Advertisement playing monitoring method and device based on virtual twin device and terminal

Abstract

The application discloses an advertisement playing monitoring method and device based on a virtual twin device and a terminal, relates to the technical field of advertisement playing control, and comprises the following steps: a server generates a unique virtual twin device instance according to the ID of each real advertisement device and constructs a corresponding virtual twin device digital model; a real advertisement device periodically reports light playing state data to the server; the server combines a preset playing task and a time axis synchronization mechanism according to the reported playing state data, simulates and reconstructs a virtual playing picture synchronized with the real device in real time at the server end; the virtual playing picture is viewed in real time; the server cross-verify based on multiple data sources, verifies the consistency between the virtual playing picture and the real playing picture of the real advertisement device; and when the virtual playing picture and the playing state data are detected to be abnormal, an alarm is triggered and corresponding personnel terminals are notified. The application provides an advertisement playing monitoring method with low bandwidth, high credibility and real-time remote monitoring.

Description

Technical Field

[0001] This invention relates to the field of advertising playback control technology, and in particular to an advertising playback monitoring method, device, server, and storage medium based on a virtual twin device. Background Technology

[0002] With the development of internet technology, digital advertising delivery systems are becoming increasingly popular.

[0003] However, existing digital advertising delivery systems have the following obvious drawbacks: 1) Inability to confirm actual playback status: The backend can usually only see task issuance, device online status, or successful playback logs, but cannot confirm whether the device has actually played the corresponding advertising content; 2) Lack of real-time visual monitoring capabilities: Users cannot remotely view the advertising screen currently playing on the device in real time, requiring manual on-site inspection; 3) High cost of video streaming solutions: If the device directly uploads the video stream, it will result in high network bandwidth consumption, high terminal hardware requirements, and high costs for concurrent monitoring of multiple devices.

[0004] Therefore, existing technologies still need to be improved and developed. Summary of the Invention

[0005] The technical problem to be solved by the present invention is to provide an advertising playback monitoring method, device, server and storage medium based on virtual twin devices, which addresses the shortcomings of the prior art. The present invention provides an advertising playback monitoring method with low bandwidth, high reliability and real-time remote monitoring capability.

[0006] The technical solution adopted by this invention to solve the problem is as follows: A method for monitoring advertising playback based on a virtual twin device, comprising: The server generates a unique virtual twin device instance based on the ID of each real advertising device and constructs a corresponding virtual twin device digital model, which is used to simulate the playback behavior and status of the corresponding real advertising device on the server side. The real advertising device receives the advertising playback task, executes the advertising playback, and periodically reports lightweight playback status data to the server; The server receives playback status data reported by the real device, and combines it with pre-set playback tasks and timeline synchronization mechanisms to simulate and reconstruct a virtual playback screen synchronized with the real device in real time on the server side. View the virtual playback screen of the digital model of the virtual twin device in real time through the client; The server performs cross-validation based on multiple data sources to verify the consistency between the virtual playback screen and the real playback screen of the real advertising device. The server continuously monitors and interprets the virtual playback screen and playback status data for any abnormalities. If any abnormalities are detected, the server will trigger an alarm and notify the corresponding personnel's terminal.

[0007] The aforementioned advertising playback monitoring method based on virtual twin devices, wherein the step prior to the step of pre-establishing a corresponding virtual twin device digital model for each real advertising device on the server side includes: When each real advertising device is first started, it registers with the server and uploads its device ID, device model, screen resolution and other basic information.

[0008] The advertising playback monitoring method based on virtual twin devices, wherein the step of pre-establishing a corresponding virtual twin device digital model for each real advertising device on the server side includes: The pre-set digital model of the virtual twin device includes the static information of the real device, as well as dynamic playback task mapping module, timeline synchronization module, playback status model, rendering and display module, anomaly detection module and history record module; The static information of the device includes: device ID, model, screen orientation, resolution, network status, and customer information.

[0009] The advertising playback monitoring method based on virtual twin devices, wherein the step of pre-establishing a corresponding virtual twin device digital model for each real advertising device on the server side includes: The virtual twin device digital model is set up using a task data + status data + timeline synchronization mechanism. By using lightweight task data and periodically reported status data, combined with timeline synchronization, the playback screen of the real device is reconstructed on the server side. The task data refers to the advertising playback instructions issued by the server to the real device, including the list of advertising materials, playback order, playback time period, duration of a single advertisement, loop strategy, and insertion rules. The status data refers to the data that the actual device periodically reports to the server during playback, including the current ad content ID, current playback progress, current timestamp, playback status, volume / screen / network status, and optional screenshots / short video clips; The timeline synchronization mechanism is used to ensure that the playback progress of the virtual twin device on the server side is consistent with the timeline of the real device.

[0010] The aforementioned advertising playback monitoring method based on virtual twin devices includes the following steps: the server performs cross-validation based on multiple data sources to verify whether the playback screen is consistent with the playback result of a real advertising device. The multi-source data includes: status data, playback logs, screenshot data, and short video clips.

[0011] The aforementioned advertising playback monitoring method based on virtual twin devices, wherein the server continuously detects and judges whether the virtual playback screen and playback status data are abnormal, and when abnormalities are detected in the virtual playback screen and playback status data, the step of controlling and triggering an alarm and notifying the corresponding personnel terminal further includes: The server continuously monitors and interprets the virtual playback screen and playback status data for any abnormalities. If any abnormalities are detected, such as black screen, stuttering, playback errors, playback not on schedule, network offline, playback interruption, or no status reporting for an extended period, the server will trigger an alarm and notify the corresponding personnel's terminals.

[0012] The advertising playback monitoring method based on virtual twin devices, wherein the step of simulating and reconstructing the virtual playback screen synchronized with the real device in real time on the server side includes: Artificial intelligence technology is introduced to perform content recognition and analysis on the screenshots of the virtual playback screen to verify the consistency between the virtual playback screen and the real playback screen of the real advertising device.

[0013] An advertising playback monitoring device based on a virtual twin device, wherein the device includes: The virtual twin device construction module is used to control the server to generate a unique virtual twin device instance based on the ID of each real advertising device, and to build a corresponding virtual twin device digital model, which is used to simulate the playback behavior and status of the corresponding real advertising device on the server side, serving as a digital substitute for the real advertising device. The playback status data reporting module is used to control the real advertising devices to receive advertising playback tasks, execute advertising playback, and periodically report lightweight playback status data to the server. The virtual playback screen reconstruction module is used to control the server to receive playback status data reported by the real device, and combine it with the pre-set playback task and timeline synchronization mechanism to simulate and reconstruct the virtual playback screen synchronized with the real device in real time on the server side. The virtual playback screen monitoring and viewing module is used to control the virtual playback screen of the virtual twin device digital model to be viewed in real time through the client. The cross-validation module is used to control the server to perform cross-validation based on multiple data sources to verify the consistency between the virtual playback screen and the real playback screen of the real advertising device; The anomaly detection and handling module is used to control the server to continuously detect and judge whether the virtual playback screen and playback status data are abnormal. When anomalies are detected in the virtual playback screen and playback status data, the system will control the triggering of an alarm and notify the corresponding personnel terminal.

[0014] A server includes a memory and one or more programs, wherein the one or more programs are stored in the memory and configured to be executed by one or more processors, the one or more programs comprising the method for performing any one of the methods.

[0015] A computer-readable storage medium, wherein, when instructions in the storage medium are executed by a processor of an electronic device, the electronic device is enabled to perform any of the methods described above.

[0016] The present invention provides a method, apparatus, server, and storage medium for monitoring advertising playback based on virtual twin devices. By establishing a virtual twin device model for each real advertising device on the server side, the real device only needs to periodically report lightweight playback status data. The server side, combined with preset playback tasks and a timeline synchronization mechanism, can reconstruct the playback screen synchronized with the real device in real time. Users can view this virtual playback screen through a remote monitoring interface, thereby gaining real-time and continuous understanding of the actual playback status of the advertisement without the need for manual on-site inspection, thus solving the problem of real-time visual monitoring. Furthermore, the present invention also has the following advantages: 1) Enables remote monitoring without video streaming, significantly reducing bandwidth costs; 2) It can realistically reproduce the ad playback process and allow continuous viewing of the ad playback status, rather than a static screenshot; 3) This invention is suitable for large-scale equipment deployment and supports the simultaneous monitoring of thousands to tens of thousands of devices; 4) This invention can detect anomalies in a timely manner, thereby improving the advertising fulfillment rate; 5) This invention can improve customer trust, allowing customers to verify in real time whether the advertisement is actually delivered. Attached Figure Description

[0017] To more clearly illustrate the technical solutions in the embodiments of the present invention 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 recorded in the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0018] Figure 1 This is a flowchart illustrating the advertising playback monitoring method based on a virtual twin device provided in Embodiment 1 of the present invention.

[0019] Figure 2 This is a flowchart illustrating the advertising playback monitoring method based on a virtual twin device provided in Embodiment 2 of the present invention.

[0020] Figure 3The principle block diagram of an embodiment of the advertising playback monitoring device based on virtual twin devices provided by the present invention.

[0021] Figure 4 This is a block diagram illustrating the internal structure of the server provided in an embodiment of the present invention. Detailed Implementation

[0022] To make the objectives, technical solutions, and advantages of this invention clearer and more explicit, the invention will be further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative of the invention and are not intended to limit the invention.

[0023] It should be noted that if the embodiments of the present invention involve directional indicators (such as up, down, left, right, front, back, etc.), the directional indicators are only used to explain the relative positional relationship and movement of the components in a certain specific posture (as shown in the figure). If the specific posture changes, the directional indicators will also change accordingly.

[0024] Traditional advertising delivery systems suffer from flawed monitoring mechanisms. Specifically, the actual playback status of ad content cannot be verified, relying solely on task issuance confirmations, device online status, and log feedback data, resulting in insufficient reliability of monitoring data. Furthermore, the lack of real-time visual monitoring capabilities necessitates manual on-site inspections for remote monitoring. Additionally, existing screenshot solutions suffer from high upload latency, low frequency, and an inability to reflect continuous playback, making it difficult to capture momentary anomalies. Video streaming solutions are uneconomical due to high network bandwidth consumption, demanding terminal hardware requirements, and high costs associated with multi-device concurrency. Ultimately, abnormal events such as device blackouts, stuttering, and content misplays are difficult to detect in a timely manner, impacting system reliability.

[0025] For example, in an advertising playback terminal network deployed in a commercial complex, multiple devices are distributed across different functional areas. When an advertising playback task is executed, the task issuance status and device online information can only be confirmed; whether the actual playback screen matches the task cannot be verified. During manual inspections, a device screen going black may be discovered, but the status data may still report it as playing. Periodic screenshots, due to long intervals, may not record momentary stuttering, leading to a disconnect between monitoring data and the actual playback status. In this scenario, the problem manifests as inconsistency between monitoring data and the actual playback process, directly affecting the accuracy of evaluating advertising effectiveness.

[0026] If the above problems are not resolved, the authenticity of advertising placements cannot be guaranteed, affecting the trust relationship between advertisers and platforms; abnormal events will continue to go undetected, resulting in wasted resources and customer complaints; the overall reliability of the system will decrease, and maintenance complexity will increase; and it may hinder the healthy development of the digital advertising industry.

[0027] In response, this application proposes a method for monitoring advertising playback based on a virtual twin device. For example... Figure 1 As shown, an advertising playback monitoring method based on a virtual twin device according to Embodiment 1 of the present invention includes the following steps: Step S100: When each real advertising device is started for the first time, it registers with the server and uploads its own device ID, device model, screen resolution and other basic information.

[0028] Step S200: The server generates a unique virtual twin device instance based on the ID of each real advertising device and constructs a corresponding virtual twin device digital model, which is used to simulate the playback behavior and status of the corresponding real advertising device on the server side. Step S300: The real advertising device receives the advertising playback task, executes the advertising playback, and periodically reports lightweight playback status data to the server. Step S400: The server receives the playback status data reported by the real device, and combines it with the pre-set playback task and timeline synchronization mechanism to simulate and reconstruct the virtual playback screen synchronized with the real device in real time on the server side. Step S500: View the virtual playback screen of the digital model of the virtual twin device in real time through the client; Step S600: The server performs cross-validation based on multiple data sources to verify the consistency between the virtual playback screen and the real playback screen of the real advertising device; Step S700: The server continuously detects and judges whether the virtual playback screen and playback status data are abnormal. When abnormality is detected in the virtual playback screen and playback status data, the server controls the triggering of an alarm and notifies the corresponding personnel terminal.

[0029] The virtual twin device in this embodiment refers to a virtual entity created in the digital space of a server that corresponds one-to-one with a physical entity (such as a real advertising device). The virtual twin device is synchronized with the physical entity through a data connection, and can reflect the state, behavior, and performance of the physical entity in real time. It can also be used to simulate, analyze, and predict the operation of the physical entity.

[0030] The digital model of the virtual twin device in this embodiment refers to the core component of the virtual twin device. It is a mathematical or computer model that includes the various attributes, behavioral logic, and operating rules of the physical entity. The digital model of the virtual twin device in this embodiment is used on the server side to simulate and reconstruct the playback behavior and state of the real advertising device.

[0031] In this embodiment, the actual advertising device refers to a hardware terminal device that is actually deployed in a physical environment to play advertising content. The actual advertising device can be a digital signage unit, LED screen, interactive kiosks, etc., which receive advertising playback tasks and execute the actual advertising playback.

[0032] The playback status data in this embodiment refers to lightweight data about its own playback status that is periodically reported to the server by the actual advertising device during the execution of the advertising playback task. The playback status data in this embodiment includes information such as the current playback content identifier, playback progress, timestamp, and device operating status.

[0033] In this embodiment, a playback task refers to a set of instructions issued by the server to a real advertising device to guide it in playing advertisements. The playback task includes detailed information such as a list of advertising materials, playback order, playback time period, duration of each advertisement, looping strategy, and insertion rules.

[0034] The timeline synchronization mechanism in this embodiment refers to the technical means to ensure that the playback progress of the virtual twin device on the server side is consistent with the timeline of the real advertising device. Through the timeline synchronization mechanism, the server can accurately reproduce the playback timeline of the real device in the virtual environment based on the playback status data of the real device and preset tasks.

[0035] The virtual playback screen in this embodiment refers to a digital image simulated and reconstructed on the server side using a virtual twin device digital model, synchronized with the actual playback screen of the real advertising device. The virtual playback screen does not rely on the video stream transmission of the real device, but is generated based on data and model calculations.

[0036] Cross-validation in this embodiment refers to the process of confirming the accuracy and consistency of information by comparing data from different sources. In this embodiment, the cross-validation is used to confirm whether the virtual playback screen is consistent with the actual playback screen of the real advertising device.

[0037] The alarm in this embodiment refers to a notification mechanism that is automatically triggered when the system detects an abnormal situation; it aims to promptly remind relevant personnel to pay attention to and handle potential problems, such as equipment failure, playback errors, or data anomalies.

[0038] This embodiment provides a method for monitoring advertising playback based on virtual twin devices. In specific implementation, when each real advertising device is started for the first time, it registers with the server and uploads its own device ID, device model, screen resolution and other basic information.

[0039] This invention involves registering and uploading basic information about the device, such as device ID, device model, and screen resolution, to the server upon initial startup of each real advertising device.

[0040] The initial startup refers to the stage where a real advertising device connects to the server and performs initialization operations for the first time after being put into use or reset. This stage is crucial for establishing an initial connection and identity verification between the device and the system, ensuring that the server accurately obtains and records the device's basic information before it is officially put into advertising playback. This provides the necessary prerequisite for the accurate construction of the subsequent digital model of the virtual twin device.

[0041] Registration refers to the process by which a real advertising device sends a request to a server to establish its identity record in the system. This process typically involves authentication and information exchange between the device and the server, enabling the server to identify and manage the device. The registration mechanism ensures that only authorized and identified devices can access the system and are assigned a unique identifier, thus laying the foundation for the creation and management of virtual twin device digital models. Registration can be achieved in various ways, such as the device sending a registration request to the server through a pre-defined API interface, or binding its identity by scanning a QR code.

[0042] Uploading basic information such as device ID, device model, and screen resolution refers to the process where a real advertising device sends data inherent to it, used to identify and describe its own characteristics, to the server. The device ID is a unique identifier used to distinguish different real advertising devices in the system and bind them to the corresponding virtual twin device digital model. It can be the device's MAC address, serial number, or a unique string assigned by the system. The device model describes the device's hardware configuration and functional characteristics, helping the server understand the device's performance limitations and supported media formats, thus enabling more accurate simulation of its behavior in the virtual model. The screen resolution indicates the pixel size of the device's display screen. This information is crucial for reconstructing the virtual playback screen, ensuring that the virtual model can be rendered with the same display specifications as the real device, thereby guaranteeing consistency between the virtual and real playback screens. This basic information is the cornerstone of building an accurate virtual twin device digital model; it provides the fundamental parameters required for the virtual model to accurately reflect the physical characteristics and display capabilities of the real device.

[0043] Then, the server generates a unique virtual twin device instance based on the ID of each real advertising device and constructs a corresponding virtual twin device digital model to simulate the playback behavior and state of the corresponding real advertising device on the server side. Specifically, in this embodiment, the server can create a corresponding virtual twin device instance for each real device based on its unique identifier (ID) using a preset template or configuration script. The virtual twin device instance can be a data structure containing parameters for simulating the behavior and state of the real device. The digital model can be constructed by inputting static information uploaded by the real device (such as device ID, model, screen size, etc.) into the server's database and assigning a unique virtual identifier to each device. In this embodiment, the virtual twin device digital model is configured to receive data from the real device and update its internal state based on this data, thereby simulating the playback behavior and state of the real device on the server side.

[0044] The real advertising device receives advertising playback tasks, executes advertising playback, and periodically reports lightweight playback status data to the server. For example, after receiving an advertising playback task, the real advertising device will execute advertising playback according to the task instructions. During playback, the real advertising device can be configured to send a simple data packet containing the current playback status to the server at fixed time intervals (e.g., every minute). The data packet may only contain a unique identifier of the currently playing advertising content and the current playback timestamp. This data reporting mechanism uses a lightweight data transmission method to reduce network bandwidth consumption.

[0045] The server receives playback status data reported by the real device and, in conjunction with pre-set playback tasks and a timeline synchronization mechanism, simulates and reconstructs a virtual playback screen synchronized with the real device in real time on the server side. Specifically, after receiving the playback status data reported by the real device, the server compares it with the pre-set playback tasks. For example, the server can calculate the current playback progress that the real device should be at based on the list of advertising materials, playback order, and duration defined in the playback task, combined with the current playback ad ID and timestamp reported by the real device. Then, the server can extract the corresponding advertising content from the advertising material library stored locally on the server and, based on the calculated playback progress, render or generate a virtual playback screen synchronized with the real device on the server side. The process of generating a virtual playback screen synchronized with the real device in this embodiment does not involve the real device uploading actual images; instead, it is based on data and logic for reconstruction.

[0046] The process involves viewing the virtual playback screen of a digital model of a virtual twin device in real time via a client. For example, the client can connect to the server via a network and request to view the virtual playback screen of a specific virtual twin device's digital model. Upon receiving the request, the server will send the currently simulated and reconstructed virtual playback screen to the client in the form of images or short animation sequences. After receiving these screens, the client can display them locally, thereby achieving remote real-time monitoring of the playback status of the real advertising device.

[0047] The server performs cross-validation based on multiple data sources to verify the consistency between the virtual playback screen and the actual playback screen of the real advertising device. For example, the server can utilize its existing data sources for cross-validation. The server can compare the playback status data reported by the real device (such as the current playing ad ID and playback progress) with the expected playback status calculated by the server based on the playback task. If the two are inconsistent, it may indicate an anomaly. This verification method mainly relies on logical comparison to determine whether the virtual playback screen is consistent with the actual playback screen of the real advertising device.

[0048] The server continuously monitors and interprets the virtual playback screen and playback status data for anomalies. If anomalies are detected, the server triggers an alarm and notifies the corresponding personnel's terminal. Specifically, the server can be configured with a scheduled task to periodically check the virtual playback screen and received playback status data. For example, the server can check whether the playback status data is reported on time or whether the virtual playback screen matches the preset playback task logic. When simple anomalies such as missing data or inconsistent playback logic are detected, the server can trigger an alarm event and send the alarm information to the corresponding management personnel's terminal via a preset notification method (e.g., email or SMS).

[0049] The technical solution of the present invention will be explained in more detail below through a more specific example: For example, a commercial center may have a large number of physical advertising devices deployed, such as digital signage located on different floors. Traditional monitoring methods struggle to monitor the actual playback status of these devices in real time, and anomaly detection is often delayed.

[0050] In this embodiment of the method, firstly, when these real advertising devices are put into use for the first time, the server generates a unique virtual twin device instance based on the unique ID of each real advertising device, such as "AD_XXXX_001" or "AD_XXXX_002". Simultaneously, the server constructs a virtual twin device digital model for each instance. The virtual twin device digital model in this embodiment includes static information of the real device (such as device model, screen resolution, installation location, etc.) and dynamic playback logic modules. For example, for "AD_XXXX_001", its digital model is configured to simulate the device's behavior of receiving advertising tasks, performing playback, and reporting status.

[0051] Next, the server sends an ad playback task to the real ad device with ID AD_XXXX_001. This task specifies the order and duration of playing ad creatives A, B, and C within a specific time period. Upon receiving the task, the real ad device AD_XXXX_001 begins ad playback. During playback, it periodically, for example, every 30 seconds, reports lightweight playback status data to the server. This data includes only the ID of the currently playing ad creative (e.g., "AD_X_ID"), the current playback progress (e.g., "15 seconds played"), and a timestamp.

[0052] After receiving the playback status data reported by the real advertising device AD_XXXX_001, the server will immediately combine the pre-set playback task and timeline synchronization mechanism to simulate and reconstruct the virtual playback screen corresponding to the playback screen of the real advertising device AD_XXXX_001 in real time on the server side. For example, when the server receives the status data that the real advertising device AD_XXXX_001 is playing "AD_X_ID" and has been playing for 15 seconds, the server will retrieve the material of advertisement X from its material library and render the screen of advertisement X at the 15th second of playback as the virtual playback screen of the real advertising device AD_XXXX_001.

[0053] At this point, administrators can view the virtual playback screen of the virtual twin device digital model of the real advertising device AD_XXXX_001 in real time through the client. The client displays the screen of advertisement A reconstructed on the server side up to the 15th second, allowing administrators to remotely understand the device's playback status without being physically present.

[0054] To ensure the accuracy of the virtual playback, the server performs cross-validation based on multiple data sources. For example, in addition to comparing the reported playback status data with preset tasks, the server can also verify the data using playback log data. If the playback status data shows that advertisement A is playing, but the playback log records that advertisement D is playing, the system will identify the potential inconsistency.

[0055] Finally, the server continuously monitors and interprets the virtual playback screen and playback status data for any abnormalities. For example, if the server does not receive playback status data from the real advertising device AD_XXXX_001 for an extended period, or if the virtual playback screen displays a black screen (based on logical judgment rather than actual screen recognition), or if the virtual playback screen is significantly inconsistent with the playback task, the server will immediately determine it to be an abnormal situation. Once an abnormality is detected, the system in this embodiment of the invention will trigger an alarm and send the alarm information to the terminal of the maintenance personnel responsible for the area via SMS or application notification, for example, notifying "Maintenance Personnel A" that the real advertising device AD_XXXX_001 may have a playback interruption problem.

[0056] Through the above method, this embodiment can effectively solve the problems of traditional monitoring solutions, such as the inability to confirm the actual playback situation, lack of real-time visualization monitoring capabilities, low efficiency of screenshot solutions, high cost of video stream solutions, and lag in anomaly detection. It provides a low-bandwidth, highly reliable, and real-time remote monitoring technology solution for advertising playback supervision.

[0057] Based on the above examples, the advertising playback monitoring method based on virtual twin devices proposed in this embodiment has the following advantages compared to the prior art: First, addressing the issues of existing technologies' inability to verify actual playback status and lack of real-time visualization monitoring capabilities, this embodiment constructs a digital model of a virtual twin device and simulates and reconstructs the virtual playback screen in real time on the server side. This allows administrators to remotely view the device's playback screen in real time via a client. This avoids the limitations of traditional solutions that rely solely on logs or manual inspections, providing intuitive and real-time monitoring capabilities.

[0058] Secondly, addressing the issues of low efficiency in existing screenshot solutions and high cost in video streaming solutions, this embodiment uses real advertising devices to periodically report lightweight playback status data, instead of transmitting high-bandwidth screenshots or video streams. The server reconstructs the virtual playback screen based on this lightweight data and preset tasks, significantly reducing network bandwidth usage and terminal hardware requirements. It also avoids the drawbacks of screenshot solutions, such as high latency and inability to reflect continuous playback, achieving low-cost, high-efficiency real-time monitoring.

[0059] Furthermore, addressing the issue of delayed anomaly detection in existing technologies, this embodiment continuously monitors and interprets virtual playback screen and playback status data for anomalies via a server, triggering alarm notifications promptly. This proactive, real-time anomaly detection mechanism can quickly identify issues such as black screens, stuttering, playback errors, and playback not on schedule, thereby significantly shortening response time and improving operational efficiency.

[0060] In summary, this embodiment, by introducing virtual twin technology, reconstructs the playback status and screen of the real device on the server side in a data-driven manner, achieving low-bandwidth, highly reliable, and real-time remote monitoring of advertising playback. It effectively solves many pain points in existing technologies and provides an innovative and efficient solution for the operation and management of the digital advertising industry.

[0061] In a further embodiment of this application, a pre-set virtual twin device digital model is proposed, including static device information of the real device, and a dynamic playback task mapping module, timeline synchronization module, playback status model, rendering display module, anomaly detection module, and history record module; wherein, the static device information includes: device ID, model, screen orientation, resolution, network status, and client information.

[0062] The static device information refers to the relatively fixed attribute data of the real advertising device under normal operating conditions, such as device ID, model, screen orientation, resolution, as well as the device's physical characteristics and configuration parameters. This information provides the basic physical and configuration context for the virtual twin device digital model, ensuring that the virtual model can accurately simulate the hardware environment and initial settings of the real device. For example, the static device information can be automatically uploaded to the server and stored by the real device during its initial registration, or it can be initialized through manual configuration or import from the device management system.

[0063] The playback task mapping module is responsible for parsing and converting the ad playback task instructions issued by the server into playback logic executable by the virtual twin device. This module ensures that the virtual twin device can execute ad playback tasks according to the same rules and order as the real device, forming the foundation for virtual playback screen reconstruction. For example, the playback task mapping module may include a parser to interpret task data (such as ad creative list, playback order, playback time period, single ad duration, loop strategy, insertion rules, etc.) and generate a virtual playback sequence; or it may configure a rule engine to associate task data with preset playback strategies, guiding the virtual playback process.

[0064] The timeline synchronization module ensures that the playback progress of the virtual twin device is consistent with the timeline of the real device. This module is crucial for synchronizing the virtual playback with the real device, guaranteeing the real-time performance and accuracy of the virtual simulation. For example, the timeline synchronization module can receive the current timestamp and playback progress reported by the real device and adjust the virtual playback timeline accordingly; or it can use NTP (Network Time Protocol) or other time synchronization protocols to ensure that the server's time base is consistent with the real device, and make fine adjustments based on the playback progress data.

[0065] The playback state model is used to describe and manage various states of the virtual twin device during playback, such as playing, paused, buffering, and error. The playback state model updates the virtual model's state based on the state data reported by the real device and provides a basis for anomaly detection. For example, the playback state model can be built based on a finite state machine (FSM), defining various playback states and the transition rules between them; or it can use a data structure to store the current playback content ID, playback progress, volume, network status, etc., and provide interfaces for other modules to query and update.

[0066] The rendering and display module is responsible for generating a visual virtual playback screen based on the playback task and status of the virtual twin device. The module presents the virtual simulation playback behavior as images for the client to view and verify for consistency. For example, the rendering and display module can use graphics rendering libraries (such as OpenGL, DirectX) or web front-end technologies (such as HTML5Canvas) to generate the screen in real time based on the source material and playback progress; or it can use pre-rendering or caching techniques to pre-generate keyframes or short video clips based on the playback task and display them on demand during playback.

[0067] The anomaly detection module continuously monitors the virtual playback screen and playback status data, identifying abnormal situations that deviate from expectations or indicate potential problems. This module is crucial for automatic alarms and notifications, improving the efficiency and timeliness of monitoring. For example, the anomaly detection module can be configured with thresholds and rules to detect prolonged periods without status reporting, playback stagnation, or discrepancies between the displayed content and expectations; or it can employ machine learning models to learn normal playback patterns and identify abnormal behaviors or data patterns.

[0068] The history recording module stores historical information such as playback tasks, status data, abnormal events, and virtual playback screens of the virtual twin device. This module provides a basis for troubleshooting, data analysis, playback effect evaluation, and auditing. For example, the history recording module can store all relevant data in a relational database and create indexes for quick retrieval; or it can use a distributed file system or NoSQL database to store large amounts of historical data and process it in conjunction with log analysis tools. The static device information includes device ID, model, screen orientation, resolution, network status, and associated customer information. This information collectively constitutes the unique identifier and basic configuration of the real device, forming the basis for the accurate simulation of the real device by the virtual twin device digital model. For example, the device ID uniquely identifies the device, the model and resolution determine the size and adaptability of the playback screen, the screen orientation affects the content layout, the network status reflects the device's connectivity, and the associated customer information is used for management and access control. This information is typically entered during device registration or configuration and stored as metadata in the virtual twin device digital model.

[0069] This application's solution involves meticulously constructing a digital model of a virtual twin device, transforming it from a simple simulator into a digital surrogate with complete device attributes and dynamic operational capabilities. Specifically, when constructing the virtual twin device's digital model, the server first incorporates the static information of the real device. This provides an accurate physical and configuration foundation for the virtual model, ensuring a high degree of consistency between the virtual environment and the real device. Based on this, a dynamic playback task mapping module accurately converts the advertising playback task instructions issued by the server into a playback sequence executable by the virtual device. Meanwhile, a timeline synchronization module compares and adjusts the data with the timestamps and playback progress reported by the real device, ensuring complete synchronization of the virtual and real playback time dimensions. A playback status model reflects the virtual device's operational status in real time, providing a basis for subsequent rendering and detection. The rendering and display module uses this information, combined with advertising materials, to generate a virtual playback screen highly consistent with the real device's playback. Once the virtual playback screen and playback status data are generated, an anomaly detection module continuously monitors this data. Upon detecting anomalies such as black screens, stuttering, or playback errors, it can promptly identify and trigger alarms. Meanwhile, the history module archives all playback tasks, status data, abnormal events, and virtual playback screen information, providing comprehensive data support for subsequent tracing, analysis, and optimization. Through this structured digital model, the server can more accurately and comprehensively simulate and monitor the operation of real advertising devices, significantly improving the practicality and reliability of virtual twin devices in advertising playback monitoring.

[0070] The following is a concrete example. In one implementation, when a new digital signage device (a real advertising device) is first put into use, it registers with the server and uploads its device ID (e.g., DS-001), model (e.g., a brand's 65-inch 4K monitor), screen orientation (e.g., landscape), resolution (e.g., 3840x2160), current network status (e.g., Wi-Fi connection), and customer information (e.g., a large shopping mall). This data is received and stored by the server, constituting the static device information of the digital twin device's digital model. Subsequently, the server issues an advertising playback task, which includes a list of advertising materials (e.g., image A, video B), playback order, playback time period, and loop strategy. The playback task mapping module in the virtual twin device's digital model parses this task data and generates a virtual playback sequence. During playback, the playback status data periodically reported by the real device (e.g., currently playing image A, progress 50%, current timestamp) is received by the timeline synchronization module and used to adjust the playback progress of the virtual twin device's digital model to maintain precise synchronization with the real device. The playback status model updates the virtual device's current state based on this data. The rendering and display module then generates a virtual playback screen simulating the real device playing image A in real time, based on the virtual twin device's digital model's static device information, playback task mapping results, and synchronized playback status. If the real device suddenly loses network access during playback, resulting in no status data reporting for an extended period, the anomaly detection module immediately identifies this network offline anomaly and triggers an alarm. All these playback tasks, status updates, virtual screen generation, and anomaly events are recorded in detail by the history module for subsequent querying and analysis.

[0071] Through the aforementioned technical solutions, this application significantly improves the accuracy and comprehensiveness of virtual twin device simulation of real advertising equipment by constructing a refined and modular digital model of the virtual twin device. The introduction of static device information ensures that the virtual model accurately reflects the physical characteristics and configuration of the real device, laying the foundation for high-fidelity simulation. The inclusion of dynamic modules (such as the playback task mapping module, timeline synchronization module, playback state model, and rendering display module) enables the virtual twin device to simulate the dynamic playback behavior and state of the real device in real time and accurately, solving the problem of difficulty in comprehensively grasping the details of device operation in traditional monitoring. In particular, the integration of the anomaly detection module and the historical record module allows the system to proactively detect and alert on playback anomalies, providing detailed data support for subsequent troubleshooting and effect evaluation, thereby greatly improving the efficiency and reliability of advertising playback monitoring and effectively avoiding advertising playback accidents and losses caused by opaque or delayed information.

[0072] In another embodiment of the present invention, a mechanism for synchronizing task data, status data, and timeline data in the digital model of the virtual twin device is further proposed. Through lightweight task data and periodically reported status data, combined with timeline synchronization, the playback screen of the real device is reconstructed on the server side. Here, task data refers to the advertising playback instructions issued by the server to the real device. The advertising playback details specify the content and strategy of the advertisements. For example, it may include an ad creative list, specifying all the advertising content to be played; a playback order, specifying the playback sequence of the advertisements; a playback time period, limiting the effective playback period of the advertisements; a single ad duration, specifying the playback duration of each advertisement; a loop strategy, defining the rules for repeated playback of advertisements; and insertion rules, used to handle the insertion of urgent or higher-priority advertisements.

[0073] The task data is distributed in the form of structured data, such as JSON or XML, and transmitted to the actual advertising device via HTTP requests or message queues. The status data refers to the data periodically reported by the actual device to the server during playback. This data reflects the current playback status and operating condition of the actual device. For example, it may include the current advertisement content ID, identifying the specific advertisement currently playing; the current playback progress, indicating the percentage or time point of the current advertisement played; the current timestamp, recording the device's local time when the status data was reported; the playback status, such as playing, paused, or stopped; volume, screen, and network status, reflecting the device's hardware and network environment; and optional screenshots or short video clips to provide more intuitive evidence of the playback.

[0074] The status data is reported in the form of lightweight structured data packets via MQTT protocol or HTTP POST requests at a preset frequency (e.g., every few seconds) or when critical events occur (e.g., during ad switching). The timeline synchronization mechanism is used to ensure that the playback progress of the virtual twin device on the server side is consistent with the timeline of the real device; it is the key to achieving accurate reconstruction of the virtual playback screen. It can be achieved in various ways, such as periodically calibrating the time between the server and the real device, such as using Network Time Protocol (NTP) to synchronize the system clock; or, including a reference timestamp in the task data, which both the real device and the server use as a reference point to calculate the playback progress; or, the server calculates the precise playback status of the real device at a specific moment based on the current timestamp and playback progress reported by the real device, combined with the task data.

[0075] This application's solution introduces task data, status data, and a timeline synchronization mechanism, enabling the server to efficiently and accurately reconstruct the playback screen of the real device. Specifically, the server first sends task data, containing detailed information such as the list of advertising materials, playback order, playback time period, single ad duration, loop strategy, and insertion rules, to the real advertising device. This task data constitutes a blueprint or instruction set for the real device's playback behavior. During the execution of the ad playback task, the real advertising device periodically reports lightweight status data to the server. This status data includes key information such as the current ad content ID, current playback progress, current timestamp, playback status, and device operating status, reflecting the real-time progress of the real device's playback.

[0076] After receiving this status data, the server does not directly receive the complete video stream or a large amount of image data. Instead, it combines pre-delivered task data with time information calibrated through a timeline synchronization mechanism. On the server side, utilizing its rendering capabilities, it precisely simulates and renders a virtual playback screen that is completely identical to the current playback screen on the real device, based on the advertisement content and playback rules defined in the task data and the current playback point indicated in the status data. For example, the server knows which advertisement the real device is playing (by the current advertisement content ID), the current playback point (by the current playback progress and timestamp), and the complete playback logic of the currently playing advertisement (by the task data). In this way, the server does not need to transmit large amounts of video data; it can efficiently reconstruct the playback screen of the real device on the server side using only lightweight task data and periodically reported status data, combined with timeline synchronization. This solves the technical challenge of accurately reconstructing dynamic images using only lightweight data.

[0077] The following is a concrete example. When a server needs to monitor the playback of advertisements on a real advertising device 100, the server first sends task data to the device. This task data can be a JSON file containing a list of advertising materials, such as ["ad_A.mp4", "ad_B.mp4"]; the playback order is [0,1]; the playback time period is "08:00-22:00"; the duration of a single advertisement is "30s"; the loop strategy is "sequential loop"; and the insertion rule is "none". After receiving this task data, the real advertising device 100 begins playing advertisements according to the instructions. During playback, the device reports status data to the server every 5 seconds. For example, when the real advertising device 100 is playing `ad_A.mp4` at the 10th second, its reported status data might include "current_ad_id":"ad_A.mp4", "current_progress":"10s", "timestamp":"2023-10-27T10:00:10Z", "playback_status":"playing", "volume":"50", and "network_status":"online". After receiving this status data, the server uses its internally stored `ad_A.mp4` material, combined with the playback rules defined in the task data, and a timeline calibrated via the NTP protocol, to accurately calculate the specific screen content of `ad_A.mp4` at the 10th second of `2023-10-27T10:00:10Z`, and renders it in the virtual twin device's digital model. If the real advertising device 100 also reports optional screenshots or short video clips, the server can use them as auxiliary verification information to further ensure the accuracy of the virtual playback.

[0078] Through the above technical solution, this application effectively solves the technical challenge of accurately reconstructing dynamic playback images on the server side using only lightweight data. By combining detailed playback instructions (task data) with real-time playback feedback (status data) and supplementing it with a precise timeline synchronization mechanism, the server can accurately reconstruct a virtual playback image in the digital model of the virtual twin device that is completely synchronized with the real advertising device without receiving high-bandwidth video streams. This significantly reduces the burden on network transmission, improves data transmission efficiency, and ensures the high precision and real-time performance of the virtual playback image. This mechanism provides a reliable and efficient data foundation for subsequent cross-validation and anomaly detection, making the monitoring process of advertising playback more accurate and economical, thereby improving the performance and reliability of the entire monitoring system.

[0079] In some other embodiments, this application further proposes a step of cross-validation by the server based on multiple data sources to verify whether the playback screen is consistent with the playback result of a real advertising device. The multi-source data includes: status data, playback logs, screenshot data, and short video clips.

[0080] The status data refers to the data periodically reported to the server by the actual advertising device during playback, reflecting the device's current operating status and the content being played. This status data is fundamental for verifying playback consistency and provides real-time, dynamic device playback information. For example, the actual advertising device can periodically send information such as the currently playing ad ID, playback progress percentage, current timestamp, device operating status (e.g., normal, paused, error), volume settings, screen brightness, and network connection status to the server in a structured data format (e.g., JSON). Alternatively, the device can use a lightweight communication protocol (e.g., MQTT) to push this status information to the server to ensure real-time and efficient data transmission.

[0081] The playback logs refer to the detailed operations and event sequences recorded by the actual advertising device when performing advertising playback tasks. These logs are typically stored locally on the device and uploaded to a server periodically. The purpose of the playback logs is to provide playback history and event details, facilitating the tracing of playback behavior and helping to determine whether playback was executed according to the preset plan. For example, the device can record the corresponding timestamp, ad ID, event type, and playback duration each time an ad starts playing, ends playing, is skipped, or an error occurs, and then package and upload this information. Alternatively, a centralized log management system can be used to collect, store, and analyze the log data generated by the device for subsequent querying and correlation.

[0082] The screenshot data refers to the screen image captured by the real advertising device at a specific time point or when a specific event is triggered. Screenshot data provides intuitive visual evidence, directly reflecting the actual playback screen of the real device at a certain moment, and is a key basis for visual comparison with the virtual playback screen. For example, the device can be configured to take screenshots at the beginning, end, or random time points of each advertisement playback and upload the compressed image data to the server. To improve efficiency, hardware-accelerated screenshots can be performed using the device's built-in graphics processing unit (GPU), and encoded using formats such as JPEG or PNG to balance image quality and data volume.

[0083] The short video clips refer to screen playback video segments recorded by real advertising devices within a specific time period. These short video clips provide dynamic visual evidence, capturing animations, transitions, and potential issues such as stuttering and screen tearing during playback, offering richer information than a single screenshot. For example, when an anomaly is detected or a preset strategy is followed, the device can record a video clip of a specific length (e.g., 5-10 seconds), encode it (e.g., H.264), and then upload it to a server. Alternatively, the device can push a real-time video stream to a server under specific circumstances for recording or real-time analysis.

[0084] This application's solution integrates multiple data sources, including status data, playback logs, screenshot data, and short video clips, enabling the server to perform comprehensive and multi-dimensional cross-verification of the playback behavior of real advertising devices. Specifically, the status data provides real-time, structured playback progress and device status, which can be compared with the playback status model within the virtual twin device's digital model to ensure the synchronization of playback logic and timeline. The playback log provides a more detailed event sequence for verifying the actual playback order, duration, and any abnormal events, and compares it with the playback tasks issued by the server. The screenshot data and short video clips provide the most intuitive visual evidence; the server can compare screenshots uploaded from real devices with the virtual playback screen generated by the virtual twin device's digital model to verify the consistency of screen content, layout, and text. Short video clips further supplement the verification of dynamic playback effects, detecting issues such as stuttering, screen tearing, and frame skipping, which cannot be provided by a single screenshot. Through this multi-dimensional data fusion and cross-validation mechanism, the server can conduct a comprehensive and detailed check on the playback behavior of real devices from multiple levels, including logical state, event history, static visual and dynamic visual, thereby significantly improving the accuracy and reliability of playback consistency judgment.

[0085] In another embodiment of this application, the server is further proposed to continuously detect and judge whether the virtual playback screen and playback status data are abnormal. When the server detects black screen, stuttering, playback error, playback not on schedule, network offline, playback interruption, or long-term absence of status reporting abnormalities, it controls and triggers an alarm and notifies the corresponding personnel terminal.

[0086] The term "black screen" refers to a completely black virtual playback screen or a real device screen, with no content output. This can be detected by analyzing the pixel data of the virtual playback screen to determine if all pixel values ​​are below a preset brightness threshold (e.g., setting the RGB brightness threshold to 10); or by analyzing screenshots periodically uploaded by the real device to identify if the screen is pure black. The term "stuttering" refers to unsmooth playback, including pauses, frame skips, or delayed screen updates. This can be detected by analyzing the playback progress and timestamp sequence in the playback status data to determine if there are unexpected pauses or jumps; or by real-time monitoring of the virtual playback screen's frame rate, identifying stuttering when the frame rate falls below a normal playback threshold.

[0087] Playback errors refer to the failure of advertising content to load, display, or play correctly, such as displaying error codes, damaged materials, or screen tearing. Detection can be achieved by comparing the hash values ​​or key feature points of the virtual playback screen with the expected playback material to identify inconsistencies; or by parsing error information in the playback log to identify the specific type of playback failure. Playback not on schedule refers to advertising content not playing according to the preset playback task and timeline, such as playing the wrong advertisement, skipping the scheduled advertisement, or playing in the wrong order. Detection can be achieved by comparing the content of the virtual playback screen or the current advertisement content ID with the playback task schedule table stored on the server in real time to determine if there are any discrepancies.

[0088] The term "network offline" refers to an interruption in the network connection between the actual advertising device and the server, resulting in the inability to transmit data normally. This can be detected by the server failing to receive periodically reported status data from the actual device within a preset time; or by using a network heartbeat mechanism, where failure to respond to a heartbeat timeout indicates a network offline status.

[0089] Playback interruption refers to the advertisement stopping before completion, such as an advertisement suddenly stopping halfway through. Detection can be achieved by playback status data indicating that playback has stopped but the preset playback duration or playlist end has not been reached; or by the virtual playback screen suddenly stopping updating and not switching to the next normally playing content. Prolonged absence of status reporting refers to the server not receiving any playback status data reported by the real device within a preset time period, which may indicate device malfunction or network anomaly. Detection can be achieved by the server maintaining a last status reporting timestamp for each real device and setting a timer. If the difference between the current time and the last reporting timestamp exceeds a preset threshold, this anomaly is triggered.

[0090] This application's solution continuously and meticulously monitors and interprets virtual playback screens and playback status data on the server side, enabling the identification of various specific anomalies. Upon receiving playback status data reported by the real device, the server, in conjunction with pre-set playback tasks and timeline synchronization mechanisms, simulates and reconstructs a virtual playback screen synchronized with the real device in real time. Furthermore, the server does not merely determine the presence of general anomalies but analyzes the data in a more targeted manner. For example, analyzing the pixel values ​​of the virtual playback screen can determine if there are black screens; comparing the playback progress and timestamp sequences in the playback status data can identify stuttering or playback interruptions; comparing the current playback content with the preset schedule can reveal issues with playback not following the schedule; and monitoring the timeliness of status data reporting can detect network offline or prolonged periods without status reporting. This meticulous anomaly identification allows the server to accurately extract the essence of the problem from massive amounts of monitoring data, providing a clear basis for subsequent alarms and handling. In this way, the solution concretizes abstract anomalies, making the monitoring of advertising playback more precise and efficient.

[0091] As a specific implementation method, the above technical means can be implemented with reference to the following example. When the server continuously detects and interprets the virtual playback screen and playback status data, it can make specific judgments on two abnormal situations: black screen and prolonged stateless reporting. For black screen abnormality, after the server reconstructs the virtual playback screen in real time, it can periodically (e.g., every 5 seconds) perform image analysis on the virtual playback screen. The analysis module can calculate the average brightness value of all pixels in the screen, or count the proportion of pixels with brightness values ​​lower than a certain preset threshold (e.g., all RGB values ​​are less than 10). If the average brightness value is lower than a certain extremely low threshold or the proportion of low-brightness pixels exceeds 95% within multiple consecutive sampling periods, the system determines it to be a black screen abnormality. For prolonged stateless reporting abnormality, the server maintains a last heartbeat time record for each registered real advertising device. The real advertising device is configured to report lightweight playback status data to the server every 30 seconds, which includes a timestamp. After receiving the data, the server updates the last heartbeat time of the corresponding device. At the same time, the server runs a background monitoring service that iterates through the last heartbeat times of all devices every minute. If the difference between the current time and the last heartbeat time of a device exceeds 90 seconds (considering network latency and processing time), the system determines that the device has experienced a prolonged stateless reporting anomaly. Once any of the above anomalies is detected, the server will trigger an alarm mechanism and notify the corresponding operations and maintenance personnel's terminal via a preset notification method (such as SMS, email, or application push).

[0092] Through the aforementioned technical solution, the server can refine general anomalies into specific exceptions such as black screen, stuttering, playback errors, playback not on schedule, network offline, playback interruption, and prolonged periods without status reporting. This refined anomaly identification capability allows maintenance personnel to immediately understand the specific nature and possible causes of the problem upon receiving an alarm, thereby significantly shortening the time for troubleshooting and location. Compared to solutions that only provide general anomaly alarms, this application provides more instructive information, effectively avoiding blind troubleshooting due to ambiguous information, significantly improving the efficiency and accuracy of ad playback monitoring, ensuring normal and high-quality playback of ad content, and reducing operational risks and maintenance costs.

[0093] In some other implementations, the steps of simulating and reconstructing a virtual playback screen synchronized with a real device in real time on the server side are further proposed, including: introducing artificial intelligence technology to perform content recognition and analysis on screenshots of the virtual playback screen, and verifying the consistency between the virtual playback screen and the real playback screen of the real advertising device.

[0094] The introduction of artificial intelligence technology refers to integrating advanced algorithm models such as machine learning and deep learning into the advertising playback monitoring system to achieve intelligent processing of image or video content. For example, convolutional neural networks (CNNs) can be used for image classification, object detection, or semantic segmentation, or recurrent neural networks (RNNs) can be used to process video sequences. These technologies endow the system with image understanding capabilities beyond traditional rule matching, enabling it to identify specific elements, text information, layout structures, etc., in the image.

[0095] The content recognition and analysis of screenshots of the virtual playback screen refers to the system periodically or triggered by specific events capturing static images of the virtual playback screen and using a pre-trained artificial intelligence model to perform in-depth analysis of these images. Content recognition can include identifying advertising materials, brand logos, text content, product images, etc., appearing in the screen. Content analysis further evaluates whether these identified elements meet the preset playback task requirements, such as checking whether the advertising materials are complete, whether the text is correct, and whether the layout conforms to specifications. For example, optical character recognition (OCR) technology can be used to extract text information from the screen and compare it with the text content in the playback task; or image similarity algorithms can be used to compare the screenshots of the virtual screen with standard advertising materials at the pixel level or feature level. Verifying the consistency between the virtual playback screen and the real playback screen of the real advertising device refers to using artificial intelligence technology to intelligently compare the recognition and analysis results of the virtual playback screen screenshots with the content that the real advertising device should play, thereby judging the degree of matching between the two in terms of visual content. This consistency verification is not limited to data-level comparison but goes deeper into the visual presentation level, ensuring that the virtual screen is visually highly consistent with the real screen. For example, a threshold can be set, and when the matching degree of key visual elements (such as brand logo, product image, and advertising slogan) between the virtual image and the real image is lower than the threshold, it is considered that there is an inconsistency.

[0096] This application's solution involves simulating and reconstructing a virtual playback screen synchronized with the real device in real time on the server side. Then, artificial intelligence technology is introduced to perform content recognition and analysis on screenshots of the virtual playback screen to verify its consistency with the real playback screen of the actual advertising device. Specifically, after the server generates a unique virtual twin device instance based on the ID of each real advertising device and constructs a corresponding digital model of the virtual twin device, receives playback status data reported by the real device, and simulates and reconstructs the virtual playback screen in real time using pre-set playback tasks and a timeline synchronization mechanism, the system of this embodiment further takes screenshots of these virtual playback screens. These screenshots are then input into a pre-trained artificial intelligence model. The artificial intelligence model can perform deep content recognition on the screenshots, such as identifying key visual elements like advertising materials, text, and brand logos, and performing content analysis to evaluate the completeness, accuracy, and layout rationality of these elements.

[0097] In this way, the system of this invention can intelligently compare the content features of the virtual playback screen obtained from artificial intelligence analysis with the real advertising materials and layout information contained in the preset playback task. This AI-based visual content comparison makes up for the shortcomings of relying solely on non-visual data such as playback status data and playback logs for cross-validation, making the verification of the consistency between the virtual playback screen and the real playback screen of the real advertising device more refined and accurate. It can detect visual content errors that are difficult to detect by traditional methods, such as partial damage to advertising materials, typos, missing or distorted brand logos, etc., thereby significantly improving the reliability and accuracy of advertising playback monitoring.

[0098] The following is a concrete example. After simulating and reconstructing a virtual playback screen synchronized with the real device in real time on the server side, the system in this embodiment can take a screenshot of the virtual playback screen every few seconds. These screenshots are then fed into a deep learning-based computer vision model for processing. The deep learning-based computer vision model in this embodiment can be a multi-task model, including an object detection module for identifying key visual elements such as the advertising subject, product, and brand logo in the screen, an optical character recognition (OCR) module for extracting all text information in the screen, and an image classification module for determining the overall theme or style of the advertisement. For example, if the playback task requires playing a specific advertisement of a certain brand, the deep learning-based computer vision model will detect whether the brand's logo exists in the screen and verify the accuracy of the logo; at the same time, the OCR module will extract the advertising text and compare it with the preset advertising text in the playback task to check for typos or omissions. In addition, an image similarity algorithm can be used to compare the feature vectors of the screenshots of the virtual screen with the reference image of the standard advertising material and calculate their similarity score. When a logo mismatch, text error, or similarity score below a preset threshold is detected, the system of this embodiment of the invention can determine that there is an inconsistency between the virtual playback screen and the real playback screen of the real advertising device.

[0099] Through the aforementioned technical solution, this application, based on real-time simulation and reconstruction of the virtual playback screen on the server side, further introduces artificial intelligence technology to perform content recognition and analysis on screenshots of the virtual playback screen. This enables a deeper and more intelligent verification of the consistency between the virtual playback screen and the real playback screen of the actual advertising device. This method can effectively identify visual content errors that are difficult to detect through traditional data comparison, such as partial damage to advertising materials, text display errors, abnormal layout, or inconsistencies in brand logos. This significantly improves the accuracy and reliability of advertising playback monitoring, ensures the correct presentation of advertising content, avoids poor advertising results or damage to brand image due to visual content errors, and provides advertisers and operators with a more comprehensive and reliable playback quality assurance.

[0100] The present invention will be further described in detail below through another specific application embodiment: Specifically, such as Figure 2 As shown in the figure, the advertising playback monitoring method based on a virtual twin device in this specific application embodiment includes the following steps: Step S1: Real advertising devices are registered and launched. After the device is started, it sends device information to the server. The server completes the registration, suggests device profiles, and proceeds to S2. The device information includes device ID, device model, screen orientation, resolution, network status, and customer information.

[0101] Step S2: Create a virtual twin device. The server generates a unique virtual device instance based on the device ID and constructs a virtual twin device model; then proceed to step S3. In this step, a corresponding digital model is created on the server side for each real advertising device. This virtual twin device model not only includes the static information of the real device (such as device ID, model, screen orientation, resolution, etc.), but also dynamic modules such as playback task mapping, timeline synchronization, playback status model, rendering and display, anomaly detection, and history recording. Its function is to act as a digital substitute for the real device, simulating the playback behavior and status of the real device on the server side.

[0102] Step S3: Issue an advertisement playback task. The server issues a playback task to the device, and the device executes the playback task after receiving it; then proceed to step S4. The ad playback task includes a list of ad creatives, playback order, playback time period, duration of a single ad, loop strategy, and insertion rules.

[0103] Step S4: The device plays and periodically reports status data; During playback, the device reports playback status data to the server according to a preset period, and then proceeds to step S5. The playback status data includes the current ad content ID, current playback progress, current time period, playback status (such as playing / paused / abnormal), volume / screen / network status, and optional screenshots / short video clips.

[0104] Step S5: Server-side reconstruction of the playback screen (core); After receiving the status data reported by the real advertising device, the server reconstructs the playback screen synchronized with the real device on the server side, combining the playback task and timeline; then proceed to step S6; In this step, the server receives playback status data reported by the real device and combines it with a pre-set playback task and timeline synchronization mechanism. That is, based on the playback task, timeline, current progress, and status information, the server simulates and reconstructs a virtual playback screen synchronized with the real device in real time.

[0105] Step S6: Remote thinning demonstration; the client accesses the server via web / mobile terminal to view the virtual playback screen of the virtual twin device in real time.

[0106] This invention can realize single-device monitoring, multi-device simultaneous monitoring, group monitoring, historical playback, full-screen viewing, and playback statistical reports.

[0107] S7. Multi-source verification mechanism (anti-bypass key point): The server uses multiple data cross-verification to determine the authenticity and consistency of the playback result; In this embodiment of the invention, to ensure the authenticity and consistency of the playback results, the server performs cross-validation based on multiple data sources, including verification status data, playback logs, screenshot data, short video clips, and virtual playback results. This helps prevent device cheating or data anomalies.

[0108] S8. Anomaly Detection and Alarm: The server detects and judges anomalies, triggers alarms, and notifies the relevant personnel to handle them. In this embodiment of the invention, the server continuously monitors the virtual playback screen and status data to determine if any abnormalities exist. If not, playback proceeds normally and continuous monitoring continues. If a black screen, stuttering, playback errors, playback not on schedule, network offline, playback interruption, or prolonged lack of status reporting is detected, an abnormal situation is determined, and an alarm notification is immediately triggered, such as via SMS / email / APP push / WeChat / DingTalk, etc. Relevant personnel are notified to perform manual processing / remote troubleshooting, problem repair / device restart / task replacement, etc.

[0109] As can be seen from the above, this invention addresses the problem of low efficiency in existing screenshot solutions. Traditional screenshot solutions suffer from drawbacks such as high upload latency, inability to reflect continuous playback, inability to detect stuttering, and insufficient realism due to low image frequency. The virtual twin device of this invention can continuously simulate the playback process, providing a continuous virtual playback screen. Users can view the advertisement playback status as if watching a video, rather than relying on static, low-frequency screenshots, significantly improving the efficiency and realism of monitoring.

[0110] To address the issue of high costs associated with video streaming solutions, directly uploading video streams consumes significant network bandwidth, places high demands on terminal hardware, and incurs substantial costs for concurrent monitoring across multiple devices. This invention reconstructs the playback content on the server side using a mechanism that integrates task data, status data, and timeline synchronization. This avoids directly transmitting the video stream, transmitting only a small amount of task and status data, thereby significantly reducing bandwidth costs and lowering the requirements for terminal hardware, making large-scale device deployment possible.

[0111] To address the issue of delayed anomaly detection, traditional systems struggle to promptly identify issues such as black screens, stuttering, and incorrect content playback. This solution's virtual twin device incorporates an anomaly detection module. The server continuously monitors the virtual playback screen and received status data. Upon detecting any anomalies (such as black screens, stuttering, playback errors, playback not on schedule, network offline, playback interruption, or prolonged lack of status reporting), an alarm is immediately triggered, and relevant personnel are notified. This enables timely anomaly detection and handling, improving ad fulfillment rates.

[0112] In addition, by using a multi-source verification mechanism, which combines status data, playback logs, screenshot data, and other data for cross-verification, the authenticity and credibility of the playback results are further improved, enhancing customers' trust in advertising.

[0113] Exemplary device like Figure 3 As shown, this embodiment of the invention provides an advertising playback monitoring device based on a virtual twin device, comprising: The virtual twin device construction module 310 is used to control the server to generate a unique virtual twin device instance based on the ID of each real advertising device, and to build a corresponding virtual twin device digital model, which is used to simulate the playback behavior and status of the corresponding real advertising device on the server side, serving as a digital substitute for the real advertising device. The playback status data reporting module 320 is used to control the real advertising device to receive advertising playback tasks, execute advertising playback, and periodically report lightweight playback status data to the server. The virtual playback screen reconstruction module 330 is used to control the server to receive playback status data reported by the real device, and combine it with the pre-set playback task and timeline synchronization mechanism to simulate and reconstruct the virtual playback screen synchronized with the real device in real time on the server side. The virtual playback screen monitoring and viewing module 340 is used to control the virtual playback screen of the virtual twin device digital model to be viewed in real time through the client. The cross-validation module 350 is used to control the server to perform cross-validation based on multiple data sources to verify the consistency between the virtual playback screen and the real playback screen of the real advertising device. The anomaly detection and processing module 360 ​​is used to control the server to continuously detect and judge whether the virtual playback screen and playback status data are abnormal. When anomalies are detected in the virtual playback screen and playback status data, the alarm is triggered and the corresponding personnel terminal is notified, as described above.

[0114] Based on the above embodiments, the present invention also provides a server, the principle block diagram of which can be as follows: Figure 4 As shown, the server can be a computer server or a cloud server. The server includes a processor, memory, network interface, display screen, and database connected via a system bus. The processor provides computing and control capabilities. The memory includes non-volatile storage media and internal memory. The non-volatile storage media stores the operating system and computer programs. The internal memory provides an environment for the operation of the operating system and computer programs in the non-volatile storage media. The network interface is used to communicate with external terminals via a network connection. When the computer program is executed by the processor, it implements an advertising playback monitoring method based on a virtual twin device. The server's database stores the advertising playback monitoring program based on the virtual twin device.

[0115] Those skilled in the art will understand that Figure 4 The block diagram shown is merely a partial structural diagram related to the present invention and does not constitute a limitation on the server to which the present invention is applied. A specific server may include more or fewer components than those shown in the figure, or combine certain components, or have different component arrangements.

[0116] In one embodiment, a server is provided, including memory and one or more programs, wherein the one or more programs are stored in memory and configured to be executed by one or more processors. The one or more programs include instructions for performing the following operations: The server generates a unique virtual twin device instance based on the ID of each real advertising device and constructs a corresponding virtual twin device digital model, which is used to simulate the playback behavior and status of the corresponding real advertising device on the server side. The real advertising device receives the advertising playback task, executes the advertising playback, and periodically reports lightweight playback status data to the server; The server receives playback status data reported by the real device, and combines it with pre-set playback tasks and timeline synchronization mechanisms to simulate and reconstruct a virtual playback screen synchronized with the real device in real time on the server side. View the virtual playback screen of the digital model of the virtual twin device in real time through the client; The server performs cross-validation based on multiple data sources to verify the consistency between the virtual playback screen and the real playback screen of the real advertising device. The server continuously monitors and interprets the virtual playback screen and playback status data for any abnormalities. When an abnormality is detected in the virtual playback screen and playback status data, the server controls the triggering of an alarm and notifies the corresponding personnel terminal, as described above.

[0117] The step prior to the step of pre-establishing a corresponding virtual twin device digital model for each real advertising device on the server side includes: When each real advertising device is first started, it registers with the server and uploads its device ID, device model, screen resolution and other basic information.

[0118] The step of pre-establishing a corresponding virtual twin device digital model for each real advertising device on the server side includes: The pre-set digital model of the virtual twin device includes the static information of the real device, as well as dynamic playback task mapping module, timeline synchronization module, playback status model, rendering and display module, anomaly detection module and history record module; The static information of the device includes: device ID, model, screen orientation, resolution, network status, and customer information.

[0119] The step of pre-establishing a corresponding virtual twin device digital model for each real advertising device on the server side includes: The virtual twin device digital model is set up using a task data + status data + timeline synchronization mechanism. By using lightweight task data and periodically reported status data, combined with timeline synchronization, the playback screen of the real device is reconstructed on the server side. The task data refers to the advertising playback instructions issued by the server to the real device, including the list of advertising materials, playback order, playback time period, duration of a single advertisement, loop strategy, and insertion rules. The status data refers to the data that the actual device periodically reports to the server during playback, including the current ad content ID, current playback progress, current timestamp, playback status, volume / screen / network status, and optional screenshots / short video clips; The timeline synchronization mechanism is used to ensure that the playback progress of the virtual twin device on the server side is consistent with the timeline of the real device.

[0120] The steps for the server to perform cross-validation based on multiple data sources to verify whether the playback screen is consistent with the playback result of a real advertising device include: The multi-source data includes: status data, playback logs, screenshot data, and short video clips.

[0121] The step of the server continuously detecting and judging whether the virtual playback screen and playback status data are abnormal, and controlling the triggering of an alarm and notifying the corresponding personnel terminal when abnormalities are detected, further includes: The server continuously monitors and interprets the virtual playback screen and playback status data for any abnormalities. If any abnormalities are detected, such as black screen, stuttering, playback errors, playback not on schedule, network offline, playback interruption, or no status reporting for an extended period, the server will trigger an alarm and notify the corresponding personnel's terminals.

[0122] The step of simulating and reconstructing the virtual playback screen synchronized with the real device in real time on the server side includes: Artificial intelligence technology is introduced to perform content recognition and analysis on the screenshots of the virtual playback screen to verify the consistency between the virtual playback screen and the real playback screen of the real advertising device, as described above.

[0123] In other embodiments, this application proposes a computer-readable storage medium that, when the instructions in the storage medium are executed by the processor of an electronic device, enables the electronic device to perform the above-described method.

[0124] Those skilled in the art will understand that all or part of the processes in the methods of the above embodiments can be implemented by a computer program instructing related hardware. The computer program can be stored in a non-volatile computer-readable storage medium. When executed, the computer program can include the processes of the embodiments of the above methods. Any references to memory, storage, databases, or other media used in the embodiments provided by this invention can include non-volatile and / or volatile memory. Non-volatile memory can include read-only memory (ROM), programmable ROM (PROM), electrically programmable ROM (EPROM), electrically erasable programmable ROM (EEPROM), or flash memory. Volatile memory can include random access memory (RAM) or external cache memory. By way of illustration and not limitation, RAM is available in various forms, such as static RAM (SRAM), dynamic RAM (DRAM), synchronous DRAM (SDRAM), dual data rate SDRAM (DDRSDRAM), enhanced SDRAM (ESDRAM), synchronous link DRAM (SLDRAM), Rambus direct RAM (RDRAM), direct memory bus dynamic RAM (DRDRAM), and memory bus dynamic RAM (RDRAM), etc.

[0125] The above description is merely an embodiment of this application and is not intended to limit the scope of protection 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 scope of protection of this application.

Claims

1. A method for monitoring advertising playback based on a virtual twin device, characterized in that, include: The server generates a unique virtual twin device instance based on the ID of each real advertising device and constructs a corresponding virtual twin device digital model, which is used to simulate the playback behavior and status of the corresponding real advertising device on the server side. The real advertising device receives the advertising playback task, executes the advertising playback, and periodically reports lightweight playback status data to the server; The server receives playback status data reported by the real device, and combines it with pre-set playback tasks and timeline synchronization mechanisms to simulate and reconstruct a virtual playback screen synchronized with the real device in real time on the server side. View the virtual playback screen of the digital model of the virtual twin device in real time through the client; The server performs cross-validation based on multiple data sources to verify the consistency between the virtual playback screen and the real playback screen of the real advertising device. The server continuously monitors and interprets the virtual playback screen and playback status data for any abnormalities. If any abnormalities are detected, the server will trigger an alarm and notify the corresponding personnel's terminal.

2. The advertising playback monitoring method based on a virtual twin device according to claim 1, characterized in that, Before the step of pre-creating a corresponding virtual twin device digital model for each real advertising device on the server side, the following steps are included: When each real advertising device is first started, it registers with the server and uploads its device ID, device model, screen resolution and other basic information.

3. The advertising playback monitoring method based on a virtual twin device according to claim 1, characterized in that, The step of pre-creating a corresponding virtual twin device digital model for each real advertising device on the server side includes: The pre-set digital model of the virtual twin device includes the static information of the real device, as well as dynamic playback task mapping module, timeline synchronization module, playback status model, rendering and display module, anomaly detection module and history record module; The static information of the device includes: device ID, model, screen orientation, resolution, network status, and customer information.

4. The advertising playback monitoring method based on a virtual twin device according to claim 1, characterized in that, The step of pre-creating a corresponding virtual twin device digital model for each real advertising device on the server side includes: The virtual twin device digital model is set up using a task data + status data + timeline synchronization mechanism. By using lightweight task data and periodically reported status data, combined with timeline synchronization, the playback screen of the real device is reconstructed on the server side. The task data refers to the advertising playback instructions issued by the server to the real device, including the list of advertising materials, playback order, playback time period, duration of a single advertisement, loop strategy, and insertion rules. The status data refers to the data that the actual device periodically reports to the server during playback, including the current ad content ID, current playback progress, current timestamp, playback status, volume / screen / network status, and optional screenshots / short video clips; The timeline synchronization mechanism is used to ensure that the playback progress of the virtual twin device on the server side is consistent with the timeline of the real device.

5. The advertising playback monitoring method based on a virtual twin device according to claim 1, characterized in that, The server performs cross-validation based on multiple data sources to verify whether the playback screen matches the playback result of a real advertising device. The steps include: The multi-source data includes: status data, playback logs, screenshot data, and short video clips.

6. The advertising playback monitoring method based on a virtual twin device according to claim 1, characterized in that, The step of the server continuously detecting and judging whether the virtual playback screen and playback status data are abnormal, and controlling the triggering of an alarm and notifying the corresponding personnel terminal when abnormality is detected, further includes: The server continuously monitors and interprets the virtual playback screen and playback status data for any abnormalities. If any abnormalities are detected, such as black screen, stuttering, playback errors, playback not on schedule, network offline, playback interruption, or no status reporting for an extended period, the server will trigger an alarm and notify the corresponding personnel's terminals.

7. The advertising playback monitoring method based on a virtual twin device according to claim 1, characterized in that, The steps of simulating and reconstructing the virtual playback screen synchronized with the real device in real time on the server side include: Artificial intelligence technology is introduced to perform content recognition and analysis on the screenshots of the virtual playback screen to verify the consistency between the virtual playback screen and the real playback screen of the real advertising device.

8. An advertising playback monitoring device based on a virtual twin device, characterized in that, The device includes: The virtual twin device construction module is used to control the server to generate a unique virtual twin device instance based on the ID of each real advertising device, and to build a corresponding virtual twin device digital model, which is used to simulate the playback behavior and status of the corresponding real advertising device on the server side, serving as a digital substitute for the real advertising device. The playback status data reporting module is used to control the real advertising devices to receive advertising playback tasks, execute advertising playback, and periodically report lightweight playback status data to the server. The virtual playback screen reconstruction module is used to control the server to receive playback status data reported by the real device, and combine it with the pre-set playback task and timeline synchronization mechanism to simulate and reconstruct the virtual playback screen synchronized with the real device in real time on the server side. The virtual playback screen monitoring and viewing module is used to control the virtual playback screen of the virtual twin device digital model to be viewed in real time through the client. The cross-validation module is used to control the server to perform cross-validation based on multiple data sources to verify the consistency between the virtual playback screen and the real playback screen of the real advertising device; The anomaly detection and handling module is used to control the server to continuously detect and judge whether the virtual playback screen and playback status data are abnormal. When anomalies are detected in the virtual playback screen and playback status data, the system will control the triggering of an alarm and notify the corresponding personnel terminal.

9. A server, characterized in that, It includes a memory and one or more programs, wherein one or more programs are stored in the memory and configured to be executed by one or more processors, wherein the one or more programs include methods for performing any one of claims 1-7.

10. A computer-readable storage medium, characterized in that, When the instructions in the storage medium are executed by the processor of the electronic device, the electronic device is able to perform the method as described in any one of claims 1-7.

Images