Intelligent product after-sales management method and system

Abstract

The invention relates to the technical field of product management, in particular to an intelligent product after-sales management method and system. The method comprises the following steps: obtaining an intelligent product after-sales problem record set, wherein the intelligent product after-sales problem record set comprises at least one or more intelligent product after-sales problem records; performing customer demand degree quantification on the after-sales problem record of the intelligent product, and generating after-sales demand emergency degree data of the intelligent product; comparing the after-sales demand emergency degree data of the intelligent product with a preset standard demand emergency degree threshold value, and generating priority response demand data and delay response demand data; and performing fault processing association analysis on the after-sales problem record set of the intelligent product through the priority response demand data and the delay response demand data to generate a product fault processing automatic solution. According to the invention, through demand quantification, fault correlation analysis, feedback diagnosis and trend evaluation, the after-sales management efficiency and time efficiency are improved.

Description

Technical Field

[0001] The present invention relates to the technical field of product management, and in particular to an intelligent product after-sales management method and system. Background Art

[0002] With the rapid development of modern science and technology, intelligent products have been widely used in all walks of life. From early single-function electronic devices to today's multifunctional intelligent devices, the complexity of products and the diversity of user needs have increased significantly. Traditional after-sales management methods, such as manual repair, telephone support and paper records, are gradually unable to meet users' expectations for efficient services due to their inefficiency, slow response and error-proneness. In order to optimize the after-sales management process, companies have begun to introduce information technology, such as customer relationship management systems (CRM) and enterprise resource planning systems (ERP), which achieve partial automation of the process by digitally storing and managing customer information. However, this method is still based on static data processing and cannot dynamically respond to the large amount of real-time data requirements of intelligent products. In recent years, with the maturity of the Internet of Things (IoT), artificial intelligence (AI) and big data analysis technologies, the after-sales management of intelligent products has entered a new stage. However, at present, traditional after-sales processing is mainly based on manual analysis, which is prone to missed diagnosis or misjudgment. At the same time, in the existing after-sales process, failure processing often lacks effective feedback and diagnosis mechanisms, resulting in low efficiency of repeated repairs, which in turn leads to low efficiency and time efficiency of after-sales management. Summary of the invention

[0003] Based on this, it is necessary to provide an intelligent product after-sales management method and system to solve at least one of the above technical problems.

[0004] To achieve the above object, an intelligent product after-sales management method is provided, the method comprising the following steps: Step S1: obtaining a set of after-sales problem records of intelligent products, wherein the set of after-sales problem records of intelligent products includes at least one or more after-sales problem records of intelligent products; quantifying customer demand for after-sales problem records of intelligent products, and generating urgency data of after-sales demand of intelligent products; Step S2: Compare the intelligent product after-sales demand urgency data with the preset standard demand urgency threshold to generate priority response demand data and delayed response demand data; perform fault processing association analysis on the intelligent product after-sales problem record set through the priority response demand data and the delayed response demand data to generate a product fault processing automation solution; perform the first remote processing feedback result collection on the product fault processing automation solution to obtain the first after-sales failure processing result; Step S3: Based on the after-sales failure processing result, a functional damage assessment is performed on the intelligent product to generate product functional damage assessment data; based on the product functional damage assessment data, an interventional fault diagnosis is performed on the after-sales failure processing result to generate interventional fault diagnosis data; a second remote processing feedback result is collected through the interventional fault diagnosis data to obtain a second after-sales failure processing result; after-sales maintenance scheduling is performed on the second after-sales failure result, thereby generating after-sales maintenance service scheduling data; Step S4: Perform a phased after-sales trend analysis on the first after-sales failure processing result, the second after-sales failure processing result, and the after-sales repair service scheduling data to generate phased after-sales trend data; visualize the after-sales quality assessment of the phased after-sales quality trend data to perform intelligent product after-sales management optimization operations.

[0005] The present invention helps to fully grasp the types and urgency of after-sales problems, realize the priority division of problems, ensure the reasonable allocation of resources, and improve the efficiency of after-sales response by obtaining the record set of after-sales problems of intelligent products and quantifying customer needs. Through intelligent analysis to generate priority response and delayed response demand data, and perform fault processing association analysis, the key points of the problem can be accurately identified and automated solutions can be quickly formulated, reducing human intervention and improving the success rate and efficiency of initial remote processing. Functional damage assessment and intervention fault diagnosis are carried out based on the after-sales failure results, and the problem location is further refined. The solution is optimized through multiple rounds of remote processing feedback, and maintenance resources are reasonably scheduled to improve the success rate of maintenance and the efficiency of solving repeated problems. By comprehensively analyzing multiple processing results and scheduling data, generating phased after-sales trend data and visually displaying it, it is helpful to discover the potential laws and trends of after-sales problems, provide data support and decision-making basis for after-sales management optimization, and improve service quality and customer satisfaction. Each step, from the acquisition, analysis, diagnosis to trend evaluation of after-sales problem records, forms a closed-loop management of the entire process, significantly improving the efficiency, accuracy and service quality of after-sales processing, and helping the continuous optimization of after-sales management of intelligent products. Therefore, the present invention improves the efficiency and time efficiency of after-sales management through demand quantification, fault correlation analysis, feedback diagnosis and trend evaluation.

[0006] Preferably, step S1 comprises the following steps: Step S11: obtaining a smart product after-sales problem record set, wherein the smart product after-sales problem record set includes at least one or more smart product after-sales problem records; Step S12: classifying the intelligent product after-sales problem record set into record types to generate text type records and voice type records; performing voice recognition on the intelligent product after-sales problem record according to the voice type record to generate after-sales demand voice recognition data; Step S13: Perform user demand speech speed analysis on the after-sales demand voice recognition data to generate user after-sales demand speech speed data; perform semantic recognition on the intelligent product after-sales problem records through text type records to generate after-sales demand semantic recognition data; Step S14: extract modal particles based on the after-sales demand semantic recognition data to obtain after-sales demand modal particle data; quantify customer demand for intelligent product after-sales problem records through the after-sales demand modal particle data and user after-sales demand speech speed data to generate intelligent product after-sales demand urgency data.

[0007] The present invention distinguishes between text and voice type records, processes different forms of data separately (such as voice recognition and text semantic recognition), makes full use of the information in the after-sales problem records, and improves the comprehensiveness and accuracy of data processing. The speech speed analysis of user needs in voice records can identify the urgency of user expressions; combined with the extraction of modal particles in the text, the user's emotional state and the urgency of needs can be further inferred, providing a key reference for after-sales service. Through the combined analysis of speech speed and modal particles, user needs are quantified and after-sales demand urgency data is generated, which can provide a scientific basis for subsequent priority processing and improve after-sales service efficiency. Based on the generated urgency data, the after-sales service team can quickly respond to high-priority needs, optimize resource allocation, achieve more accurate and personalized services, and improve customer satisfaction. Systematically record and analyze after-sales problems to help companies discover potential product defects or service bottlenecks, and provide data support for product improvement and service process optimization.

[0008] Preferably, quantifying customer demand for after-sales problem records of intelligent products by using after-sales demand modal particle data and user after-sales demand speech speed data includes: Perform tone feature analysis on the tone particle data of after-sales demand, so as to obtain customer tone feature data; perform speech rate change feature extraction on the speech rate data of user after-sales demand, so as to obtain user speech rate feature data; Emotional fluctuation analysis is performed based on the user speech speed feature data and the customer tone feature data, thereby obtaining customer emotional fluctuation data; after-sales problem priorities are preliminarily quantified based on customer emotional fluctuation data, thereby obtaining quantitative data on after-sales problem priorities; The after-sales problem records of intelligent products are collected in time series to obtain time series after-sales records; based on the time series after-sales records, a comprehensive demand urgency analysis is performed on the quantitative data of after-sales problem priorities to obtain the after-sales demand urgency data of intelligent products.

[0009] The present invention extracts emotional information (such as impatience, happiness, anxiety, etc.) in the customer's tone through modal particle data, providing in-depth information for evaluating customer emotions. The user's speech speed change feature extraction is used to capture the sense of urgency or the degree of slowness of speech speed in the user's expression, and further improve the urgency judgment of user needs. Emotional fluctuation analysis based on user speech speed characteristics and tone feature data can quantify the change amplitude and trend of user emotions, providing a scientific basis for the priority quantification of after-sales problems. Emotional fluctuation data provides more dimensional information for after-sales service personnel, helping to more accurately judge the urgency of user needs. Using customer emotional fluctuation data, after-sales problems are quickly prioritized, providing a basis for efficient processing. Through the time-series after-sales record collection and combined with priority quantification data, the urgency of the problem is comprehensively analyzed to avoid the deviation caused by single-dimensional evaluation and improve the accuracy of demand urgency judgment. Through intelligent priority evaluation, the after-sales team can quickly respond to high-urgency problems, optimize resource allocation, shorten problem solving time, and thus improve customer satisfaction. Dynamic update and adjustment of priorities can ensure the flexibility and effectiveness of the after-sales service process.

[0010] Preferably, step S2 comprises the following steps: Step S21: Compare the intelligent product after-sales demand urgency data with the preset standard demand urgency threshold. When the intelligent product after-sales demand urgency data is greater than or equal to the preset standard demand urgency threshold, give priority response to the corresponding intelligent product after-sales problem record and generate priority response demand data; Step S22: when the intelligent product after-sales demand urgency data is less than the preset standard demand urgency threshold, a delayed response is performed on the corresponding intelligent product after-sales problem record to generate delayed response demand data; product failure prediction is performed on the intelligent product after-sales problem record set by using the priority response demand data and the delayed response demand data to generate product failure prediction data; Step S23: construct a product knowledge graph; import product failure prediction data into the product knowledge graph to perform failure handling association analysis and generate an automated solution for product failure handling; Step S24: perform the first remote processing feedback data collection on the product fault handling automation solution. When the corresponding remote processing feedback data is not collected, repeat the data collection until the remote processing feedback data is received; divide the feedback results of the remote processing feedback data, eliminate the first after-sales successful processing result, and obtain the first after-sales failed processing result.

[0011] The present invention distinguishes high-priority issues from low-priority issues by comparing the urgency of after-sales demand with the standard threshold. High-priority issues are responded to in a timely manner (generating priority response demand data), ensuring that key issues are handled quickly and improving customer satisfaction. Low-priority issues are responded to in a delayed manner (generating delayed response demand data), rationally arranging resources and reducing operating costs. Priority response demand data and delayed response demand data are analyzed to predict potential product failure problems, so that measures can be taken in advance to avoid further expansion of the problem. The foresight and preventive nature of after-sales service are improved, and customer complaints and potential risks are reduced. By constructing a product knowledge graph, the fault prediction data is associated with existing knowledge, the cause of the fault is quickly located, and the problem diagnosis time is shortened. Based on the association analysis of the knowledge graph, an automated fault handling solution is generated to reduce manual participation and improve processing efficiency and accuracy. A closed-loop feedback mechanism is formed by collecting and classifying the first remote processing feedback data (successful processing and failed processing). The failure processing results are further analyzed and optimized to ensure that the success rate of subsequent processing is gradually improved. When the feedback data is not collected, the system repeats the collection to ensure the integrity and reliability of the data and provide high-quality data support for subsequent decision-making.

[0012] Preferably, predicting product failures for the intelligent product after-sales problem record set by using the priority response demand data and the delayed response demand data includes: Perform after-sales problem time series analysis based on priority response demand data and delayed response demand data, thereby obtaining after-sales problem time series data; perform demand frequency distribution extraction on the after-sales problem time series data, thereby obtaining after-sales problem demand frequency distribution data; The priority response demand data and the delayed response demand data are classified by the frequency distribution data of after-sales problem demand, so as to obtain the priority demand characteristic data and the delayed demand characteristic data; the priority demand characteristic data and the delayed demand characteristic data are clustered by the problem record concentration feature analysis, so as to obtain the problem record feature clustering data; Potential failure mode matching is performed through problem record feature clustering data to obtain potential product failure mode data; failure probability prediction calculation is performed on the intelligent product after-sales problem record set based on the potential product failure mode data to obtain product failure prediction data.

[0013] The present invention can accurately capture the time regularity and frequency characteristics of after-sales problems through time series analysis and demand frequency distribution extraction, thereby improving the prediction accuracy of the trend of after-sales problems. The priority response demand and delayed response demand are classified separately, and the demand characteristics are classified in combination with the frequency distribution data, so that the problem classification results are more targeted, providing a high-quality data basis for subsequent problem clustering and fault analysis. Through problem record feature clustering analysis and potential fault mode matching, the potential rules in the problem records can be effectively identified, and hidden fault modes can be mined, thereby achieving in-depth analysis of complex after-sales problems. Fault probability prediction calculation based on potential fault mode data can provide quantitative risk assessment results for the after-sales team, help formulate countermeasures in advance, and reduce the uncontrollability of fault occurrence. Providing demand feature data with clear priority and potential fault prediction data allows the after-sales team to respond quickly according to the priority demand characteristics, while reasonably allocating resources to solve delayed demands and improve the efficiency of after-sales problem solving. By analyzing potential fault modes and predicting fault probabilities, feedback data can be provided for product design and manufacturing links, so that common fault points can be improved to improve product reliability and user experience.

[0014] Preferably, step S23 includes the following steps: Step S231: Obtain a product basic information data set and a historical intelligent product after-sales problem record set; extract common failure modes and corresponding solutions from the historical intelligent product after-sales problem record set to generate a historical failure solution data set; use graph structure construction technology to construct an initial product knowledge graph for the product basic information data set and the historical failure solution data set to generate an initial product knowledge graph; Step S232: semantic enhancement processing is performed on the initial product knowledge graph to generate a product knowledge graph; the core elements in the product fault prediction data are analyzed, including potential fault types, occurrence probabilities, impact ranges, and time ranges, to generate a fault prediction factor data set; the fault prediction factor data set is matched with nodes and edges in the product knowledge graph, and the association relationship between the fault prediction data and existing knowledge is established through similarity analysis to generate a product dynamic knowledge graph; Step S233: Perform path analysis on the product dynamic knowledge graph to extract functional modules, components and historical processing methods related to potential faults, thereby building a fault processing association model; use the fault processing association model to perform fault processing association mining on product fault prediction data to generate an initial product fault processing automation solution; Step S234: Perform step-by-step and logic verification on the initial product fault handling automation solution, thereby generating a product fault handling automation solution.

[0015] The present invention integrates the product basic information data set and the historical fault solution data set, and uses the graph structure construction technology to construct the initial product knowledge graph, which provides comprehensive and systematic knowledge support for product problem analysis and solutions. Semantic enhancement processing makes the knowledge graph more accurate in expressing product information and fault relationships, and provides high-quality semantic association data for subsequent analysis. The fault prediction element data set analyzes the core elements of potential faults (such as type, probability, impact range and time range), and establishes the association relationship between fault prediction data and knowledge graph through similarity analysis, thereby generating a dynamic knowledge graph and realizing the knowledge interpretation of the prediction results. The real-time update of the dynamic knowledge graph enhances the timeliness and accuracy of knowledge and adapts to the changes in product status and fault mode. Through the path analysis of the dynamic knowledge graph, the functional modules, components and historical processing methods related to potential faults are extracted, and the fault processing association model is constructed, which provides a data-driven logical basis for the analysis and solution of complex faults. The fault processing association model is used for association mining to generate an initial automated solution, and through the step-by-step and logical verification of the solution, an efficient and accurate automated solution is finally output, which greatly reduces the need for manual intervention. By integrating fault prediction data with dynamic knowledge graphs, the weights of associated nodes in the semantically enhanced knowledge graph are updated, forming a dynamic optimization closed-loop mechanism for the knowledge graph, which continuously enhances the fault handling capability.

[0016] Preferably, step S3 comprises the following steps: Step S31: performing a functional damage assessment on the intelligent product based on the after-sales failure processing result, and generating product functional damage assessment data; Step S32: performing interventional fault diagnosis on the after-sales failure processing result according to the product function damage assessment data to generate interventional fault diagnosis data; performing a second remote processing feedback data collection based on the interventional fault diagnosis data, and when the corresponding remote processing feedback data is not collected, repeating the data collection until the remote processing feedback data is received; Step S33: dividing the feedback results of the remote processing feedback data, eliminating the second after-sales successful processing result, and obtaining the second after-sales failed processing result; performing after-sales maintenance scheduling on the second after-sales failure result, thereby generating after-sales maintenance service scheduling data.

[0017] The present invention can accurately identify and quantify the functional problems of intelligent products by evaluating the damage of product functions based on the after-sales failure processing results, and provide clear basic data for subsequent fault diagnosis and maintenance. Functional damage assessment provides a systematic perspective for further fault analysis, reducing misjudgment and omission of product problems. Analysis of functional damage assessment data helps to accurately diagnose the root cause of the problem, thereby reducing unnecessary troubleshooting and optimizing the maintenance process. The collection of remote processing feedback data provides a basis for dynamically adjusting the after-sales strategy to ensure that each fault diagnosis can be optimized based on the latest data. The repeated data collection mechanism ensures that the integrity and accuracy of the information can be ensured regardless of whether the feedback is timely. The feedback result division can accurately distinguish between successful and failed after-sales processing, avoiding the interference of erroneous data on after-sales decision-making. By eliminating successful processing results and retaining only failed results, subsequent processing is more focused on actual problems. The generation of after-sales maintenance service scheduling can automatically schedule maintenance resources based on the second after-sales failure processing result, making the maintenance process more efficient and in line with the priority and severity of actual product failures.

[0018] Preferably, performing intervention fault diagnosis on after-sales failure processing results according to product function damage assessment data includes: Mark the damaged modules of intelligent products according to the product function damage assessment data to obtain the damaged module marking data of intelligent products; perform module failure intervention analysis on the after-sales failure processing results based on the damaged module marking data of intelligent products to generate intervention analysis result data; Use product knowledge graph to expand the fault chain of intervention analysis result data to generate extended fault diagnosis data; prioritize fault modes of extended fault diagnosis data, identify optimal processing paths, and generate fault diagnosis path data; perform simulation verification on fault diagnosis path data to generate optimized fault diagnosis data; The optimized fault diagnosis data is integrated into the product knowledge graph for fault extension node mapping, thereby generating intervention fault diagnosis data.

[0019] The present invention accurately identifies and marks the specific damaged modules in the product by analyzing the functional damage assessment data of the intelligent product, which provides a clear direction and focus for subsequent fault diagnosis and reduces the blind investigation of problems. Accurate damaged module marking helps to improve the efficiency of fault diagnosis and avoid unnecessary time wasted on irrelevant modules. Based on the damaged module marking data, the after-sales failure processing results are accurately intervened and analyzed. The generated intervention analysis result data can quickly locate the fault module, and further confirm the cause of the problem through detailed analysis to ensure the pertinence and effectiveness of the fault processing. By extending the fault chain of the intervention analysis results through the product knowledge graph, the potential association between the fault modules can be revealed, thereby identifying the impact range and the source of the fault. This process provides a comprehensive perspective for fault diagnosis and reduces the neglect of other modules by a single module failure. Fault chain expansion helps to discover hidden problems, improve the depth and breadth of fault diagnosis, and avoid missed diagnosis or misdiagnosis. Fault mode priority sorting and optimal processing path identification can sort different fault modes according to the severity and probability of the fault, helping the after-sales team to prioritize high-risk and high-priority faults and ensure that critical faults can be resolved as soon as possible. After identifying the optimal processing path, it can provide the maintenance team with a clear and direct fault resolution route, reduce redundant steps and improve repair efficiency.

[0020] Preferably, step S4 comprises the following steps: Step S41: performing a phased after-sales trend analysis on the first after-sales failure processing result, the second after-sales failure processing result, and the after-sales repair service scheduling data to generate phased after-sales trend data; performing a after-sales quality evaluation on the phased after-sales quality trend data to generate intelligent product after-sales quality evaluation data; Step S42: Visualize the intelligent product after-sales quality assessment data to generate an intelligent product after-sales quality assessment report to perform intelligent product after-sales management optimization operations.

[0021] The present invention generates phased after-sales trend data through comprehensive analysis of the first after-sales failure processing result, the second after-sales failure processing result and the after-sales maintenance service scheduling data, which can comprehensively show the trend changes and distribution patterns of after-sales problems. After-sales quality evaluation is based on the above trend data for in-depth analysis, generates intelligent product after-sales quality evaluation data, accurately reveals the core issues affecting the quality of after-sales service, and provides data support for optimization. By visualizing the after-sales quality evaluation data of intelligent products to generate an evaluation report, the after-sales service performance is presented in an intuitive way, including key indicators such as trend charts, failure rate distribution, service response time and success rate. The data visualization presentation method is easy to understand, which not only provides management with the ability to quickly understand problems, but also facilitates communication and coordination between departments. The after-sales quality evaluation report of intelligent products provides a clear guiding direction for after-sales management optimization and supports the formulation of improvement strategies, such as strengthening reliability testing of key modules, optimizing service processes or enhancing resource allocation. Phased trend analysis helps management predict potential problems, improve the response speed to after-sales needs, deploy resources in advance, and avoid concentrated outbreaks of problems.

[0022] In this specification, an intelligent product after-sales management system is provided, which is used to execute the above-mentioned intelligent product after-sales management method, and the intelligent product after-sales management system includes: The urgency analysis module is used to obtain a set of intelligent product after-sales problem records, wherein the set of intelligent product after-sales problem records contains at least one or more intelligent product after-sales problem records; quantify customer demand for the intelligent product after-sales problem records, and generate intelligent product after-sales demand urgency data; The automated processing module is used to compare the urgency data of after-sales demand of intelligent products with the preset standard demand urgency threshold, and generate priority response demand data and delayed response demand data; perform fault processing association analysis on the after-sales problem record set of intelligent products through the priority response demand data and delayed response demand data, and generate a product fault processing automated solution; perform the first remote processing feedback result collection on the product fault processing automated solution, and obtain the first after-sales failure processing result; The intervention processing module is used to evaluate the functional damage of the intelligent product based on the after-sales failure processing result and generate product functional damage evaluation data; perform intervention fault diagnosis on the after-sales failure processing result according to the product functional damage evaluation data and generate intervention fault diagnosis data; collect the second remote processing feedback result through the intervention fault diagnosis data to obtain the second after-sales failure processing result; perform after-sales maintenance scheduling on the second after-sales failure result, thereby generating after-sales maintenance service scheduling data; The after-sales quality evaluation module is used to perform a phased after-sales trend analysis on the first after-sales failure processing result, the second after-sales failure processing result and the after-sales repair service scheduling data to generate phased after-sales trend data; and to visualize the after-sales quality evaluation of the phased after-sales quality trend data to perform intelligent product after-sales management optimization operations.

[0023] The beneficial effect of the present invention is that the urgency analysis module can clearly distinguish the urgency of different problems by quantifying customer needs for after-sales problem records, thereby assigning a reasonable priority to each problem. This mechanism ensures that key problems can be handled in the shortest time, avoiding customer dissatisfaction caused by processing delays. By comparing with the standard demand urgency threshold, it can ensure that important problems are responded to first, reducing the response time for lower priority problems, thereby improving service efficiency. The automation processing module can reduce manual intervention and reduce the risk of human error by automatically generating an automated solution for product fault processing and providing the first remote processing feedback, while improving the accuracy of fault processing. The automated solution can improve the consistency and efficiency of processing, shorten the processing cycle, ensure that the problem can be effectively solved at the first remote processing, and reduce the need for subsequent manual intervention. The intervention processing module can accurately locate the cause of the fault through a functional damage assessment based on the first after-sales failure processing result and further intervention fault diagnosis. This process can provide a more accurate repair solution by combining historical data with fault chain expansion, reducing unnecessary rework and improving overall maintenance efficiency. Through multiple rounds of remote processing feedback collection, the system can track the progress of fault repair in real time and ensure that the repair plan is verified and improved at multiple stages. After analyzing the after-sales trend data, the after-sales quality assessment module can generate a detailed after-sales quality assessment report, which can help the company's management clearly understand the problem by visually displaying the after-sales service quality and further optimize the after-sales service process. By analyzing the periodic after-sales quality trend data, service bottlenecks or areas of quality decline can be quickly discovered, so that the company can make timely adjustments in strategy and enhance customer trust and satisfaction. Therefore, the present invention improves the efficiency and time efficiency of after-sales management through demand quantification, fault correlation analysis, feedback diagnosis and trend evaluation. BRIEF DESCRIPTION OF THE DRAWINGS

[0024] Figure 1 A schematic diagram of the steps of an intelligent product after-sales management method; Figure 2 for Figure 1 Detailed implementation steps of step S2 in the flowchart; Figure 3 for Figure 1 Detailed implementation steps of step S3 in FIG. The realization of the purpose, functional features and advantages of the present invention will be further explained in conjunction with embodiments and with reference to the accompanying drawings. DETAILED DESCRIPTION

[0025] The following is a clear and complete description of the technical method of the present invention in conjunction with the accompanying drawings. Obviously, the described embodiments are 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 technicians in this field without creative work are within the scope of protection of the present invention.

[0026] In addition, the accompanying drawings are only schematic illustrations of the present invention and are not necessarily drawn to scale. The same reference numerals in the figures represent the same or similar parts, and their repeated description will be omitted. Some of the block diagrams shown in the accompanying drawings are functional entities and do not necessarily correspond to physically or logically independent entities. The functional entities can be implemented in software form, or implemented in one or more hardware modules or integrated circuits, or implemented in different networks and / or processor methods and / or microcontroller methods.

[0027] It should be understood that although the terms "first", "second", etc. may be used herein to describe various units, these units should not be limited by these terms. These terms are used only to distinguish one unit from another. For example, without departing from the scope of the exemplary embodiments, a first unit may be referred to as a second unit, and similarly, a second unit may be referred to as a first unit, and the term "and / or" used herein includes any and all combinations of one or more of the listed associated items.

[0028] To achieve this, please refer to Figures 1 to 3 , an intelligent product after-sales management method, the method comprising the following steps: Step S1: obtaining a set of after-sales problem records of intelligent products, wherein the set of after-sales problem records of intelligent products includes at least one or more after-sales problem records of intelligent products; quantifying customer demand for after-sales problem records of intelligent products, and generating urgency data of after-sales demand of intelligent products; Step S2: Compare the intelligent product after-sales demand urgency data with the preset standard demand urgency threshold to generate priority response demand data and delayed response demand data; perform fault processing association analysis on the intelligent product after-sales problem record set through the priority response demand data and the delayed response demand data to generate a product fault processing automation solution; perform the first remote processing feedback result collection on the product fault processing automation solution to obtain the first after-sales failure processing result; Step S3: Based on the after-sales failure processing result, a functional damage assessment is performed on the intelligent product to generate product functional damage assessment data; based on the product functional damage assessment data, an interventional fault diagnosis is performed on the after-sales failure processing result to generate interventional fault diagnosis data; a second remote processing feedback result is collected through the interventional fault diagnosis data to obtain a second after-sales failure processing result; after-sales maintenance scheduling is performed on the second after-sales failure result, thereby generating after-sales maintenance service scheduling data; Step S4: Perform a phased after-sales trend analysis on the first after-sales failure processing result, the second after-sales failure processing result, and the after-sales repair service scheduling data to generate phased after-sales trend data; visualize the after-sales quality assessment of the phased after-sales quality trend data to perform intelligent product after-sales management optimization operations.

[0029] The present invention helps to fully grasp the types and urgency of after-sales problems, realize the priority division of problems, ensure the reasonable allocation of resources, and improve the efficiency of after-sales response by obtaining the record set of after-sales problems of intelligent products and quantifying customer needs. Through intelligent analysis to generate priority response and delayed response demand data, and perform fault processing association analysis, the key points of the problem can be accurately identified and automated solutions can be quickly formulated, reducing human intervention and improving the success rate and efficiency of initial remote processing. Functional damage assessment and intervention fault diagnosis are carried out based on the after-sales failure results, and the problem location is further refined. The solution is optimized through multiple rounds of remote processing feedback, and maintenance resources are reasonably scheduled to improve the success rate of maintenance and the efficiency of solving repeated problems. By comprehensively analyzing multiple processing results and scheduling data, generating phased after-sales trend data and visually displaying it, it is helpful to discover the potential laws and trends of after-sales problems, provide data support and decision-making basis for after-sales management optimization, and improve service quality and customer satisfaction. Each step, from the acquisition, analysis, diagnosis to trend evaluation of after-sales problem records, forms a closed-loop management of the entire process, significantly improving the efficiency, accuracy and service quality of after-sales processing, and helping the continuous optimization of after-sales management of intelligent products. Therefore, the present invention improves the efficiency and time efficiency of after-sales management through demand quantification, fault correlation analysis, feedback diagnosis and trend evaluation.

[0030] In the embodiment of the present invention, reference Figure 1 The above is a schematic diagram of the steps of an intelligent product after-sales management method of the present invention. In this example, the intelligent product after-sales management method includes the following steps: Step S1: obtaining a set of after-sales problem records of intelligent products, wherein the set of after-sales problem records of intelligent products includes at least one or more after-sales problem records of intelligent products; quantifying customer demand for after-sales problem records of intelligent products, and generating urgency data of after-sales demand of intelligent products; In an embodiment of the present invention, after-sales problem records related to intelligent products are collected through the enterprise after-sales service system and customer feedback platform (such as online customer service, emails, telephone records, etc.). Data directly related to product failures, usage problems, maintenance requests, etc. are screened out to ensure the integrity and accuracy of the after-sales problem record set. The collected records are standardized, including deduplication, format unification and semantic cleaning, to generate a structured after-sales problem record set. Natural language processing (NLP) technology is used to analyze the text content of the after-sales problem records and extract key fields that reflect the urgency of customer needs (such as "immediately", "serious", "unavailable", etc.). The frequency of occurrence and contextual semantic relationships of keywords are calculated to generate preliminary customer demand measurement indicators. Define multiple evaluation dimensions that affect the urgency of demand, such as: problem type weight: product function loss, performance problems, appearance problems, etc. Customer level weight: VIP customers, ordinary customers, etc. Problem impact range: single device problems or batch equipment problems. Score each record based on these dimensions to form a preliminary demand quantification model. Calculate the comprehensive score using a weighted model: ;in is the urgency index, is the speech rate data, is the frequency of occurrence of modal particles, and The weight coefficients are respectively (adjusted according to actual applications). The urgency data of after-sales demand for intelligent products is generated according to the calculation results, and the after-sales problem records are graded according to the comprehensive scores (such as "high priority", "medium priority", and "low priority"). The urgency data is bound to the original problem record and stored as an urgency data set for after-sales demand for intelligent products, which is convenient for subsequent processing and decision support.

[0031] Step S2: Compare the intelligent product after-sales demand urgency data with the preset standard demand urgency threshold to generate priority response demand data and delayed response demand data; perform fault processing association analysis on the intelligent product after-sales problem record set through the priority response demand data and the delayed response demand data to generate a product fault processing automation solution; perform the first remote processing feedback result collection on the product fault processing automation solution to obtain the first after-sales failure processing result; In an embodiment of the present invention, by defining the standards for urgency classification, for example: high priority threshold: records with a score of more than 90 points; medium priority threshold: between 60 and 90 points; low priority threshold: records with a score of less than 60 points. The after-sales demand urgency data is compared with the threshold one by one: if the urgency score ≥ the high priority threshold, it is marked as "priority response demand data"; otherwise, it is marked as "delayed response demand data". The classification results are stored in two sets of data sets in layers: the priority response demand data set contains all problem records that require quick response and immediate processing, and the delayed response demand data set contains problem records that can be processed later for subsequent batch analysis and optimization. A rule base is established between product failure types and processing solutions, and the rule base includes: causes of failure (hardware, software, operational problems, etc.); processing steps (remote debugging, parts replacement, on-site repair, etc.). Use data mining methods (such as association rule analysis, decision trees, etc.) combined with the characteristic data of after-sales problem records to identify the association patterns between problems and historical solutions. Based on the priority response demand data and delayed response demand data, generate an automated solution for product fault handling according to the following logic: prioritize historical solutions in the rule library; if there is no direct match, use machine learning models (such as random forests, XGBoost) to predict the optimal solution; generate an automated fault handling solution, including fault classification, cause analysis, and recommended processing procedures. Based on the generated automated fault handling solution, perform the following remote processing operations: push software patches or configuration adjustment instructions to customer devices; remotely debug fault modules and record processing logs; request customers to provide more diagnostic information (such as uploading device logs, taking photos of problems, etc.). Collect feedback data after remote processing, mainly including: recording the time when the problem was successfully solved, the steps for executing the solution, and recording the reasons for failure (such as unsuccessful execution of instructions, problems beyond the scope of remote capabilities, etc.). Store the failure processing results as "the first after-sales failure processing results" for further analysis.

[0032] Step S3: Based on the after-sales failure processing result, a functional damage assessment is performed on the intelligent product to generate product functional damage assessment data; based on the product functional damage assessment data, an interventional fault diagnosis is performed on the after-sales failure processing result to generate interventional fault diagnosis data; a second remote processing feedback result is collected through the interventional fault diagnosis data to obtain a second after-sales failure processing result; after-sales maintenance scheduling is performed on the second after-sales failure result, thereby generating after-sales maintenance service scheduling data; In the embodiment of the present invention, after-sales failure processing result data is collected, including product model, fault description, fault occurrence time, processing measures and results, etc. The functional module list and corresponding performance indicators of the intelligent product are extracted. Based on the fault description and processing results, the affected functional modules are identified and the degree of damage is marked. The degree of functional damage of each damaged module (e.g., mild damage, moderate damage, complete failure) is evaluated using historical data and machine learning models (such as classification models or regression models). The evaluation results are integrated to generate product functional damage evaluation data, including the damage level and damage cause of each module. According to the functional damage evaluation data, the specific damaged module is located, the relevant fault information is extracted, and the damaged module marking data is generated. The module fault intervention analysis is performed on the marked data, and detailed fault causes and associated modules are output. The damaged module is matched with the fault cause by using the product knowledge graph to generate fault chain extension data. The fault modes are sorted according to the probability of occurrence and the scope of influence, and the optimal processing path is selected. The optimal fault diagnosis path is simulated and verified to test the effectiveness and reliability of the path. The diagnostic path is optimized and the final intervention fault diagnosis data is generated. Establish a data collection process to monitor remote processing results, including maintenance operation records, user feedback, and performance test data. If not enough remote processing feedback data is collected, repeat the collection process until complete data is obtained. Classify the collected data, eliminate successful processing results, and filter out the second failed processing results. Conduct in-depth analysis of the second failed processing results to identify the types and areas of faults that need to be handled first. Prioritize faults according to the urgency and scope of impact of the faults, and use scheduling optimization algorithms (such as linear programming, genetic algorithms, etc.) to allocate resources to maintenance tasks, determine maintenance time and execution team. Output after-sales maintenance service scheduling data containing scheduling plans, including maintenance team allocation, task priority, estimated completion time, etc.

[0033] Step S4: Perform a phased after-sales trend analysis on the first after-sales failure processing result, the second after-sales failure processing result, and the after-sales repair service scheduling data to generate phased after-sales trend data; visualize the after-sales quality assessment of the phased after-sales quality trend data to perform intelligent product after-sales management optimization operations.

[0034] In an embodiment of the present invention, the first after-sales failure processing result, the second after-sales failure processing result, and the after-sales maintenance service scheduling data are integrated into a phased after-sales data set. The frequency and cause distribution of failed maintenance are counted by time period (such as week, month, quarter), and time trend data is generated. Based on fault classification (such as hardware, software, interface problems), the frequency and proportion of each type of fault are counted to generate fault type trend data. The distribution of maintenance failures is counted by region or market partition to generate regional trend data. Linear regression, moving average and other methods are used to predict future maintenance failure trends. For example, ARIMA or Prophet models are used to analyze long-term trends and cyclical fluctuations. The analysis results are integrated to generate phased after-sales trend data, including time trends, fault type trends, and regional trends. Based on the phased after-sales trend data, various evaluation indicators are calculated to form intelligent product after-sales quality evaluation data, wherein the various evaluation indicators include: the proportion of successful repairs, the average time from fault reporting to repair completion, the quality of after-sales service measured by feedback scores, and the recurrence ratio of similar problems after repair. Use visualization tools such as Power BI and Tableau to visualize the periodic after-sales quality trend data for after-sales quality assessment, and integrate the visualization charts into the intelligent product after-sales quality assessment report.

[0035] Preferably, step S1 comprises the following steps: Step S11: obtaining a smart product after-sales problem record set, wherein the smart product after-sales problem record set includes at least one or more smart product after-sales problem records; Step S12: classifying the intelligent product after-sales problem record set into record types to generate text type records and voice type records; performing voice recognition on the intelligent product after-sales problem record according to the voice type record to generate after-sales demand voice recognition data; Step S13: Perform user demand speech speed analysis on the after-sales demand voice recognition data to generate user after-sales demand speech speed data; perform semantic recognition on the intelligent product after-sales problem records through text type records to generate after-sales demand semantic recognition data; Step S14: extract modal particles based on the after-sales demand semantic recognition data to obtain after-sales demand modal particle data; quantify customer demand for intelligent product after-sales problem records through the after-sales demand modal particle data and user after-sales demand speech speed data to generate intelligent product after-sales demand urgency data.

[0036] In an embodiment of the present invention, historical after-sales service records are extracted from the enterprise after-sales system, including text records of customer feedback and voice call records. Ensure that the data covers a variety of intelligent product categories and problem types, such as equipment failure, functional abnormality, user guidance, etc. Construct an intelligent product after-sales problem record set, which must contain the timestamp, product model, problem description and basic customer information of each record. The data in the intelligent product after-sales problem record set is automatically classified into text type records and voice type records according to the content format. Use a format detection algorithm, such as a file extension (such as .txt or .mp3), file size, data header parsing and other technologies to complete rapid classification. For voice type records, a speech recognition algorithm (such as an ASR model based on deep learning) is used to extract the text content in the voice to generate after-sales demand voice recognition data. During the recognition process, it is necessary to handle the multi-language support and background noise interference problems in the voice to ensure the accuracy of text transcription. Using the speech recognition results, extract the time length of the voice segment and the number of transcribed words, calculate the speech speed index of each record, and generate the user after-sales demand speech speed data. Perform cluster analysis on speech rate data to distinguish between normal speech rate, faster speech rate, and rapid speech rate to reflect the urgency of the customer. For text type records, use natural language processing (NLP) technology to perform semantic analysis, identify key problem descriptions and demand intentions in customer feedback, and generate after-sales demand semantic recognition data. Extract specific keywords, sentiment tendencies (such as positive, negative, neutral), and problem categories (such as hardware failure, software problems, operational consultation, etc.). Perform further modal particle analysis on the after-sales demand semantic recognition data, extract words that reflect the customer's tone (such as "urgent", "immediately", "as soon as possible", etc.), and generate after-sales demand modal particle data. Use a predefined modal particle dictionary and context dependency analysis method to accurately extract tone-related information. Combine the speech rate data with the modal particle data, and calculate the urgency of the customer's needs through a weighted algorithm: ;in is the urgency index, is the speech rate data, is the frequency of occurrence of modal particles, and The weight coefficients are respectively (adjusted according to the actual application) and the urgency data of after-sales demand for intelligent products is generated according to the calculation results. The data are classified according to the urgency (such as normal, relatively urgent, and urgent) to guide the priority sorting of after-sales service.

[0037] Preferably, quantifying customer demand for after-sales problem records of intelligent products by using after-sales demand modal particle data and user after-sales demand speech speed data includes: Perform tone feature analysis on the tone particle data of after-sales demand, so as to obtain customer tone feature data; perform speech rate change feature extraction on the speech rate data of user after-sales demand, so as to obtain user speech rate feature data; Emotional fluctuation analysis is performed based on the user speech speed feature data and the customer tone feature data, thereby obtaining customer emotional fluctuation data; after-sales problem priorities are preliminarily quantified based on customer emotional fluctuation data, thereby obtaining quantitative data on after-sales problem priorities; The after-sales problem records of intelligent products are collected in time series to obtain time series after-sales records; based on the time series after-sales records, a comprehensive demand urgency analysis is performed on the quantitative data of after-sales problem priorities to obtain the after-sales demand urgency data of intelligent products.

[0038] In an embodiment of the present invention, by using the data of after-sales demand modal particles, the frequency and distribution of particles are counted. A modal particle feature vector is constructed, for example, by extracting feature weights through TF-IDF (term frequency-inverse document frequency). A sentiment analysis model (such as a sentiment classification model based on BERT) is used to combine modal particles with context, analyze the customer's tone emotional tendency (such as anxiety, anger, satisfaction, etc.), and generate customer tone feature data, including emotional category, emotional intensity, and tone type. By calculating the rate of change of speech speed (such as the difference in speech speed sentence by sentence or paragraph by paragraph), the trend of speech speed change and the amplitude of fluctuation are extracted. The speech speed change features are classified into three categories: "stable", "slight fluctuation" and "violent fluctuation", and user speech speed feature data is generated to reflect the anxiety or calm state in the customer's expression. Combined with the user speech speed feature data and the customer tone feature data, a decision tree, a random forest and other machine learning models are used to analyze the user's emotional fluctuation pattern. According to the feature model, the customer's emotional fluctuation intensity score (such as 0-100, the higher the score, the greater the fluctuation) is output, and the customer's emotional fluctuation data is generated, including the fluctuation type (such as stable, fluctuation), fluctuation intensity and trend. Combine the emotional fluctuation data with the after-sales demand classification data, and make preliminary quantification according to the urgency rules of customer problems: ;in, Quantify the priority score. The intensity of emotional fluctuations, is the question category weight, and To adjust the parameters. Output the preliminary priority score of each after-sales problem and generate quantitative data on the priority of after-sales problems. Based on the customer's multiple after-sales records, sort the problem records in chronological order to form a time-series after-sales record, marking the time of occurrence and progress of the problem. Use time-series analysis technology to extract the changing trend of customer needs in the record and predict subsequent emergency situations. Based on the time-series after-sales records, combine the preliminary priority quantitative data with the time-series information to analyze whether there is a need to increase the priority of the problem, generate the urgency data of after-sales needs of intelligent products, and classify them according to the urgency level (such as ordinary, more urgent, and urgent).

[0039] As an example of the present invention, refer to Figure 2As shown, in this example, step S2 includes: Step S21: Compare the intelligent product after-sales demand urgency data with the preset standard demand urgency threshold. When the intelligent product after-sales demand urgency data is greater than or equal to the preset standard demand urgency threshold, give priority response to the corresponding intelligent product after-sales problem record and generate priority response demand data; Step S22: when the intelligent product after-sales demand urgency data is less than the preset standard demand urgency threshold, a delayed response is performed on the corresponding intelligent product after-sales problem record to generate delayed response demand data; product failure prediction is performed on the intelligent product after-sales problem record set by using the priority response demand data and the delayed response demand data to generate product failure prediction data; Step S23: construct a product knowledge graph; import product failure prediction data into the product knowledge graph to perform failure handling association analysis and generate an automated solution for product failure handling; Step S24: perform the first remote processing feedback data collection on the product fault handling automation solution. When the corresponding remote processing feedback data is not collected, repeat the data collection until the remote processing feedback data is received; divide the feedback results of the remote processing feedback data, eliminate the first after-sales successful processing result, and obtain the first after-sales failed processing result.

[0040] In an embodiment of the present invention, after-sales demand urgency data and a preset standard demand urgency threshold are obtained for intelligent products. The after-sales demand urgency data are compared with the standard demand urgency threshold. When the urgency data is greater than or equal to the threshold, the corresponding after-sales problem record is marked as a priority response demand, and priority response demand data is generated for rapid processing of after-sales problems. When the urgency data is less than the threshold, the corresponding after-sales problem record is marked as a delayed response demand. The priority response demand data and delayed response demand data are summarized, all after-sales problem records are analyzed, and a fault prediction algorithm (such as time series analysis or machine learning model) is applied to generate product fault prediction data. Product life cycle data is summarized, including information on design, production, sales, use, after-sales and other links. Relationship nodes between product characteristics, common fault types, fault causes and processing methods are defined. A dynamic knowledge graph is established using a graph database (such as Neo4j). The product fault prediction data is imported into the knowledge graph to analyze the correlation between the fault and the processing method. A rule reasoning algorithm (such as an OWL framework based on logical reasoning) is applied to automatically generate an automated solution for product fault processing. Perform remote processing operations on the product fault handling automation solution and monitor feedback data in real time through the API interface. If no feedback data is received, set up a retry mechanism and loop the collection until the remote processing feedback data is received. Perform result analysis on the received feedback data: If the processing is successful, record it as the "first after-sales successful processing result". If the processing fails, after removing the successful records, organize the remaining data as the "first after-sales failed processing result".

[0041] Preferably, predicting product failures for the intelligent product after-sales problem record set by using the priority response demand data and the delayed response demand data includes: Perform after-sales problem time series analysis based on priority response demand data and delayed response demand data, thereby obtaining after-sales problem time series data; perform demand frequency distribution extraction on the after-sales problem time series data, thereby obtaining after-sales problem demand frequency distribution data; The priority response demand data and the delayed response demand data are classified by the frequency distribution data of after-sales problem demand, so as to obtain the priority demand characteristic data and the delayed demand characteristic data; the priority demand characteristic data and the delayed demand characteristic data are clustered by the problem record concentration feature analysis, so as to obtain the problem record feature clustering data; Potential failure mode matching is performed through problem record feature clustering data to obtain potential product failure mode data; failure probability prediction calculation is performed on the intelligent product after-sales problem record set based on the potential product failure mode data to obtain product failure prediction data.

[0042] In an embodiment of the present invention, time series modeling is performed on priority response demand data and delayed response demand data based on the time dimension. A suitable time series analysis method (such as ARIMA model, LSTM model, etc.) is selected to generate after-sales problem time series data. The time series trend and periodic changes of after-sales problems are analyzed to identify the peak and trough time periods of demand fluctuations. The frequency of demand occurrence in the after-sales problem time series is counted, and the frequency distribution characteristics are extracted according to the time window (such as day, week or month). The frequency distribution curve or histogram is used for data visualization to clarify the distribution characteristics of priority demand and delayed demand. The priority response demand and delayed response demand are characterized by classification algorithms (such as KNN, decision tree or SVM). The key features of the demand (such as problem type, occurrence area, urgency, processing time, etc.) are extracted to generate priority demand feature data and delayed demand feature data. Cluster analysis is performed on the problem records based on the feature data, and the feature similarity of the problem records is found by using K-means, DBSCAN or hierarchical clustering algorithm. The clustering dimensions include problem occurrence frequency, equipment model, geographical distribution, etc., and problem record feature clustering data is generated. Build a potential failure mode database (including historical failure modes and problem records). Based on the clustering results and the failure mode database, use pattern matching algorithms (such as nearest neighbor algorithms or semantic similarity matching) to identify potential failure modes related to the problem record clustering data. Use Bayesian probability models or machine learning models (such as random forests, XGBoost) to perform correlation analysis and failure probability calculation on potential product failure modes and problem record sets. Output the probability value of each potential failure mode to identify high-risk potential failures.

[0043] Preferably, step S23 includes the following steps: Step S231: Obtain a product basic information data set and a historical intelligent product after-sales problem record set; extract common failure modes and corresponding solutions from the historical intelligent product after-sales problem record set to generate a historical failure solution data set; use graph structure construction technology to construct an initial product knowledge graph for the product basic information data set and the historical failure solution data set to generate an initial product knowledge graph; Step S232: semantic enhancement processing is performed on the initial product knowledge graph to generate a product knowledge graph; the core elements in the product fault prediction data are analyzed, including potential fault types, occurrence probabilities, impact ranges, and time ranges, to generate a fault prediction factor data set; the fault prediction factor data set is matched with nodes and edges in the product knowledge graph, and the association relationship between the fault prediction data and existing knowledge is established through similarity analysis to generate a product dynamic knowledge graph; Step S233: Perform path analysis on the product dynamic knowledge graph to extract functional modules, components and historical processing methods related to potential faults, thereby building a fault processing association model; use the fault processing association model to perform fault processing association mining on product fault prediction data to generate an initial product fault processing automation solution; Step S234: Perform step-by-step and logic verification on the initial product fault handling automation solution, thereby generating a product fault handling automation solution.

[0044] In the embodiment of the present invention, by collecting basic information related to the product, such as model, configuration, production batch, key technical parameters, etc., intelligent fault problems and their solutions recorded in the past product after-sales service are obtained to form a historical fault solution data set. The fault mode in the historical after-sales problem record is analyzed, and the corresponding solution is extracted to generate a historical fault solution data set. The graph structure construction technology is adopted to integrate the product basic information data set and the historical fault solution data set, and various attributes, fault modes, solutions, etc. of the product are associated as nodes and edges of the graph to construct an initial product knowledge graph. The initial product knowledge graph is semantically enhanced, and the knowledge graph is made more in-depth and accurate by adding more associated information, context understanding and reasoning rules, and finally a product knowledge graph is generated. The product fault prediction data is analyzed, and key factors such as potential fault types, occurrence probability, impact range and time range are extracted to generate a fault prediction element data set. The elements in the fault prediction element data set are matched with the nodes (such as functional modules, components, historical processing methods, etc.) and edges (such as association relationships) in the product knowledge graph, and the association relationship between the fault prediction data and the existing knowledge is established through similarity analysis to generate a product dynamic knowledge graph. Perform path analysis on the product dynamic knowledge graph to extract functional modules, components and historical fault handling methods related to potential faults. These paths will help to deeply understand the root causes and solutions of faults. Build a fault handling association model by analyzing the relationship between fault modes and solutions to guide the logic of fault handling. Use the constructed fault handling association model to mine fault prediction data and generate an initial product fault handling automation solution. Process the initial product fault handling automation solution step by step to clarify the operation requirements and processing methods of each step. Verify the automation solution to ensure that each step can be effectively executed in actual applications and the processing logic is correct, and finally generate a complete and effective product fault handling automation solution.

[0045] As an example of the present invention, refer to Figure 3 As shown, in this example, step S3 includes: Step S31: performing a functional damage assessment on the intelligent product based on the after-sales failure processing result, and generating product functional damage assessment data; Step S32: performing interventional fault diagnosis on the after-sales failure processing result according to the product function damage assessment data to generate interventional fault diagnosis data; performing a second remote processing feedback data collection based on the interventional fault diagnosis data, and when the corresponding remote processing feedback data is not collected, repeating the data collection until the remote processing feedback data is received; Step S33: dividing the feedback results of the remote processing feedback data, eliminating the second after-sales successful processing result, and obtaining the second after-sales failed processing result; performing after-sales maintenance scheduling on the second after-sales failure result, thereby generating after-sales maintenance service scheduling data.

[0046] In an embodiment of the present invention, by collecting and analyzing past after-sales failure processing records, including the type of failure, repaired or replaced parts, processing time and cost, etc., a functional damage assessment model is constructed based on the after-sales failure processing result data. The model comprehensively considers factors such as product usage status, failure frequency, component damage degree, repair history, etc., and evaluates the degree of functional damage of the product. The functional damage of the product is output by the evaluation model to generate product functional damage assessment data. The evaluation data will help identify the fault modules and subsequent processing steps that need to be focused on. According to the product functional damage assessment data, interventional fault diagnosis is performed on each product with problems. The diagnosis is based on the damaged functional module and historical fault information, analyzes the root cause of the fault, and generates interventional fault diagnosis data. According to the interventional fault diagnosis data, the remote technical support team will perform a second remote processing operation. After processing, remote feedback data is collected, which includes whether the fault processing is successful or not, the effectiveness of the processing method, etc. If sufficient remote feedback data is not obtained in the first collection, the data collection process needs to be repeated. The system will continue to monitor until valid remote processing feedback data is received. Classify the collected remote processing feedback data, remove the "second after-sales successful processing results", and retain the "second after-sales failed processing results". This step will help identify products that failed to solve the problem. For the second after-sales failed processing results, schedule the after-sales maintenance service according to the type of failure, the scope of impact, and the functional damage assessment of the product. The scheduling arrangement will assign the relevant products to the appropriate maintenance site or technician, and determine the priority, schedule, etc. of the maintenance. Based on the scheduling plan, generate after-sales maintenance service scheduling data to ensure that each fault handling link can be carried out on time and effectively, so as to improve the efficiency and quality of after-sales service.

[0047] Preferably, performing intervention fault diagnosis on after-sales failure processing results according to product function damage assessment data includes: Mark the damaged modules of intelligent products according to the product function damage assessment data to obtain the damaged module marking data of intelligent products; perform module failure intervention analysis on the after-sales failure processing results based on the damaged module marking data of intelligent products to generate intervention analysis result data; Use product knowledge graph to expand the fault chain of intervention analysis result data to generate extended fault diagnosis data; prioritize fault modes of extended fault diagnosis data, identify optimal processing paths, and generate fault diagnosis path data; perform simulation verification on fault diagnosis path data to generate optimized fault diagnosis data; The optimized fault diagnosis data is integrated into the product knowledge graph for fault extension node mapping, thereby generating intervention fault diagnosis data.

[0048] In an embodiment of the present invention, the damage of each module is extracted from the product function damage assessment data. The assessment data includes information such as the number of failures of each module, the failure probability, and the scope of influence (such as whether it affects the functions of other modules). The assessment data is processed by an algorithm (such as a machine learning model such as a decision tree or a random forest), and each module is marked to indicate whether it is damaged, and a priority is assigned to each damaged module (for example, according to the coreness of the module, the severity of the damage, and the frequency of occurrence). The marking data of each module is integrated into "intelligent product damaged module marking data", which includes the fault state and priority of each module (for example: module 1, damaged, priority 1). By analyzing historical fault data (such as fault type, repair method, etc.), the fault mode of the damaged module is analyzed. For example, if module 1 often has power failure, its root cause can be analyzed (such as power management chip failure, voltage instability, etc.). A preset processing scheme is proposed for each fault mode, the efficiency and cost of different processing schemes are analyzed, and a preliminary intervention analysis result is generated for each fault mode (for example, the repair method for the power failure of module 1 is to replace the power management chip). The intervention analysis result data is output to record the diagnosis and recommended processing scheme for each fault mode. According to the product's knowledge graph, the fault chain is extended using the nodes in the graph (module type, fault cause, historical repair method, etc.). First, the fault mode in the intervention analysis result data is matched with the relevant nodes in the graph (for example, the power failure is associated with the "battery module" or "power chip" node). The extended fault chain will contain multi-level information. For example, the failure of a module is not only "power management chip damage", but also related to "unstable battery voltage" or "software configuration problem", generating a diagnostic extension data containing all potential fault sources and repair paths (for example, power failure chain: battery damage -> power management chip failure -> power supply module failure). A priority sorting algorithm based on the severity and scope of the fault mode is used. For example, if the failure of the power management chip will cause the entire device to paralyze, while the low battery power will only affect the device's battery life, the power management chip failure can be listed as a high priority. Based on the priority sorting results, the fault diagnosis path data is generated. The path includes multiple steps, such as first checking the power module, then checking the battery, and finally checking the system settings. Each step is sorted according to importance to ensure that high-priority faults are handled first. Use simulation tools (such as fault tree analysis, Monte Carlo simulation, etc.) to verify the fault diagnosis path. Evaluate the success rate and time cost of each treatment plan by simulating the impact of different processing paths on device fault recovery. For example, simulate the repair path for low battery and power management chip failure to observe which path can restore the normal operation of the device faster. Optimize the fault diagnosis path based on the simulation results. If the simulation shows that some paths are more efficient than others, you can adjust the diagnosis strategy to generate a final optimized version of the fault diagnosis data.Integrate the optimized fault diagnosis data into the existing product knowledge graph, so that when a similar fault occurs again, the system can quickly reference historical data and give the best repair suggestions. For example, when a power management chip fails, the system can refer to the historical repair records and quickly recommend replacing the chip or checking the battery module. Create a mapping for each new fault extension node in the product knowledge graph. For example, after adding a diagnostic path for "power management chip failure", the fault path will be marked as a new fault extension node in the knowledge graph, forming a new relationship with other related nodes (such as batteries, power supply modules, etc.). Through node mapping, ensure the expansion and upgrade of the product knowledge graph. Finally, integrate the optimized diagnostic data into the product knowledge graph to form the final "intervention fault diagnosis data", which will be used for future fault diagnosis, intelligent processing path recommendation and fault pattern identification.

[0049] Preferably, step S4 comprises the following steps: Step S41: performing a phased after-sales trend analysis on the first after-sales failure processing result, the second after-sales failure processing result, and the after-sales repair service scheduling data to generate phased after-sales trend data; performing a after-sales quality evaluation on the phased after-sales quality trend data to generate intelligent product after-sales quality evaluation data; Step S42: Visualize the intelligent product after-sales quality assessment data to generate an intelligent product after-sales quality assessment report to perform intelligent product after-sales management optimization operations.

[0050] In an embodiment of the present invention, the first after-sales failure processing result, the second after-sales failure processing result, and the after-sales maintenance service scheduling data are collected and integrated. Data collection comes from multiple channels, such as customer feedback, service center records, maintenance history, etc. These data are divided according to the time dimension (for example, by month, by quarter, etc.), and the after-sales failure processing results and maintenance service scheduling in each time period will be summarized. Data mining and analysis methods (such as moving average method, time series analysis, trend line analysis, etc.) are used to analyze the data of each time period. The goal of the analysis is to identify the common after-sales failure types, frequency changes, maintenance response time and other trends in a certain stage. For example: in a certain quarter, the battery failure rate increases, the user feedback processing speed is slow, or a certain module (such as a display screen) has frequent failures. Based on the analysis results, the staged after-sales trend data is generated. The data will show the fluctuation trend of after-sales problems, such as the change in the frequency of failure types, the time period of increased customer complaints, and the response time of after-sales service. According to the after-sales trend data, a series of after-sales quality evaluation indicators are set. These indicators may include: the frequency of failures within a certain period of time, the average time from receiving the fault report to the completion of the repair, the customer satisfaction score for after-sales processing, which is usually obtained through customer questionnaires or return visits, and the proportion of customer complaints handled. Use the phased after-sales trend data to calculate the values ​​of various evaluation indicators. You can set weights (for example, failure rate accounts for 40%, repair efficiency accounts for 30%, and user satisfaction accounts for 30%), and calculate the final after-sales quality score by weighted average. Summarize and calculate all indicators to finally generate after-sales quality evaluation data. For example, the after-sales quality score for a period of time is 85 points, indicating that the after-sales quality is good at this stage. Based on the after-sales quality evaluation data, select an appropriate visualization method. For example: Line charts are used to display after-sales trend data, such as the trend of failure rate over time. Bar charts are used to display the frequency of different types of failures, or the distribution of after-sales service response time. Pie charts are used to display the proportion of various types of failures in total failures. Radar charts are used to display the comprehensive performance of multiple evaluation indicators, highlighting the advantages and disadvantages of different dimensions. Add data labels to visualizations to indicate the specific value and time of each data point (e.g., battery failure rate was 15% in a certain month). At the same time, you can add comments to indicate the cause of abnormal fluctuations (e.g., display failures increased during a certain period of time, specifically due to quality issues in new model batches). Use visualization tools (such as Power BI, Tableau, Matplotlib and Seaborn in Python, etc.) to automatically generate reports and export them to PDF or Excel format for further analysis and sharing.

[0051] In this specification, an intelligent product after-sales management system is provided, which is used to execute the above-mentioned intelligent product after-sales management method, and the intelligent product after-sales management system includes: The urgency analysis module is used to obtain a set of intelligent product after-sales problem records, wherein the set of intelligent product after-sales problem records contains at least one or more intelligent product after-sales problem records; quantify customer demand for the intelligent product after-sales problem records, and generate intelligent product after-sales demand urgency data; The automated processing module is used to compare the urgency data of after-sales demand of intelligent products with the preset standard demand urgency threshold, and generate priority response demand data and delayed response demand data; perform fault processing association analysis on the after-sales problem record set of intelligent products through the priority response demand data and delayed response demand data, and generate a product fault processing automated solution; perform the first remote processing feedback result collection on the product fault processing automated solution, and obtain the first after-sales failure processing result; The intervention processing module is used to evaluate the functional damage of the intelligent product based on the after-sales failure processing result and generate product functional damage evaluation data; perform intervention fault diagnosis on the after-sales failure processing result according to the product functional damage evaluation data and generate intervention fault diagnosis data; collect the second remote processing feedback result through the intervention fault diagnosis data to obtain the second after-sales failure processing result; perform after-sales maintenance scheduling on the second after-sales failure result, thereby generating after-sales maintenance service scheduling data; The after-sales quality evaluation module is used to perform a phased after-sales trend analysis on the first after-sales failure processing result, the second after-sales failure processing result and the after-sales repair service scheduling data to generate phased after-sales trend data; and to visualize the after-sales quality evaluation of the phased after-sales quality trend data to perform intelligent product after-sales management optimization operations.

[0052] The beneficial effect of the present invention is that the urgency analysis module can clearly distinguish the urgency of different problems by quantifying customer needs for after-sales problem records, thereby assigning a reasonable priority to each problem. This mechanism ensures that key problems can be handled in the shortest time, avoiding customer dissatisfaction caused by processing delays. By comparing with the standard demand urgency threshold, it can ensure that important problems are responded to first, reducing the response time for lower priority problems, thereby improving service efficiency. The automation processing module can reduce manual intervention and reduce the risk of human error by automatically generating an automated solution for product fault processing and providing the first remote processing feedback, while improving the accuracy of fault processing. The automated solution can improve the consistency and efficiency of processing, shorten the processing cycle, ensure that the problem can be effectively solved at the first remote processing, and reduce the need for subsequent manual intervention. The intervention processing module can accurately locate the cause of the fault through a functional damage assessment based on the first after-sales failure processing result and further intervention fault diagnosis. This process can provide a more accurate repair solution by combining historical data with fault chain expansion, reducing unnecessary rework and improving overall maintenance efficiency. Through multiple rounds of remote processing feedback collection, the system can track the progress of fault repair in real time and ensure that the repair plan is verified and improved at multiple stages. After analyzing the after-sales trend data, the after-sales quality assessment module can generate a detailed after-sales quality assessment report, which can help the company's management clearly understand the problem by visually displaying the after-sales service quality and further optimize the after-sales service process. By analyzing the periodic after-sales quality trend data, service bottlenecks or areas of quality decline can be quickly discovered, so that the company can make timely adjustments in strategy and enhance customer trust and satisfaction. Therefore, the present invention improves the efficiency and time efficiency of after-sales management through demand quantification, fault correlation analysis, feedback diagnosis and trend evaluation.

[0053] Therefore, the embodiments should be regarded as illustrative and non-restrictive from all points, and the scope of the present invention is limited by the appended claims rather than the above description, and it is therefore intended that all changes falling within the meaning and range of equivalent elements of the application documents are included in the present invention.

[0054] The above description is only a specific embodiment of the present invention, so that those skilled in the art can understand or implement the present invention. Various modifications to these embodiments will be apparent to those skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of the present invention. Therefore, the present invention will not be limited to the embodiments shown herein, but should conform to the widest scope consistent with the principles and novel features invented herein.

Claims

1. An intelligent product after-sales management method, characterized in that: The following steps are involved: Step S1: obtaining a set of after-sales problem records of intelligent products, wherein the set of after-sales problem records of intelligent products includes at least one or more after-sales problem records of intelligent products; quantifying customer demand for after-sales problem records of intelligent products, and generating urgency data of after-sales demand of intelligent products; Step S2: Compare the intelligent product after-sales demand urgency data with the preset standard demand urgency threshold to generate priority response demand data and delayed response demand data; Through the priority response demand data and delayed response demand data, the fault handling correlation analysis of the intelligent product after-sales problem record set is carried out to generate an automated solution for product fault handling; Collect the first remote processing feedback results of the product failure handling automation solution and obtain the first after-sales failure handling result; Step S3: Based on the after-sales failure processing result, the intelligent product is evaluated for functional damage, and product functional damage evaluation data is generated; based on the product functional damage evaluation data, interventional fault diagnosis is performed on the after-sales failure processing result to generate interventional fault diagnosis data; and the second remote processing feedback result is collected through the interventional fault diagnosis data to obtain a second after-sales failure processing result; Performing after-sales maintenance scheduling for the second after-sales failure result, thereby generating after-sales maintenance service scheduling data; Step S4: performing a phased after-sales trend analysis on the first after-sales failure processing result, the second after-sales failure processing result, and the after-sales repair service scheduling data to generate phased after-sales trend data; Visualize the after-sales quality assessment of periodic after-sales quality trend data to perform intelligent product after-sales management optimization operations.

2. The intelligent product after-sales management method according to claim 1, characterized in that: Step S1 includes the following steps: Step S11: obtaining a smart product after-sales problem record set, wherein the smart product after-sales problem record set includes at least one or more smart product after-sales problem records; Step S12: classifying the intelligent product after-sales problem record set into record types to generate text type records and voice type records; performing voice recognition on the intelligent product after-sales problem record according to the voice type record to generate after-sales demand voice recognition data; Step S13: Perform user demand speech speed analysis on the after-sales demand voice recognition data to generate user after-sales demand speech speed data; perform semantic recognition on the intelligent product after-sales problem records through text type records to generate after-sales demand semantic recognition data; Step S14: extract modal particles based on the after-sales demand semantic recognition data to obtain after-sales demand modal particle data; quantify customer demand for intelligent product after-sales problem records through the after-sales demand modal particle data and user after-sales demand speech speed data to generate intelligent product after-sales demand urgency data.

3. The intelligent product after-sales management method according to claim 2, characterized in that: The customer demand quantification of intelligent product after-sales problem records through after-sales demand modal particle data and user after-sales demand speech speed data includes: Perform tone feature analysis on the tone particle data of after-sales demand, so as to obtain customer tone feature data; perform speech rate change feature extraction on the speech rate data of user after-sales demand, so as to obtain user speech rate feature data; Emotional fluctuation analysis is performed based on the user speech speed feature data and the customer tone feature data, thereby obtaining customer emotional fluctuation data; after-sales problem priorities are preliminarily quantified based on customer emotional fluctuation data, thereby obtaining quantitative data on after-sales problem priorities; The after-sales problem records of intelligent products are collected in time series to obtain time series after-sales records; based on the time series after-sales records, a comprehensive demand urgency analysis is performed on the quantitative data of after-sales problem priorities to obtain the after-sales demand urgency data of intelligent products.

4. The intelligent product after-sales management method according to claim 1, characterized in that: Step S2 includes the following steps: Step S21: Compare the intelligent product after-sales demand urgency data with the preset standard demand urgency threshold. When the intelligent product after-sales demand urgency data is greater than or equal to the preset standard demand urgency threshold, give priority response to the corresponding intelligent product after-sales problem record and generate priority response demand data; Step S22: when the intelligent product after-sales demand urgency data is less than the preset standard demand urgency threshold, a delayed response is performed on the corresponding intelligent product after-sales problem record to generate delayed response demand data; product failure prediction is performed on the intelligent product after-sales problem record set by using the priority response demand data and the delayed response demand data to generate product failure prediction data; Step S23: construct a product knowledge graph; import product failure prediction data into the product knowledge graph to perform failure handling association analysis and generate an automated solution for product failure handling; Step S24: perform the first remote processing feedback data collection on the product fault handling automation solution. When the corresponding remote processing feedback data is not collected, repeat the data collection until the remote processing feedback data is received; divide the feedback results of the remote processing feedback data, eliminate the first after-sales successful processing result, and obtain the first after-sales failed processing result.

5. The intelligent product after-sales management method according to claim 4, characterized in that: The product failure prediction of intelligent product after-sales problem record set by using priority response demand data and delayed response demand data includes: Perform after-sales problem time series analysis based on priority response demand data and delayed response demand data, thereby obtaining after-sales problem time series data; perform demand frequency distribution extraction on the after-sales problem time series data, thereby obtaining after-sales problem demand frequency distribution data; The priority response demand data and the delayed response demand data are classified by the frequency distribution data of after-sales problem demand, so as to obtain the priority demand characteristic data and the delayed demand characteristic data; the priority demand characteristic data and the delayed demand characteristic data are clustered by the problem record concentration feature analysis, so as to obtain the problem record feature clustering data; Potential failure mode matching is performed through problem record feature clustering data to obtain potential product failure mode data; failure probability prediction calculation is performed on the intelligent product after-sales problem record set based on the potential product failure mode data to obtain product failure prediction data.

6. The intelligent product after-sales management method according to claim 4, characterized in that: Step S23 includes the following steps: Step S231: Obtain a product basic information data set and a historical intelligent product after-sales problem record set; extract common failure modes and corresponding solutions from the historical intelligent product after-sales problem record set to generate a historical failure solution data set; use graph structure construction technology to construct an initial product knowledge graph for the product basic information data set and the historical failure solution data set to generate an initial product knowledge graph; Step S232: semantic enhancement processing is performed on the initial product knowledge graph to generate a product knowledge graph; the core elements in the product fault prediction data are analyzed, including potential fault types, occurrence probabilities, impact ranges, and time ranges, to generate a fault prediction factor data set; the fault prediction factor data set is matched with nodes and edges in the product knowledge graph, and the association relationship between the fault prediction data and existing knowledge is established through similarity analysis to generate a product dynamic knowledge graph; Step S233: Perform path analysis on the product dynamic knowledge graph to extract functional modules, components and historical processing methods related to potential faults, thereby building a fault processing association model; use the fault processing association model to perform fault processing association mining on product fault prediction data to generate an initial product fault processing automation solution; Step S234: Perform step-by-step and logic verification on the initial product fault handling automation solution, thereby generating a product fault handling automation solution.

7. The intelligent product after-sales management method according to claim 6, characterized in that: Step S3 includes the following steps: Step S31: performing a functional damage assessment on the intelligent product based on the after-sales failure processing result, and generating product functional damage assessment data; Step S32: performing interventional fault diagnosis on the after-sales failure processing result according to the product function damage assessment data to generate interventional fault diagnosis data; performing a second remote processing feedback data collection based on the interventional fault diagnosis data, and when the corresponding remote processing feedback data is not collected, repeating the data collection until the remote processing feedback data is received; Step S33: dividing the feedback results of the remote processing feedback data, eliminating the second after-sales successful processing result, and obtaining the second after-sales failed processing result; performing after-sales maintenance scheduling on the second after-sales failure result, thereby generating after-sales maintenance service scheduling data.

8. The intelligent product after-sales management method according to claim 7, characterized in that: Interventional fault diagnosis of after-sales failure processing results based on product function damage assessment data includes: Mark the damaged modules of intelligent products according to the product function damage assessment data to obtain the damaged module marking data of intelligent products; perform module failure intervention analysis on the after-sales failure processing results based on the damaged module marking data of intelligent products to generate intervention analysis result data; Use product knowledge graph to expand the fault chain of intervention analysis result data to generate extended fault diagnosis data; prioritize fault modes of extended fault diagnosis data, identify optimal processing paths, and generate fault diagnosis path data; perform simulation verification on fault diagnosis path data to generate optimized fault diagnosis data; The optimized fault diagnosis data is integrated into the product knowledge graph for fault extension node mapping, thereby generating intervention fault diagnosis data.

9. The intelligent product after-sales management method according to claim 1, characterized in that: Step S4 includes the following steps: Step S41: performing a phased after-sales trend analysis on the first after-sales failure processing result, the second after-sales failure processing result, and the after-sales repair service scheduling data to generate phased after-sales trend data; performing a after-sales quality evaluation on the phased after-sales quality trend data to generate intelligent product after-sales quality evaluation data; Step S42: Visualize the intelligent product after-sales quality assessment data to generate an intelligent product after-sales quality assessment report to perform intelligent product after-sales management optimization operations.

10. An intelligent product after-sales management system, characterized in that: Used to execute the intelligent product after-sales management method according to claim 1, the intelligent product after-sales management system comprises: The urgency analysis module is used to obtain a set of intelligent product after-sales problem records, wherein the set of intelligent product after-sales problem records contains at least one or more intelligent product after-sales problem records; quantify customer demand for the intelligent product after-sales problem records, and generate intelligent product after-sales demand urgency data; The automated processing module is used to compare the urgency data of after-sales demand of intelligent products with the preset standard demand urgency threshold, and generate priority response demand data and delayed response demand data; perform fault processing association analysis on the after-sales problem record set of intelligent products through the priority response demand data and delayed response demand data, and generate a product fault processing automated solution; perform the first remote processing feedback result collection on the product fault processing automated solution, and obtain the first after-sales failure processing result; The intervention processing module is used to evaluate the functional damage of the intelligent product based on the after-sales failure processing result and generate product functional damage evaluation data; perform intervention fault diagnosis on the after-sales failure processing result according to the product functional damage evaluation data and generate intervention fault diagnosis data; collect the second remote processing feedback result through the intervention fault diagnosis data to obtain the second after-sales failure processing result; perform after-sales maintenance scheduling on the second after-sales failure result, thereby generating after-sales maintenance service scheduling data; The after-sales quality evaluation module is used to perform a phased after-sales trend analysis on the first after-sales failure processing result, the second after-sales failure processing result and the after-sales repair service scheduling data to generate phased after-sales trend data; and to visualize the after-sales quality evaluation of the phased after-sales quality trend data to perform intelligent product after-sales management optimization operations.

Images