A personnel identification and sending system for smart work order calls

Abstract

The present invention relates to the technical field of call work order processing, and specifically discloses a personnel identification and sending system for smart work order calls, including a voice collection module, a voice conversion module, an information processing module, a personnel identification module, a task scheduling module, and a work order allocation module. This solution automatically generates work orders and sets priorities by identifying call voices, reduces the workload of manual entry and management, intelligently identifies the task type in the work order and extracts relevant information, uses the most relevant maintenance type as the task entity of the work order, reduces the risk of misjudgment, comprehensively captures various requirements of the work order, and provides a detailed basis for subsequent personnel matching and task allocation; by mapping the task entity and the personnel entity and the relationship between entities, an intelligent and dynamically updated knowledge graph is constructed to improve the accuracy of task matching, has strong adaptability and flexibility, and ensures the optimal use of resources and efficient completion of tasks.

Description

Technical Field

[0001] The present invention relates to the technical field of call work order processing, and specifically refers to a personnel identification and sending system for smart work order calls. Background Art

[0002] The smart work order call system is a scenario where a large number of customer requests need to be processed efficiently, such as customer service, maintenance service, and support center. General call order processing relies on manual entry, which is prone to entry errors, delays, and increased management burden. The use of simple allocation rules leads to unreasonable priority allocation, lack of adjustment, and low efficiency. General task matching and resource optimization methods are difficult to fully integrate task and personnel information, and data updates slowly, resulting in information lags and inaccurate resource matching, as well as poor adaptability and waste of resources. Summary of the invention

[0003] In view of the above situation, in order to overcome the defects of the prior art, the present invention provides a personnel identification and sending system for smart work order calls. In view of the fact that general call work order processing relies on manual input, which is prone to input errors, delays and increased management burden, and the use of simple allocation rules has the problems of unreasonable priority allocation, lack of adjustment and low efficiency, this solution recognizes the call voice, automatically generates work orders and sets priorities, reduces the workload of manual input and management, intelligently identifies the task type in the work order and extracts relevant information, takes the maintenance type with the highest correlation as the task entity of the work order, reduces the risk of misjudgment, comprehensively captures the various requirements of the work order, and provides detailed basis for subsequent personnel matching and task allocation; in view of the fact that general matching methods are difficult to fully integrate dynamic resources, information is lagging, and resource matching is inaccurate, resulting in poor personnel identification accuracy, this solution maps the task entity and the personnel entity and the relationship between entities, constructs an intelligent and dynamically updated knowledge graph, improves the accuracy of task matching, and dynamically adjusts the personnel candidate set according to actual conditions. It has strong adaptability and flexibility, ensuring the optimal utilization of resources and efficient completion of tasks.

[0004] The present invention provides a personnel identification and sending system for intelligent work order calls, comprising a voice collection module, a voice conversion module, an information processing module, a personnel identification module, a task scheduling module, and a work order distribution module;

[0005] The voice collection module collects traffic voice in real time, sets an emotional state category for the traffic voice, and sends the traffic voice and the emotional state category to the voice conversion module;

[0006] The speech conversion module converts the traffic speech into traffic text, pre-processes the traffic text and automatically generates a work order, sets the priority of the work order according to the emotional state category, and sends the work order and its priority to the information processing module and the work order allocation module;

[0007] The information processing module decomposes the work order, constructs the task entity and its attributes, and sends the task entity and its attributes to the personnel identification module;

[0008] The personnel identification module reads the personnel database containing all maintenance personnel information, extracts personnel entities and their attributes, screens the maintenance personnel candidate set, and sends the maintenance personnel candidate set to the task scheduling module;

[0009] The task scheduling module is provided with a functional unit, which selects the best matching personnel in combination with the maintenance personnel information updated in real time, and sends the best matching personnel information to the work order allocation module;

[0010] The work order distribution module sends the work order to the best matching person, who confirms the work order after receiving it, and the system continuously tracks the processing progress of the work order.

[0011] Furthermore, the information processing module decomposes the work order and constructs the task entity and its attributes, including the following steps:

[0012] Step S1: predefine a maintenance type set;

[0013] Step S2: Decompose the work order content into individual words, identify and normalize synonyms and near-synonyms;

[0014] Step S3: Calculate the importance index of all words in each work order, set a similarity threshold, and the words with an importance index higher than the similarity threshold are the keywords of the work order, and the keywords are not less than one;

[0015] Step S4: Correlation analysis, calculate the correlation between the keywords in the work order and various maintenance types, and match the work order with the maintenance type with the highest correlation. The formula used is as follows:

[0016] ;

[0017] In the formula, Indicates keywords, represents any maintenance type in the maintenance type set, represents the weight, Indicates keywords and repair type The relevance of Indicates keywords In repair type The probability distribution in Indicates keywords Overall probability across all repair types;

[0018] Step S5: Construct a task entity, take the maintenance type with the highest correlation as the task entity of the work order, and match the attributes of the task entity from the vocabulary of the work order.

[0019] Furthermore, the personnel identification module filters the maintenance personnel candidate set by constructing a knowledge graph, including the following steps:

[0020] Step A1: Initialize the knowledge graph, take task entities and personnel entities as nodes, and use labels to distinguish node types; calculate the matching relationship between entities based on the attributes of task entities and personnel entities, and use the matching relationship as the edge connecting the nodes;

[0021] Step A2: Initialize entities and matching relations into vector form, represented as entity embedding and relation embedding respectively;

[0022] Step A3: Encoder construction, using the encoder architecture to learn semantic aggregation embedding, structural aggregation embedding, and relation aggregation embedding from the input entities and matching relations;

[0023] Step A4: Decoder construction, using the decoder architecture to build a knowledge graph based on semantic aggregation embedding, structural aggregation embedding, and relational aggregation embedding;

[0024] Step A5: Candidate set screening, calculating the similarity between the task entity and the personnel entity, and storing the information of the top k maintenance personnel with the highest similarity in the maintenance personnel candidate set.

[0025] Further, in step A3, the encoder is constructed, comprising the following steps:

[0026] Step A31: Construct an entity semantic aggregation model, including an RNN network, a node aggregation layer, and an embedding merging layer. The RNN network captures the semantic similarity between the task entity and the person entity. The node aggregation layer aggregates nodes according to the type of edge. The embedding merging layer transfers the semantic information of the person entity to the task entity. The initial entity embedding is input and the final semantic aggregation embedding is output. The formula used is as follows:

[0027] ;

[0028] In the formula, Indicates Semantic aggregation embedding of layers, represents the activation function, Represents a matching relationship Sum, represents the person entity embedding, represents relation embedding, Represents the task entity All connected person entities And matching relationship To sum, represents the number of encoder layers, represents the normalization constant, Represents the entity semantic aggregation model The weight matrix of the layer, Representing Person Entity Embedding After RNN network processing The output of the layer;

[0029] Step A32: Construct an entity structure aggregation model, including a GNN network, a multi-head attention mechanism layer and a node aggregation layer. The GNN network learns the joint embedding of entities and matching relationships. The multi-head attention mechanism layer enhances the structural representation relationship. The node aggregation layer aggregates the structural information obtained from the person entity and the matching relationship. The initial entity embedding and relationship embedding are input, and the final structural aggregation embedding is output. The formula used is as follows:

[0030] ;

[0031] In the formula, Indicates The structure of the layers is aggregated and embedded. Indicates The attention weights of the layers, Represents the entity structure aggregation model The weight matrix of the layer, Representing Person Entity Embedding and Person entity Connected Relation Embedding Combination operation of

[0032] Step A33: Build an explicit relation aggregation model, use the GRU network to enhance the representation of matching relations, and input relation embedding , the final relation aggregation embedding is obtained through the activation function, and the formula used is as follows:

[0033] ;

[0034] In the formula, Indicates Layer-wise relation aggregation embedding, Indicates the explicit relationship aggregation model The weight matrix of the layer, Representing relation embeddings After GRU network processing The output of the layer.

[0035] The beneficial effects achieved by the present invention using the above scheme are as follows:

[0036] (1) The processing of general call work orders relies on manual input, which is prone to input errors, delays and increased management burden. In addition, the use of simple allocation rules has problems such as unreasonable priority allocation, lack of adjustment and low efficiency. This solution recognizes the call voice, automatically generates work orders and sets priorities, reduces the workload of manual input and management, intelligently identifies the task type in the work order and extracts relevant information, and uses the most relevant maintenance type as the task entity of the work order, reducing the risk of misjudgment. It fully captures the various requirements of the work order and provides a detailed basis for subsequent personnel matching and task allocation.

[0037] (2) In view of the problem that general matching methods are difficult to fully integrate dynamic resources, information is lagging, and resource matching is inaccurate, resulting in poor accuracy in personnel identification, this solution maps task entities and personnel entities and the relationships between entities to build an intelligent and dynamically updated knowledge graph to improve the accuracy of task matching. It dynamically adjusts the candidate set of personnel according to actual conditions and has strong adaptability and flexibility, ensuring the optimal utilization of resources and efficient completion of tasks. BRIEF DESCRIPTION OF THE DRAWINGS

[0038] Figure 1 A schematic diagram of a personnel identification and sending system for intelligent work order calls proposed by the present invention;

[0039] Figure 2 Schematic diagram of the knowledge graph construction process in the person identification module.

[0040] The accompanying drawings are used to provide further understanding of the present invention and constitute a part of the specification. They are used to explain the present invention together with the embodiments of the present invention and do not constitute a limitation of the present invention. DETAILED DESCRIPTION

[0041] The technical solutions in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, rather than all the embodiments; based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative work are within the scope of protection of the present invention.

[0042] Example 1, see Figure 1 , a personnel identification and sending system for smart work order traffic provided by the present invention includes a voice collection module, a voice conversion module, an information processing module, a personnel identification module, a task scheduling module, and a work order allocation module;

[0043] The voice collection module collects traffic voice in real time, sets an emotional state category for the traffic voice, and sends the traffic voice and the emotional state category to the voice conversion module;

[0044] The speech conversion module converts the traffic speech into traffic text, pre-processes the traffic text and automatically generates a work order, sets the priority of the work order according to the emotional state category, and sends the work order and its priority to the information processing module and the work order allocation module;

[0045] The information processing module decomposes the work order, constructs the task entity and its attributes, and sends the task entity and its attributes to the personnel identification module;

[0046] The personnel identification module reads the personnel database containing all maintenance personnel information, extracts personnel entities and their attributes, screens the maintenance personnel candidate set, and sends the maintenance personnel candidate set to the task scheduling module;

[0047] The task scheduling module is provided with a functional unit, which selects the best matching personnel in combination with the maintenance personnel information updated in real time, and sends the best matching personnel information to the work order allocation module;

[0048] The work order distribution module sends the work order to the best matching person, who confirms the work order after receiving it, and the system continuously tracks the processing progress of the work order.

[0049] Example 2, see Figure 1 This embodiment is based on the above embodiment. The speech collection module divides the emotional state into three categories: high, medium and low. The pre-trained Wav2Vec model is used to recognize the intonation, timbre, volume and rhythm characteristics in the traffic speech, and the emotional state category of the traffic speech is set.

[0050] Example 3, see Figure 1 This embodiment is based on the above embodiment. The speech conversion module uses a pre-trained Whisper model to convert the traffic speech into traffic text. According to semantic segmentation, noise words and stop words are removed, and punctuation is processed to automatically generate a work order. The work order content includes the work order number, user account, mobile phone number, fault address, fault cause, and work order source; the priority of the work order is set according to the emotional state, and the priority 1 represents the emotional state category of high, the priority 2 represents the emotional state category of medium, and the priority 3 represents the emotional state category of low.

[0051] Example 4, see Figure 1 This embodiment is based on the above embodiment. The information processing module decomposes the work order and constructs the task entity and its attributes, including the following steps:

[0052] Step S1: predefine a set of maintenance types, including equipment maintenance, equipment upgrade, parts replacement, network fault repair, and regular maintenance;

[0053] Step S2: Decompose the work order content into individual words, identify and normalize synonyms and near-synonyms;

[0054] Step S3: Calculate the word frequency-inverse text frequency index of the vocabulary, and set the similarity threshold to 0.7. The vocabulary with a word frequency-inverse text frequency index higher than the similarity threshold is the keyword of the work order. The formula used is as follows:

[0055] ;

[0056] ;

[0057] ;

[0058] In the formula, Represents each individual word, Indicates a work order. Words to express In the work order The frequency of occurrence in Words to express In the work order The number of times it appears in Indicates a work order The total number of words, Indicates all work orders. Words to express In all work orders Rareness in Indicates the total number of work orders. Indicates that it contains words The number of work orders, Words to express In the work order word frequency in - inverse text frequency index;

[0059] Step S4: Correlation analysis, calculate the correlation between the keywords in the work order and various maintenance types, and match the work order with the maintenance type with the highest correlation. The formula used is as follows:

[0060] ;

[0061] In the formula, Indicates keywords, represents any maintenance type in the maintenance type set, represents the weight, Indicates keywords and repair type The relevance of Indicates keywords In repair type The probability distribution in Indicates keywords Overall probability across all repair types;

[0062] Step S5: Construct a task entity, take the maintenance type with the highest correlation as the task entity of the work order, and match the attributes of the task entity from the vocabulary of the work order, including task name, start time, end time, and priority.

[0063] Example 5, see Figure 1 This embodiment is based on the above embodiment, and the personnel identification module extracts the personnel entity and its attributes. The attributes of the personnel entity include the maintenance personnel's number, name, location, skills, and current task status.

[0064] By performing the above operations, the processing of general call work orders relies on manual entry, which is prone to entry errors, delays and increased management burden, and the use of simple allocation rules has problems such as unreasonable priority allocation, lack of adjustment and low efficiency. This solution recognizes call voice, automatically generates work orders and sets priorities, reduces the workload of manual entry and management, intelligently identifies the task type in the work order and extracts relevant information, takes the most relevant maintenance type as the task entity of the work order, reduces the risk of misjudgment, comprehensively captures the various requirements of the work order, and provides a detailed basis for subsequent personnel matching and task allocation.

[0065] Example 6, see Figure 1 and Figure 2 This embodiment is based on the above embodiment. The personnel identification module selects a candidate set of maintenance personnel by constructing a knowledge graph, including the following steps:

[0066] Step A1: Initialize the knowledge graph, take task entities and personnel entities as nodes, and use labels to distinguish node types; calculate the matching relationship between entities based on the attributes of task entities and personnel entities, and use the matching relationship as the edge connecting the nodes;

[0067] Step A2: Initialize entities and matching relations into vector form, represented as entity embedding and relation embedding respectively;

[0068] Step A3: Encoder construction, using the encoder architecture to learn semantic aggregation embedding, structural aggregation embedding, and relation aggregation embedding from the input entities and matching relations, including the following steps:

[0069] Step A31: Construct an entity semantic aggregation model, including an RNN network, a node aggregation layer, and an embedding merging layer. The RNN network captures the semantic similarity between the task entity and the person entity. The node aggregation layer aggregates nodes according to the type of edge. The embedding merging layer transfers the semantic information of the person entity to the task entity. The initial entity embedding is input and the final semantic aggregation embedding is output. The formula used is as follows:

[0070] ;

[0071] In the formula, Indicates Semantic aggregation embedding of layers, represents the activation function, Represents a matching relationship Sum, represents the person entity embedding, represents relation embedding, Represents the task entity All connected person entities And matching relationship To sum, represents the number of encoder layers, represents the normalization constant, Represents the entity semantic aggregation model The weight matrix of the layer, Representing Person Entity Embedding After RNN network processing The output of the layer;

[0072] Step A32: Construct an entity structure aggregation model, including a GNN network, a multi-head attention mechanism layer and a node aggregation layer. The GNN network learns the joint embedding of entities and matching relationships. The multi-head attention mechanism layer enhances the structural representation relationship. The node aggregation layer aggregates the structural information obtained from the person entity and the matching relationship. The initial entity embedding and relationship embedding are input, and the final structural aggregation embedding is output. The formula used is as follows:

[0073] ;

[0074] In the formula, Indicates The structure of the layers is aggregated and embedded. Indicates The attention weights of the layers, Represents the entity structure aggregation model The weight matrix of the layer, Representing Person Entity Embedding and Person entity Connected Relation Embedding Combination operation of

[0075] Step A33: Build an explicit relation aggregation model, use the GRU network to enhance the representation of matching relations, and input relation embedding , the final relation aggregation embedding is obtained through the activation function, and the formula used is as follows:

[0076] ;

[0077] In the formula, Indicates Layer-wise relation aggregation embedding, Indicates the explicit relationship aggregation model The weight matrix of the layer, Representing relation embeddings After GRU network processing The output of the layer;

[0078] Step A4: Decoder construction. Use the decoder architecture to build a knowledge graph based on semantic aggregation embedding, structural aggregation embedding, and relational aggregation embedding, including embedding merging layer, convolution layer, and fully connected layer. The specific steps are as follows:

[0079] Step A41: The embedding merging layer merges the semantic aggregation embedding and the structural aggregation embedding to form the final representation of the entity. The formula used is as follows:

[0080] ;

[0081] In the formula, represents the final representation of the entity, Represents the operation of merging semantic aggregation embedding and structural aggregation embedding;

[0082] Step A42: The convolutional layer performs a convolution operation on the merged embedding to capture the local pattern of the embedding;

[0083] Step A43: The fully connected layer completes the missing relations and entities in the knowledge graph;

[0084] Step A5: Candidate set screening, calculate the cosine similarity between the task entity and the personnel entity, and store the information of the top k maintenance personnel with the highest cosine similarity in the maintenance personnel candidate set. The formula used is as follows:

[0085] ;

[0086] In the formula, Represents a task entity and Person entity The cosine similarity between Represents the index of the task entity, Represents the index of the person entity.

[0087] By performing the above operations, the general matching methods are difficult to fully integrate dynamic resources, information is delayed, and resource matching is inaccurate, resulting in poor accuracy in personnel identification. This solution maps task entities and personnel entities and the relationship between entities to build an intelligent and dynamically updated knowledge graph to improve the accuracy of task matching. It dynamically adjusts the candidate set of personnel according to actual conditions, has strong adaptability and flexibility, and ensures optimal utilization of resources and efficient completion of tasks.

[0088] Embodiment 7, see Figure 1 This embodiment is based on the above embodiment, and the task scheduling module sets the functional unit, including the following steps:

[0089] Step T1: Set up the work order life cycle, including the functions of adding a new work order, temporarily storing a work order, issuing a work order, and transferring a work order;

[0090] Step T2: sorting optimization, checking the workload and customer distance of the maintenance personnel in the candidate set of maintenance personnel, and optimizing the sorting of the candidate personnel;

[0091] Step T3: Screening of the best matching personnel. The best matching personnel are screened in order. If the current matching personnel has task conflicts and resource allocation conflicts, they will be replaced by the next matching personnel. If all candidate personnel cannot be matched, they will be reported to the administrator for manual intervention.

[0092] It should be noted that, in this article, relational terms such as first and second, etc. are only used to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Moreover, the terms "include", "comprise" or any other variants thereof are intended to cover non-exclusive inclusion, so that a process, method, article or device including a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, method, article or device.

[0093] Although embodiments of the present invention have been shown and described, it will be appreciated by those skilled in the art that various changes, modifications, substitutions and variations may be made to the embodiments without departing from the principles and spirit of the present invention, and that the scope of the present invention is defined by the appended claims and their equivalents.

[0094] The present invention and its embodiments are described above, and such description is not restrictive. The drawings show only one embodiment of the present invention, and the actual structure is not limited thereto. In short, if ordinary technicians in the field are inspired by it, without departing from the purpose of the invention, they can design a structure and embodiment similar to the technical solution without creativity, which should belong to the protection scope of the present invention.

Claims

1. A personnel identification and sending system for smart work order traffic, characterized by: It includes voice collection module, voice conversion module, information processing module, personnel identification module, task scheduling module and work order allocation module; The voice collection module collects traffic voice in real time, sets an emotional state category for the traffic voice, and sends the traffic voice and the emotional state category to the voice conversion module; The speech conversion module converts the traffic speech into traffic text, pre-processes the traffic text and automatically generates a work order, sets the priority of the work order according to the emotional state category, and sends the work order and its priority to the information processing module and the work order allocation module; The information processing module decomposes the work order, constructs the task entity and its attributes, and sends the task entity and its attributes to the personnel identification module; The personnel identification module reads the personnel database containing all maintenance personnel information, extracts personnel entities and their attributes, screens the maintenance personnel candidate set, and sends the maintenance personnel candidate set to the task scheduling module; The task scheduling module is provided with a functional unit, which selects the best matching personnel in combination with the maintenance personnel information updated in real time, and sends the best matching personnel information to the work order allocation module; The work order distribution module sends the work order to the best matching person, who confirms the work order after receiving it, and the system continuously tracks the processing progress of the work order; The personnel identification module selects a candidate set of maintenance personnel by constructing a knowledge graph, including the following steps: Step A1: Initialize the knowledge graph, take task entities and personnel entities as nodes, and use labels to distinguish node types; calculate the matching relationship between entities based on the attributes of task entities and personnel entities, and use the matching relationship as the edge connecting the nodes; Step A2: Initialize entities and matching relations into vector form, represented as entity embedding and relation embedding respectively; Step A3: Encoder construction, using the encoder architecture to learn semantic aggregation embedding, structural aggregation embedding, and relation aggregation embedding from the input entities and matching relations; Step A4: Decoder construction, using the decoder architecture to build a knowledge graph based on semantic aggregation embedding, structural aggregation embedding, and relational aggregation embedding; Step A5: Candidate set screening, calculating the similarity between the task entity and the personnel entity, and storing the information of the top k maintenance personnel with the highest similarity in the maintenance personnel candidate set; In step A3, the encoder is constructed, comprising the following steps: Step A31: Construct an entity semantic aggregation model, including an RNN network, a node aggregation layer, and an embedding merging layer. The RNN network captures the semantic similarity between the task entity and the person entity. The node aggregation layer aggregates nodes according to the type of edge. The embedding merging layer transfers the semantic information of the person entity to the task entity. The initial entity embedding is input and the final semantic aggregation embedding is output. The formula used is as follows: ； In the formula, Indicates Semantic aggregation embedding of layers, represents the activation function, Represents a matching relationship Sum, represents the person entity embedding, represents relation embedding, Represents the task entity All connected person entities And matching relationship To sum, represents the number of encoder layers, represents the normalization constant, Represents the entity semantic aggregation model The weight matrix of the layer, Representing Person Entity Embedding After RNN network processing The output of the layer; Step A32: Construct an entity structure aggregation model, including a GNN network, a multi-head attention mechanism layer and a node aggregation layer. The GNN network learns the joint embedding of entities and matching relationships. The multi-head attention mechanism layer enhances the structural representation relationship. The node aggregation layer aggregates the structural information obtained from the person entity and the matching relationship. The initial entity embedding and relationship embedding are input, and the final structural aggregation embedding is output. The formula used is as follows: ； In the formula, Indicates The structure of the layers is aggregated and embedded. Indicates The attention weights of the layers, Represents the entity structure aggregation model The weight matrix of the layer, Representing Person Entity Embedding and Person entity Connected Relation Embedding Combination operation of Step A33: Build an explicit relation aggregation model, use the GRU network to enhance the representation of matching relations, and input relation embedding , the final relation aggregation embedding is obtained through the activation function, and the formula used is as follows: ； In the formula, Indicates Layer-wise relation aggregation embedding, Indicates the explicit relationship aggregation model The weight matrix of the layer, Representing relation embeddings After GRU network processing The output of the layer.

2. According to claim 1, a personnel identification and sending system for intelligent work order traffic is characterized in that: The information processing module decomposes the work order and constructs the task entity and its attributes, including the following steps: Step S1: predefine a maintenance type set; Step S2: Decompose the work order content into individual words, identify and normalize synonyms and near-synonyms; Step S3: Calculate the importance index of all words in each work order, set a similarity threshold, and the words with an importance index higher than the similarity threshold are the keywords of the work order, and the keywords are not less than one; Step S4: Correlation analysis, calculate the correlation between the keywords in the work order and various maintenance types, and match the work order with the maintenance type with the highest correlation. The formula used is as follows: ； In the formula, Indicates keywords, represents any maintenance type in the maintenance type set, represents the weight, Indicates keywords and repair type The relevance of Indicates keywords In repair type The probability distribution in Indicates keywords Overall probability across all repair types; Step S5: Construct a task entity, take the maintenance type with the highest correlation as the task entity of the work order, and match the attributes of the task entity from the vocabulary of the work order.

Images