Management server of automobile part remanufacturing-based repair system for determining recommended repair shops
A management server for automobile repair systems provides objective shop recommendations using image analysis and performance data, addressing subjective selection issues and enhancing repair transparency and efficiency.
Patent Information
- Application Number
- KR1020250195371
- Authority / Receiving Office
- KR · KR
- Patent Type
- Patents
- Current Assignee / Owner
- Filing Date
- 2025-12-10
- Publication Date
- 2026-11-11
- Estimated Expiration
- 2045-12-10
AI Technical Summary
The selection of repair shops for automobile damage is often subjective and lacks objective data-based recommendations, leading to variations in repair quality, potential over-repairs, and disputes due to cost reductions, with limited integration of remanufactured parts considerations.
A management server that manages repair requests, determines repair difficulty, and recommends shops based on image analysis of damage, past performance data, and remanufactured parts availability, providing a transparent and efficient repair process.
Ensures accurate and reliable repair shop recommendations, reduces unnecessary repairs, and enhances transparency by integrating image analysis with repair shop performance data, reducing disputes and improving repair efficiency.
Smart Images

Figure 112025139632735-PAT00013_ABST
Abstract
Description
Technology Field
[0001] The present invention relates to a repair system based on the remanufacturing of automobile parts, and more particularly, to a management server of a system that receives a repair request for a specific damaged part of an automobile, determines an appropriate repair shop responding to the repair request, and provides the repair request for the damaged part to the repair shop. Background Technology
[0002] During the automobile accident repair process, the party requesting the repair (vehicle owner or insurance company) must select an appropriate repair shop from among multiple options. Traditionally, the selection of a repair shop was often based on geographical distance, simple price comparisons, or subjective reputations, making it difficult to provide quantitative recommendations based on objective data such as the actual extent of vehicle damage, the complexity of the required repairs, the shop's technical capabilities, and past repair performance. Consequently, there are issues such as significant variations in repair quality, potential for over-repairs or unnecessary parts replacements, and frequent customer complaints and disputes between repair shops and insurance companies arising from cost reductions during the insurance claim process.
[0003] Furthermore, while technical attempts are being made to reduce costs or shorten repair times by utilizing remanufactured parts, these efforts remain limited to a simple parts supply perspective or are not integrated with repair shop recommendation algorithms. Consequently, there is currently no system in place to support optimal decision-making by comprehensively considering factors such as the availability of remanufactured parts, insurance reduction rates, and repair difficulty. The problem to be solved
[0004] The objective of the present invention to solve the above-mentioned problems is to provide a management server for a system that receives a repair request for a specific damaged part of a vehicle, determines an appropriate repair shop responding to the repair request, and provides the repair request for the damaged part to the repair shop. means of solving the problem
[0005] One aspect of the present invention for achieving the above objective relates to a management server of an automobile parts remanufacturing-based repair system that manages repair requesters and repair shops, wherein the management server comprises an information management unit configured to manage repair shop information associated with repair shops, a repair request receiving unit configured to receive a repair request from a repair requester and an image of a damaged part, a damage degree determination unit configured to generate damage characteristic information indicating the degree of damage from the image of the damaged part, a repair difficulty determination unit configured to generate repair difficulty information indicating the repair difficulty required for repairing the damaged part based on the damage characteristic information, and a recommended repair shop determination unit configured to determine a recommended repair shop suitable for repairing the damaged part based on the information of repair shops and the repair difficulty information. Effects of the invention
[0006] According to embodiments of the present invention, there is an effect of receiving a repair request for a specific damaged part of a vehicle, determining an appropriate repair shop corresponding to the repair request, and providing the repair request for the damaged part to the repair shop. Brief explanation of the drawing
[0007] FIG. 1 shows an automobile parts remanufacturing-based repair system according to one embodiment of the present invention. FIGS. 2 to 4 illustrate the operation of a management server according to embodiments of the present invention. FIG. 5 shows a management server according to embodiments of the present invention. FIG. 6 shows a damage degree determination unit according to embodiments of the present invention. FIG. 7 shows a repair difficulty determination unit according to embodiments of the present invention. FIG. 8 shows a recommended repair shop determination unit according to embodiments of the present invention. FIG. 9 is a diagram illustrating the hardware configuration of a management server according to embodiments of the present invention. Specific details for implementing the invention
[0008] The present invention is susceptible to various modifications and may have various embodiments; specific embodiments are illustrated in the drawings and described in detail in the detailed description. However, this is not intended to limit the invention to specific embodiments, and it should be understood that the invention includes all modifications, equivalents, and substitutions that fall within the spirit and scope of the invention. Similar reference numerals have been used for similar components in the description of each drawing.
[0009] Terms such as first, second, A, B, etc., may be used to describe various components, but said components should not be limited by said terms. These terms are used solely for the purpose of distinguishing one component from another. For example, without departing from the scope of the present invention, the first component may be named the second component, and similarly, the second component may be named the first component. The term "and / or" includes a combination of a plurality of related described items or any of a plurality of related described items.
[0010] When it is stated that one component is "connected" or "connected" to another component, it should be understood that while it may be directly connected or connected to that other component, there may also be other components in between. On the other hand, when it is stated that one component is "directly connected" or "directly connected" to another component, it should be understood that there are no other components in between.
[0011] The terms used in this application are used merely to describe specific embodiments and are not intended to limit the invention. The singular expression includes the plural expression unless the context clearly indicates otherwise. In this application, terms such as "comprising" or "having" are intended to specify the presence of the features, numbers, steps, actions, components, parts, or combinations thereof described in the specification, and should be understood as not precluding the existence or addition of one or more other features, numbers, steps, actions, components, parts, or combinations thereof.
[0012] Unless otherwise defined, all terms used herein, including technical or scientific terms, have the same meaning as generally understood by those skilled in the art to which the present invention pertains. Terms such as those defined in commonly used dictionaries should be interpreted as having a meaning consistent with their meaning in the context of the relevant technology, and should not be interpreted in an ideal or overly formal sense unless explicitly defined in this application.
[0013] Hereinafter, preferred embodiments according to the present invention will be described in detail with reference to the attached drawings.
[0014] FIG. 1 illustrates an automobile parts remanufacturing-based repair system according to an embodiment of the present invention. Referring to FIG. 1, the automobile parts remanufacturing-based repair system (hereinafter, the system (10)) includes a management server (100), a client terminal (201) of a repair client, and a repair shop terminal (203) of a repair shop. At this time, the management server (100) and the terminals (201, 203) can exchange data with each other over a wireless network.
[0015] The management server (100) is a server configured to perform overall management and operation of the system (10), and receives a repair request from a requester such as a repair requester or an insurance company, determines the difficulty of the repair based on images of the damaged parts of the requested vehicle and repair-related information, and determines and proposes a repair shop that can most appropriately handle the repair details. In addition, when a repair shop is determined (automatically or by the repair requester), the management server (100) can provide the details of the repair request to the determined repair shop so that the repair can be carried out.
[0016] Furthermore, the management server (100) can manage the repair progress, insurance claim details, and settlement status in real time, and can perform the role of mediating and managing the repair process between the client and the repair company.
[0017] The management server (100) is connected to the repair shop terminal (203) and the client terminal (201) via a network (not shown) and may be operated in the form of an application or a web-based service. For example, a repair client may use their client terminal (201) (e.g., using an application provided by the management server (100)) to send a repair request to the management server (100) along with an image of the damaged part taken at the accident scene. The management server (100) receives this, calculates the repair difficulty, and determines a recommended repair shop suitable for the repair case by synthesizing information such as repair shop information and insurance claim reduction rate information, and provides a list of recommended repair shops to the client terminal (201).
[0018] Furthermore, the management server (100) can provide information (e.g., photos or comments) that can be checked for each repair in progress to the client terminal (201).
[0019] The client terminal (201) may be a user terminal such as an insurance company or a car owner. For example, when requesting vehicle repair, the insurance company may directly select a repair shop or select one of the recommended repair shops provided by the system. Alternatively, the management server (100) may be configured to automatically determine a specific shop among the recommended repair shops as the optimal shop and automatically transmit the repair request to the repair shop terminal (203).
[0020] The repair shop terminal (203) may be an operating terminal of each repair shop, specialized company, or repair network, or a terminal owned by a repairer of the repair shop. It can receive a repair request transmitted by the management server (100) and respond with information such as whether the repair is possible, the estimated time required, and the cost, or upload the real-time repair progress status. This information can be provided to the client terminal (201) through the management server (100). Through this, the management server (100) can establish a transparent repair process by continuously recording and managing photos of each repair process stage, work details, insurance claim amounts, and settlement amounts.
[0021] Unlike conventional simple recommendation methods based on distance and price, the automobile repair system (10) according to the present invention provides a quantitative and data-based repair shop recommendation technology based on image analysis of the actual damaged parts of the accident vehicle and past repair performance data of the repair shop. Accordingly, it is possible to ensure repair quality, prevent unnecessary over-repairs, reduce the insurance claim reduction rate, shorten the repair period, and ensure transparency of repair costs.
[0022] In addition, since the system of the present invention can manage the entire process from repair request to completion, insurance claim, and settlement in real time on a single platform, the repair requester (vehicle owner or insurance company), the repair shop, and the settlement process can be efficiently integrated and operated.
[0023] FIGS. 2 to 4 illustrate the operation of a management server according to embodiments of the present invention. Referring to FIGS. 2 to 4, a screen of a web service or application service provided by the management server is shown.
[0024] Referring to FIG. 2, the client terminal (201) may display input items related to a repair request, such as vehicle number, vehicle name, insurance company information, counterparty insurance company information, wheel quantity, and estimated delivery date. Additionally, input fields for the repair shop name and repair shop address are included, and location information required for a repair request can be registered along with an address search function. At the bottom, a photo registration button for uploading images of damage to the accident vehicle may be included. When the client enters the necessary information into the input items, the entered repair request data and images of the damaged parts are transmitted to the management server (100), and the management server (100) can perform a process of determining the repair difficulty and deciding on a recommended repair shop based on this.
[0025] Referring to FIG. 3, a list of multiple repair shops corresponding to the repair request conditions and the recommendation results of the management server (100) is displayed on the client terminal (201). Each repair shop is provided in a list format including information such as the shop name, region, and recent repair throughput, and can be sorted according to the recommendation score calculated by the management server (100) and displayed to the user in a ranking format. Each shop item in the list includes a selection button that allows the client to proceed with a repair request to the corresponding shop, and when the client selects a specific shop, the management server (100) can automatically transmit the repair request data to the corresponding repair shop terminal (203).
[0026] Referring to FIG. 4, the client terminal (201) visually displays the repair progress stage collected by the management server (100) from the repair shop terminal (203). The repair status can be displayed in the form of a circular stage display UI, categorized by the progress stages of the repair process, such as scheduled pickup, arrival at the shop, repair in progress, and repair completed. Each stage is displayed as either active or inactive depending on the current progress status, and additional information such as a detailed description of the stage, carrier information, insurance company review status, and estimated arrival time can be displayed at the bottom of the screen. Through this, the client can check the repair progress in real time, and the management server (100) updates the status changes through information linkage with the repair shop and insurance company and provides them to the client terminal.
[0027] FIG. 5 illustrates a management server according to embodiments of the present invention. Referring to FIG. 5, the management server (100) includes an information management unit (110), a repair request receiving unit (120), a damage degree determination unit (130), a repair difficulty determination unit (140), and a recommended repair company determination unit (150).
[0028] The information management unit (110) stores and manages information necessary for the operation of the management server (100). For example, the information management unit (110) may be a database included within the management server (100) or accessible to the management server (100).
[0029] The information management unit (110) stores and manages information regarding basic information, past repair records, past repair cost claims, etc. for multiple repair shops. For example, the information management unit (110) stores repair shop information for each repair shop, including information on the repair part, vehicle, repair period, repair difficulty, claimed cost, settlement amount, customer rating, etc.
[0030] The repair request receiving unit (120) receives a repair request from the client terminal (201). According to embodiments, the repair request receiving unit (120) receives a repair request including information about the damaged part and an image of the damaged part. Additionally, the repair request receiving unit (120) may receive additional information about the damaged vehicle, the location of the accident, the type of accident, etc. The repair request receiving unit (120) may transmit the received information to the damage degree determination unit (130) and the repair difficulty determination unit (140) so that it can be used as input data for subsequent processing.
[0031] The damage degree determination unit (130) determines the degree of damage by analyzing the damage status of the vehicle from the damaged part image input through the repair request receiving unit (120). The damage degree determination unit (130) generates damage characteristic information indicating the degree of damage by performing preprocessing on the input image, extraction of damage-related feature information, and determination of the damage status. Through this, it is possible to make a quantitative judgment based on actual images without relying on simple user descriptions or the mechanic's visual judgment.
[0032] Damage characteristic information determined by the damage degree determination unit (130) can be transmitted to the repair difficulty determination unit (140).
[0033] The repair difficulty determination unit (140) determines the repair difficulty required to repair the damaged part by synthesizing the damage characteristic information transmitted from the damage degree determination unit (130), vehicle information, and damaged part information. The repair difficulty determination unit (140) evaluates the elements necessary for determining the repair difficulty, such as the estimated repair time, technical level, and repair complexity, calculates repair difficulty information indicating the repair difficulty of the accident vehicle, and can provide this to the recommended repair shop determination unit (150).
[0034] The recommended repair company determination unit (150) can determine the optimal repair company for a repair request by comprehensively considering the repair difficulty information calculated by the repair difficulty judgment unit (140) and the repair company information managed by the information management unit (110). The recommended repair company determination unit (150) can calculate a recommendation score by considering the suitability between the repair difficulty information and the actual processing capacity of the repair company, as well as reflecting the discrepancy between the billed cost and the actual settled amount, and can rank multiple repair companies according to this recommendation score and transmit them to the client terminal (201).
[0035] Accordingly, according to the management server (100) of the present invention, by integrally performing quantitative damage analysis based on images of damaged parts of the vehicle, prediction of repair difficulty based on repair shop performance data, and judgment of appropriateness based on settlement history, it becomes possible to recommend an optimal repair shop based on objective evidence rather than a simple distance or price comparison method.
[0036] FIG. 6 illustrates a damage degree determination unit according to embodiments of the present invention. Referring to FIG. 6, the damage degree determination unit (130) includes an image preprocessing unit (131), a damage feature extraction unit (133), and a damage feature vector calculation unit (135).
[0037] The image preprocessing unit (131) performs basic preprocessing operations for damage feature analysis on an image of a damaged part of an accident vehicle input through a repair request receiving unit (120). The image preprocessing unit (131) performs preprocessing operations such as resolution normalization, brightness and contrast adjustment, noise removal, and pixel value normalization of the input image, thereby enabling stable computation during the CNN-based feature extraction process for the image transmitted to the damage feature extraction unit (133). Additionally, the image preprocessing unit (131) can extract a region of interest automatically or according to user settings if necessary, and convert it into image data centered on the region containing the damaged part.
[0038] The damage feature extraction unit (133) takes an image processed by the image preprocessing unit (131) as input and performs multiple stages of convolutional and pooling operations based on a Convolutional Neural Network (CNN) to generate a damage feature map that constitutes visual features meaningful for damage judgment. The damage feature extraction unit (133) according to the present embodiment hierarchically extracts low-level and intermediate-level visual features such as edges, textures, crack shapes, geometric structures of deformation, and paint damage areas from the input image, and can perform pooling operations or attention-based importance reflection operations to emphasize features that contribute to the judgment of the severity of damage. Such a feature extraction process is not intended for simple image classification or object identification, but rather for generating a feature representation that comprehensively reflects the intensity, extent, and pattern of vehicle damage.
[0039] The damage characteristic vector calculation unit (135) receives the damage characteristic map generated by the damage characteristic extraction unit (133) and generates a multidimensional damage characteristic vector that quantitatively represents the damage state. The damage characteristic vector calculation unit (135) performs intensive operations such as flattening or global average pooling on the feature map to generate a damage characteristic vector that concisely represents only the key features significant for determining the degree of damage, and the generated damage characteristic vector is subsequently provided as input to the repair difficulty judgment unit (140). At this time, each component of the damage characteristic vector may represent a damage indicator from various aspects of the corresponding damaged area. For example, the damage characteristic vector may include multiple dimensional elements reflecting the presence or absence of a specific damage pattern, the distribution of the damaged area, and the damage intensity, and each element is used as a quantitative judgment criterion in the process of calculating the repair difficulty.
[0040] That is, the damage degree determination unit (130) of the present invention performs an in-depth characteristic analysis of the damage state based on an actual image of the damaged area, thereby enabling an analysis of the damage degree based on objective and quantitative criteria, unlike the conventional method in which the severity of damage is determined by visual evaluation by a repair expert or qualitative criteria.
[0041] FIG. 7 illustrates a repair difficulty determination unit according to embodiments of the present invention. Referring to FIG. 7, the repair difficulty determination unit (140) includes a repair characteristic vector generation unit (141) and a repair difficulty vector calculation unit (143).
[0042] The repair characteristic vector generation unit (141) combines the damage characteristic vector input from the damage degree determination unit (130) with the vehicle basic information and repair part information stored in the information management unit (110) to generate a repair characteristic vector, which is an input vector for determining the difficulty of repair.
[0043] For example, vehicle information includes the vehicle model, year of manufacture, body material (steel plate / aluminum / composite material, etc.), and vehicle type classification (sedan, SUV, etc.), while repair part information includes the location of damage to the accident vehicle (e.g., front / rear bumpers, door panels, fenders, radiator, wheels, etc.) or type information of the component to be repaired.
[0044] The mathematical characteristic vector generation unit (141) combines these various attribute data into a single input vector form to produce a result, and uses it for the neural network-based prediction of the mathematical difficulty judgment unit (140).
[0045] The mathematical difficulty vector calculation unit (143) receives the mathematical characteristic vector generated by the mathematical characteristic vector generation unit (141) as input, applies a deep learning-based prediction model for determining mathematical difficulty, and calculates the mathematical difficulty vector as output.
[0046] The mathematical difficulty vector calculation unit (143) may include a deep learning model such as a multilayer perceptron (Deep Neural Network, DNN) or a recurrent neural network (RNN), and learns non-linear relationships and weighted correlations between multiple features within an input vector to estimate mathematical difficulty components that are difficult to derive through simple numerical calculations. For example, by using a fully-connected neural network that includes multiple hidden layers to progressively compress and transform the input dimension, a latent representation that highlights key factors determining mathematical difficulty can be generated.
[0047] Additionally, when the input vector includes past similar accident history and repair performance data based on time change, the repair difficulty vector calculation unit (143) may use an RNN-based structure such as a Long Short-Term Memory (LSTM) or a Gated Recurrent Unit (GRU) to effectively model the time series pattern of the input vector. This structure is advantageous for predicting the trend of difficulty change of a specific repair part or the deviation of repair difficulty by vehicle type by reflecting the change in repetitive or cumulative repair results.
[0048] The numerical difficulty vector, which is the output of the numerical difficulty vector calculation unit (143), is a multidimensional vector that quantitatively expresses multiple elements constituting numerical difficulty. For example, the numerical difficulty vector calculation unit (143) calculates a numerical difficulty vector according to the mathematical formula 1 below.
[0049]
[0050] Here, RDV is a repair difficulty vector, RTime is the estimated repair time (e.g., in days), RCost is the estimated repair cost, and RLev is the graded repair skill level required for the repair (e.g., the value of any one of the 10 graded skill levels).
[0051] The calculated repair difficulty vector is then transmitted to the recommended repair company determination unit (150) and used as a standard for calculating a recommendation score by comparing each repair company's actual processing ability and past repair effectiveness.
[0052] Accordingly, the repair difficulty vector calculation unit (143) of the present invention enables quantitative repair difficulty calculation that reflects complex correlations between damage characteristics and vehicle characteristics through a deep learning-based prediction model, thereby providing a repair shop recommendation result that is more accurate and reliable than existing experience-based or simple estimation methods.
[0053] FIG. 8 illustrates a recommended repair shop determination unit according to embodiments of the present invention. Referring to FIG. 8, the recommended repair shop determination unit (150) includes an expertise vector generation unit (151), a fee weighting calculation unit (153), and a recommendation score calculation unit (155).
[0054] The expertise vector generation unit (151) generates an expertise vector that quantitatively expresses the repair capabilities of each repair company based on past repair history data of multiple repair companies stored in the information management unit (110). The expertise vector generation unit (151) can construct the expertise vector by utilizing multiple performance indicators, such as the actual repair time, repair skill level, and billing cost in the repair work performed by each repair company. To this end, the expertise vector generation unit (151) can be designed so that a specific component of the expertise vector is evaluated more highly than that of a company that stably handles difficult tasks, by applying a weighted average-based aggregation operation or a performance-based normalization operation to multiple repair records.
[0055] For example, the expertise vector generation unit (151) can generate an expertise vector for each repair company according to the following mathematical formula 2.
[0056]
[0057] Here, STime is the representative repair time of each repair shop, SCost is the representative billing cost of each repair shop, and SLev is the representative repair skill level of each repair shop.
[0058] STime is calculated from the repair time spent on each repair case by the repair company and the level of repair skill for each repair case. For example, STime can be calculated according to the following mathematical formula 3.
[0059]
[0060] Here, i is an element of set I consisting of repair cases, and is an index representing each repair case. iis the graded repair skill level of the i-th repair case, and t i is the time required for the i-th repair case.
[0061] SCost is calculated from the repair costs, repair skill level, and repair time required for each repair case at the respective repair company. For example, SCost can be calculated according to the following mathematical formula 4.
[0062]
[0063] Here, l i is the graded repair skill level of the i-th repair case, and t i is the time required for the i-th repair case, and c i is the cost charged for the i-th repair case.
[0064] SLev is calculated based on the repair skill level and the number of corresponding repairs for each repair case of the relevant repair company. For example, SLev can be calculated according to the following mathematical formula 5.
[0065]
[0066] Here, L is the set of repair skill levels of the cases repaired by the company, l is the repair skill level, nl is the number of repair cases corresponding to a given repair skill l, and a and b are positive integers.
[0067] The fee weight calculation unit (153) analyzes the relationship between the amount claimed by each repair shop for each repair case (e.g., the amount claimed from the insurance company) and the actual settlement amount received, and calculates a fee weight indicating the appropriateness of the repair cost claim. The fee weight calculation unit (153) generates an indicator that reflects the appropriateness of the claim or the degree of overcharging tendency based on the ratio or difference between the amount claimed and the settlement amount, and calculates a weight by performing a non-linear transformation on the indicator. Through this, the fee weight can be configured to gradually decrease for companies that have a history of overcharging compared to the actual settlement amount or being repeatedly reduced by the insurance company, and can be set to maintain a high weight for companies that have consistently maintained appropriate billing practices.
[0068] For example, the rate weight calculation unit (153) can calculate the rate weight according to the following mathematical formula 6.
[0069]
[0070] Here, W is the rate weight, k is a positive constant, |I| is the size of the set of all repair cases, i.e., the total number of repair cases, and c i is the claimed amount for the i-th repair case, and b i is the settlement amount settled for the i-th repair case.
[0071] The recommendation score calculation unit (155) calculates a recommendation score for each repair company using the suitability between the repair difficulty vector calculated by the repair difficulty judgment unit (140) and the expertise vector calculated by the expertise vector generation unit (151), and the fee weight calculated by the fee weight calculation unit (153). Subsequently, the repair companies with the highest recommendation scores or those of the highest authority may be proposed as recommended companies.
[0072] For example, the recommendation score calculation unit (155) calculates recommendation scores for repair companies according to the following mathematical formula 7.
[0073]
[0074] Here, S is the recommendation score, sim is a function that calculates the cosine similarity between two vectors, and q is positive.
[0075] That is, the recommendation score calculation unit (155) applies a vector similarity calculation algorithm that calculates the degree of alignment based on the difference in direction or alignment between two vectors, thereby not only considering the technical suitability for repair work of the required difficulty level but also additionally considering cost transparency by reflecting the fee weighting.
[0076] The recommendation score calculation unit (155) ranks multiple repair companies based on the final recommendation score and provides a list of top recommended companies to the repair request terminal (201).
[0077] Accordingly, the recommended repair shop determination unit (150) of the present invention can provide high reliability and rationality compared to conventional repair shop selection methods based on simple distance or cost comparison by performing objective data-based judgments by comprehensively reflecting the performance, cost, and technical level of the repair shop. Furthermore, technical effects can be obtained that reduce cost disputes between the insurer, the vehicle owner, and the repair shop, reduce quality variations in repair results, and improve the transparency of the entire repair process.
[0078] According to the embodiments, the recommendation score calculation unit (155) may calculate the recommendation score by additionally considering whether to apply part remanufacturing according to the client's preference.
[0079] FIG. 9 is a diagram illustrating an exemplary hardware configuration of a management server according to embodiments of the present invention. The electronic device (300) of FIG. 9 may represent the management server (100) described with reference to FIG. 5 to FIG. 8.
[0080] Referring to FIG. 9, the electronic device (300) may include at least one processor (310) and a memory (320) that stores instructions that instruct the at least one processor (310) to perform at least one operation.
[0081] The above at least one operation is interpreted to include at least one of the operations of the aforementioned electronic device (300) or the operations of the functional part, and a specific description is omitted to prevent redundant explanation.
[0082] Here, at least one processor (310) may mean a central processing unit (CPU), a graphics processing unit (GPU), or a dedicated processor on which methods according to embodiments of the present invention are performed.
[0083] The memory (320) may be composed of at least one of a volatile storage medium and a non-volatile storage medium. For example, the memory (320) may be composed of at least one of a read-only memory (ROM) and a random access memory (RAM).
[0084] Additionally, the electronic device (300) may include a transceiver (330) that performs communication via a wireless network. Additionally, the electronic device (300) may further include an input interface device (340), an output interface device (350), a storage device (360, which may be referred to interchangeably with internal storage), etc. Each component included in the electronic device (300) may be connected by a bus (370) to communicate with one another.
[0085] The methods according to the present invention may be implemented in the form of program instructions that can be executed through various computer means and recorded on a computer-readable medium. The computer-readable medium may include program instructions, data files, data structures, etc., either alone or in combination. The program instructions recorded on the computer-readable medium may be those specifically designed and configured for the present invention, or they may be those known and available to those skilled in the art of computer software.
[0086] Examples of computer-readable media may include hardware devices specifically configured to store and execute program instructions, such as ROM, RAM, and flash memory. Examples of program instructions may include machine code, such as that generated by a compiler, as well as high-level language code that can be executed by a computer using an interpreter, etc. The aforementioned hardware devices may be configured to operate as at least one software module to perform the operation of the present invention, and vice versa.
[0087] In addition, the above-described method or device may be implemented by combining all or part of its configuration or function, or by implementing it separately.
[0088] Although the present invention has been described above with reference to preferred embodiments, those skilled in the art will understand that various modifications and changes can be made to the invention without departing from the spirit and scope of the invention as described in the following claims.
Claims
Claim 1 In a management server of an automotive parts remanufacturing-based repair system that manages repair requesters and repair shops, the system comprises: an information management unit configured to manage repair shop information associated with repair shops; a repair request receiving unit configured to receive a repair request from a repair requester and an image of a damaged part; a damage degree determination unit configured to generate damage characteristic information indicating the degree of damage from the image of the damaged part; and a repair difficulty determination unit configured to generate repair difficulty information indicating the repair difficulty required to repair the damaged part based on the damage characteristic information. The system includes a recommended repair shop determination unit configured to determine a recommended repair shop suitable for repairing the damaged part based on information about the repair shops and information on the repair difficulty, wherein the repair shop information includes, for each of the repair shops, information on the repair part, vehicle, repair period, repair difficulty, billing cost, and settlement amount per repair case, and the damage degree determination unit includes an image preprocessing unit configured to perform preprocessing on the image of the damaged part; a damage feature extraction unit configured to extract a damage feature map representing visual features significant for damage determination by performing multiple stages of convolutional and pooling operations based on a Convolutional Neural Network (CNN) on the preprocessed image of the damaged part; and a damage feature vector calculation unit configured to receive the damage feature map and generate a damage feature vector representing damage indicators that quantitatively express the state of damage, and the repair difficulty determination unit includes a repair feature vector generation unit configured to generate a repair feature vector, which is an input vector for determining the repair difficulty, by combining the damage feature vector and information related to the damaged part.and includes a repair difficulty vector calculation unit configured to receive the above repair characteristic vector as input, apply a deep learning-based prediction model for determining repair difficulty, and calculate as output a repair difficulty vector representing the repair difficulty of the damaged part, wherein the repair difficulty vector is determined according to the following mathematical formula 1, [Mathematical Formula 1]; (Here, RDV is a repair difficulty vector, RTime is the estimated repair time, RCost is the estimated repair cost, and RLev represents the graded repair skill level required for the repair.) The recommended repair shop determination unit comprises: a professionalism vector generation unit that generates a professionalism vector for each repair shop based on the past repair history data of the repair shops; a fee weight calculation unit that calculates a fee weight indicating the appropriateness of the repair cost claim based on the relationship between the insurance claim amount claimed by each of the repair shops for each repair case and the settlement amount actually received; and a recommendation score calculation unit that calculates a recommendation score for each of the repair shops by applying the fee weight to the degree of alignment calculated based on the directional difference or degree of alignment between the repair difficulty vector and the professionalism vector; wherein the fee weight is determined according to the following mathematical formula 2, [Mathematical Formula 2] (In the above mathematical formula 2, W is the fee weight, k is a positive constant, |I| is the total number of repair cases, and c i is the insurance claim amount for the i-th repair case, and b i is the settlement amount settled for the i-th repair case) management server. Claim 2 delete Claim 3 delete
Citation Information
Patent Citations
Method and system for providing costomized garage information
KR1020220094516A
Apparatus for Diagnosing Replacement or Restoration of Broken Glass in Vehicle
KR1020250110577A
