An edge cloud cooperation-based material whole-process trusted management and control system

Abstract

This invention discloses a reliable end-to-end material management system based on edge cloud collaboration, belonging to the field of material management technology. It includes a material requisition module: used to receive and generate electronic material requisition forms via edge-side terminal devices and upload them to the cloud; and a process approval module: used to deploy a preset approval rule base and organizational rule base on the cloud, combine real-time operational data to generate a dynamic approval path for the electronic material requisition form, and issue approval tasks to the approver's terminal. This invention achieves intelligent routing and adaptive optimization of the approval process through a dynamic approval relationship network and rule-driven path generation, improving the response speed and approval efficiency for urgent needs; through an operation and maintenance decision model, it calculates available inventory and shortage risks in real time and feeds them back to the approval stage to execute differentiated reservation strategies, achieving early warning of shortage risks and precise intelligent allocation of resources, preventing improper resource allocation.

Description

Technical Field

[0001] This invention relates to the field of material management technology, specifically to a reliable end-to-end material management and control system based on edge cloud collaboration. Background Technology

[0002] In industrial manufacturing, engineering construction, and large-scale project management, materials management is a core element for ensuring production continuity and controlling project costs. For a long time, materials management has relied primarily on manual operations and paper-based document processing, resulting in inherent drawbacks such as low efficiency, high error rates, information lag, and difficulty in traceability. Although subsequent information systems based on client / server or single cloud architectures have emerged, achieving data digitization, they still fall short of meeting the modern enterprise's demands for real-time, accurate, and reliable control.

[0003] For example, invention patent CN113469558A discloses a material management method and system. This system includes: a material information maintenance module for maintaining material inventory information; a requisition module for receiving material requisition information from applicants; an approval module for generating approval information based on stored information and material requisition information; and a financial information processing module for generating corresponding financial data after material requisition is completed. Applicants in each work group can fill in material requisition information in advance on the system. The administrator confirms the materials needed by each work group based on the requisition information and distributes them to designated locations. Work groups only need to collect materials at designated times and locations each day, reducing daily waiting time. Furthermore, through information-based requisition forms, daily warehouse inflows and outflows are automatically recorded in the system, improving the efficiency of subsequent account maintenance and data analysis for administrators.

[0004] However, the above and similar technical solutions still have the following shortcomings: the approval path is rigid, making it difficult to deal with special situations in actual business, resulting in slow response to urgent needs and low approval efficiency; the approval process lacks deep integration with inventory information, making it impossible to identify and warn of potential shortage risks in a timely manner during the approval process, resulting in improper allocation of resources. Summary of the Invention

[0005] The purpose of this invention is to provide a reliable material management and control system based on edge cloud collaboration to solve the problems mentioned in the background art.

[0006] To achieve the above objectives, the present invention provides the following technical solution: a reliable material management and control system based on edge cloud collaboration, comprising: Material requisition module: Used to receive and generate electronic material requisition forms through edge terminal devices and upload them to the cloud; Workflow approval module: This module is used to deploy a pre-set approval rule base and organizational rule base on the cloud, combine real-time operation data, generate a dynamic approval path for electronic material requisition forms, and issue approval tasks to the approver's terminal. Warehouse Management Module: Based on the warehouse optimization model, it continuously analyzes the core characteristics of materials to automatically allocate storage locations for incoming materials and plan dynamic paths for inventory count operations; based on the operation and maintenance decision model, it performs inventory verification and reservation according to the approval process status and automatically warns of shortage risks. The material issuance module generates a unique material issuance voucher for approved electronic material requisition forms; it verifies the voucher to complete the material issuance and synchronizes inventory data to the cloud in real time based on the actual issuance quantity. Data Analysis and Traceability Module: Used to aggregate and analyze data from the entire material process, generate multi-dimensional analysis reports, and build a digital responsibility chain based on blockchain-style logs.

[0007] Furthermore, the process approval module specifically includes: A dynamic approval relationship network is constructed based on the organizational rule base and real-time operational data; The system automatically parses and extracts key feature information from electronic material requisition forms, and determines their process type based on an approval rule base; the process types include standard processes, emergency processes, and temporary borrowing processes. Based on the approval rule base and dynamic approval relationship network, dynamic approval paths are automatically generated according to the determined process type. Based on the generated dynamic approval path, the approval tasks are driven to execute sequentially or in parallel along the path nodes, and the global process status is updated and synchronized in real time. Through real-time monitoring and periodic analysis, abnormal conditions at different levels can be identified, and corresponding intervention mechanisms can be automatically triggered.

[0008] Furthermore, the method for constructing the warehouse optimization model includes: Based on the preset core feature dimensions of materials, a dynamic feature vector is constructed for each material; the core feature dimensions of materials include: material information, frequency of entry and exit from the warehouse, correlation between material requisition, physical size and storage requirements, and frequency of supplier entry into the warehouse; Based on the dynamic feature vectors of each material, the materials are divided into several storage strategy clusters, and corresponding storage location area allocation rules are generated for each strategy cluster. When materials are put into storage, the real-time feature vector of the material is calculated, and it is classified into a target storage strategy cluster based on similarity, and the corresponding target storage location area is obtained accordingly. Within the target storage area, a unique recommended storage location is output based on a predefined multi-objective optimization function in real time. The multi-objective optimization function is used to balance picking efficiency, space utilization, and workload balance. Based on the dynamic feature vectors of materials, the confidence level of inventory data for each material is calculated, and an inventory count task list is dynamically generated; through reinforcement learning algorithms, inventory count routes are planned for warehouse managers.

[0009] Furthermore, the method for constructing the operation and maintenance decision model includes: When approval is triggered, the current inventory data is obtained in real time, and the theoretical real-time available total amount of required materials is calculated by combining the allocated quantity and the inventory in transit. The theoretical real-time available total is compared with the dynamic safety threshold. When it falls below the threshold, an inventory shortage warning is automatically generated and sent to the applicant and the current approver. The theoretical real-time available total amount of materials and inventory shortage early warning signals are fed back to the current approval process node in real time, and the approval rules are dynamically adjusted. Once approved, a differentiated reservation strategy will be implemented based on the priority of the electronic material requisition form attributes.

[0010] Furthermore, the content of implementing a differentiated reservation strategy based on the priority of electronic material requisition attributes includes: Based on a pre-set reserve rule library, priority levels are automatically assigned and reserve strategies are matched according to the key feature information of the electronic material requisition form; Implement differentiated reservations; Lock and reserve the first priority electronic material requisition form, immediately deduct the available inventory and generate a reservation mark; Implement flexible reservation for second-priority electronic material requisition forms, update the committable quantity, and set priority tags and expiration periods; Continuously monitor reservation records, automatically switch reservation modes and dynamically adjust reservation quantities based on inventory status and pre-defined business rules, and establish a full-process tracking record from reservation creation to termination.

[0011] Furthermore, the execution issuance module specifically includes: A unique encrypted material requisition voucher is generated for each approved electronic material requisition form, and the lifecycle status of the material requisition voucher is tracked throughout the entire process; the lifecycle status includes generation, verification, and cancellation. Verify the validity of material requisition vouchers and send picking instructions to the warehouse manager; automatically record the actual quantity issued through intelligent data collection devices and update inventory data in real time; During the material issuance process, the status of the material requisition voucher and the deviation in the issuance quantity are monitored in real time, and hierarchical alarm information is automatically generated for any detected anomalies.

[0012] Furthermore, the electronic material requisition form includes material code, name, application quantity, application type, and intended use information; the real-time operational data includes the approver's on-duty status, current pending task load, and historical average approval time.

[0013] Compared with the prior art, the beneficial effects of the present invention are: A reliable material management and control system based on edge cloud collaboration enables intelligent allocation of storage locations and inventory route planning through a warehouse optimization model and dynamic feature analysis, thereby improving warehouse space utilization and operational efficiency. Through full-process data analysis and a blockchain-based log chain, it achieves business transparency and accountability, supporting continuous optimization and decision support.

[0014] Meanwhile, through a dynamic approval relationship network and rule-driven path generation, intelligent routing and adaptive optimization of the approval process are achieved, improving the response speed and approval efficiency for urgent needs. Through the operation and maintenance decision model, available inventory and shortage risks are calculated in real time and fed back to the approval process to implement differentiated reservation strategies, achieving early warning of shortage risks and precise intelligent allocation of resources, preventing improper resource allocation. Attached Figure Description

[0015] Figure 1 This is a schematic diagram of the material control system of the present invention; Figure 2 This is a schematic diagram of the approval process of the present invention; Figure 3 This is a schematic diagram of the warehouse optimization model of the present invention; Figure 4 This is a schematic diagram of the operation and maintenance decision-making model of the present invention; Figure 5 This is a schematic diagram of the material dispensing method of the present invention. Detailed Implementation

[0016] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0017] like Figure 1 As shown, the present invention provides a technical solution: a reliable material management and control system based on edge cloud collaboration, comprising: Material requisition module: Used to receive and generate electronic material requisition forms through edge terminal devices and upload them to the cloud.

[0018] The applicant initiates the material requisition application on the edge terminal, collecting identity data through the terminal's integrated biometric sensor or RFID / NFC reader. The collected identity information is then compared with a trusted identity directory synchronized in the cloud to complete identity verification. Upon successful verification, application permissions are granted. The applicant enters basic material information (such as code, name, type), application quantity, total value, urgency, and application type by scanning material labels, selecting from a standardized list, or manually inputting data. The system automatically associates the applicant's department and project information based on their logged-in identity. After information entry, the terminal executes front-end validation logic to instantly verify the data's completeness, format (e.g., quantity must be a positive integer), and reasonableness. Upon successful verification, the terminal serializes the form data into a structured data object (such as JSON format) and uploads it to the cloud platform's receiving interface. After successfully receiving and persistently storing the data, the cloud platform returns a globally unique material requisition form number to the terminal. The edge terminal then updates its local application status based on this number, completing the electronic material requisition form generation process.

[0019] Workflow approval module: It is used to deploy a preset approval rule base and organizational rule base on the cloud, combine real-time operation data, generate dynamic approval paths for electronic material requisition forms, and issue approval tasks to the approver's terminal.

[0020] like Figure 2 As shown, the present invention provides a process approval method; Specifically: A dynamic approval relationship network is constructed based on the organizational rule base and real-time operational data; Automatically parse and extract key feature information from electronic material requisition forms, and determine their process type based on the approval rule base; Based on the approval rule base and dynamic approval relationship network, dynamic approval paths are automatically generated according to the determined process type. Based on the generated dynamic approval path, the approval tasks are driven to execute sequentially or in parallel along the path nodes, and the global process status is updated and synchronized in real time. Through real-time monitoring and periodic analysis, abnormal conditions at different levels can be identified, and corresponding intervention mechanisms can be automatically triggered.

[0021] It is important to note the construction of the dynamic approval relationship network. The system first accesses and parses the organizational rule base to obtain explicit definitions of the enterprise's organizational structure, job positions, and permission allocation. Based on these rules, the system identifies all positions or roles granted approval permissions and establishes a list of permission roles. Using this list, it queries and retrieves information on the specific personnel currently holding these positions or roles from one or more integrated organizational data sources (such as the enterprise's human resource management system, Active Directory, etc.). These personnel constitute the initial nodes of the dynamic approval network, each node being associated with the personnel's basic attributes, such as their department, approval permissions, job title, and professional skills. Based on the approval logic defined in the organizational rule base, such as reporting relationships, budget approval paths, departmental collaboration processes, and approval delegation rules, the system establishes directed connections, i.e., edges, between the created personnel nodes. For example, if a rule defines that an employee's material requisition application requires approval from their direct department head, the system will establish an edge between the applicant and the corresponding department head node. These edges constitute the possible transmission paths of the approval process. In the initial stage, the weight of each edge is assigned according to the preset priority or default approval order in the organizational rules.

[0022] To ensure the network reflects real-time status, the system continuously receives and processes real-time operational data to calculate and update the weight of each edge. This real-time operational data includes: the on-duty status of approvers (e.g., on-duty, busy, off-duty), the current workload of tasks awaiting approval, and the historical average approval time. The system captures this data in real-time by monitoring terminal status events and task status changes. Edge weights are calculated using a pre-defined deterministic weighting model. This model integrates the above factors and dynamically outputs a quantified weight value representing the current expected processing time for each edge in the network. This dynamic approval relationship network is stored in a cloud-based central database in a graph structure and is kept updated through periodic data stream processing tasks, ensuring the network reflects the current status of approval resources in real time.

[0023] Process Type Determination. Once the electronic material requisition form data from the edge is successfully received and loaded by the cloud platform, the system automatically parses and extracts key feature information based on a predefined data structure template. This data structure template is a standardized material requisition form data model defined internally by the cloud platform. It maps or deserializes the raw data of the electronic material requisition form (e.g., JSON format) onto this template-defined model, accurately and efficiently extracting key feature information. This key feature information includes the requisitioning department, the type and quantity of the requested materials, the total value, the urgency level, and whether it is a temporary loan. Then, this key feature information is input into a rule-based classifier, which embeds process determination rules loaded from the approval rule base. The system executes these rules and automatically outputs the final process type of the electronic material requisition form based on the input key feature information, including standard process, urgent process, and temporary loan process. For example, if the urgency level is marked as "urgent," it is classified as an urgent process; if the application type is "temporary loan," it is classified as a "temporary loan process"; if the total value exceeds a certain threshold, it needs to enter a multi-level approval branch within the standard process.

[0024] Approval path generation. Based on the determined process type, the system queries the approval rule base to obtain the approval role sequence corresponding to that process type. For each approval role in the sequence, the system filters out all candidate approvers who meet the role requirements and whose current status satisfies the basic approval conditions (e.g., on-duty) from the dynamic approval relationship network. Subsequently, the system runs a graph search algorithm, which aims to minimize the total path weight. Based on the constraints of the approval role sequence, it searches for an optimal approval path from the start point to the end point. The dynamic nature of the weights allows the path to automatically avoid approvers who are currently overloaded, have low approval efficiency, or are not on duty. It prioritizes approvers with rapid responses for urgent processes and ensures the participation of key nodes for high-value processes. Finally, it outputs an executable dynamic approval path consisting of the names of specific approvers.

[0025] Approval task execution. Based on the generated dynamic approval path, the workflow engine automatically, sequentially, or in parallel creates and assigns approval tasks to the terminals of each approver along the path. Task assignment is achieved through the system's built-in message notification mechanism, which may include desktop reminders, mobile application push notifications, etc. Each time the approval status changes, the system generates an approval operation log at the cloud platform layer and persistently stores this log record in the cloud central database. This log record includes: approval operation timestamp, operator, electronic material requisition form number, previous approval node, current approval node, and operation opinion (approved, rejected, transferred). Approval status change events are also instantly published to the system's message bus. Any user interface or external system authorized to subscribe to this process status can listen to these messages to obtain and refresh the overall progress of the process in real time, achieving global process status synchronization and transparency.

[0026] Monitoring and Automatic Intervention. The intervention mechanism includes: execution anomaly identification and dynamic rerouting based on real-time event monitoring: The system establishes a real-time event monitoring engine to continuously monitor the status changes of the electronic material requisition approval process and the real-time updates of the approver's status; the abnormal conditions identified by this engine include: task node timeout, where the time taken for any approval task from issuance to completion exceeds the maximum time threshold preset for that node; and approver status change, where the real-time on-duty status of the currently pending approver changes to "off-duty". When any of the above abnormal conditions is triggered in real time, the system automatically starts dynamic rerouting; based on the latest dynamic approval relationship network data, starting from the currently interrupted node, it recalculates and generates an optimal approval path for the subsequent part of the process, and automatically jumps the approval task to the next available approver on the new path.

[0027] Process status intervention based on externally related events: The system integrates mechanically with external business systems (such as project management and financial systems) through predefined interfaces and message subscription mechanisms, receiving relevant status change events in real time. This mechanism responds to changes in the external business environment related to the process, including: project status changes (e.g., the project status associated with the material requisition form changes to "paused" or "cancelled") and budget freezes (e.g., receiving notification that the budget for the requesting department or corresponding project has been frozen). When the above external events are received, the system will automatically suspend the approval process for the relevant material requisition form and mark its status as "pending." Simultaneously, the system will notify the applicant and relevant administrators of the reason for the process interruption due to the external event. Subsequently, the system will automatically resume the approval process upon receiving the corresponding event (e.g., project restart or budget unfreezing).

[0028] The system employs a periodic data analysis-based approach to identify and adaptively adjust rule performance anomalies. It aggregates, mines, and statistically analyzes historical operational data across the entire process at preset intervals (e.g., weekly). Anomalies identified by this mechanism primarily manifest as biases in the evaluation of the approval rule base's performance, including: abnormal rejection rates (e.g., material requisition forms for specific types, departments, or value ranges consistently have an average rejection rate exceeding a preset reasonable threshold); and abnormal approval efficiency (e.g., the average approval time for specific approval steps or applications with specific attributes is significantly longer than the benchmark). When periodic analysis identifies these systemic anomalies, an adaptive adjustment mechanism for process rules is automatically triggered. This mechanism generates warnings and rule optimization suggestions. Within preset security policies, the system can automatically adjust rule parameters, such as adjusting the automatic approval value threshold for specific material categories or enabling / disabling necessary review steps for specific requesting departments.

[0029] When the system performs dynamic rerouting or process status intervention, a system intervention log is automatically generated on the cloud platform and persistently stored in the cloud central database. The log includes: intervention timestamp, the electronic material requisition number being intervened, intervention type, the specific reason for triggering the intervention, and the process path before and after the intervention.

[0030] Warehouse Management Module: Based on the warehouse optimization model, it continuously analyzes the core characteristics of materials to automatically allocate storage locations for incoming materials and plan dynamic paths for inventory operations; based on the operation and maintenance decision model, it performs inventory verification and reservation according to the approval process status and automatically issues early warnings for shortage risks.

[0031] like Figure 3 As shown, this invention provides a warehouse optimization model; Specifically: Based on the preset core feature dimensions of materials, a dynamic feature vector is constructed for each material; the core feature dimensions of materials include: material information, frequency of entry and exit from the warehouse, correlation between material requisition, physical size and storage requirements, and frequency of supplier entry into the warehouse; Based on the dynamic feature vectors of each material, the materials are divided into several storage strategy clusters, and corresponding storage location area allocation rules are generated for each strategy cluster. When materials are put into storage, the real-time feature vector of the material is calculated, and it is classified into a target storage strategy cluster based on similarity, and the corresponding target storage location area is obtained accordingly. Within the target storage area, a unique recommended storage location is output based on a predefined multi-objective optimization function in real time. The multi-objective optimization function is used to balance picking efficiency, space utilization, and workload balance. Based on the dynamic feature vectors of materials, the confidence level of inventory data for each material is calculated, and an inventory count task list is dynamically generated; through reinforcement learning algorithms, inventory count routes are planned for warehouse managers.

[0032] It is important to note that the warehouse optimization model is deployed on a cloud platform. Through data-driven methods, it intelligently determines the storage location of materials and inventory plans to maximize warehouse space utilization and operational efficiency. The specific methods for constructing the warehouse optimization model include: Construction and updating of dynamic feature vectors for materials. The system retrieves material master data from the cloud-based central database, including the material's unique code, name, physical dimensions, weight, and special storage requirements (such as temperature and humidity). Through time series analysis, it calculates the total number of times and the total quantity of each material issued within the most recent rolling time window (e.g., 30 days) to determine its dynamic inbound / outbound frequency. Using association rule mining algorithms (such as the Apriori algorithm), it analyzes historical material requisition data to identify frequently used material combinations and calculates the requisition correlation between each pair of materials. The system aggregates the above data to construct a feature vector for each material, which is dynamically updated through cloud-based batch processing tasks (e.g., daily / nightly) or triggered by specific events (e.g., sudden changes in inbound / outbound frequency).

[0033] Storage strategy cluster partitioning and rule generation. The cloud platform aggregates the dynamic feature vectors of all materials and uses unsupervised clustering algorithms (such as K-Means) to perform machine learning analysis on these vectors. This clustering algorithm automatically divides materials with similar characteristics into several storage strategy clusters, such as high turnover rate and small volume clusters, clusters strongly correlated with a certain material and medium turnover rate clusters, and low turnover rate and large volume clusters. After completing the calculation, the clustering algorithm outputs a central vector for each generated storage strategy cluster, which represents the average feature of all material feature vectors within the cluster. For each generated storage strategy cluster and its corresponding central vector, the corresponding storage location area allocation rule is automatically generated through an operations research optimization model. For example, the rule for the high turnover rate and small volume cluster is "allocated to light-duty shelf position a at the front of the picking aisle in area A". The system persistently stores the central vector of each storage strategy cluster and its corresponding storage location area allocation rule in the database, forming a strategy mapping table for real-time querying during inbound operations.

[0034] Inbound material strategy cluster matching and region positioning. When new material enters the warehouse, the warehouse edge node (such as the inbound workstation) triggers a storage location allocation request and sends the material code to the cloud platform. The cloud platform calculates its feature vector in real time based on the material code. The system loads the center vectors of all storage strategy clusters from the database and compares the real-time feature vector of the material with the center vectors of each storage strategy cluster through vector similarity calculation (such as cosine similarity). Based on the principle of highest similarity, the material is assigned to a target storage strategy cluster. The system queries the strategy mapping table to obtain the target storage location region corresponding to the target storage strategy cluster.

[0035] Storage location determination. Within the target storage location area, the system obtains a list of all currently available storage locations and their attributes (such as coordinates, associated shelf, and distance from the shipping area); it constructs a multi-objective optimization function, which aims to simultaneously maximize, but is not limited to, the following objectives: picking efficiency, minimizing the weighted distance from the candidate storage location to the storage location of frequently associated materials, and the distance to the main warehouse aisle; space utilization, selecting the storage location that best matches the material size to avoid space waste; workload balancing, avoiding concentrating high-turnover materials on a few pickers or shelves, and balancing the workload of each area based on historical task data; a heuristic algorithm (such as a genetic algorithm) is used to solve the optimization function in real time, calculating a unique recommended storage location with the highest overall score from all candidate storage locations. This recommended storage location information is sent to the terminal devices of the warehouse edge nodes to guide warehouse managers in putting the goods into storage. After each time a final storage location is assigned to an incoming material, the system generates a storage location assignment log. This log includes: the assignment timestamp, the material code and batch number of the operation, the inbound order number, the assigned storage location code, the strategy cluster ID that recommended the storage location, and the personnel who triggered the assignment operation.

[0036] Dynamic Inventory Count Tasks and Path Planning. The system periodically (e.g., every 24 hours) runs an inventory count task generator. This generator filters materials whose inventory confidence scores are below a preset threshold or whose inventory quantity is close to zero, dynamically generating an inventory count task list. The inventory confidence score is calculated in real-time for each material based on its dynamic feature vectors, particularly its inbound / outbound frequency and historical inventory error rate, using a pre-trained confidence calculation model. The system treats all storage locations in the inventory count task list as nodes in a graph network and trains an agent using reinforcement learning algorithms (such as Q-Learning) to find the optimal path to traverse all nodes. The planned optimal inventory count path is distributed to the warehouse manager's handheld edge terminal. The terminal device guides the manager through a graphical interface to complete the inventory count sequentially and uploads the data to the cloud platform's inventory database in real-time.

[0037] like Figure 4 As shown, this invention provides an operation and maintenance decision-making model; Specifically: When approval is triggered, the current inventory data is obtained in real time, and the theoretical real-time available total amount of required materials is calculated by combining the allocated quantity and the inventory in transit. The theoretical real-time available total is compared with the dynamic safety threshold. When it falls below the threshold, an inventory shortage warning is automatically generated and sent to the applicant and the current approver. The theoretical real-time available total amount of materials and inventory shortage early warning signals are fed back to the current approval process node in real time, and the approval rules are dynamically adjusted. Once approved, a differentiated reservation strategy will be implemented based on the priority of the electronic material requisition form attributes.

[0038] It is important to note that the aforementioned operation and maintenance decision-making model adopts an edge-cloud collaborative architecture, connecting the approval process and inventory status in real time to achieve intelligent commitment to material requirements, risk warnings, and dynamic resource allocation. The specific construction method of the operation and maintenance decision-making model includes: Theoretical real-time available total quantity calculation. When the approval process begins or any approval node is completed, the process approval module sends an event notification to the operation and maintenance decision model. The model then obtains the current available inventory quantity of the material in real time through the inventory query interface. The model summarizes the quantities of all materials with the status "reserved" and "hard reserved" from the inventory reservation records; this is the allocated quantity. The model combines this with the in-transit inventory quantity (ordered but not yet delivered) to calculate the theoretical available total quantity. The calculation formula is as follows: Theoretical real-time available total quantity = Current available inventory quantity + In-transit inventory quantity - Allocated quantity. This calculation process is performed on the cloud platform to ensure global data consistency.

[0039] Inventory shortage risk assessment and early warning. The dynamic safety threshold is dynamically calculated by a data analysis module deployed on a cloud platform based on the historical consumption trend analysis and prediction of the material, using a time series analysis model. When the theoretically real-time available quantity falls below the dynamic safety threshold, the system automatically generates an inventory shortage early warning notification and sends it simultaneously to the applicant and the terminal devices of the current pending approval node via message push service; the notification content includes material information, shortage quantity, and expected impact.

[0040] Context integration and dynamic adjustment of the approval process. The theoretical real-time available quantity and inventory shortage warning signals are used as key contextual information, which is fed back in real-time and highlighted on the approver's task interface. This contextual information triggers dynamic adjustments to the approval rules. For example, for non-urgent, low-priority material requisition requests, the system can provide a "recommend rejection" or "partial approval" prompt on the approval interface; for high-priority requests, it can accelerate process routing or require approvers to fill in high-risk approval reasons.

[0041] Implementation of the differentiated reservation strategy after approval. Inventory reservation is automatically triggered after the electronic material requisition form is approved. Specific steps include: The system uses a rule engine deployed on a cloud platform to analyze the key features of electronic material requisition forms. These key features include: project level, material criticality, urgency, application amount, and purpose description. The rule engine matches these key features with a pre-defined reserved rule library to output the priority level of the electronic material requisition forms. The reserved rule library adopts a binary priority architecture, dividing all electronic material requisition forms into "first priority" and "second priority." The core logic of this classification lies in assessing the directness and urgency of the material requisition needs' impact on the company's core operations. Specifically, the first priority is determined when the key features of the material requisition form meet one or more of the following core dimensions: strategic relevance (the material requisition need is directly related to company-level strategic projects or national key projects); production assurance (the material requisition need is used to ensure the continuous operation of mainstream product production lines or fulfill key customer orders, and involves critical materials); event urgency (the material requisition need is used to deal with sudden emergencies such as production interruptions, emergency fault repairs, and quality traceability); and high asset value (the total amount of materials involved in a single application exceeds a preset threshold). The second priority includes all routine, planned, or supportive material requisition needs that do not meet the conditions of the first priority. It has a certain degree of flexibility and includes needs such as general production planning, sample making, preventive maintenance, and administrative logistics.

[0042] Reserved strategy matching and execution. Based on the priority level of the electronic material requisition, the corresponding reserved strategy is matched and executed through a predefined, configurable mapping relationship: For first-priority electronic material requisitions, the "locked reservation" strategy is matched, and the system immediately performs a locking operation in the inventory database, directly deducting the corresponding physical available inventory quantity, and generating a reservation record with a status of "hard reservation", ensuring that this part of the inventory is verified and locked; for second-priority electronic material requisitions, the "flexible reservation" strategy is matched, physical inventory is not immediately deducted, and a reservation record with a status of "reserved" is generated in the inventory database. This record updates the committed quantity of the material and includes the priority level mark and expiration period.

[0043] Reserved Status Monitoring and Adaptive Adjustment. The system continuously scans the status, inventory level changes, and validity period of all reserved records through a background monitoring service. Based on pre-defined business rules, it automatically performs dynamic adjustments. For example, when the validity period of a "flexible reserved" record expires, the monitoring service automatically updates its status to "cancelled." When the inventory of "flexible reserved" materials becomes sufficient, the system automatically upgrades its status from "reserved" to "hard reserved." When a new "first priority" request lacks inventory, it automatically reduces or cancels existing low-priority "flexible reserved" records to release resources and ensure core needs. The system tracks and records every reserved operation throughout the entire process, generating a reserved tracking log on the cloud platform and persistently storing this log in the cloud central database. The log includes: reserved operation timestamp, operator, associated material requisition number, priority level, reserved strategy, operation type (create, upgrade, cancel), quantity change, and the rule that triggered the operation.

[0044] The material issuance module generates a unique material issuance voucher for approved electronic material requisition forms; the material issuance is completed by verifying the voucher, and inventory data is synchronized to the cloud platform in real time based on the actual issuance quantity.

[0045] like Figure 5 As shown, the present invention provides a material dispensing method; Specifically: A unique encrypted material requisition voucher is generated for each approved electronic material requisition form, and the lifecycle status of the material requisition voucher is tracked throughout the entire process; the lifecycle status includes generation, verification, and cancellation. Verify the validity of material requisition vouchers and send picking instructions to the warehouse manager; automatically record the actual quantity issued through intelligent data collection devices and update inventory data in real time; During the material issuance process, the status of the material requisition voucher and the deviation in the issuance quantity are monitored in real time, and hierarchical alarm information is automatically generated for any detected anomalies.

[0046] It is important to note the generation of material requisition vouchers. The moment the electronic material requisition form is approved on the cloud platform, the voucher generation service is automatically triggered. This service takes key information such as the material requisition form number, material code, requested quantity, and validity period as input, and generates a unique, unpredictable material requisition voucher using an encryption algorithm. The voucher can take the form of a QR code / barcode, a unique alphanumeric code, or an NFC / RFID data packet. After the material requisition voucher is generated, its lifecycle status in the database is initialized to "generated," and a state machine model is established for it. Its status will follow the business process, switching between "generated," "verified," and "cancelled," and recording the timestamp and operation context of each status change.

[0047] Voucher verification and issuance execution. Warehouse administrators use edge-side smart terminals to scan, read, or input material requisition vouchers. The terminal sends the voucher data to the cloud platform's verification interface via network request for validity, anti-reuse, and expiration date checks. If any verification fails, the system returns a verification failure message and reason to the edge terminal. After successful verification, the cloud platform sends the picking list and warehouse location navigation information corresponding to the material requisition to the warehouse administrator's edge terminal interface, guiding them to perform the picking operation. After issuing materials, the warehouse administrator uses the interactive interface on the smart terminal to automatically collect the actual issued quantity through a connected smart counting device. After confirmation by the warehouse administrator, the edge terminal rebinds the actual issued quantity to the material requisition voucher and sends it to the cloud via HTTPS protocol. Upon receiving the data, the cloud platform first updates the voucher status to "verified," and then immediately updates the inventory based on the actual issued quantity.

[0048] When verifying vouchers, the system generates a voucher verification log on the cloud platform and persistently stores the log record in the cloud central database. The log record includes: verification timestamp, verification terminal number, verifier, material requisition voucher, and verification result. When inventory is updated, the system generates an inventory change log, which includes: change timestamp, associated material requisition voucher, material code, change type (outbound / inbound), quantity before change, quantity changed, and quantity after change.

[0049] Real-time monitoring and tiered alerting during the material issuance process. The system monitors key indicators in real time during material issuance, specifically including: monitoring for abnormal status transitions in the voucher status stream; calculating |actual issuance quantity - requested quantity| and comparing it with a preset allowable deviation threshold. When an anomaly is detected, the system automatically triggers different levels of alerts based on the severity of the anomaly: Level 1 alerts, detecting security anomalies such as "invalid voucher" or "duplicate voucher verification," immediately lock the relevant terminal operation permissions and send an immediate alert to the warehouse supervisor and safety officer; Level 2 alerts, detecting business anomalies such as "actual issuance quantity exceeds the allowable deviation range," pop up a confirmation prompt for the warehouse administrator, requiring them to input the reason for the deviation, and notify their direct supervisor; Level 3 alerts, detecting efficiency anomalies such as "excessive issuance operation time," record in the log and selectively notify the supervisor. Through real-time monitoring and tiered alerting, the material issuance process is transformed from passive recording to proactive risk control, effectively preventing problems such as mis-issuance and theft of materials.

[0050] Data Analysis and Traceability Module: Used to aggregate and analyze data from the entire material process, generate multi-dimensional analysis reports, and build a digital responsibility chain based on blockchain-style logs.

[0051] It is important to note that the data analysis and traceability module is deployed on a cloud platform. Cloud data is aggregated and cleaned. The system extracts business data covering the entire lifecycle of material management from the central cloud database through a predefined ETL process. The extracted data includes, but is not limited to: electronic material requisition form data, approval path data and approval time data at each node, inventory status data, warehouse location optimization operation records and dynamic inventory task data, material requisition voucher verification records, and actual issuance quantity data. The extracted heterogeneous data is cleaned and transformed. Missing and abnormal data are handled through predefined data cleaning rules and standardized mapping tables, and the coding system for key business operations such as materials, departments, and projects is unified to build a consistent data model suitable for analysis. The processed data is loaded into the analytical data warehouse on the cloud platform, and fact tables and dimension tables (such as time dimension, material dimension, department dimension, and project dimension) are built using dimensional modeling methods to support efficient multidimensional analysis. The analytical data warehouse can be implemented using a relational database with an MPP architecture or a cloud-native data warehouse to support high-performance queries of massive historical data.

[0052] Multi-dimensional intelligent analysis in report generation. Based on a data warehouse, the system performs multi-dimensional intelligent analysis through a built-in computing engine, including: Material consumption trend analysis and forecasting: Using time series analysis algorithms, historical consumption data is used as input to train a prediction model; the system automatically identifies the trends, seasonality, and periodicity of the data, generating a material demand forecast report for a specific future period (e.g., the next 4 weeks) and providing a prediction confidence interval, providing data support for procurement decisions. Inventory health analysis: The system automatically extracts relevant fields from financial and inventory data, calculates and ranks inventory turnover rates by material category or individual material. The inventory turnover rate calculation formula is as follows: Inventory Turnover Rate = Cost of Goods Sold During the Period / Average Inventory Value, where: Cost of Goods Sold During the Period is obtained from the enterprise's financial system through system integration; Average Inventory Value is obtained by calculating the arithmetic mean of the beginning and ending inventory values ​​based on the inventory quantity and unit standard cost of materials maintained by the warehouse management module. The system automatically scans inventory data through preset rules (e.g., "Last outbound date is more than 365 days ago"), generates a list of obsolete inventory, and calculates its percentage of total inventory value. Dynamic ABC Classification Analysis: Every cycle (e.g., monthly), the system automatically calculates the total consumption of all materials over a past period (e.g., 12 months), sorts them from highest to lowest, calculates the cumulative percentage, and automatically classifies the top 70% of materials into Category A (critical and important items), the next 20% into Category B (generally important items), and the last 10% into Category C (unimportant items), generating a dynamic ABC classification report. The classification results can be automatically fed back to the warehouse management module to guide warehouse location optimization strategies. Approval and Process Efficiency Analysis: The system calculates the average dwell time at each approval node and uses process mining algorithms to visualize the actual path of the approval process, comparing it with the preset ideal path. It automatically identifies bottleneck nodes and abnormal jumps, generating an approval process efficiency analysis report that points out efficiency bottlenecks and provides suggestions for process optimization. The system has a built-in report template engine. Administrators can pre-design report templates, and through scheduled tasks, the system automatically drives the calculation engine at a specified time to populate the template with the results, generating standardized PDF or Excel format reports.

[0053] A blockchain-based digital responsibility chain is constructed. Key operational events for each module are extracted from the operation logs, including: material requisition form creation, approval actions (approval / rejection), inventory reservation, voucher generation, voucher verification, and inventory deduction; as well as core elements such as the operator, timestamp, operation object (e.g., order number), and operation content corresponding to each event. The system generates a hash value for the first operation log; when the second operation log is generated, the system combines the original data of the second log with the hash value of the previous log, recalculates the hash value, and stores this new hash value in the second log record. This process continues, forming a linked hash chain. Any tampering with historical records will cause a change in its hash value, thereby disrupting the chain relationship with subsequent records, making tampering immediately detectable. The system periodically anchors the hash value of the latest log to a public blockchain or stores it through a trusted timestamp service.

[0054] Although embodiments of the invention have been shown and described, it will be understood by those skilled in the art that various changes, modifications, substitutions and alterations can be made to these embodiments without departing from the principles and spirit of the invention, the scope of which is defined by the appended embodiments and their equivalents.

Claims

1. A reliable material management and control system based on edge cloud collaboration, characterized in that, include: Material requisition module: Used to receive and generate electronic material requisition forms through edge terminal devices and upload them to the cloud; Workflow approval module: This module is used to deploy a pre-set approval rule base and organizational rule base on the cloud, combine real-time operation data, generate a dynamic approval path for electronic material requisition forms, and issue approval tasks to the approver's terminal. Warehouse Management Module: Based on the warehouse optimization model, it automatically allocates storage locations to incoming materials by continuously analyzing the core characteristics of materials, and plans dynamic paths for inventory operations. Based on the operation and maintenance decision-making model, inventory verification and reservation are performed according to the approval process status, and automatic warnings are given for shortage risks. The material issuance module generates a unique material issuance voucher for approved electronic material requisition forms; it verifies the voucher to complete the material issuance and synchronizes inventory data to the cloud in real time based on the actual issuance quantity. Data Analysis and Traceability Module: Used to aggregate and analyze data from the entire material process, generate multi-dimensional analysis reports, and build a digital responsibility chain based on blockchain-style logs.

2. The material end-to-end trusted management and control system based on edge cloud collaboration according to claim 1, characterized in that: The process approval module specifically includes: S1. Construct a dynamic approval relationship network based on the organizational rule base and real-time operational data; S2. Automatically parse and extract key feature information of electronic material requisition forms, and determine their process type based on the approval rule base; the process type includes standard process, emergency process, and temporary borrowing process; S3. Based on the approval rule base and dynamic approval relationship network, automatically generate dynamic approval paths according to the determined process type; S4. Based on the generated dynamic approval path, drive the approval tasks to be executed sequentially or in parallel along the path nodes, and update and synchronize the global process status in real time. S5. Through real-time monitoring and periodic analysis, identify abnormal conditions at different levels and automatically trigger corresponding intervention mechanisms.

3. The material end-to-end trusted management and control system based on edge cloud collaboration according to claim 1, characterized in that: The method for constructing the warehouse optimization model includes: M1. Based on the preset core feature dimensions of materials, construct dynamic feature vectors for each material; the core feature dimensions of materials include: material information, inbound and outbound frequency, requisition correlation between materials, physical size and storage requirements, and supplier inbound frequency. M2. Based on the dynamic feature vectors of each material, the materials are divided into several storage strategy clusters, and corresponding storage location area allocation rules are generated for each strategy cluster. M3. When materials are put into storage, calculate the real-time feature vector of the material, classify it into a target storage strategy cluster based on similarity, and obtain the corresponding target storage location area accordingly. M4. Within the target storage area, perform real-time calculations based on a predefined multi-objective optimization function to output a unique recommended storage location; the multi-objective optimization function is used to balance picking efficiency, space utilization, and workload balance. M5 calculates the confidence level of inventory data for each material based on the material's dynamic feature vector and dynamically generates an inventory count task list; it also plans inventory count routes for warehouse managers through reinforcement learning algorithms.

4. The material end-to-end trusted management and control system based on edge cloud collaboration according to claim 1, characterized in that: The method for constructing the operation and maintenance decision model includes: N1. When approval is triggered, the current inventory data is obtained in real time, and the theoretical real-time available total amount of required materials is calculated by combining the allocated quantity and the inventory in transit. N2. Compare the theoretical real-time available total with the dynamic safety threshold. If it falls below the threshold, automatically generate and send an inventory shortage warning notification to the applicant and the current approver. N3. The theoretical real-time available total amount of materials and inventory shortage early warning signals are fed back to the current approval process node in real time, and the approval rules are dynamically adjusted. N4. After approval, a differentiated reservation strategy will be implemented based on the priority of the electronic material requisition form attributes.

5. A material end-to-end trusted management and control system based on edge cloud collaboration as described in claim 4, characterized in that: The N4 step includes: Q1. Based on the preset reservation rule library, the priority level is automatically assigned and the reservation strategy is matched according to the key feature information of the electronic material requisition form. Q2. Implement differentiated reservations; Q21. Lock and reserve for the first priority electronic material requisition form, immediately deduct available inventory and generate a reservation mark; Q22. Implement flexible reservation for second-priority electronic material requisition forms, update the committable quantity, and set priority tags and expiration periods; Q3. Continuously monitor reservation records, and automatically switch reservation modes and dynamically adjust reservation quantities based on inventory status and pre-defined business rules, and establish a full-process tracking record from reservation creation to termination.

6. The material end-to-end trusted management and control system based on edge cloud collaboration according to claim 1, characterized in that: The execution and issuance module specifically includes: P1. Generate a uniquely encrypted material requisition voucher for each approved electronic material requisition form, and track the lifecycle status of the material requisition voucher throughout its entire lifecycle; the lifecycle status includes generation, verification, and cancellation. P2. Verify the validity of the material requisition voucher and send picking instructions to the warehouse manager; automatically record the actual quantity issued through intelligent data collection devices and update inventory data in real time; P3. During the material issuance process, the status of the material requisition voucher and the deviation of the issued quantity are monitored in real time, and hierarchical alarm information is automatically generated for detected anomalies.

7. A material end-to-end trusted management and control system based on edge cloud collaboration as described in claim 1, characterized in that: The electronic material requisition form includes material code, name, quantity requested, application type, and intended use information; the real-time operation data includes the status of the approver, the current workload of tasks awaiting approval, and the historical average approval time.

Images