel coordination, and long-term dynamic storage of information in actual operation, the company's operations will be delayed in decision-making, low work efficiency, and even cause negative situations such as internal conflicts. Traditional enterprise resource planning systems can integrate the information flow and business processes of various departments within the enterprise. Through a unified database, the ERP system integrates multiple business modules such as finance, manufacturing, supply chain, sales, procurement, and human resources into one platform to improve the company's operational efficiency and decision-making level.

[0003] After searching, it was found that the invention patent with application publication number CN117709452A provides a set of enterprise intelligent body construction and sharing platform and method, including language model module, multimodal knowledge reading module, knowledge base management module, knowledge graph interactive semi-automatic construction module, knowledge graph module, vector database module, AI Agent intelligent body module, LangChai n framework module and knowledge security and sharing module. This invention, through the powerful understanding and generation capabilities of the large language model, helps enterprises to automatically build and maintain knowledge bases, and helps developers quickly build applications based on the large language model. At the same time, this new knowledge base solution can handle the knowledge needs of enterprises more efficiently and flexibly, and can achieve more powerful and intelligent knowledge interaction and utilization. This patent is based on the large language model, AI The shared platform is described in terms of agent intelligent bodies, vector databases, etc., and a general description is given of how to help enterprises automatically build and maintain knowledge bases, and how to quickly build applications based on large language models. However, this general architecture is not suitable for enterprises that integrate production and research based on drones. No special architecture design has been carried out for this purpose, and it does not involve the overall integrated management of production, manufacturing, R&D and testing, authority management during collaboration, and high-risk event warning management. The actual business problems cannot be solved from the perspective of architectural design.

[0004] After searching, the invention patent with application publication number CN117455186A was found, which proposed an enterprise management method and system based on artificial intelligence. The enterprise management method based on artificial intelligence includes: extracting the task flow chart corresponding to the enterprise management task, extracting the task processing logic and the departments involved in the task processing through the task flow chart; setting up data management nodes for the departments involved in each enterprise management task; real-time monitoring of whether new enterprise management tasks appear, and when new enterprise management tasks appear, using the data management nodes corresponding to each department to perform data management and task process monitoring on the task processing of the department. The patent is based on enterprise task flow, department data management nodes, etc. for monitoring, and to a certain extent realizes intelligent management. However, its design and module do not involve strong artificial intelligence. In essence, it still belongs to a process-based management module, and fails to introduce the empowerment of management by intelligent agents and LLM.

[0005] Therefore, general drone R&D / manufacturing companies will generate a lot of data, reports, reviews and emails during the R&D testing and manufacturing stages. These contents and information involve many stakeholders, and the interaction logic is complex and secretive. It is very easy to have risks and loopholes in the management process. Special personnel are required to be responsible for the long-term sorting, archiving, monitoring, review, verification and allocation of various types of information. The collection, management and review of information bring a lot of extra workload to various grassroots departments. Therefore, the company management based on the ERP system often has the following problems in actual implementation:

[0006] A. Information storage may be lost over time or updated in a timely manner. Different departments need to spend a lot of time searching and calibrating, and there may be information errors.

[0007] B. Unnecessary waste of resources or wrong solutions due to information asymmetry

[0008] C. Failure to quickly resolve user issues may lead to a decrease in customer satisfaction

[0009] D. Due to the lack of rapid communication between different departments, some resolutions cannot be implemented quickly.

[0010] In summary, it is necessary to propose a digital intelligent management system and method for UAV R&D and manufacturing enterprises to solve the above problems. Summary of the invention

[0011] The purpose of the present invention is to provide a digital intelligent management system and method for drone R&D and manufacturing enterprises to solve the problems raised in the above background technology.

[0012] To achieve the above object, the present invention provides the following technical solutions:

[0013] A digital intelligent management system for drone R&D and manufacturing companies, including knowledge management module, data embedding processing module, LLM center model, AI agent, enterprise resource planning module, user interaction feedback module, and safety compliance module;

[0014] The knowledge management module is used to provide internal knowledge base, encryption authority management and real-time update;

[0015] The data embedding processing module is used for vector database and multimodal embedding;

[0016] The LLM core model is used for natural language understanding, decision support and providing intelligent customer service;

[0017] The AI ​​agent is used for production management, inventory management and quality control;

[0018] The enterprise resource planning module is used for project management, supply chain coordination and market sales analysis;

[0019] The user interaction feedback module is used for user interaction, feedback optimization and providing education and training resources to employees;

[0020] The security compliance module is used for data encryption, compliance monitoring and risk assessment.

[0021] Preferably, the knowledge management module further includes an internal knowledge base, an encryption authority management unit and a real-time update unit;

[0022] The internal knowledge base is used to store documents and data on drone design, testing, and manufacturing;

[0023] The encryption rights management unit is used to ensure that only authorized personnel can access sensitive information and protect intellectual property rights and business secrets;

[0024] The real-time updating unit is used to automatically update the knowledge base content to ensure the timeliness and accuracy of the information.

[0025] Preferably, the data embedding processing module further includes a multimodal embedding unit and a vector database;

[0026] The multimodal embedding unit is used to process and convert text, image, video and audio data to facilitate machine learning model analysis;

[0027] The vector database is used to store the embedded data in the form of low-dimensional vectors to improve data retrieval and processing speed.

[0028] Preferably, the LLM central model further includes a natural language understanding unit, a decision support unit and an intelligent customer service unit;

[0029] The natural language understanding unit is used to parse engineering documents, emails and chat records to extract key information;

[0030] The decision support unit is used to generate decision suggestions and operation steps based on the extracted information;

[0031] The intelligent customer service unit is used to automatically answer technical support questions and reduce manpower requirements.

[0032] Preferably, the AI ​​agent further includes a production management unit, an inventory management unit and a quality control unit;

[0033] The production management unit is used to monitor the status of the production line, predict material requirements, and schedule production tasks;

[0034] The inventory management unit is used to track inventory levels in real time, automatically place orders for replenishment, and prevent production interruptions;

[0035] The quality control unit detects production quality through machine vision and sensor data for real-time feedback of problems.

[0036] Preferably, the enterprise resource planning module further includes a project management unit, a supply chain coordination unit and a market sales analysis unit;

[0037] The project management unit is used to track R&D progress, coordinate the work of various departments, and optimize resource allocation;

[0038] The supply chain coordination unit is used for real-time data exchange between upstream and downstream enterprises and coordination of supply chain activities;

[0039] The market sales analysis unit is used to analyze market data, predict sales trends, and guide marketing strategies.

[0040] Preferably, the user interaction feedback module further includes an interaction unit, a feedback optimization unit and a teaching and training unit;

[0041] The interactive unit is used to provide an intuitive operation interface for engineers and managers;

[0042] The feedback optimization unit is used to adjust the agent behavior and system response according to user feedback;

[0043] The teaching and training unit is used to provide online training resources to help employees better use the system.

[0044] Preferably, the security compliance module further includes a data encryption unit, a compliance monitoring unit and a risk assessment unit;

[0045] The data encryption unit is used to encrypt the transmission and storage of enterprise data to prevent unauthorized access;

[0046] The compliance monitoring unit is used to ensure that all operations comply with industry standards and regulatory requirements;

[0047] The risk assessment unit is used to regularly assess system security risks and formulate response strategies.

[0048] Based on the above system, the present invention also proposes a digital intelligent management method for drone R&D and manufacturing enterprises, including the following steps:

[0049] S1. Data storage: Users upload meeting minutes, production data, and document work content to the AI ​​platform, which parses, records, and learns through its own multimodal model. The AI ​​platform provides long-term stable storage for data and enables rapid search and call when users need it. As users continue to input into the knowledge base, the AI ​​platform updates the knowledge base in real time by covering previously repeated content. Since the AI ​​platform is deployed locally, company privacy data will not be uploaded to the Internet or leaked to other companies, ensuring company data security.

[0050] S2. Propose a demand: Users ask questions based on the information they need for their work, including drone production estimates. Users input text content into the AI ​​platform: I need to produce 100 drones. When will it be completed? The AI ​​platform model uses text embedding and semantic understanding to analyze what information is needed for the user's question. Take drone production as an example:

[0051] Semantic analysis: "I need to produce 100 drones. When will it be completed?" The model extracts information from the keywords "production", "100 drones", and "time" to retrieve the data from the production department for estimation.

[0052] S3. Decision analysis: After the AI ​​platform retrieves relevant information based on the problem, it reports and analyzes the real-time data related to production, which is specifically manifested as follows:

[0053] "We currently have sufficient inventory for most parts, only 10 cameras are missing. We have notified the purchasing department to place an order as soon as possible."

[0054] "The relevant personnel in the production department have to take leave tomorrow. The estimated delivery date may be delayed by one day."

[0055] “…”

[0056] After analysis, the AI ​​platform will list the current situation of the problem to the user and make decision suggestions based on the content of its own learning and integration. The user will also ask further questions based on the reply content to gradually improve the plan. The content provided by the AI ​​platform helps users to quickly collect and organize information and eliminate the information islands and communication delays between independent departments.

[0057] S4. Customer service response: The system receives after-sales questions from users. The large language model embeds the questions into text, analyzes the semantics, and combines the previously trained data with its own language ability to output the corresponding response.

[0058] When a user asks a question that is not within the scope of the current knowledge base, the AI ​​platform will calm the customer's emotions while informing the relevant professional technicians to provide a manual response;

[0059] After the technicians provide professional answers, the AI ​​platform will record, update and learn the questions and answers of this after-sales conversation to enrich its own knowledge base and achieve a richer and more comprehensive database.

[0060] Compared with the prior art, the beneficial effects of the present invention are: the system can realize a series of operations such as data input, storage, learning, calling, and updating to realize an intelligent work system. Employees within the company can exchange information with zero delay, and combined with intelligent auxiliary analysis of data, one-stop work content can be realized, avoiding redundant and repetitive invalid delays and misjudgments caused by information mismatch. With the cooperation of this system, work can be completed in a timely, accurate, and comprehensive manner. Customers can quickly obtain relevant after-sales services, thereby improving user satisfaction and product brand reputation to a certain extent. The company's internal privacy data will also be securely protected, and the database will be continuously expanded, bringing more possibilities to the future data market. Therefore, the present invention effectively improves the flow of information within the company, breaks through the problem of information islands in various departments, improves data confidentiality within the enterprise, and enhances management risk early warning capabilities. BRIEF DESCRIPTION OF THE DRAWINGS

[0061] Figure 1 This is the architecture diagram of the digital intelligent management system for drone R&D and manufacturing enterprises of the present invention;

[0062] Figure 2 This is a flow chart of the digital intelligent management method for drone R&D and manufacturing enterprises of the present invention. DETAILED DESCRIPTION

[0063] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0064] See also Figure 1 , the present invention provides a technical solution:

[0065] A digital intelligent management system for drone R&D and manufacturing companies, including knowledge management module, data embedding processing module, LLM center model, AI agent, enterprise resource planning module, user interaction feedback module, and safety compliance module;

[0066] The knowledge management module is used to provide internal knowledge base, encryption rights management and real-time updates;

[0067] In the specific implementation process, the knowledge base provided by the knowledge management module refers to the knowledge base system established within the company, based on the company's own knowledge system, and with strong confidentiality;

[0068] Different from the knowledge base in the public domain, it focuses more on the knowledge and data that are unique to the company and have differentiating significance for competition;

[0069] Combining private domain knowledge base with intelligent management means integrating the company's internal expertise and data with advanced intelligent technologies to provide highly customized support for the company's management decisions and operations;

[0070] Another important function of private domain knowledge base is information protection;

[0071] Based on the characteristics of the company's sensitive information and knowledge, the system will use encryption, permission management and other means to ensure that sensitive information is well protected;

[0072] Only authorized personnel can access relevant data and knowledge, thereby strengthening company information security on an intelligent basis;

[0073] The data embedding processing module is used for vector database and multimodal embedding;

[0074] Specifically, the working principle of embedding technology is to map data into a low-dimensional space, representing complex data structures, such as text, images, and even audio, in the form of low-dimensional vectors;

[0075] These vectors are encoded into continuous, meaningful representations in a low-dimensional space, making data processing more efficient;

[0076] The LLM center model is used for natural language understanding, decision support and intelligent customer service. In actual application, the LLM center model plays an important role in the company's intelligent management;

[0077] Its working principle is based on deep learning and natural language processing technology. It builds a powerful language model through large-scale corpus training to understand and generate natural language text.

[0078] LLM can help companies understand and analyze massive amounts of text data, including documents, emails, chat records, etc., extract useful information from them, and provide in-depth information support for decision-making;

[0079] The intelligent customer service system built on LLM can realize natural language communication and exchange, help companies conduct intelligent dialogues and decision-making support with customers, employees, etc., and provide more convenient and efficient services and intelligent data analysis;

[0080] AI agents are used for production management, inventory management, and quality control. Specifically, AI agents are intelligent systems implemented through artificial intelligence technology, which have learning, reasoning, and decision-making capabilities, and can help companies achieve automation and intelligence in decision-making, operations, and customer service.

[0081] Based on machine learning, deep learning and other technologies, it can understand and respond to diverse management scenarios through learning and pattern recognition of large-scale data;

[0082] Agents can access information, analyze data, make decisions, and even interact with employees and customers in real time;

[0083] Ability to continuously optimize itself based on accumulated data and experience to adapt to changing management needs;

[0084] Enterprise resource planning modules are used for project management, supply chain coordination, and market sales analysis;

[0085] The user interaction feedback module is used for user interaction, feedback optimization and providing education and training resources to employees;

[0086] The security and compliance module is used for data encryption, compliance monitoring, and risk assessment;

[0087] In summary, the AI ​​intelligent management system combines the above parts to establish a safe, stable, accurate, rapid, intelligent and humanized company management system;

[0088] Users can upload proprietary knowledge to the system in a multimodal form for long-term storage and updating. These data are converted into low-dimensional vectors through the embedding model to form a vector database, which is then input into the large language model for enterprise-specific fine-tuning. The AI ​​agent will automatically retrieve information from the database according to the user's instructions and integrate the large language model to form a complete, accurate and reliable decision, opinion or response.

[0089] On the enterprise operation side, the AI ​​intelligent system can grasp the work progress of each department in real time and send out alerts to department heads in a timely manner;

[0090] If the production progress between departments is not as expected, it will be updated to the upstream and downstream departments simultaneously, and solutions will be recommended based on the work arrangements of other departments such as R&D progress, project scheduling, supply chain progress, marketing, etc.

[0091] Within a department, if there are functional changes or interface changes caused by version updates between R&D personnel, the AI ​​intelligent system can grasp the latest progress based on the codes archived by R&D personnel every day, and promptly inform upstream and downstream personnel of the project of the changed interface information;

[0092] It should also be noted that the knowledge management module of this system also includes an internal knowledge base, an encryption authority management unit and a real-time update unit. The internal knowledge base is used to store documents and data on drone design, testing and manufacturing. The encryption authority management unit is used to ensure that only authorized personnel can access sensitive information and protect intellectual property rights and trade secrets. The real-time update unit is used to automatically update the knowledge base content to ensure the timeliness and accuracy of the information.

[0093] It should also be noted that the data embedding processing module of the present system also includes a multimodal embedding unit and a vector database. The multimodal embedding unit is used to process and convert text, image, video and audio data to facilitate machine learning model analysis. The vector database is used to store the embedded data in the form of low-dimensional vectors to improve data retrieval and processing speed.

[0094] It should also be noted that the LLM center model of this system also includes a natural language understanding unit, a decision support unit and an intelligent customer service unit. The natural language understanding unit is used to parse engineering documents, emails and chat records to extract key information. The decision support unit is used to generate decision suggestions and operation steps based on the extracted information.

[0095] Intelligent customer service units are used to automatically answer technical support questions, reducing manpower requirements;

[0096] It should also be noted that the AI ​​agent of this system also includes a production management unit, an inventory management unit and a quality control unit. The production management unit is used to monitor the status of the production line, predict material requirements, and schedule production tasks. The inventory management unit is used to track inventory levels in real time, automatically place orders for replenishment, and prevent production interruptions. The quality control unit detects production quality through machine vision and sensor data and is used to provide real-time feedback on problems.

[0097] It should also be noted that the enterprise resource planning module of this system also includes a project management unit, a supply chain coordination unit and a market sales analysis unit. The project management unit is used to track R&D progress, coordinate the work of various departments, and optimize resource allocation. The supply chain coordination unit is used to exchange data between upstream and downstream companies in real time and coordinate supply chain activities. The market sales analysis unit is used to analyze market data, predict sales trends, and guide marketing strategies.

[0098] It should also be noted that the user interaction feedback module of this system also includes an interaction unit, a feedback optimization unit and a training unit. The interaction unit is used to provide an intuitive operation interface for engineers and managers, the feedback optimization unit is used to adjust the agent behavior and system response according to user feedback, and the training unit is used to provide online training resources to help employees better use the system.

[0099] It should also be noted that the security compliance module of this system also includes a data encryption unit, a compliance monitoring unit and a risk assessment unit. The data encryption unit is used to encrypt the transmission and storage of enterprise data to prevent unauthorized access. The compliance monitoring unit is used to ensure that all operations comply with industry standards and regulatory requirements. The risk assessment unit is used to regularly assess system security risks and formulate response strategies.

[0100] See also Figure 2 Based on the above system, the present invention also proposes a digital intelligent management method for drone R&D and manufacturing enterprises. Specifically, the steps are as follows:

[0101] S1. Data storage:

[0102] Users can upload work contents such as meeting minutes, production data, and instruction documents to the system, which will analyze, record, and learn through its own multimodal model.

[0103] This system can provide long-term stable storage of data and realize rapid search and call when users need it;

[0104] As users continue to input into the knowledge base, the system can overwrite previously repeated content to update the knowledge base in real time;

[0105] Since this system is deployed locally, the company's private data will not be uploaded to the Internet or leaked to other companies, ensuring the security of company data;

[0106] S2. Make a request:

[0107] The user asks questions based on the information required for his or her work content. For example, for drone production estimation, the user enters text content into the system: I need to produce 100 drones. When can it be completed?

[0108] This system model uses text embedding and semantic understanding to analyze what information is needed for the questions raised by users. Take drone production as an example:

[0109] Semantic analysis: "I need to produce 100 drones. When can I finish the production?" The model extracts information from keywords such as 'production', '100 drones', and 'time' to retrieve the data from the production department for estimation.

[0110] S3.Decision Analysis:

[0111] After the system retrieves relevant data according to the problem, it reports and analyzes the real-time data related to production, as shown below:

[0112] "We currently have sufficient inventory for most parts, only 10 cameras are missing. We have notified the purchasing department to place an order as soon as possible."

[0113] "The relevant personnel in the production department have to take leave tomorrow. The estimated delivery date may be delayed by one day."

[0114] “…”

[0115] After the analysis, the system will list the current situation of the problem to the user and make decision suggestions based on the content of its own learning and integration. The user can also ask further questions based on the response content to gradually improve the plan;

[0116] The content provided by this system helps users to quickly collect and organize information, eliminating the information islands and communication delays that exist between independent departments;

[0117] S4.Customer service reply:

[0118] The working principle is similar to the process of facilitating collaboration between companies;

[0119] When receiving after-sales questions from users on WeChat official accounts and other systems, the big language model will analyze the questions by embedding text and understanding semantics, and combine previously trained data with its own language ability to output corresponding responses;

[0120] When the questions raised by users are not within the scope of the current knowledge base, this system can inform relevant professional technicians to provide manual responses while stabilizing the customer's emotions;

[0121] After the technicians provide professional answers, the system will record, update and learn the questions and answers of this after-sales conversation to enrich its own knowledge base and achieve a richer and more comprehensive database;

[0122] In summary, through the above working process, this system can realize a series of operations such as data input, storage, learning, calling, and updating to realize an intelligent working system;

[0123] Internal employees of the company can exchange information without delay, and combined with intelligent data analysis, work content can be completed in one stop, avoiding redundant and repetitive invalid delays and misjudgments caused by mismatched information;

[0124] Complete work in a timely, accurate, and comprehensive manner with the collaboration of this system;

[0125] Customers can quickly get relevant after-sales services, which can improve user satisfaction and product brand reputation to a certain extent;

[0126] In addition, the company's internal privacy data will be securely protected and the database will be continuously expanded, bringing more possibilities to the future data market.

[0127] Although embodiments of the present invention have been shown and described, it will be appreciated by those skilled in the art that various changes, modifications, substitutions and variations may be made to the embodiments without departing from the principles and spirit of the present invention, and that the scope of the present invention is defined by the appended claims and their equivalents.

Claims

1. A digital intelligent management system for drone R&D and manufacturing enterprises, characterized by: Including knowledge management module, data embedding processing module, LLM center model, AI agent, enterprise resource planning module, user interaction feedback module, and security compliance module; The knowledge management module is used to provide internal knowledge base, encryption authority management and real-time update; The data embedding processing module is used for vector database and multimodal embedding; The LLM core model is used for natural language understanding, decision support and providing intelligent customer service; The AI ​​agent is used for production management, inventory management and quality control; The enterprise resource planning module is used for project management, supply chain coordination and market sales analysis; The user interaction feedback module is used for user interaction, feedback optimization and providing education and training resources to employees; The security compliance module is used for data encryption, compliance monitoring and risk assessment.

2. The digital intelligent management system for drone R&D and manufacturing enterprises according to claim 1 is characterized by: The knowledge management module also includes an internal knowledge base, an encryption authority management unit and a real-time update unit; The internal knowledge base is used to store documents and data on drone design, testing, and manufacturing; The encryption rights management unit is used to ensure that only authorized personnel can access sensitive information and protect intellectual property rights and business secrets; The real-time updating unit is used to automatically update the knowledge base content to ensure the timeliness and accuracy of the information.

3. The digital intelligent management system for drone R&D and manufacturing enterprises according to claim 2 is characterized by: The data embedding processing module also includes a multimodal embedding unit and a vector database; The multimodal embedding unit is used to process and convert text, image, video and audio data to facilitate machine learning model analysis; The vector database is used to store the embedded data in the form of low-dimensional vectors to improve data retrieval and processing speed.

4. The digital intelligent management system for drone R&D and manufacturing enterprises according to claim 3 is characterized by: The LLM center model also includes a natural language understanding unit, a decision support unit and an intelligent customer service unit; The natural language understanding unit is used to parse engineering documents, emails and chat records to extract key information; The decision support unit is used to generate decision suggestions and operation steps based on the extracted information; The intelligent customer service unit is used to automatically answer technical support questions and reduce manpower requirements.

5. The digital intelligent management system for drone R&D and manufacturing enterprises according to claim 4 is characterized by: The AI ​​agent also includes a production management unit, an inventory management unit and a quality control unit; The production management unit is used to monitor the status of the production line, predict material requirements, and schedule production tasks; The inventory management unit is used to track inventory levels in real time, automatically place orders for replenishment, and prevent production interruptions; The quality control unit detects production quality through machine vision and sensor data for real-time feedback of problems.

6. The digital intelligent management system for drone R&D and manufacturing enterprises according to claim 5 is characterized by: The enterprise resource planning module also includes a project management unit, a supply chain coordination unit and a market sales analysis unit; The project management unit is used to track R&D progress, coordinate the work of various departments, and optimize resource allocation; The supply chain coordination unit is used for real-time data exchange between upstream and downstream enterprises and coordination of supply chain activities; The market sales analysis unit is used to analyze market data, predict sales trends, and guide marketing strategies.

7. The digital intelligent management system for drone R&D and manufacturing enterprises according to claim 6 is characterized by: The user interaction feedback module also includes an interaction unit, a feedback optimization unit and a teaching and training unit; The interactive unit is used to provide an intuitive operation interface for engineers and managers; The feedback optimization unit is used to adjust the agent behavior and system response according to user feedback; The teaching and training unit is used to provide online training resources to help employees better use the system.

8. The digital intelligent management system for drone R&D and manufacturing enterprises according to claim 7 is characterized by: The security compliance module also includes a data encryption unit, a compliance monitoring unit and a risk assessment unit; The data encryption unit is used to encrypt the transmission and storage of enterprise data to prevent unauthorized access; The compliance monitoring unit is used to ensure that all operations comply with industry standards and regulatory requirements; The risk assessment unit is used to regularly assess system security risks and formulate response strategies.

9. A digital intelligent management method for drone R&D and manufacturing enterprises, according to the digital intelligent management system for drone R&D and manufacturing enterprises according to claims 1-8, characterized in that: The following steps are involved: S1. Data storage: Users upload meeting minutes, production data, and document work content to the AI ​​platform, which parses, records, and learns through its own multimodal model. The AI ​​platform provides long-term stable storage for data and enables rapid search and call when users need it. As users continue to input into the knowledge base, the AI ​​platform updates the knowledge base in real time by covering previously repeated content. Since the AI ​​platform is deployed locally, company privacy data will not be uploaded to the Internet or leaked to other companies, ensuring company data security. S2. Propose a demand: Users ask questions based on the information they need for their work, including drone production estimates. Users input text content into the AI ​​platform: I need to produce 100 drones. When will it be completed? The AI ​​platform model uses text embedding and semantic understanding to analyze what information is needed for the user's question. Take drone production as an example: Semantic analysis: "I need to produce 100 drones. When will it be completed?" The model extracts information from the keywords "production", "100 drones", and "time" to retrieve the data from the production department for estimation. S3. Decision analysis: After the AI ​​platform retrieves relevant information based on the problem, it reports and analyzes the real-time data related to production, which is specifically manifested as follows: "We currently have sufficient stock of most parts, only 10 cameras are missing. We have notified the purchasing department to place an order as soon as possible." "The relevant personnel in the production department have to take leave tomorrow. The estimated delivery date may be delayed by one day." "……" After analysis, the AI ​​platform will list the current situation of the problem to the user and make decision suggestions based on the content of its own learning and integration. The user will also ask further questions based on the reply content to gradually improve the plan. The content provided by the AI ​​platform helps users to quickly collect and organize information and eliminate the information islands and communication delays between independent departments. S4. Customer service response: The system receives after-sales questions from users. The large language model embeds the questions into text, analyzes the semantics, and combines the previously trained data with its own language ability to output the corresponding response. When a user asks a question that is not within the scope of the current knowledge base, the AI ​​platform will calm the customer's emotions while informing the relevant professional technicians to provide a manual response; After the technicians provide professional answers, the AI ​​platform will record, update and learn the questions and answers of this after-sales conversation to enrich its own knowledge base and achieve a richer and more comprehensive database.

Images