Component whole-process tracking management system based on RFID

Abstract

The invention relates to the field of intelligent construction, and discloses an RFID-based component full-process tracking management system, which comprises a data acquisition module, a data transmission module, a data processing center module and an application display module, and is characterized in that the data acquisition module acquires component identity, position, environment and precision data through an RFID tag, a reader-writer and a sensor; the transmission module completes data encryption and lightweight transmission through a wireless network and a cloud gateway; the data processing center module screens data anomalies in real time through an RFID data quality management algorithm, transportation anomalies are judged by a GPS dominated scene, tags are automatically updated after verification of a PDA and a sensor, authenticity is discriminated through cross verification of equipment and data, false anomalies are automatically repaired, and real anomalies are pushed in a layered mode; and the application display module realizes three-dimensional visual full-process tracking and abnormity early warning, the problem that data in a prefabricated building is easy to make mistakes and difficult to trace is solved, and full-process precise and intelligent management is realized.

Description

Technical Field

[0001] This invention relates to the field of intelligent construction technology, specifically to an RFID-based full-process component tracking and management system. Background Technology

[0002] In the field of prefabricated buildings, the tracking and management of prefabricated components throughout the entire process has always been a focus of industry attention. Traditional management methods rely heavily on manual recording, barcode or QR code identification, which suffers from problems such as delayed information updates and easy errors. With the development of Internet of Things (IoT) technology, RFID (Radio Frequency Identification) technology has been gradually introduced into the management of building components due to its non-contact, batch reading and strong environmental adaptability.

[0003] The existing published literature 1 (Research on the Application of RFID Technology in the Life Cycle Management of Prefabricated Building Components, 2024) proposes that RFID technology can realize the collection and transmission of basic information of components from production, transportation to installation, significantly improving the automation level of information processing. It also reveals the centrality and density of each unit in the supply chain through social network analysis (SNA) method. However, the literature also points out that the existing RFID system still has significant shortcomings in data quality management, fails to achieve collaborative verification of multi-source data, and has difficulty distinguishing between real anomalies and equipment errors, affecting management efficiency and decision reliability.

[0004] Existing public document 2 (Research on the Application of BIM and RFID Technology in Prefabricated Buildings, 2018) proposes a component tracking and management system comprising four modules: monitoring information sharing, monitoring data management, early warning, and RFID data management. This system includes a safety early warning module, comprising two parts: early warning settings and early warning management. Early warning settings mainly concern the configuration of early warning displays, including highlight colors and shape displays. Early warning management primarily enables the linked viewing of early warning data and models, allowing for rapid location of early warning components and facilitating rapid construction management to mitigate potential risks, such as… Figure 2 As shown. However, existing component tracking and management systems are insufficient in terms of scene adaptation, failing to dynamically adjust monitoring strategies according to different stages of the component. Especially in the transportation process, GPS data and RFID scanning data are often processed independently, lacking integrated analysis, making it difficult to accurately determine whether the component has deviated from the predetermined route or entered an abnormal area.

[0005] Therefore, there is an urgent need for an intelligent tracking system that can achieve cross-validation of multi-source information, hierarchical handling of anomalies, and dynamic rule invocation. Summary of the Invention

[0006] To overcome the aforementioned deficiencies of existing technologies, this invention provides an RFID-based component end-to-end tracking and management system. This system integrates multi-source data through an RFID data quality management algorithm, establishes a dynamic scene recognition and tag-rule linkage mechanism, and achieves collaborative verification of multi-source data throughout the component's lifecycle, as well as adaptive scene control. It also screens for data anomalies in real time and distinguishes between genuine and false anomalies through "equipment status verification + data cross-verification." False anomalies are automatically repaired and their causes recorded, while genuine anomalies are handled in a tiered manner according to risk level, with the handling process recorded in a closed loop. This addresses the problems mentioned in the background section.

[0007] To achieve the above objectives, the present invention provides the following technical solution: An RFID-based component end-to-end tracking and management system includes a data acquisition module, a data transmission module, and an application display module, as well as a data processing center module. The data processing center module receives transportation data from the data transmission module. Through an RFID data quality management algorithm, it screens for anomalies in the component end-to-end data in real time. For transportation data, GPS positioning data is used as the primary dynamic scene judgment, combined with the operation type of handheld PDA scanning and sensor data for auxiliary verification. After scene changes are confirmed, the RFID data quality management algorithm automatically updates the tags according to the 'scene feature-tag mapping library' and synchronizes the updates to the data processing center's backend database, the application display module, and the on-site PDAs. The RFID data quality management algorithm also queries the 'tag-rule relationship' in real time. The "Linked Tables" feature adapts to the new rules corresponding to the new tags. If transportation data anomalies occur, the RFID data quality management algorithm distinguishes between genuine and false anomalies through "equipment status verification + data cross-validation." If a false anomaly is detected, the algorithm automatically repairs the data, records the cause, and removes the anomaly warning. For genuine anomalies, tiered handling is implemented. Low-risk anomalies are automatically pushed with intervention suggestions, while high-risk anomalies are pushed to the responsible party through the application display module, along with BIM positioning and handling guidelines. After the responsible party completes the handling, they upload the handling results via a handheld PDA. The RFID data quality management algorithm then restarts "equipment status verification + data cross-validation." Once the anomaly is confirmed to be resolved, the anomaly warning is removed, the handling process is recorded in the full-process log, and the risk assessment report for component tracking is updated.

[0008] As a further aspect of the present invention, the data acquisition module is used to comprehensively collect various types of data about the component throughout the entire process, including anti-metal, high-temperature and high-humidity resistant industrial-grade UHF RFID tags. Each tag has a globally unique ID (EPC code) permanently bound to a specific component, serving as its "identity card" in the digital world. A handheld RFID reader (PDA) is used to accurately record the operator's actions (such as "outbound loading," "on-site acceptance," and "installation") when scanning component or area tags with the PDA. This data is then compared with the timestamp of the scan and the data obtained from the PDA's built-in GPS module. The system automatically binds real-time latitude and longitude coordinates to ensure that every operation has a clear spatiotemporal context; it adds sensor groups to critical or vulnerable components (such as prestressed concrete components and precision steel nodes), such as temperature and humidity sensors that can monitor the environment of the components to prevent premature solidification of concrete or condensation corrosion of steel; triaxial vibration and impact sensors that can record events such as bumps and sudden braking during transportation to assess transportation quality; tilt sensors that can monitor the posture of components to prevent irreversible tilting or deformation during hoisting or transportation; and precision measurement sensors that can be linked with equipment such as total stations during the installation phase to collect installation positioning data.

[0009] The equipment used in the data acquisition module does not work in isolation. When the PDA scans a component tag, it can simultaneously trigger and read the latest data from the sensors on that component, and package the tag ID, sensor data, GPS location, operation type, and timestamp into a data packet to ensure the spatiotemporal consistency and correlation of the data at the source.

[0010] As a further aspect of the present invention, the data transmission module is used to securely, reliably, and efficiently transmit the collected raw data to the data processing center. This module can automatically select the optimal network channel based on the site environment. In areas with Wi-Fi coverage, such as factory areas or construction site warehouses, Wi-Fi transmission is prioritized to save bandwidth. During transportation, it automatically switches to 4G / 5G mobile networks. In network signal blind spots (such as remote mountain tunnels), the data acquisition devices (handheld PDAs, sensors) temporarily store the raw data locally. After the network is restored, the data is resumed to the cloud gateway. The cloud gateway first removes invalid data (such as misread data from extremely weak signals) and duplicate data, and then uses TLS / SSL encryption protocols for transmission to ensure data security during transmission and prevent information leakage. It also converts the raw, redundant data packets into standardized, structured JSON or XML formats, significantly reducing network bandwidth usage and the computational pressure on the subsequent processing center. The pre-processed "clean" data is then pushed to the data processing center module via a high-speed network, forming an efficient, secure, and standardized "collection-transmission-preprocessing" data pipeline.

[0011] As a further aspect of this invention, the core of the data processing center module is an RFID data quality management algorithm. This algorithm performs real-time intelligent processing on massive amounts of incoming data and builds a "scene feature-tag mapping library," which predefines all typical scenarios that a component may experience (such as "production completed," "in-plant storage," "loading and shipping," "transit in transit," "arrival at construction site," "on-site storage," "lifting preparation," and "installation completed"). Each scenario has its own feature vector, and a set of dedicated data verification rules and anomaly judgment rules are configured for each scenario in the "tag-rule association table." For example, the rules for the "transit in transit" scenario focus on GPS trajectory deviation and vibration data; while the rules for the "on-site storage" scenario focus more on whether the inventory time has expired and whether the temperature and humidity meet the standards.

[0012] The RFID data quality management algorithm employs a dynamic scene recognition and tag update mechanism. It continuously receives data packets from the data transmission module and uses GPS positioning data as the primary driver for dynamic scene judgment. The system pre-sets electronic geofences for key scenes on an electronic map. When a component's GPS coordinates continuously leave the current fence or enter a new fence for a period of time, the model initially triggers a "scene change alert." Subsequently, the model retrieves PDA operation data and sensor data from the same time period for auxiliary verification. For example, if the GPS shows that the vehicle has left the factory, the model will check for recent PDA scan operations related to "loading and shipping"; and whether the sensors show that the component has changed from a static state to a state of continuous vibration. Only after multiple sources of evidence corroborate each other is the scene change finally confirmed. Once confirmed, the model immediately updates the "scene identifier" field in the component's RFID electronic tag based on the "scene feature-tag mapping library," and synchronizes this update to the backend database, the application display module interface, and the PDAs of on-site personnel. Thus, all systems and personnel have a unified and real-time understanding of the component's current status.

[0013] As a further aspect of the present invention, when data anomalies occur during system operation, the RFID data quality management algorithm immediately invokes the "equipment status verification + data cross-validation" anomaly diagnosis sub-process to distinguish between genuine and fake anomalies and locate the root cause of the problem through a three-layer progressive mechanism.

[0014] The first layer: The model monitors the health status of all terminal devices in real time. If it finds that a GPS device loses signal, a sensor's battery is below 10%, or a PDA has no connection for 1 hour, it will immediately mark all data collected by the device during that period as "low confidence" and notify the maintenance personnel to repair it. It will not trigger business anomaly warnings for the time being, effectively avoiding false alarms caused by device failure.

[0015] The second layer involves deep fusion analysis of data on normal equipment status, using geographic matching, pattern matching, logical matching, and behavioral matching. This includes comparing GPS coordinates with the location of the most recent RFID scan, ensuring the distance is within a reasonable range; comparing vibration sensor data with a transportation road condition knowledge base to determine if the vibration pattern is abnormal; analyzing whether temperature and humidity data exceed the tolerance threshold of the component material; and analyzing whether the operator ID in the PDA operation log has the authority to execute the current operation, whether the operation type matches the current scenario, and whether the operation time sequence conforms to the workflow.

[0016] The third layer involves determining the root cause of the anomaly. If there is a data conflict and the equipment status is found to be abnormal, it is determined to be a equipment failure. If the equipment is normal but the operation logic is incorrect, it is determined to be human error. When both the equipment and the operation are normal, but the GPS continues to deviate from the route, the vibration / tilt data seriously exceeds the threshold, and multiple sources of evidence all point to the same unfavorable conclusion, it is finally diagnosed as a real anomaly.

[0017] For "genuine anomalies," the RFID data quality management algorithm handles them differently and more precisely based on their risk level. For low-risk anomalies, the system automatically generates optimization suggestions and pushes them to the PDAs of on-site drivers or operators, prompting them to pay attention and make minor adjustments. The system automatically records but does not escalate the warning. For medium-risk anomalies, in addition to pushing warnings, the system automatically records a complete set of data for the anomaly period, generates a preliminary analysis report, and notifies the area administrator for manual review and judgment. For high-risk anomalies, the system immediately sends a red warning to multiple roles, including the project manager and safety director, through the application display module. The warning information includes highlighting the location of the abnormal component in the BIM 3D model, the specific value of the anomaly, the threshold, and the duration, and provides preliminary handling suggestions based on historical cases and expert knowledge base.

[0018] After receiving the alert via PDA, the responsible person can view the details and take action. After the action is completed, the action result must be uploaded via PDA. After receiving the feedback, the model will restart the "equipment status check + data cross-validation" to confirm whether the anomaly has been truly resolved. Only after the resolution is confirmed will the system lift the alert and store the entire lifecycle of this anomaly in the full process log.

[0019] As a further aspect of the present invention, an RFID-based component end-to-end tracking and management system maintains a dynamically updated risk assessment report for each component. This report records historical abnormal events, integrates key data indicators at each stage of the component's entire lifecycle, and is deeply bound to the BIM model. In the 3D visualization interface, the project manager can click on the digital twin of any component to view its current status and the historical versions and real-time content of its risk assessment report.

[0020] As a further embodiment of the present invention, the application display module creates a corresponding digital twin for each physical component based on the project's BIM model. Users can seamlessly switch perspectives in the 3D model via a web terminal or mobile APP to view the component's location, status (color coding indicates status: green for normal, yellow for warning, and red for abnormal) and detailed data, historical trajectory (replaying the complete movement route of the component from production to the present), and environmental parameters (real-time display of temperature, humidity, and other data transmitted by sensors).

[0021] When component data anomalies occur, the digital twin of the abnormal component will immediately highlight and flash in the model, and a warning window will pop up, integrating all key information (type, level, responsible person, guidelines). Simultaneously, the module provides powerful historical data query and statistical analysis functions, capable of generating various reports, offering strong data-driven decision support for project managers to optimize processes, evaluate partners, and prevent future risks.

[0022] The technical effects and advantages of the RFID-based component end-to-end tracking and management system of this invention are as follows: Through "dynamic scene recognition" and "tag-rule linkage", the system can understand the specific stage of the component and adaptively adopt the most appropriate monitoring strategy. The management granularity is refined from "component" to "the status of the component in a specific scenario", realizing refined and intelligent management of the entire process. The three-layer progressive diagnostic mechanism of "equipment status verification + data cross-validation" solves the persistent problems of false alarms and missed alarms in RFID applications. The system can intelligently distinguish between equipment failure, human error and real anomaly, ensuring that every warning issued has been rigorously verified, which greatly improves the credibility of data and the reliability of decision-making. Attached Figure Description

[0023] Figure 1 This is an architecture diagram of a component end-to-end tracking management system based on RFID according to the present invention; Figure 2 This is a structural monitoring and management system based on existing technologies such as BIM and RFID.

[0024] Figure 3 This is a framework diagram for RFID data quality management algorithms.

[0025] Figure 4 A flowchart illustrating the three-tiered progressive mechanism for handling data anomalies.

[0026] Figure 5 This is a schematic diagram of an abnormality tiered handling mechanism.

[0027] Figure 6 This is a screenshot of the app's early warning center interface. Detailed Implementation

[0028] 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.

[0029] Example 1 In this invention, a component end-to-end tracking and management system based on RFID, four modules form a closely interconnected whole, such as... Figure 1 As shown, the data acquisition module, acting as the information source, collects raw data from multiple sources, including component identity, spatiotemporal location, environmental status, and manual operations, through RFID tags, handheld PDAs, and temperature, humidity, and vibration sensors. It ensures the spatiotemporal correlation of the data at the source. The collected raw data is then transmitted to the data transmission module, which acts as a "bridge" for data flow. This module automatically switches between Wi-Fi / 4G / 5G network channels based on the site environment. The cloud gateway performs invalid data removal, TLS / SSL encryption, and lightweight format conversion. The data transmission module pushes standardized "pre-processed clean data" to the data processing center module. As the core of the system, the data processing center module performs in-depth processing of the clean data based on RFID data quality management algorithms. It uses GPS data to drive dynamic scene judgment, combines PDA operation and sensor data to verify scene changes, and uses a "scene feature-tag mapping library" for analysis. The component labels are updated, and the "label-rule association table" is used to adapt the monitoring and anomaly judgment rules for the corresponding scenarios. If data anomalies occur, the authenticity is distinguished by "equipment status verification + data cross-validation". Genuine anomalies are handled in layers according to risk level. The clean data, early warning information and risk assessment report generated after processing are synchronized to the application display module on the one hand and fed back to the on-site PDA on the other. The application display module, as the interaction and display entry point, realizes three-dimensional visualization based on the BIM model, and displays the entire process trajectory, status and anomaly details of the component in real time. The responsible person receives the early warning and uploads the handling results through the PDA. The results are sent back to the data processing center module for secondary verification. After confirming that the anomaly is resolved, the early warning is lifted, the entire process log and risk assessment report are updated, and finally a closed-loop link of "collection-transmission-processing-display-feedback-verification" is formed to realize accurate and intelligent tracking and management of the entire process of the component.

[0030] As a further aspect of the present invention, the data acquisition module is used to comprehensively collect various types of data about the component throughout the entire process, including anti-metal, high-temperature and high-humidity resistant industrial-grade UHF RFID tags. Each tag has a globally unique ID (EPC code) permanently bound to a specific component, serving as its "identity card" in the digital world. A handheld RFID reader (PDA) is used to accurately record the operator's actions (such as "outbound loading," "on-site acceptance," and "installation") when scanning component or area tags with the PDA. This data is then compared with the timestamp of the scan and the data obtained from the PDA's built-in GPS module. The system automatically binds real-time latitude and longitude coordinates to ensure that every operation has a clear spatiotemporal context; it adds sensor groups to critical or vulnerable components (such as prestressed concrete components and precision steel nodes), such as temperature and humidity sensors that can monitor the environment of the components to prevent premature solidification of concrete or condensation corrosion of steel; triaxial vibration and impact sensors that can record events such as bumps and sudden braking during transportation to assess transportation quality; tilt sensors that can monitor the posture of components to prevent irreversible tilting or deformation during hoisting or transportation; and precision measurement sensors that can be linked with equipment such as total stations during the installation phase to collect installation positioning data.

[0031] The handheld PDA serves as a key data acquisition terminal for the system, supporting various operation types closely related to the component process, including outbound loading, on-site acceptance, installation, warehousing, hoisting preparation, departure confirmation, en route inspection, anomaly reporting, handling feedback, and photo uploading. Each type of operation is automatically bound to a timestamp, GPS location, and operator identity.

[0032] As a further aspect of the present invention, the data transmission module is used to securely, reliably, and efficiently transmit the collected raw data to the data processing center. This module can automatically select the optimal network channel based on the site environment. In areas with Wi-Fi coverage, such as factory areas or construction site warehouses, Wi-Fi transmission is prioritized to save bandwidth. During transportation, it automatically switches to 4G / 5G mobile networks. In network signal dead zones (such as remote mountain tunnels), the data acquisition devices (handheld PDAs, sensors) temporarily store the raw data locally. After the network is restored, the data is resumed from the interruption point to the cloud gateway. The cloud gateway first removes invalid data (such as misread data from extremely weak signals) and duplicate data, and then uses TLS / SSL encryption protocols for transmission to ensure data security during transmission and prevent information leakage. It also converts the raw, redundant data packets into standardized, structured JSON or XML formats, significantly reducing network bandwidth usage and the computational pressure on the subsequent processing center. The pre-processed "clean" data is then pushed to the data processing center module via a high-speed network, forming an efficient, secure, and standardized "collection-transmission-preprocessing" data pipeline.

[0033] As a further aspect of this invention, the core of the data processing center module is an RFID data quality management algorithm. This algorithm performs real-time intelligent processing on massive amounts of incoming data and builds a "scene feature-tag mapping library," which predefines all typical scenarios that a component may experience (such as "production completed," "in-plant storage," "loading and shipping," "transit in transit," "arrival at construction site," "on-site storage," "lifting preparation," and "installation completed"). Each scenario has its own feature vector, and a set of dedicated data verification rules and anomaly judgment rules are configured for each scenario in the "tag-rule association table." For example, the rules for the "transit in transit" scenario focus on GPS trajectory deviation and vibration data; while the rules for the "on-site storage" scenario focus more on whether the inventory time has expired and whether the temperature and humidity meet the standards.

[0034] The RFID data quality management algorithm employs a dynamic scene recognition and tag update mechanism. It continuously receives data packets from the data transmission module and uses GPS positioning data as the primary driver for dynamic scene judgment. The system pre-sets electronic geofences for key scenes on an electronic map. When a component's GPS coordinates continuously leave the current fence or enter a new fence for a period of time, the model initially triggers a "scene change alert." Subsequently, the model retrieves PDA operation data and sensor data from the same time period for auxiliary verification. For example, if the GPS shows that the vehicle has left the factory, the model will check for recent PDA scan operations related to "loading and shipping"; and whether the sensors show that the component has changed from a static state to a state of continuous vibration. Only after multiple sources of evidence corroborate each other is the scene change finally confirmed. Once confirmed, the model immediately updates the "scene identifier" field in the component's RFID electronic tag based on the "scene feature-tag mapping library," and synchronizes this update to the backend database, the application display module interface, and the PDAs of on-site personnel. Thus, all systems and personnel have a unified and real-time understanding of the component's current status.

[0035] As a further aspect of the present invention, an RFID-based component end-to-end tracking and management system maintains a dynamically updated risk assessment report for each component. This report records historical abnormal events, integrates key data indicators at each stage of the component's entire lifecycle, and is deeply bound to the BIM model. In the 3D visualization interface, the project manager can click on the digital twin of any component to view its current status and the historical versions and real-time content of its risk assessment report.

[0036] As a further embodiment of the present invention, the application display module creates a corresponding digital twin for each physical component based on the project's BIM model. Users can seamlessly switch perspectives in the 3D model via a web terminal or mobile APP to view the component's location, status (color coding indicates status: green for normal, yellow for warning, and red for abnormal) and detailed data, historical trajectory (replaying the complete movement route of the component from production to the present), and environmental parameters (real-time display of temperature, humidity, and other data transmitted by sensors).

[0037] When component data anomalies occur, the digital twin of the abnormal component will immediately highlight and flash in the model and pop up an alert window, integrating all key information (type, level, responsible person, guidelines). At the same time, the application display module provides powerful historical data query and statistical analysis functions, which can generate various reports.

[0038] Example 2 This embodiment uses the "Prefabricated Housing Industrial Park PC Component Full-Process Tracking Project" as the application scenario. The project has a total construction area of ​​150,000 square meters, covering 10 prefabricated residential buildings, and requires the production and transportation of a total of 5,000 prefabricated composite slabs, prefabricated beams, prefabricated columns and other components.

[0039] In this embodiment, the core equipment of a component full-process tracking management system based on RFID meets the requirements of anti-interference and harsh environment resistance in prefabricated building scenarios. The equipment types and uses are shown in Table 1 below. All equipment has passed industrial-grade certification to ensure stable operation in high temperature, high humidity and dust environments.

[0040] Table 1 Equipment Types and Uses

[0041] This system comprises three front-end application modules: a mobile APP that provides functions such as tag scanning, operation input, early warning viewing, and handling feedback; a web-based management platform with capabilities such as full-process tracking, BIM visualization, report statistics, and system configuration; and a monitoring center dashboard that can display the progress of component transfer and anomaly statistical analysis in real time.

[0042] In the early stages of system application, based on the CGCS2000 national geodetic coordinate system, electronic fences were drawn for key areas of the project. Specific details included: a polygonal fence covering three production workshops in the factory production area, with coordinates ranging from 30°15′20″–30°15′30″ N and 120°30′40″–120°30′50″ E; a rectangular fence dividing the factory storage area into ten storage locations, each 10m × 20m in size; a polygonal fence covering the gate, storage area, and hoisting area in the construction site area, with coordinates ranging from 30°20′10″–30°20′25″ N and 120°35′15″–120°35′30″ E; and circular fences with a radius of 50m covering key transportation nodes.

[0043] In the early stages of system application, a "Scenario-Label-Rule Mapping Table" was created based on the MySQL database, pre-setting verification rules and anomaly judgment thresholds for 8 core scenarios to ensure dynamic adaptive management of the system. The specific details are shown in Table 2. Table 2 "Scene-Label-Rule Mapping Table"

[0044] In the early stages of system application, an "Anomaly Risk Level Classification Table" was created to clarify the judgment criteria, handling methods, and response time limits for different risk levels, ensuring accurate and efficient anomaly handling, as shown in Table 3: Table 3 "Classification of Abnormal Risk Levels"

[0045] In this embodiment, the system achieves adaptive adjustment of the monitoring strategy through dynamic scene recognition and tag update. The triggering conditions for scene recognition include: PDA operation type change (such as from "in storage" to "loading"), GPS location exceeding the current electronic fence for 10 consecutive seconds, or sensor data characteristic change (such as vibration from ≤0.1g to >0.3g).

[0046] Taking the switch from "in-plant warehousing" to "loading and delivery" as an example, such as Figure 3As shown, after detecting the PDA "loading" operation, the system immediately initiates the recognition process: extracting multi-source features such as the current GPS location, vibration data, and operation type, and matching them with the scene feature vector library. If the matching degree exceeds a set threshold (e.g., 95% > 85%), a scene switching candidate log is generated, followed by an auxiliary verification stage. Within 10 seconds, the system continuously verifies whether the GPS is stable within the expected fence, whether the sensor data matches the new scene characteristics, and whether subsequent operations are logically consistent. After verification, the system confirms the scene switch and automatically updates the component label fields according to the scene-label mapping relationship. The updated label data is synchronized in real time to the application server database, the on-site PDA terminal, and the BIM digital twin model, driving the monitoring screen and APP interface to update dynamically. At the same time, the system automatically switches to the verification rules under this scene according to the label-rule association table.

[0047] In this embodiment, the system accurately distinguishes between "equipment failure," "human error," and "genuine anomalies" through a three-layer mechanism of "equipment status verification → data cross-validation → root cause determination." Figure 4 As shown, the specific content is as follows: First layer: Equipment status check (ruling out hardware failures) 1. The system monitors key indicators of all terminal devices in real time, collecting device status data every 5 seconds through the cloud gateway: RFID reader: communication status (online / offline), read success rate (≥95% is normal); GPS module: positioning accuracy (≤5m is normal), signal strength (≥-85dBm is normal); Sensor: communication status (online / offline), battery level (≥20% is normal), data update frequency (meeting the preset sampling rate is normal); PDA: battery level (≥20% is normal), network signal strength (≥-90dBm is normal), operation log completeness (no missing logs are normal); 2. The device is considered "faulty" if any of the following conditions are met: the device is offline for more than 10 seconds; the GPS positioning accuracy is greater than 5m for more than 30 seconds; the sensor battery level is less than 20% or the data update frequency deviates from the preset value by more than 50%; the PDA battery level is less than 20% or the network signal strength is less than -90dBm. 3. Equipment Failure Handling Procedure: For example, at 9:40 AM on May 12, 2024, the system receives GPS data (positioning accuracy 10m, duration 30 seconds) from component DB-2024-0001 and initiates equipment status verification; the system queries the status data of the GPS module (integrated into driver D001's PDA): signal strength -92dBm (< -85dBm), battery level 18% (< 20%); the system determines this as "equipment failure" and marks the GPS data during this period (9:40-9:45) as "low reliability". The system will not trigger any abnormal warnings for the time being. An automatic equipment maintenance notification will be pushed to the PDA of the maintenance personnel (employee number Y001): "The GPS signal of vehicle T001 PDA (ID: P-001) is weak and the battery is low. Please have it repaired in time." After the maintenance personnel arrive, they will replace the PDA battery (charge it to 80%), adjust the GPS antenna position, and the equipment status will return to normal (GPS signal strength -75dBm, positioning accuracy 2m). The system will detect that the equipment has returned to normal, generate an "Equipment Fault Repair" log, and remove the low confidence mark.

[0048] Second layer: Data cross-validation (eliminating data conflicts) When the equipment is in normal condition, the system performs "four-dimensional matching verification" on multi-source data at the same point in time to ensure data consistency. The specific verification dimensions are shown in Table 3. Table 3 Four-Dimensional Matching Validation

[0049] In this embodiment, the data cross-validation process takes "GPS deviation during transit" as an example: At 10:00 on May 12, 2024, the system received GPS data (30°18′00″N, 120°33′00″E) from component DB-2024-0001. The current location of the planned route R001 should be (30°17′30″N, 120°32′30″E), with a deviation distance of 600m. The first layer of equipment status verification: the PDAGPS module is online, the signal strength is -70dBm, the positioning accuracy is 2m, and the sensor communication is normal, indicating that the equipment is fault-free. The second layer of data cross-validation is initiated: the most recent RFID scan location (10 minutes ago, 30°17′25″N, 120°33′00″E) is queried through geographic matching. The distance between the current GPS data (20°32′25″) and the current GPS distance is 600m > 50m, indicating a geographical mismatch. Pattern matching of the vibration sensor data (0.3g) shows a 90% ≥ 80% match with the typical vibration pattern (0.2-0.4g) of "county roads" in the transportation road condition knowledge base, indicating a successful pattern match. Logical matching of the temperature and humidity (25℃ / 50%RH) is within the tolerance range of C30 concrete. The operation sequence is "loading → departure → en route," which is logically sound, indicating a successful logical match. Behavioral matching shows that the current operator D001 has the authority to perform "en route transportation" related operations, and the operation type is normal, indicating a successful behavioral match. Based on the cross-validation results, if one item is unqualified (geographical matching) and three items are qualified, the system determines that "data conflicts exist and further root cause analysis is required."

[0050] Third layer: Root cause determination (identifying the type of anomaly) Based on cross-validation results and human feedback, the system determines the root cause of anomalies through "rule-based reasoning + case matching." The specific determination logic is shown in Table 4. Table 4. Logic for Anomaly Detection

[0051] In this embodiment, false anomalies include: equipment malfunctions, which require that the signal, power, communication, or data frequency of devices such as GPS, RFID readers, sensors, and PDAs are substandard, such as weak GPS signals causing false location deviations; human error, where the device is normal but the operation is not standardized, such as PDA operation type not matching the scene or exceeding permission limits; and data transmission / environmental interference, where data distortion is caused by network blind spots, electromagnetic interference, or cloud gateway preprocessing errors, such as no signal in a tunnel causing false location loss. The determination requires a process of "initial triggering of anomaly prompt → priority verification of device status → cross-verification if the device is normal → manual review if necessary → confirmation of closed loop after repair". When the system determines that the anomaly is due to non-real situations such as equipment malfunction, the RFID data quality management algorithm will initiate an automatic repair process: first, the data generated by the faulty device in that period is marked as "low confidence" and isolated, thereby immediately lifting the false alarm warning; at the same time, the detailed cause of the fault (such as device number and fault type) is automatically recorded; subsequently, the system will attempt to compensate with the last valid data or through the associated data of other normal devices to ensure the continuity and reliability of the data chain.

[0052] In the "Full-Process Tracking Project of PC Components in Prefabricated Housing Industrial Park," the system confirms a "real anomaly" through a three-tiered progressive mechanism of "equipment status verification → data cross-validation → root cause determination," and then initiates a tiered handling process, such as... Figure 5 As shown, the specific content is as follows: The handling of low-risk anomalies is centered on operator-led operation with automatic system assistance. Taking the temperature and humidity exceeding the standard of component DB-2024-0003 as an example, when the system detects that the humidity of the precast composite slab exceeds the threshold of 8%RH for 15 minutes and is diagnosed as a low-risk anomaly, a pop-up window with optimization suggestions and a line graph of temperature and humidity changes will be pushed to operator D002's PDA within 5 minutes. The operator needs to complete the on-site adjustment and upload feedback within 30 minutes. The system will then initiate a second cross-validation within 15 minutes after the handling, generate an anomaly handling log after confirming that the data has returned to the threshold, and update the component risk assessment report, recording the handling details and subsequent prevention suggestions, thus achieving a rapid closed loop for low-risk anomalies.

[0053] The handling of medium-risk anomalies adds a review step at the management level. Taking component DB-2024-0015 with GPS trajectory deviation and excessive vibration as an example, after a medium-risk anomaly is diagnosed, the system will send a dual warning to the operator and the regional administrator within 10 minutes, and automatically generate a preliminary analysis report containing a snapshot of the abnormal data. The regional administrator needs to review the data and provide guidance within 15 minutes. After the operator handles the situation according to the guidance, the administrator needs to review and confirm it again. Finally, the system will store the entire process in the log and highlight the key points for follow-up tracking in the risk assessment report, which not only ensures the standardization of the handling, but also strengthens risk traceability.

[0054] High-risk anomaly handling emphasizes multi-role collaboration and technical support. Taking component DB-2024-0028 with an excessive tilt angle and visible damage as an example, after diagnosing a high-risk anomaly, if... Figure 6 As shown in the APP early warning center interface diagram, the system pushes early warnings to project managers, safety administrators, and the technical department within 5 minutes through multiple channels such as APP pop-ups, SMS, and full-screen prompts on the web. At the same time, it links the BIM model to highlight abnormal components and display key information. The technical department formulates a handling plan within 15 minutes. After the operator executes the plan, the supervisor needs to review it on-site and upload the inspection report. After the project manager makes the final confirmation, the early warning is closed. The system then integrates data from all stages to generate a handling file and updates the risk assessment report to ensure the rigor and safety of handling high-risk anomalies.

[0055] In this embodiment, the system dynamically generates an independent risk assessment report for each component, enabling full lifecycle risk traceability. The report content covers six core modules: basic component information, full lifecycle anomaly records, sensor data trend charts, operation log summary, risk level assessment, and improvement suggestions.

[0056] The above description is merely a specific embodiment of this application, but the scope of protection of this application is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the scope of the technology disclosed in this application should be included within the scope of protection of this application. Therefore, the scope of protection of this application should be determined by the scope of the claims.

[0057] In conclusion, the above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

Claims

1. An RFID-based component whole-process tracking management system, comprising a data acquisition module, a data transmission module and an application display module, characterized in that, The data processing center module receives the transportation data of the data transmission module, and the data processing center module screens the abnormal conditions of the component whole-process data in real time through an RFID data quality management algorithm. For the transportation data, the GPS positioning data is used to determine the dynamic scene, and the operation type scanned by the handheld PDA and the sensor data are used for auxiliary verification. After the scene change is confirmed, the RFID data quality management algorithm automatically updates the tag according to the'scene feature-tag mapping library', and synchronizes to the background database of the data processing center, the application display module and the on-site PDA. The RFID data quality management algorithm queries the 'tag-rule association table' in real time, adapts to the new rules corresponding to the new tag, and if the transportation data is abnormal, the RFID data quality management algorithm distinguishes the true and false of the abnormality through 'equipment state check + data cross verification'. If the abnormality is false, the RFID data quality management algorithm automatically repairs the data, records the reason and removes the abnormality warning. If the abnormality is true, the RFID data quality management algorithm processes the abnormality in layers. The low-risk abnormality automatically pushes the intervention suggestion, the high-risk abnormality is pushed to the person in charge through the application display module, and the BIM positioning and disposal guide is attached. After the person in charge completes the disposal, the disposal result is uploaded through the handheld PDA, the RFID data quality management algorithm starts again 'equipment state check + data cross verification', confirms that the abnormality has been solved, removes the abnormality warning, stores the disposal process into the whole-process log, and updates the risk assessment report of the component tracking.

2. The RFID-based component whole-process tracking management system according to claim 1, wherein, In the process of the RFID data quality management algorithm processing the transportation data abnormality, the GPS positioning data is used as the core, the electronic geofencing of the preset key scene is used to check whether the component real-time position is out of the current fence or enters a new fence, the track deviation degree is compared with the planned transportation route, the speed abnormality is analyzed by referring to the reasonable speed range of different scenes, the scene change prompt is preliminarily triggered, then the PDA scans the component and the area RFID tag, and the component state is input, the operation time stamp is synchronized with the GPS data, the matching of the manual operation and the scene is verified, the temperature, humidity, vibration and inclination data are collected by the sensor, and the component environment and the component state are verified.

3. The RFID-based component whole-process tracking management system according to claim 1, wherein, The RFID data quality management algorithm relies on a 'tag-rule linkage mechanism', a plurality of typical scenes are preset, each scene corresponds to a set of data acquisition rules and abnormality judgment rules, when the RFID tag is scanned by the reader, the current GPS position, time stamp and operation type are used to dynamically identify the scene, the component tag is updated according to the'scene feature-tag mapping library', and the corresponding rules in the 'tag-rule association table' are called to perform data verification and abnormality judgment. If the scene changes, the scene identifier in the component tag is automatically updated, and the rule set under the new scene is switched to, so that dynamic adaptive management is realized. The 'tag-rule linkage mechanism' ensures that the most suitable monitoring strategy is adopted in different stages.

4. The RFID-based component whole-process tracking management system according to claim 1, wherein, The "device state verification + data cross-validation" specifically includes that the RFID data quality management algorithm monitors the device running state of the RFID reader, GPS receiving module, environmental sensor and handheld PDA in real time, and if it is found that the device communication is interrupted, the power is lower than the threshold or the signal strength is insufficient, the data collected in this period is automatically marked as low credibility data, and the abnormal early warning is not triggered, at the data level, the system performs consistency comparison on multi-source data at the same time point, including geographic matching of GPS positioning coordinates and RFID latest scanning position, pattern matching of vibration and inclination data collected by the sensor and transportation road condition knowledge base, logical matching of temperature and humidity data and component material tolerance threshold, the system checks whether the operation type, operator identity and timestamp in the PDA operation log match the current transportation stage and electronic fence area, and if data conflict is found, the abnormal diagnosis sub-process is started to distinguish device failure, human error or real abnormality, and only the real abnormality verified passes to start the subsequent disposal process.

5. The RFID-based component whole-process tracking management system according to claim 1, characterized in that The abnormal diagnosis sub-process includes distinguishing the abnormal source through a three-layer progressive mechanism, first, the device state is verified in real time, if the GPS signal is lost, the sensor is powered off or the reader fails, it is determined as device failure; if the device is normal, the operation log is checked, when the operation site and time sequence do not match, the personnel permission exceeds the boundary or the process jumps the logical error, it is determined as human error; only when the device and operation are normal, but the GPS trajectory continuously deviates, the sensor data exceeds the threshold and the multi-source evidence chain is mutually verified, it is finally diagnosed as a real abnormality, thereby triggering a graded warning.

6. The RFID-based component whole-process tracking management system according to claim 1, characterized in that For the "real abnormality" output by "device state verification + data cross-validation", a hierarchical disposal mechanism is adopted, and a differentiated disposal scheme is formulated according to the risk level of the abnormality: for low-risk abnormality, the system automatically generates optimization suggestions and pushes them to the PDA of the on-site operator, prompting him to pay attention and make slight adjustments; for medium-risk abnormality, the system will automatically record the abnormal period data, generate a preliminary analysis report, and notify the regional administrator for review; for high-risk abnormality, the system immediately sends a warning to the project manager and safety administrator through the application display module, with BIM model accurate positioning, abnormal data snapshot and disposal guide, and requires the person in charge to feedback the disposal result within the specified time, all abnormal disposal processes are recorded and used to update the component risk assessment report.

7. The RFID-based component whole-process tracking management system according to claim 1, wherein, The component risk assessment report adopts a dynamic updating mechanism, and the report content includes abnormal disposal results, and is also associated with key data indicators in each stage of the component life cycle. The risk assessment report is deeply bound with the BIM model of the application display module, and in the three-dimensional visualization interface, the historical version and real-time update content of the corresponding risk assessment report can be viewed by clicking the component digital twin.

8. The RFID-based component whole-process tracking management system according to claim 1, wherein, The data acquisition module takes multi-device collaborative acquisition as the core, carries RFID tags, handheld RFID readers and auxiliary sensors, and can collect component basic identity information, real-time spatial position information, environmental state information and installation precision data. The handheld PDA also synchronously collects manual operation data and GPS positioning data for real-time matching, ensuring the scene correlation and integrity of the collected information.

9. The RFID-based component whole-process tracking management system according to claim 1, characterized in that, The data transmission module is used for transmitting and preprocessing the original data. It uses multiple network systems as data transmission channels, and first uploads the component identity, position, environment and operation original data obtained by the data acquisition module to the cloud gateway; The cloud gateway preprocesses the received original data, completes data filtering, encryption and lightweight format conversion in turn, and then transmits the cleaned data to the data processing center module. The cleaned data provides standardized data input for the subsequent abnormal screening and rule verification of the RFID data quality management algorithm, forming a coherent data link of "collection-transmission-preprocessing".

10. The RFID-based component whole-process tracking management system according to claim 1, wherein, The application display module supports three-dimensional visualization of the whole process tracking and abnormal early warning, including building a component three-dimensional digital twin based on a BIM model, real-time display of component position, state, historical trajectory and environmental parameters, user viewing of component whole life cycle information from production, transportation to installation through mobile APP or Web, highlighting of abnormal components in the three-dimensional model when an abnormality occurs, and pop-up of a warning window to show abnormal type, risk level, disposal suggestion and responsible person information.

Images