Long-term care insurance service process auditing and verifying method based on multi-source spatiotemporal fingerprints

Abstract

The application provides a long-term care insurance service process auditing and verifying method based on multi-source space-time fingerprints, and belongs to the technical field of long-term care insurance service supervision. The method solves the problem that the existing technology cannot confirm whether the nursing staff is actually present and the service behavior. The method comprises the following steps: generating an irregular electronic fence matched with the physical boundary of a service site; periodically collecting wireless environment signals around the mobile terminal of the nursing staff after the terminal enters the fence; collecting human activity state data through physical sensors in the service site; finally, performing consistency coupling verification based on the pre-designed service duration, wireless environment signals and human activity state data to determine the authenticity of the nursing service. The application fuses wireless environment fingerprints and human activity data of physical sensors to construct a dual verification mechanism of "device presence" and "personnel activity", which can effectively prevent fraud and realize high-credibility auditing of the whole service process.

Description

Technical Field

[0001] This application relates to the field of service supervision technology for long-term care insurance, and in particular to a method for auditing and verifying the long-term care insurance care service process based on multi-source spatiotemporal fingerprints. Background Technology

[0002] With the promotion of long-term care insurance, the demand for home-based care services is increasing. However, the regulation of such services faces significant challenges, especially in verifying the authenticity of the services.

[0003] In existing technologies, a common monitoring method is to use mobile devices held by caregivers for location tracking and check-in. For example, the location function of the mobile device can be used to determine whether the caregiver is within a pre-defined geofence and record the time spent within that geofence, serving as proof of service delivery. However, this method has significant drawbacks. First, location signals are easily simulated or forged. Caregivers could use technical means to "remotely check in," meaning the system can only confirm that the caregiver's device is in a designated location, but cannot guarantee the caregiver's actual presence, posing a risk of fraud through "human-machine separation." Second, even if the caregiver is present, existing technology cannot effectively monitor the service process, failing to confirm whether the service duration meets the standard or whether the service content actually occurred; the entire service process remains a "black box" for regulators. Furthermore, while some technologies attempt to introduce electronic fences, they often use fixed circular fences, which cannot accurately match the irregular building outlines of older residential areas or hospital wards, easily leading to misjudgments or omissions. Traditional auditing methods, such as manual spot checks and telephone follow-ups, are not only costly and inefficient, but also have limited coverage, making it difficult to create an effective deterrent and unable to build a complete and credible chain of evidence to deal with potential service disputes.

[0004] Therefore, existing technologies still have shortcomings in confirming the actual presence of personnel and in automating and reliably verifying the entire service process, and a new technical solution is urgently needed to solve the above problems. Summary of the Invention

[0005] The purpose of this application is to provide a method for auditing and verifying the long-term care insurance nursing service process based on multi-source spatiotemporal fingerprints. This method aims to solve the technical problem in existing long-term care insurance service supervision technologies, which can only verify the location of service equipment but cannot confirm the actual presence of caregivers and their actual service behavior, thus making it difficult to effectively prevent fraudulent behaviors such as "separation of person and machine" and insufficient service time.

[0006] To achieve the above objectives, this application provides a method for auditing and verifying long-term care insurance nursing service processes based on multi-source spatiotemporal fingerprints, comprising: obtaining a preset planned service duration corresponding to the nursing service; generating an irregular polygonal electronic fence that matches the physical boundary of the preset service location; periodically collecting wireless environmental signals around the mobile terminal after the caregiver's mobile terminal enters the electronic fence; collecting human activity status data within the service location through at least one physical sensor deployed within the service location; and performing consistency coupling verification based on the preset planned service duration, the wireless environmental signals, and the human activity status data to determine the authenticity of the nursing service.

[0007] Optionally, the wireless environment signal includes at least one of Wi-Fi signal fingerprint, Bluetooth signal fingerprint, and cellular network base station information.

[0008] Optionally, the physical sensor is a millimeter-wave radar sensor, and the human activity status data includes at least one of the following: human movement trajectory, duration of stillness, respiratory and heart rate information, or fall event.

[0009] Optionally, the consistency coupling verification includes: comparing the dwell time of the mobile terminal within the electronic fence with a preset duration requirement, wherein the preset duration requirement is determined based on the preset planned service duration, and the dwell time is determined based on the wireless environment signal.

[0010] Furthermore, the preset duration requirement is no less than 80% of the preset planned service duration.

[0011] Optionally, the consistency coupling verification includes: during the period when the mobile terminal is located within the electronic fence, combining the human activity status data, and by analyzing the periodically collected wireless environmental signals, verifying whether the mobile terminal has spatiotemporal continuity.

[0012] Optionally, the consistency coupling verification includes: during the period when the mobile terminal is within the electronic fence, by analyzing the human activity status data in conjunction with the wireless environmental signal, verifying whether the human activity status data conforms to at least one preset valid activity pattern.

[0013] Optionally, the consistency coupling verification includes: acquiring service operation logs recorded on the mobile terminal, and performing spatiotemporal matching verification between the service operation logs and the collected wireless environment signals and human activity status data.

[0014] Optionally, the method further includes: generating an encrypted evidence containing verification digest information based on the result of the consistency coupling verification, and automatically outputting an audit conclusion or marking cases of verification inconsistency as requiring manual review based on the evidence.

[0015] Compared with the prior art, this application has the following beneficial effects: 1. This application constructs a dual verification mechanism of "device presence" and "personal activity" by integrating wireless environmental signals collected by mobile terminals with human activity data collected by physical sensors in the service area. This effectively solves the fraud loophole of "human-machine separation" in the traditional location check-in method and greatly improves the reliability of presence verification.

[0016] 2. This application, through continuous analysis of wireless signal fingerprints and continuous monitoring of human activity status, can determine whether caregivers are "present throughout the entire process" and whether "service behavior actually occurs" during the service period, thus expanding supervision from a single point in time to the supervision of the entire service process.

[0017] 3. Adopting irregular electronic fences that match the physical boundaries of the service area can effectively reduce misjudgments and omissions caused by inaccurate fence ranges; at the same time, the automated multi-dimensional coupled verification process can handle the vast majority of audit tasks, allowing manual review efforts to focus on a few high-risk events, effectively improving audit efficiency and reducing operating costs.

[0018] 4. This application can generate consistency verification results containing multi-dimensional information such as time, space, environment, and behavior, forming a highly credible electronic evidence chain, providing strong support for the adjudication and handling of service disputes. Attached Figure Description

[0019] To more clearly illustrate the technical solutions in the embodiments of this application or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0020] Figure 1 A schematic diagram illustrating multi-source spatiotemporal data acquisition and fusion provided in an embodiment of this application; Figure 2 A layered architecture diagram of the audit system provided in the embodiments of this application; Figure 3 A flowchart illustrating the audit method provided in this application embodiment; Figure 4 This is a timing diagram of the signaling interactions between entities in one embodiment of this application. Detailed Implementation

[0021] To make the objectives, technical solutions, and advantages of this application clearer, the technical solutions in the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. It is to be understood that the specific embodiments described herein are for illustrative purposes only and are not intended to limit the scope of this application.

[0022] Example 1 This embodiment provides a complete implementation scheme for a long-term care insurance nursing service process audit and verification method based on multi-source spatiotemporal fingerprints. Specifically, the scheme integrates irregular electronic fences, multi-source wireless environmental signals, and millimeter-wave radar sensors, aiming to achieve highly reliable automated auditing of the entire nursing service process.

[0023] Please see Figure 2 This is a schematic diagram of the layered architecture of an audit system provided in one embodiment of this application. Physically, the system may include an audit backend server deployed in the cloud, a mobile terminal held by caregivers, and millimeter-wave radar installed in the service location. The audit backend server can be logically divided into multiple layers and modules. The bottom layer is the data source and support capability layer, responsible for providing basic data and capability support; above it is the data processing and storage layer, used to receive, clean, format, and persistently store the collected data. The core business processing layer of the system includes a dynamic attribute management module, a polymorphic perception acquisition module, a spatiotemporal preset comparison engine, a service behavior coupling verification module, a lightweight evidence generator, and a performance intelligent decision center. The top layer of the system is the application layer, providing an interactive interface for system administrators, auditors, and other users.

[0024] Please refer to the following: Figure 3 This is a flowchart illustrating an audit method provided in one embodiment of this application. For ease of understanding, this embodiment will use a specific nursing service scenario as an example to illustrate the process in detail.

[0025] The scenario is as follows: Mr. Zhang, the person being cared for, resides in Unit 1, Building A, in a residential community in a certain city. He has applied for a home care service with a preset service duration of 120 minutes through long-term care insurance. The system assigns caregiver Ms. Li to perform this service.

[0026] First, in step S301, an irregular polygonal electronic fence is generated. Before the service begins, the administrator of the audit system accesses the audit backend server through the application layer. On the map interface provided by the dynamic attribute management module of the server, the administrator enters Mr. Zhang's service address. The system loads the online map corresponding to the address and can preferably overlay and display the vector building outline data of the area. The administrator then uses a polygon drawing tool to accurately draw an irregular polygonal electronic fence that matches the height of the physical boundary of the apartment by clicking multiple vertices along the building exterior wall outline of Unit 1, Room 801, Building A displayed on the map. The geographic coordinate information of this electronic fence, such as an array of multiple latitude and longitude coordinate points, is saved and associated with Mr. Zhang's service order. Subsequently, the dynamic attribute management module sends the electronic fence data to the caregiver application installed on Ms. Li's mobile terminal via the network.

[0027] Next, physical sensors are deployed. To collect data on human activity within the service area, at least one physical sensor needs to be deployed beforehand. In this embodiment, as a preferred option, a 60 GHz millimeter-wave radar is installed in Mr. Zhang's living room or main activity area. This millimeter-wave radar connects to the internet via an indoor wireless network, enabling it to upload its detection data to the audit backend server in real-time or near real-time. It is understood that this millimeter-wave radar possesses non-contact detection capabilities, its advantage being the ability to perceive the presence, movement trajectory, posture (such as standing, sitting, falling), and minute vital signs (such as chest rise and fall caused by breathing and heartbeat) of people in an indoor space without collecting image information or infringing on personal privacy.

[0028] Subsequently, in steps S302 and S303, the operation of entering the fence and triggering multi-source spatiotemporal data fusion acquisition is performed. Caregiver Ms. Li arrives near Mr. Zhang's home at the agreed time. When she enters Unit 1, Room 801, Building A with her mobile terminal, the caregiver application on the mobile terminal detects through its own positioning function (such as GPS) that the current location has entered the geographical area of ​​a pre-issued irregular electronic fence.

[0029] Once the device enters the fenced area, the polymorphic sensing and data acquisition module on the mobile terminal (this module can be understood as functional logic within the application) is automatically triggered, starting to perform a data acquisition task in the background at a preset cycle (e.g., every 5 minutes). Figure 1As shown in the figure, this diagram visually illustrates the source of data collection. In each collection task, specific operations may include: 1. Collecting wireless environmental signals and motion data: The mobile terminal scans the surrounding wireless environment through its communication module and reads the step count data recorded by the mobile terminal's built-in sensors. Specifically, it collects Wi-Fi fingerprints, i.e., scans the service set identifiers, basic service set identifiers, and their respective received signal strength indicators of all detectable Wi-Fi access points in the vicinity; it collects Bluetooth signal fingerprints, i.e., scans surrounding Bluetooth beacons and records their universally unique identifiers, primary values, secondary values, and received signal strength indicators, which in this embodiment is represented as Bluetooth RSSI; it collects cellular network base station information, which, by collecting step count data, can further assist in determining the dynamic activity level of the caregiver within the fence. Specifically, it obtains information such as the location area code, cell identification code, and signal strength of the cellular network base station currently connected to the mobile terminal, which in this embodiment is represented as base station LAC / CI. It should be noted that these three types of wireless signals together constitute a highly unique "spatiotemporal fingerprint" within a specific time and specific micro-space. 2. Collect Human Activity Data: Simultaneously, the millimeter-wave radar deployed in Mr. Zhang's home continues to operate, sensing the indoor environment at its own detection frequency (e.g., several times per second). It processes and packages the detected human activity data and uploads comprehensive activity information (such as the presence of a human target, the sequence of target movement points, cumulative stillness duration, average respiratory rate, and whether a fall alarm occurred) to the audit backend server at the same interval (e.g., every 5 minutes). 3. Record Service Operation Logs: During the service process, caregiver Ms. Li may perform several key operations on the mobile application, such as clicking "Start Feeding" or "Complete Cleaning" and taking photos to upload. These operations, along with their timestamps and geographical location information, are recorded in the service operation log.

[0030] Figure 4 This is a sequence diagram of the signaling interactions between the entities in this embodiment. The audit backend server first sends fence data to the caregiver's APP. When the caregiver's APP enters the fence, it reports an entry event. Combined with... Figure 4 As shown, during the monitoring process after the service begins, the caregiver's app periodically reports the collected wireless fingerprints and human activity data. Until the service termination condition is triggered, when the caregiver's app leaves the fence, a departure event is reported at the end of the signaling cycle after the service begins.

[0031] In step S304, consistency coupling verification is performed. After the preset 2-hour service period ends, or after receiving an event that a caregiver has left the fence, the performance intelligent decision center of the audit backend server initiates the verification process. This process is supported by the core algorithm provided by the spatiotemporal preset comparison engine, and the service behavior coupling verification module performs multi-dimensional consistency coupling verification on the collected multi-source data to comprehensively determine the authenticity of the nursing service. Specifically, the verification dimensions may include: 1. On-site duration verification: The performance intelligent decision center first calculates the total stay time of the mobile terminal within the electronic fence. This duration can be determined based on the reported wireless environmental signal sequence, for example, from the time point of the first successful fingerprint match within the fence to the time point of the last match. If the calculated total stay time is 115 minutes, the system will obtain the preset planned service duration (120 minutes) for the service and compare it with a preset duration requirement. In this embodiment, the requirement is set to be no less than 80% of the preset planned service duration, i.e., 96 minutes. Since 115 minutes is greater than 96 minutes, the on-site duration verification passes. 2. Spatiotemporal Continuity Verification: The spatiotemporal preset comparison engine analyzes 23 consecutive wireless environmental signal samples collected during the service period (115 minutes / 5 minutes ≈ 23). Location stability is determined by calculating the similarity of wireless environmental fingerprints between two adjacent time points (e.g., t and t+5 minutes). The similarity calculation can be a comprehensive algorithm, such as calculating the Jaccard similarity coefficient for the list of basic service set identifiers in the Wi-Fi fingerprint, and calculating the cosine similarity or Pearson correlation coefficient for the sequence of received signal strength indicators of common basic service set identifiers. If the similarity between all adjacent samples is higher than a preset threshold (e.g., 90%), it indicates that the mobile terminal's location was stable during the service period, without abnormal location drift such as leaving and returning to the service location, thus indirectly proving the caregiver's continuous presence during the service period. 3. Human Activity Validity Verification: The intelligent decision-making center then analyzes human activity status data received from millimeter-wave radar to verify whether this data conforms to preset valid activity patterns. For example, an effective activity pattern can be defined as "during the service period, at least one human target with normal vital signs is continuously detected, and there is no continuous absence or purely static state for more than 15 minutes." The system checked the radar data and found stable human movement trajectories and normal breathing and heartbeat information records throughout the service period, without any abnormal events such as prolonged absence or loss of vital signs. Therefore, the validity of human activity was verified. This proves that there are indeed people active in the service area, and combined with the aforementioned verification, it can be inferred with a high probability that the caregiver is present and active. 4. Operation behavior matching verification: The system obtains the service operation logs submitted by caregiver Ms. Li on her mobile terminal, such as the "completed feeding" operation recorded at 10:30 am.The system performs spatiotemporal matching verification between the timestamp of this operation and the collected wireless environmental signals and human activity status data. The verification includes: checking whether 10:30 AM is within the valid service duration; checking whether the wireless environmental fingerprint collected at this time matches the stable fingerprint of the service location; and checking whether the millimeter-wave radar also detected corresponding interactive activities before and after this time point. If all matches are successful, it proves that the critical service node was completed at the correct time and location.

[0032] Finally, in step S305, the audit conclusion is output and a certificate of authenticity is generated. Since the consistency coupling verification of all the above dimensions has passed, the performance intelligent decision center ultimately determines that the nursing service has passed the audit, meaning it has been genuinely and effectively performed. The system will automatically generate an encrypted certificate of authenticity containing verification digest information using a lightweight certificate of authenticity generator. This certificate of authenticity can be a structured data packet containing the service order ID, caregiver ID, patient ID, planned service time, actual on-site duration, wireless environment fingerprint digests of each sampling point (e.g., generated using a hash algorithm), the validity verification conclusion of human activity, key operation logs and their verification results, etc. This certificate of authenticity packet is encrypted (e.g., signed using a secure hash algorithm) to ensure its immutability. Finally, the system outputs the conclusion "audit passed" and can synchronize this conclusion and the certificate of authenticity information to the upper-level insurance claims system as a reliable basis for cost settlement. Accordingly, if any verification fails, the system will mark the service as "pending manual review" and push it to the auditor's work queue, along with detailed anomaly data and preliminary analysis, for further manual review.

[0033] In summary, the technical solution provided in this embodiment can effectively solve the problems of "separation of man and machine" and "no service provided on site" in traditional supervision methods, and realize comprehensive, automated and highly reliable audit of nursing services from location, duration to process behavior.

[0034] Example 2 This embodiment is used to illustrate the adaptability of the technical solution of this application, especially in scenarios where the types of wireless environmental signal sources are limited, it can still achieve effective auditing. The main difference between this embodiment and Embodiment 1 is that the collected wireless environmental signals do not include Bluetooth signals.

[0035] Assume the nursing service is being provided in a newly built residential community where no public or private Bluetooth beacon devices are deployed. Therefore, the caregiver's mobile device will not be able to scan for available Bluetooth signals.

[0036] In this scenario, the auditing method of this application will be adjusted as follows: The preliminary steps, such as fence generation and physical sensor deployment, will remain consistent with Example 1. The key difference lies in the data acquisition and verification stages.

[0037] During the data acquisition phase, when the caregiver's mobile terminal enters the irregular electronic fence, its multi-modal sensing acquisition module will only collect and report Wi-Fi fingerprints and cellular network base station information during periodic scanning. At this time, the Bluetooth scanning function can be disabled or its scan result will be empty. It should be noted that the millimeter-wave radar's operation and data reporting process are completely unaffected, and it continues to collect and report human activity status data.

[0038] Correspondingly, during the consistency coupling verification phase, the verification logic of the audit backend server is also adaptively adjusted: 1. On-site duration verification: This verification is unaffected because the duration of the mobile terminal's stay within the fence can still be determined by analyzing the sequence of Wi-Fi fingerprints and base station information 103. 2. Spatiotemporal continuity verification: At this time, the spatiotemporal preset comparison engine will rely solely on Wi-Fi fingerprints and base station information 103 to calculate the similarity of the location fingerprint. For example, the algorithm will compare whether the list of Wi-Fi access points (basic service set identifier set) of adjacent sampling points is stable, and whether the cell identification code of the connected base station has changed abruptly. Although the dimensionality of the fingerprint is reduced, in most indoor environments, the uniqueness of the Wi-Fi signal is sufficient to provide a reliable judgment of location stability. If the caregiver leaves the service location for a considerable distance, the base station they are connected to is likely to change, which will cause a sharp drop in fingerprint similarity, thus being identified as an anomaly by the system. 3. Human activity validity verification and operational behavior matching verification: These two verifications mainly rely on millimeter-wave radar data and operation logs, combined with time information and remaining wireless signals for spatiotemporal anchoring. Since the collection of millimeter-wave radar and operation logs is unaffected, these two verifications can proceed as usual.

[0039] Ultimately, even without the Bluetooth signal dimension, the system can still comprehensively determine the authenticity of the service by coupling and verifying Wi-Fi signals, base station information, and human activity data from millimeter-wave radar.

[0040] Therefore, this embodiment fully demonstrates the robustness of the solution proposed in this application. The "wireless environment signal" defined in the technical solution can be a combination of one or more signals. The system can dynamically adjust its verification model according to the actual available signal sources, thereby working effectively in environments with different signal coverage conditions. This provides a broader scope of protection for the technical solution claimed in this application.

[0041] Example 3 This embodiment is used to illustrate the universality of the core idea of ​​this application, namely the concept of combining "wireless environmental fingerprinting" with "physical environmental perception," and is not limited to specific millimeter-wave radar technology. This embodiment will demonstrate the use of other types of physical sensor combinations to replace millimeter-wave radar in order to monitor human activity.

[0042] In this embodiment, millimeter-wave radar is not deployed within the service area. As an optional implementation, a sensor network composed of different types of physical sensors can be used. The specific deployment is as follows: 1. A passive infrared motion sensor is installed at the entrance to the living room or main passageway of the patient's residence. This sensor can detect changes in infrared radiation caused by human movement, thereby determining whether someone has passed by or is active in the area. 2. In key areas involved in the service, such as the kitchen and bathroom, a privacy-protecting sound sensor is installed. It is understood that this sensor does not record the raw sound content, but rather uses built-in edge computing capabilities to analyze the sound signal in real time and extract sound event patterns. For example, it can identify the sound of flowing water, clattering dishes, television sounds, or human voices, and only uploads the tags and timestamps of these events.

[0043] Once the nursing service begins, the data collection and verification process is adjusted accordingly: During the data collection phase, the process for the mobile terminal to collect wireless environmental signals is the same as in Example 1 or Example 2. The difference is that human activity status data is no longer provided by millimeter-wave radar, but by a network of passive infrared sensors and sound sensors. These sensors are also connected to the Internet and upload the events they detect (e.g., “PIR_motion_detected, timestamp: T1” or “sound_event: water_flow, location: bathroom, timestamp: T2”) to the audit backend server.

[0044] During the consistency coupling verification phase, the logic for validating human activity in the intelligent decision-making center for contract fulfillment was modified as follows: 1. The system first checks the data from the passive infrared sensors to verify whether motion detection signals were continuously or intermittently received throughout the entire service time window. If there are no motion signals for a long period (e.g., more than 30 minutes), the system may determine it as abnormal, indicating that there may be no activity in the service area. 2. The system then analyzes the sound events uploaded by the sound sensors. This verification is correlated with the content of the service work order. For example, if the service work order lists items such as "assisting with washing up" and "preparing lunch," the system expects to receive a "water flow sound" event from the sound sensor in the bathroom and related sound events such as "clinking dishes" from the sound sensor in the kitchen during the corresponding time period. If the recorded operation log shows that the caregiver completed "assisting with washing up" at a certain time, but the system does not receive any related sound events from the sound sensor in the bathroom before or after that time, an inconsistency alarm will be generated.

[0045] In this way, even without using millimeter-wave radar, the system can still determine the authenticity of service behavior by analyzing data indirectly related to human activity from different physical sensors and coupling it with service plans, operation logs, and wireless environmental fingerprints.

[0046] This embodiment demonstrates that the concept of "physical sensor" in this application has broad applicability. It can be millimeter-wave radar, or any combination of devices that can directly or indirectly reflect the physical environment or human activity status within a service area, such as passive infrared sensors, sound sensors, pressure sensors, and door magnetic sensors. This provides solid implementation support for the generalization of this technical feature.

[0047] Example 4 This embodiment further demonstrates the flexibility and scalability of the technical solution of this application in practical applications. Specifically, it introduces a differentiated verification strategy based on service risk level to achieve a balance between audit rigor and deployment cost.

[0048] In actual operation, different nursing services vary in importance, complexity, and historical fraud risk. For example, a simple "medication reminder" service carries a significantly different risk level than a complex "full body bath" service. Therefore, different audit intensities can be assigned to different services.

[0049] As an optional implementation, a risk level management function can be added to the system configuration of the audit backend server. The system administrator or business rule engine can mark each service order as "high risk" or "low risk" based on factors such as the content of the service project, the cost, the health status of the service recipient, and the caregiver's historical credit history.

[0050] When a nursing service is triggered, the intelligent decision-making center for fulfillment will first obtain the risk level of the service and then execute different verification processes based on the level: 1. Low-Risk Service Validation Mode: For services marked as "low-risk," such as "blood pressure measurement" or "medication reminders," these services are typically short-duration and simple to operate. The system can adopt a "lightweight" validation mode. In this mode, the system only mandates the use of irregular geofences and the collection of multi-source wireless environmental signals. The focus of consistency coupling validation will be on "presence duration verification" and "spatiotemporal continuity verification." In other words, the core of this mode is to ensure that the caregiver's mobile terminal arrives at the designated location at the predetermined time, stays there for a sufficient period of time, and does not move abnormally during this period. For data from physical sensors (such as millimeter-wave radar), the system may not mandate collection, or even if collected, it may only be used as an auxiliary reference and not as a strong basis for determining service invalidity. This mode reduces the requirements for hardware deployment at service locations and is suitable for large-scale, low-cost, and rapid deployment.

[0051] 2. High-Risk Service Verification Mode: For services marked as "high-risk," such as "full-body sponge bath," "pressure ulcer care," or services for elderly people at severe risk of disability, the system will activate a "strong verification" mode. In this mode, the system will utilize all the functions described in this application. That is, in addition to mandatory wireless environmental signal acquisition, physical sensors (such as millimeter-wave radar) must be deployed and activated at the service location to monitor human activity. Consistency coupling verification will perform rigorous checks across all four dimensions: presence duration verification, spatiotemporal continuity verification, human activity validity verification, and operational behavior matching verification. In "strong verification" mode, failure in any dimension may directly result in the service being marked as "awaiting manual review." In particular, if the millimeter-wave radar detects a serious anomaly, such as during the claimed "full-body sponge bath" period, radar data showing "unmanned state" or "single person remaining stationary for a long time," the system will immediately raise the risk level of the event to the highest level and trigger a real-time alarm to auditors.

[0052] By implementing this risk-based verification strategy, the technical solution of this application can allocate audit resources more intelligently and efficiently. It allows the most stringent and comprehensive monitoring methods to be concentrated on the high-risk services that require the most attention, while adopting lower-cost and easier-to-deploy verification methods for a large number of low-risk services. Thus, while ensuring that core risks are controllable, it achieves the optimization of overall operating costs and management efficiency, demonstrating the solution's good commercial applicability and scalability.

[0053] The above description is merely a preferred embodiment of this application and is not intended to limit the scope of this application. Various modifications and variations can be made to this application by those skilled in the art. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of this application should be included within the protection scope of this application.

Claims

1. A method for auditing and verifying the long-term care insurance nursing service process based on multi-source spatiotemporal fingerprints, characterized in that, include: Obtain the preset planned service duration corresponding to the nursing service; Generate irregular polygonal electronic fences that match the physical boundaries of the preset service locations; After the caregiver's mobile terminal enters the electronic fence, the wireless environmental signals around the mobile terminal are periodically collected. Data on human activity status within the service area are collected by at least one physical sensor deployed within the service area. The authenticity of the nursing service is determined by performing a consistency coupling verification based on the preset planned service duration, the wireless environment signal, and the human activity status data.

2. The method according to claim 1, characterized in that, The wireless environment signal includes at least one of the following: Wi-Fi signal fingerprint, Bluetooth signal fingerprint, and cellular network base station information.

3. The method according to claim 1, characterized in that, The physical sensor is a millimeter-wave radar sensor, and the human activity status data includes at least one of the following: human movement trajectory, duration of stillness, respiratory and heart rate information, or fall event.

4. The method according to claim 1, characterized in that, The consistency coupling verification includes: The dwell time within the electronic fence is compared with a preset duration requirement, wherein the preset duration requirement is determined based on the preset planned service duration, and the dwell time is determined based on the wireless environmental signal.

5. The method according to claim 4, characterized in that, The preset duration requirement is no less than 80% of the preset planned service duration.

6. The method according to claim 1, characterized in that, The consistency coupling verification includes: During the period when the mobile terminal is located within the electronic fence, by combining the human activity status data and analyzing the periodically collected wireless environmental signals, it is verified whether the mobile terminal has spatiotemporal continuity.

7. The method according to claim 1, characterized in that, The consistency coupling verification includes: During the period when the mobile terminal is within the electronic fence, the human activity status data is analyzed in conjunction with the wireless environmental signal to verify whether the human activity status data conforms to at least one preset valid activity mode.

8. The method according to claim 1, characterized in that, The consistency coupling verification includes: Obtain the service operation log recorded on the mobile terminal, and perform spatiotemporal matching verification between the service operation log and the collected wireless environment signal and human activity status data.

9. The method according to claim 1, characterized in that, Also includes: Based on the results of the consistency coupling verification, an encrypted certificate containing verification digest information is generated, and an audit conclusion is automatically output based on the certificate, or cases of verification inconsistency are marked as requiring manual review.

Images