An integrated AI project risk management and control workbench and method

Abstract

The application provides an integrated AI project risk management and control workbench and method, and belongs to the technical field of risk management and control, and specifically comprises the following steps: determining an identified control point in a collaborative inspection risk point based on a monitoring device in the collaborative inspection risk point and in combination with the association of the monitoring device with other monitoring devices of the collaborative inspection risk point; determining an inspection control scheme of the collaborative inspection risk point according to the association of the identified control point with the monitoring device; performing identification and processing of a construction quality risk of the collaborative inspection risk point by using the inspection control scheme; and determining a risk control method of the collaborative inspection risk point according to the identification and processing result of the construction quality risk of the collaborative inspection risk point and the identification and processing result of the identified control point associated with the monitoring device, so that the reliability of the control and processing of the quality risk of the construction project is improved.

Description

Technical Field

[0001] This invention belongs to the field of risk management technology, and in particular relates to a project risk management workbench and method integrating AI. Background Technology

[0002] Current safety risk management for construction projects typically relies on manual inspections and data reports, making it difficult for project managers to accurately assess the project's status and consequently leading to frequent safety accidents.

[0003] To address the aforementioned technical problems, the existing technical solution, as described in invention patent application CN202411781075.8 "A Smart Construction Site Safety and Quality Inspection Method Based on Big Data," utilizes real-time monitoring and risk assessment. This smart construction site system can improve the safety management level of construction sites, reduce accidents, and predict potential safety risks using predictive maintenance and accident prevention technologies. Furthermore, it leverages big data and intelligent optimization algorithms to optimize resource allocation and construction processes, thereby improving construction efficiency. However, the aforementioned technical solution suffers from the following technical problems:

[0004] When conducting monitoring and analysis of construction quality risks, using a single monitoring device may not be effective in monitoring and analyzing construction quality risks. Therefore, how to determine a collaborative monitoring method for monitoring devices based on construction quality risk points, and realize joint monitoring and scheduling of monitoring devices at multiple construction quality risk points to improve the reliability of construction quality risk monitoring has become an urgent technical problem to be solved.

[0005] Therefore, there is an urgent need for a project risk management platform and methodology that integrates AI. Summary of the Invention

[0006] To achieve the objectives of this invention, the following technical solution is adopted:

[0007] Specifically, this application provides a project risk management method integrating AI, which includes:

[0008] S1 uses the detection data of construction quality risk points in the construction project by the monitoring device based on the integrated AI model to determine the historical identification results of the construction quality risk points in different monitoring devices. Based on the identification data of different monitoring devices in the risk event, the collaborative inspection risk points in the construction management risk points are determined.

[0009] S2, based on the monitoring devices in the collaborative inspection risk points and in conjunction with the association between the monitoring devices and the monitoring devices of other collaborative inspection risk points, determines the identification and control points in the collaborative inspection risk points, and determines the inspection and control scheme for the collaborative inspection risk points according to the association between the monitoring devices of the identification and control points and the monitoring devices of the identification and control points.

[0010] S3 uses the aforementioned inspection and control scheme to identify and process construction quality risks at collaborative inspection risk points. Based on the identification and processing results of construction quality risks at collaborative inspection risk points and the identification and processing results of associated control points of the monitoring device, the risk control method for the collaborative inspection risk points is determined.

[0011] The beneficial effects of this application are as follows:

[0012] Based on the identification data of different monitoring devices in risk events, collaborative inspection risk points are identified in the construction management risk points. Those risk points that cannot be reliably monitored by a single camera due to complex monitoring conditions and must rely on the collaborative operation of multiple cameras for effective control are identified as collaborative inspection risk points, thereby ensuring the reliability of construction quality monitoring and processing.

[0013] Based on the monitoring devices in the collaborative inspection risk points and their association with other collaborative inspection risk points, identification and control points are determined. Specific risk points that, when their monitoring resources (such as camera angles) are adjusted, have the least impact on other risk points in the entire monitoring network are selected and defined as identification and control points. By analyzing the network association structure formed by each collaborative inspection risk point and its matching monitoring devices, the potential cascading effects of optimizing specific risk points are quantitatively assessed. This improves the monitoring efficiency of the target risk point while keeping the negative impact on other related risk points within an acceptable range, thus optimizing the overall efficiency of the monitoring network.

[0014] Furthermore, the construction quality risk points are determined based on the construction locations in the construction project where construction quality risks exist.

[0015] Furthermore, the detection data of the construction quality risk points are determined based on the detection results of the construction quality problems at the construction risk points.

[0016] Furthermore, the historical identification results of the construction quality risk points in different monitoring devices are determined based on the detection results of construction quality problems of the construction quality risk points in different monitoring devices.

[0017] Furthermore, the risk event is any event in which a construction quality problem is identified by any monitoring device.

[0018] Furthermore, the method for determining the collaborative inspection risk points among the construction management risk points is as follows:

[0019] Based on the identification data of different monitoring devices in a risk event, identify the monitoring devices that identified construction quality problems in the risk event and use them as matching monitoring devices.

[0020] The number of matching monitoring devices in the risk event is determined based on the data from the matching monitoring devices in the risk event.

[0021] Based on the number of matching monitoring devices in the risk events, and the risk event data of different monitoring devices belonging to matching monitoring devices, it is determined whether the construction quality risk point belongs to the collaborative inspection risk point.

[0022] Furthermore, the method for determining the risk control method for the collaborative inspection risk points is as follows:

[0023] S41 Based on the identification and processing results of the construction quality risks of the collaborative inspection risk points, determine the risk events monitored after the collaborative inspection of the collaborative inspection risk points, and treat them as monitoring risk events.

[0024] S42 Based on the monitoring risk events, determine the monitoring devices that can detect construction quality problems in different monitoring risk events, and regard the monitoring devices that can detect construction quality problems in the monitoring risk events as reliable monitoring devices.

[0025] S43 utilizes the monitored risk events and reliable monitoring devices in different monitored risk events, and combines the monitored risk events of the identification control points associated with the impact monitoring devices of the collaborative inspection risk points to determine the risk control method for the collaborative inspection risk points.

[0026] Furthermore, it is determined whether there are any monitoring risk events at the identification and control points associated with the monitoring device. If so, no dynamic switching is required. If not, the monitoring device for the impact of the collaborative inspection risk points is switched according to the preset time period, based on the angle with the highest number of construction quality risk points at the collaborative inspection risk points and identification and control points.

[0027] Secondly, this invention provides an AI-integrated project risk management workbench, employing the aforementioned AI-integrated project risk management method, specifically including:

[0028] Collaborative identification module, inspection and control module, risk control module;

[0029] The collaborative identification module is responsible for determining the collaborative inspection risk points among the construction management risk points;

[0030] The inspection and control module is responsible for determining the inspection and control plan for the collaborative inspection risk points;

[0031] The risk management module is responsible for determining the risk management methods for the collaborative inspection risk points.

[0032] Other features and advantages will be set forth in the following description, and the objects and other advantages of the invention are realized and obtained through the structures particularly pointed out in the description and the drawings.

[0033] To make the above-mentioned objects, features and advantages of the present invention more apparent and understandable, preferred embodiments are described below in detail with reference to the accompanying drawings. Attached Figure Description

[0034] The above and other features and advantages of the present invention will become more apparent from a detailed description of exemplary embodiments thereof with reference to the accompanying drawings.

[0035] Figure 1 This is a flowchart of a project risk management method that integrates AI.

[0036] Figure 2 This is a flowchart illustrating the method for determining collaborative inspection risk points in construction risk management.

[0037] Figure 3 This is a flowchart illustrating the method for identifying and determining control points within the risk points of collaborative inspections;

[0038] Figure 4 This is a framework diagram of a project risk management workbench that integrates AI. Detailed Implementation

[0039] To enable those skilled in the art to better understand the technical solutions in this specification, the technical solutions in the embodiments of this specification will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this specification, and not all embodiments. Based on the embodiments of this specification, all other embodiments obtained by those skilled in the art without creative effort should fall within the scope of protection of this specification.

[0040] Example 1

[0041] like Figure 1 As shown, this application provides a project risk management method integrating AI, specifically including:

[0042] S1 uses the detection data of construction quality risk points in the construction project by the monitoring device based on the integrated AI model to determine the historical identification results of the construction quality risk points in different monitoring devices. Based on the identification data of different monitoring devices in the risk event, the collaborative inspection risk points in the construction management risk points are determined.

[0043] Furthermore, the construction quality risk points are determined based on the construction locations in the construction project where construction quality risks exist.

[0044] Furthermore, the detection data of the construction quality risk points are determined based on the detection results of the construction quality problems at the construction risk points.

[0045] Furthermore, the historical identification results of the construction quality risk points in different monitoring devices are determined based on the detection results of construction quality problems of the construction quality risk points in different monitoring devices.

[0046] Furthermore, the risk event is any event in which a construction quality problem is identified by any monitoring device.

[0047] In this embodiment, from among numerous quality risk points in a construction project, those risk points that cannot be reliably monitored by a single camera due to complex monitoring conditions and require multi-camera collaborative operation for effective management are identified; these are called collaborative inspection risk points. The core logic follows a hierarchical and progressive judgment process: First, the activity level of the risk point is initially assessed based on the frequency of risk events; second, for active risk points, the stability and redundancy of the monitoring network are evaluated by quantitatively analyzing the recognition behavior patterns of cameras in historical monitoring events to determine collaborative inspection risk points; finally, among the identified collaborative inspection risk points, those most suitable for monitoring angle optimization and adjustment, and with the least impact on other parts of the overall monitoring network, are further screened and defined as identification and control points, thereby achieving the dual goals of refined allocation of monitoring resources and minimizing the impact of adjustments.

[0048] Specifically, such as Figure 2 As shown, the method for determining the collaborative inspection risk points in the construction management risk points is as follows:

[0049] S11 uses the identification data of different monitoring devices in a risk event to identify the monitoring device that identified the construction quality problem in the risk event and uses it as the matching monitoring device.

[0050] Monitoring device: In the specific implementation of this method, it specifically refers to a camera used for video surveillance. Identification data: refers to the structured judgment result on the existence of a specific construction quality problem output after the original video stream collected by the camera is processed by the image recognition algorithm.

[0051] Matching monitoring devices: In a specific risk event record, the identification data of the camera is confirmed to accurately reflect the construction quality problem.

[0052] In a multi-location camera network, due to factors such as viewpoint obstruction, lighting variations, and differences in algorithm sensitivity to different features, not all cameras can effectively trigger alarms when a problem occurs at a risk point. This step aims to accurately select the individuals that truly "worked" in each event from all deployed cameras, building a clean and effective "successful identification" dataset for subsequent analysis. This is a fundamental step in transforming raw surveillance data into key feature data that can be used for pattern analysis. It filters out invalid or negative surveillance records, ensuring that all subsequent calculations are based on verified, successful identification behaviors, thus making the assessment of camera collaboration capabilities more accurate and focused.

[0053] S12 determines the number of matching monitoring devices in the risk event based on the matching monitoring device data in the risk event;

[0054] Number of matched surveillance devices: refers to the number of cameras that are marked as matched surveillance devices in a single risk event.

[0055] This quantity is a direct indicator of the extent to which a single risk event is "collectively detected" by the monitoring system. It reflects how many independent camera perspectives can effectively capture the problem at the moment it occurs. This is the basic unit for evaluating the redundancy and breadth of the monitoring network's coverage of the risk point, quantifying the monitoring overlap of each event, and providing core input data for subsequent calculations of averages and identification of off-target events. It helps distinguish whether a particular problem is easily detected from multiple angles or can only be captured by a few specific perspectives.

[0056] S13 determines whether the construction quality risk point belongs to the collaborative inspection risk point based on the number of matching monitoring devices in the risk event and the risk event data of different monitoring devices belonging to matching monitoring devices.

[0057] Collaborative inspection risk points: These refer to construction quality risk points that, according to this method, require the activation of a multi-camera collaborative monitoring mechanism (such as data fusion and cross-validation).

[0058] Risk event data for different monitoring devices: This refers to the number or frequency of times each camera has been listed as a matching monitoring device in all its historical risk events for that risk point.

[0059] Determining whether a risk point requires coordination cannot be based solely on a single incident; a comprehensive assessment of its long-term, overall monitoring performance is essential. This necessitates analysis on two levels: first, examining the overall level of the number of matching cameras in historical events (breadth); and second, assessing the stability of individual camera identification (depth). Combining these two aspects is crucial for a comprehensive evaluation of the adequacy of the existing monitoring layout. This step is the core decision-making step of the entire methodology. Through a standardized analysis process, it transforms the data generated in S11 and S12 into clear and actionable classification decisions (whether or not coordinated inspection is needed), thus automating and intelligently formulating monitoring strategies.

[0060] It is understood that if the number of risk events at the construction quality risk point is less than the preset risk event number threshold, then the construction quality risk point is determined not to be a collaborative inspection risk point.

[0061] Number of risk events: refers to the total number of construction quality problem events that have occurred within a historical period for a specific construction quality risk point and have been successfully identified and recorded by any camera. Preset risk event number threshold: a frequency benchmark set to distinguish the level of risk activity.

[0062] Collaborative inspection is a enhanced monitoring strategy designed to improve the control of high-frequency, difficult-to-monitor risk points. When a risk point has historically experienced only a very few quality incidents, it indicates that the construction quality at that point is relatively stable and the risk exposure level is low. For such low-risk exposure points, investing additional collaborative monitoring resources is not cost-effective; the management focus should be on routine inspections. This step serves as a preliminary risk screening mechanism, prioritizing the allocation of management attention and collaborative monitoring resources to "high-activity" risk points that repeatedly experience problems and require close monitoring. This reflects the principle of prioritizing in risk management and improves the overall efficiency of monitoring resource utilization.

[0063] Additionally, it can be understood that if the number of risk events at the construction quality risk point is not less than a preset risk event number threshold, it includes the following:

[0064] Case 1: If the average number of matching monitoring devices in different risk events is greater than the preset threshold for the number of monitoring devices, then the construction quality risk point is determined not to be a collaborative inspection risk point.

[0065] Average number of matching monitoring devices: refers to the arithmetic mean of the number of matching cameras in all historical risk events at this risk point.

[0066] Preset threshold for the number of monitoring devices: a benchmark value used to determine whether the monitoring coverage is sufficient.

[0067] When historical data shows that each risk event can be simultaneously identified by a majority of cameras on average, it indicates that the existing camera network has already formed a highly redundant and consistent effective monitoring coverage of the risk point. In this case, even if individual cameras temporarily fail, other cameras can fill in, and the risk of the problem being missed is extremely low. Therefore, there is no need to activate additional collaborative inspection mechanisms to identify and confirm risk points whose monitoring status is already very ideal.

[0068] Scenario 2: If the average number of matching monitoring devices in different risk events is not greater than the preset threshold for the number of monitoring devices, the risk event with a number of matching monitoring devices less than the preset threshold for the number of devices is identified as a deviation risk event. If the proportion of deviation risk events in the risk events is greater than the preset threshold for the proportion of risk events, then the construction quality risk point is determined to be a collaborative inspection risk point.

[0069] Differential deviation risk events: These refer to events in the history of the risk point where the number of matching cameras is less than a certain set standard (preset device number threshold).

[0070] Deviation risk event percentage: refers to the proportion of the number of identified deviation risk events to the total number of risk events at that risk point.

[0071] When the average number of matches is low, the overall monitoring breadth of risk points is insufficient. At this point, further investigation is needed to determine if there are a large number of events with particularly weak monitoring. If the proportion of "low match" events is high, it indicates that the risk point is frequently in a "monitoring vulnerable" state, highly susceptible to missed issues due to the malfunction of a single camera. This unstable monitoring performance is a clear signal that reliability must be forcibly improved through collaborative inspections, accurately locating unstable links and potential vulnerabilities in the monitoring network. This rule ensures that reinforcement measures are taken for risk points with highly fluctuating monitoring performance, thereby improving the robustness of the overall quality monitoring system.

[0072] Case 3: If the proportion of deviation risk events in the risk events is not greater than the preset risk event proportion threshold, the monitoring matching proportion of the monitoring device is determined by the proportion of risk events in which the monitoring device belongs to the matching monitoring device. If there is no monitoring device with a monitoring matching proportion greater than the preset matching proportion threshold, then the construction quality risk point is determined to be a collaborative inspection risk point.

[0073] Monitoring Matching Ratio: Refers to the frequency with which a single camera appears as a matching monitoring device in all historical risk events for that risk point. Preset Matching Ratio Threshold: A benchmark value used to determine whether a certain camera has become the "dominant monitoring device" for that risk point.

[0074] When there are few deviation events but the average number of matches is also low, it indicates that the overall performance of the monitoring network is stable but the coverage is not wide. At this time, it is necessary to check whether a "hidden dependency" on a particular camera has formed. If no camera has a significantly high matching frequency (i.e., none of them exceed the preset matching percentage threshold), it means that the monitoring responsibility is dispersed, and no single camera can independently perform the task of reliable monitoring. Collaborative inspections are also needed to integrate the recognition capabilities of multiple cameras to form a reliable collective judgment, revealing and addressing those cameras that appear to be covered by multiple cameras but actually lack reliable coverage.

[0075] It should be noted that the monitoring device in the collaborative inspection risk point is a monitoring device that monitors the construction quality problems of the collaborative inspection risk point.

[0076] In one possible implementation: Analysis of quality risks in the reinforcement binding construction of the core tube shear wall of a high-rise building:

[0077] Construction quality risk point: "vertical reinforcement spacing exceeds the standard" in shear walls.

[0078] Monitoring equipment: Four fixed high-definition cameras, numbered CAM_A, CAM_B, CAM_C, and CAM_D, were deployed around the work surface on the core tube construction floor, monitoring from the east, south, west, and north directions respectively.

[0079] Data collection and execution of steps:

[0080] The system recorded all 15 risk events triggered by this risk point in the past two months. The event list and matching cameras are as follows: Event 1: CAM_A, CAM_C, Event 2: CAM_B, Event 3: CAM_A, CAM_D, Event 4: CAM_B, CAM_C, Event 5: CAM_A, Event 6: CAM_C, CAM_D, Event 7: CAM_A, CAM_B, Event 8: CAM_B, CAM_D, Event 9: CAM_C, Event 10: CAM_A, CAM_B, CAM_C, Event 11: CAM_B, Event 12: CAM_A, CAM_D, Event 13: CAM_C, CAM_D, Event 14: CAM_A, Event 15: CAM_B, CAM_C.

[0081] Judgment process:

[0082] Step 1 (Preliminary Judgment): If the number of risk events at this risk point is 15 and not less than the preset threshold of 10, the subsequent collaborative inspection and judgment process will begin.

[0083] Step 2 (S11 and S12): Based on the above data, determine the matching monitoring devices and their quantity for each event. For example, the matching devices for event 1 are CAM_A and CAM_C, with a quantity of 2; the matching device for event 2 is CAM_B, with a quantity of 1, and so on.

[0084] Step 3 (S13): Calculate the average number of matched monitoring devices in all 15 events as (2+1+2+2+1+2+2+2+1+3+1+2+2+1+2) / 15 ≈ 1.73. This average (1.73) is not greater than the preset threshold for the number of monitoring devices (2.5), so it does not meet condition 1, and proceeds to condition 2 analysis.

[0085] In scenario 2, the preset threshold for the number of devices is 2. Events with a matching quantity less than 2 (i.e., events with a quantity of 1) are identified as deviation risk events, including events 2, 5, 9, 11, and 14, totaling 5 events. The percentage of deviation risk events is calculated as: 5 / 15 ≈ 33.3%. This percentage (33.3%) is greater than the preset risk event percentage threshold (30%). Therefore, according to the rules of scenario 2, the construction quality risk point of "excessive spacing of vertical reinforcement in shear wall" is directly determined to be a collaborative inspection risk point.

[0086] This embodiment fully covers the entire process from determining the number of risk events (preliminary step), to identifying matching devices (S11), counting the number (S12), and finally making a final judgment by calculating the average value, identifying deviation events, calculating the deviation ratio, and applying the rules of Case 2 (S13). Although the judgment logic of Case 1 and Case 3 is not triggered in this example, their judgment paths are fully defined in the method.

[0087] This invention achieves precise evaluation and optimization of the performance of construction site camera monitoring networks through a rigorous, data-driven analysis process. By screening risk event frequencies in advance, it accurately focuses the resource-intensive strategy of collaborative inspection on truly high-frequency, high-attention quality risk points, avoiding the dispersion and waste of management resources. It transcends human experience judgment and scientifically identifies risk points that are unreliable due to limitations in monitoring perspective, environmental interference, or unstable recognition algorithms by quantitatively analyzing historical monitoring data. These risk points are the collaborative inspection risk points.

[0088] S2, based on the monitoring devices in the collaborative inspection risk points and in conjunction with the association between the monitoring devices and the monitoring devices of other collaborative inspection risk points, determines the identification and control points in the collaborative inspection risk points, and determines the inspection and control scheme for the collaborative inspection risk points according to the association between the monitoring devices of the identification and control points and the monitoring devices of the identification and control points.

[0089] It should be noted that the monitoring device in the collaborative inspection risk point is a monitoring device that monitors the construction quality problems of the collaborative inspection risk point, i.e., a matching monitoring device.

[0090] Furthermore, such as Figure 3 As shown, the method for determining the identification and control points in the collaborative inspection risk points is as follows:

[0091] Among the identified collaborative inspection risk points, those that, when their monitoring resources (such as camera angles) are adjusted, have the least impact on other risk points in the entire monitoring network, are selected and defined as identification and control points. The core logic is to analyze the network relationship structure formed by each collaborative inspection risk point and its matching monitoring devices, quantitatively assess the potential cascading effects of optimizing specific risk points, thereby improving the monitoring efficiency of the target risk point while keeping the negative impact on other related risk points within an acceptable range, thus optimizing the overall efficiency of the monitoring network.

[0092] S21 uses the monitoring device data in the collaborative inspection risk points to determine the matching monitoring device for the collaborative inspection risk points;

[0093] The above steps include the following:

[0094] Obtain the number of matching monitoring devices for the collaborative inspection risk points, and determine whether the number of matching monitoring devices for the collaborative inspection risk points is less than a preset threshold for the number of matching monitoring devices. If yes, determine that the collaborative inspection risk point belongs to the identification and control point; otherwise, proceed to step S22.

[0095] It should be noted that if the collaborative inspection risk point is an identification and control point, the matching monitoring device for the collaborative inspection risk point will be set at the angle that can monitor the largest number of construction quality risk points of the collaborative inspection risk point.

[0096] Data from monitoring devices at risk points in collaborative inspections: This refers to all cameras and their identification records that have successfully identified construction quality problems at a point during the historical analysis process of determining that the point is a risk point in collaborative inspections.

[0097] Matching monitoring device for collaborative inspection risk points: Specifically refers to a set of cameras that have successfully identified a construction quality problem at least once in the historical risk events of the collaborative inspection risk point.

[0098] Identifying which specific cameras are responsible for effectively monitoring a risk point is fundamental to assessing any adjustments made to it. Only by determining "who is monitoring" can we further analyze "who will be affected by adjustments": transforming the abstract entity of the "risk point" into a concrete, actionable "set of cameras," thus establishing a clear target for subsequent refined network correlation impact analysis.

[0099] Specific example: During the construction of the core tube of a high-rise building, regarding the issue of "excessive spacing of vertical reinforcing bars in the shear wall," which has been identified as a risk point for collaborative inspection, historical identification records show that four cameras (CAM_A, CAM_B, CAM_C, CAM_D) have successfully captured this issue. These four cameras constitute the matching monitoring device for this risk point.

[0100] Preset threshold for the number of matching monitoring devices: a benchmark value used to determine whether the number of cameras relied upon by a collaborative inspection risk point is small enough that adjusting it is simple and the scope of influence is naturally limited.

[0101] If a risk point is monitored by only a very small number of cameras, adjusting these cameras involves a narrow scope, simplifies the decision-making process, and reduces the likelihood of unintended interference with the rest of the monitoring network. Such risk points are suitable as priority candidates for rapid optimization. As an efficient fast-track mechanism, it can quickly identify risk points with simple monitoring architectures, easy implementation, and localized impact, thereby improving the efficiency of management decision-making and response.

[0102] S22 determines, based on the association between the matching monitoring device and other matching monitoring devices of collaborative inspection risk points, that the matching monitoring device belongs to the collaborative inspection risk point of the matching monitoring device and uses it as the associated inspection risk point.

[0103] Related situations: refers to the situation where the same camera is recorded as a matching monitoring device for two or more different collaborative inspection risk points.

[0104] Related inspection risk points: For the target collaborative inspection risk points currently being analyzed, these refer to other collaborative inspection risk points that share at least one identical matching monitoring device with it.

[0105] A construction site monitoring network is a resource-sharing system. A single camera typically covers multiple work areas. When adjusting a camera at a target risk point, it's crucial to anticipate whether this adjustment will weaken the camera's monitoring capabilities for other related risk points. Identifying these related points is the starting point for assessing "side effects." Introducing a systemic perspective prevents isolated optimization of a single risk point from undermining the overall synergy and effectiveness of the monitoring network, ensuring that optimization decisions are based on an understanding of the dependencies between global monitoring resources.

[0106] Specific example: Camera CAM_A is simultaneously a matching monitoring device for both "excessive spacing of shear wall reinforcement" (risk point A) and "verticality of core tube formwork" (risk point B). When considering adjusting CAM_A to optimize monitoring of risk point A, risk point B becomes an associated inspection risk point of risk point A.

[0107] The above steps include the following:

[0108] S221 uses the associated inspection risk points of the collaborative inspection risk point matching monitoring device to determine whether the number of associated inspection risk points of the collaborative inspection risk point is less than the preset risk point number threshold. If yes, proceed to step S222; otherwise, determine that the collaborative inspection risk point does not belong to the identification and control point.

[0109] Number of associated inspection risk points: refers to the total number of other different collaborative inspection risk points associated with the target collaborative inspection risk point through all its matching monitoring devices.

[0110] Preset risk point number threshold: a benchmark used to measure whether the extent to which the adjustment of a risk point may directly affect other risk points is within an acceptable level.

[0111] If a target risk point is directly related to a large number of other risk points in the network, adjusting its cameras would have a wide-ranging and direct cascading effect, making the optimization action too risky and uncertain. Risk points located at network hubs are not suitable as priority control points for adjustment. Effectively screening and excluding risk points with excessively high correlation and wide-ranging impact is crucial to ensure that selected control points have relatively independent monitoring environments or limited direct impact ranges, thereby controlling the potential risks of optimization adjustments.

[0112] Specific example: A certain collaborative inspection risk point is directly linked to multiple other risk points exceeding a preset threshold through all its matched cameras. This indicates that adjusting it would have a domino effect, so it should not be listed as an identification and control point.

[0113] S222 identifies the matching monitoring devices with associated inspection risk points as associated monitoring devices, and determines whether the proportion of associated monitoring devices in the matching monitoring devices of the collaborative inspection risk points is greater than the preset monitoring device proportion threshold. If yes, it is determined that the collaborative inspection risk point does not belong to the identification and control point; otherwise, it proceeds to step S223.

[0114] Associated monitoring devices: refers to cameras that simultaneously serve other associated inspection risk points in the set of matching monitoring devices for target collaborative inspection risk points.

[0115] The proportion of associated monitoring devices: refers to the percentage of the number of associated monitoring devices out of the total number of matching monitoring devices at that risk point.

[0116] Even if the total number of directly related risk points is not large, if most of the cameras a target risk point relies on are "shared resources" (i.e., a high proportion of associated monitoring devices), then adjusting any one of these cameras will very likely affect at least one other risk point. This high degree of shared dependency makes independent optimization of that point difficult and prone to conflict. Therefore, a thorough screening should be conducted from the perspective of the "property" or "independence" of monitoring resources. Risk points that heavily rely on public cameras for their monitoring capabilities and lack proprietary monitoring resources should be identified. These points are unsuitable for control due to the high marginal cost and significant side effects of adjustment.

[0117] Specific example: Although a risk point has multiple matching cameras, the cameras exceeding the preset proportion threshold are all simultaneously serving other risk points. This means that the point has almost no independent monitoring resources, and adjusting its cameras can easily interfere with others. Therefore, it is not suitable as an identification and control point.

[0118] S223 Based on the number of associated inspection risk points of the matching monitoring device, determine whether there is a matching monitoring device with a number of associated inspection risk points greater than the preset threshold for the number of associated risk points. If yes, proceed to step S23; otherwise, determine that the collaborative inspection risk point belongs to the identification and control point.

[0119] Number of associated inspection risk points for a matching monitoring device: This refers to the number of other matching monitoring devices that are also associated with a specific matching monitoring device for a target risk point.

[0120] Preset threshold for the number of associated risk points: a benchmark value used to identify whether a single camera has undertaken monitoring tasks far exceeding the norm and has become a "critical node" with a huge impact on the network.

[0121] It's possible that most cameras at a target risk point are not strongly correlated, but one particular camera might be responsible for monitoring a large number of other risk points. Adjusting this "critical node" camera has a significant impact and cannot be decided solely based on overall proportions. A more prudent and comprehensive assessment process is needed to accurately identify "critical single points" or "super hubs" within the monitoring network. This step ensures that the decision-making process remains highly vigilant regarding critical equipment with exceptionally large impact, preventing systemic risks from being unintentionally triggered by localized adjustments.

[0122] Specific example: Most of the cameras matching risk point Z have weak correlations, but one of its cameras, CAM_X, is associated with a far greater number of risk points than the other cameras. Due to the existence of such a highly correlated key device as CAM_X, the decision needs to proceed to S23 for comprehensive calculation and evaluation.

[0123] S23 determines whether the collaborative inspection risk point is an identification and control point based on the matching monitoring device of the collaborative inspection risk point and the associated inspection risk points of different matching monitoring devices.

[0124] It is understood that in the above steps, the average proportion of the associated monitoring devices in the matching monitoring devices of the collaborative inspection risk points, and the average proportion of the matching monitoring devices in the matching monitoring devices whose number of associated inspection risk points is greater than the preset threshold for the number of associated risk points, are used as the basis to determine the average proportion. It is then determined whether the average proportion is greater than the preset threshold for the average proportion. If it is, the collaborative inspection risk point is determined not to be an identification and control point; otherwise, the collaborative inspection risk point is determined to be an identification and control point.

[0125] When a target risk point is neither widely interconnected nor lacks a single, highly relevant key camera, a quantitative assessment method is needed to balance the "overall dependence" and the "impact of key points." By calculating the average proportion of associated monitoring devices and key cameras, a comprehensive index reflecting the "level of cascading impact of overall adjustments" at that point can be obtained, providing a refined and quantitative decision-making tool for handling complex interconnected situations. By establishing a balance calculation model between "optimization benefits" and "potential risks," it makes the selection of control points in boundary situations more scientific and reasonable.

[0126] Specific example: For risk points with a key camera CAM_X, the system will calculate the total proportion of its associated monitoring devices and the proportion of the key camera (CAM_X) itself in the matching devices, take the average of the two and compare it with a preset threshold to make a final judgment.

[0127] Example: Identification and Control Point Determination and Impact Analysis of Core Tube Construction Area for High-Rise Buildings:

[0128] Scene and settings (strictly consistent with the aforementioned embodiments):

[0129] Project Background: Undertaking the aforementioned high-rise building core tube case. Three collaborative inspection risk points have been identified in this construction area:

[0130] Collaborative inspection risk point A: "Vertical reinforcement spacing exceeds the standard" in shear walls.

[0131] Collaborative inspection risk point B: "Tightness of template joints in the core tube".

[0132] Collaborative inspection risk point C: "positioning accuracy of reserved sleeves" in floor slabs.

[0133] Monitoring setup: Six high-definition cameras (CAM_1 to CAM_6) are deployed. Assuming an initial wide-angle coverage mode, they will perform routine monitoring of all risk points, but the accuracy in capturing specific quality issues will be limited.

[0134] Related data: The association network formed by historical matching records is as follows:

[0135] Risk point A: Matching monitoring devices are [CAM_1, CAM_3, CAM_5].

[0136] Risk point B: Matching monitoring devices are [CAM_2, CAM_4, CAM_5, CAM_6].

[0137] Risk point C: Matching monitoring devices are [CAM_3, CAM_4].

[0138] Judgment process and operation:

[0139] Preliminary assessment of risk points A, B, and C in collaborative inspections:

[0140] Following the procedures S21-S222, risk points A, B, and C were all determined not to be identification and control points because the number of associated risk points was not less than the threshold (S221) or the proportion of associated monitoring devices was too high (S222). This initially indicates that adjusting the angles of the cameras at these points in a concentrated manner may have a significant impact on the associated points.

[0141] Background: Collaborative risk monitoring point E.

[0142] S21: Analysis shows that the matching monitoring devices are [CAM_2, CAM_7 (newly added dedicated cameras inside the shaft)]. CAM_2 originally monitored risk point B.

[0143] Sub-step S21: The number of matching devices is 2, which is not less than the preset threshold of 2. Proceed to S22.

[0144] S22: CAM_2 is associated with risk point B; CAM_7 is a newly added dedicated point and is not associated. Therefore, the associated inspection risk point is: B.

[0145] S221: The number of associated inspection risk points is 1, which is less than the preset threshold of 2. Proceed to S222.

[0146] S222: The associated monitoring device is CAM_2. The proportion of associated monitoring devices is 1 / 2 = 50%, which is not greater than the preset threshold of 60%. Proceed to S223.

[0147] S223: CAM_2 is associated with 1 risk point (B); CAM_7 is associated with 0. Neither exceeds the preset threshold of 2 for the number of associated risk points. Therefore, risk point E in the collaborative inspection is determined to be an identification and control point. Based on this determination, the angles of the matching monitoring devices CAM_2 and CAM_7 for risk point E can be adjusted to the optimal position for monitoring the elevator shaft safety door status. During this operation:

[0148] Adjustments to the dedicated camera CAM_7 have no negative impact.

[0149] The adjustment to the shared camera CAM_2, while potentially slightly reducing its wide-angle coverage of the original associated risk point B (template joint), is within a manageable range because the proportion of associated monitoring devices (50%) did not exceed the threshold. This indicates that monitoring of risk point E does not entirely rely on shared resources, and risk point B itself has other matching devices such as CAM_4, CAM_5, and CAM_6 (this situation was already considered in the associated monitoring device proportion assessment). This optimization achieved an effective improvement in the monitoring performance of the critical safety node (elevator shaft) at an acceptable, localized cost.

[0150] This embodiment strictly follows all the determination steps from S21 to S23. Through coherent scenarios and data, it not only demonstrates the process of identifying control points, but more importantly, the operation instructions of the embodiment clearly explain the actual impact that the specific measure of "setting the matching monitoring device at an angle that can monitor the construction quality risk of the collaborative inspection risk point" may have on the associated inspection risk point. It also demonstrates how this method quantifies, controls and ultimately accepts this impact through a series of threshold judgments (such as S221, S222, S223), thereby making the construction of the embodiment and the content of the patent description form a tight logical closed loop.

[0151] The method for identifying control points provided by this invention is a deeper and more refined application based on collaborative inspection risk point identification. Its core value lies in realizing a closed-loop management of the construction quality monitoring system from "risk identification" to "precise optimization with controllable impact." The core value of this method is not to pursue optimization with zero impact, but rather to accurately locate risk points whose cascading effects from optimization adjustments are quantifiable and within acceptable limits through systematic network correlation analysis and multi-level threshold judgment. It ensures that resource investment in identifying control points can achieve a significant improvement in the monitoring efficiency of the target area at the cost of clear and limited negative impacts, thus achieving a balanced decision between risk and benefit.

[0152] This method transforms experience-based camera adjustment decisions, which can trigger unpredictable chain reactions, into a quantitative analysis process based on objective network topology data and clearly defined rules. By assessing and quantifying potential impacts before making decisions, it effectively avoids the risk of systemic monitoring performance degradation that may result from blind optimization, significantly enhancing the scientific rigor, foresight, and overall security of operational decisions.

[0153] Furthermore, the method for determining the inspection and control plan for the collaborative inspection risk points is as follows:

[0154] For each identified collaborative inspection risk point, a refined and executable camera inspection and control plan is developed. Its core logic is: while prioritizing the monitoring needs of the identified control point (whose camera angles are already fixed at the optimal level), the available monitoring resources for each other collaborative inspection risk point are systematically assessed. By analyzing its dedicated resources, shared resources, and resource conflicts with the identified control point, a dynamic decision is made on whether each camera should adopt a fixed monitoring or dynamic scheduling strategy. This achieves optimal overall monitoring coverage and maximizes emergency response efficiency for all high-priority risk points within the constraints of a complex monitoring network.

[0155] S31 uses the association with the monitoring device of the identification and control point to determine the matching monitoring device that belongs to the matching monitoring device of the identification and control point among the matching monitoring devices of the collaborative inspection risk point, and uses it as the influencing monitoring device.

[0156] Identify control points: These are construction quality risk points that have been previously determined using specific methods, and whose matching monitoring devices have been set and fixed at the optimal monitoring angle.

[0157] Impact on monitoring devices: This refers to cameras among the matching monitoring devices of the target collaborative inspection risk points currently being analyzed, those cameras that are also used by a certain identification and control point. The angles of these cameras are already occupied and cannot be freely adjusted.

[0158] Identifying control points with the highest monitoring priority means their resources are locked. Clearly defining how many resources are "frozen" at a target risk point is an absolute prerequisite for assessing its remaining monitoring capacity and developing feasible solutions, clearly defining the boundaries of resource constraints. Identifying cameras affected by the highest priority external tasks provides accurate input for subsequent solutions based on remaining available resources, avoiding unrealistic solutions that require utilizing already occupied resources.

[0159] The above steps include the following:

[0160] S311 Based on the aforementioned impact monitoring device, determine the number of matching monitoring devices other than the impact monitoring device, and determine whether the number of matching monitoring devices other than the impact monitoring device is greater than a preset device number threshold. If yes, determine that the collaborative inspection risk point inspection control scheme is the preset control scheme; otherwise, proceed to step S312.

[0161] The number of matching monitoring devices excluding those affecting the monitoring device: refers to the number of cameras that the target risk point can freely control or at least not be directly locked by the identification and control point. The preset device number threshold: a quantitative benchmark used to judge whether the available monitoring resources are sufficient.

[0162] If, after excluding occupied resources, a target risk point still has a sufficient number of controllable cameras (exceeding the threshold), it indicates a strong monitoring foundation and minimal external constraints. In this case, a relatively simple and stable standard solution (pre-defined control scheme) can be directly adopted as an efficient and rapid triage mechanism, without the need for complex dynamic scheduling. Quickly categorizing resource-rich risk points and directly applying standardized management strategies simplifies the decision-making process and improves the overall system operating efficiency.

[0163] S312 determines whether the proportion of the impact monitoring devices in the matching monitoring devices of the collaborative inspection risk points is less than the preset impact proportion threshold. If yes, the inspection control scheme of the collaborative inspection risk points is determined to be the preset control scheme. If no, proceed to step S32.

[0164] The percentage of monitoring devices affected: This refers to the percentage of monitoring devices affected out of the total number of matching monitoring devices at the target risk point. Even if the absolute number of available cameras is small, if the percentage of locked cameras is very low (i.e., most resources are still under autonomous control), it indicates that the monitoring autonomy of this risk point is still very high, with little external dependence. It is also suitable to adopt a stable preset solution, supplementing the judgment from the perspective of the relative proportion of resource "autonomy." This ensures that even risk points with a small total number of cameras but where most resources are not controlled by others can also obtain a simple and reliable management solution, avoiding overly complex scheduling.

[0165] S32 determines the associated inspection risk points in different matching monitoring devices based on the overlapping data of matching monitoring devices with other collaborative inspection risk points.

[0166] Overlapping data: refers to network relationship data reflecting the sharing of the same camera among different collaborative inspection risk points. Associated inspection risk points: for a specific matching monitoring device of a target risk point, it refers to other collaborative inspection risk points served by the camera in addition to the current target risk point.

[0167] When a simple preset solution cannot be adopted, it is necessary to conduct an in-depth analysis of the "part-time" service relationships of each available camera. Understanding what other risk points each camera is associated with is the foundation for developing a refined sharing and scheduling strategy (the second preset control solution), and drawing a monitoring service relationship network diagram with the cameras as the hub. This is the core data basis for subsequently differentiating camera types, assessing scheduling priorities, and formulating dynamic rules.

[0168] The above steps include the following:

[0169] S321 uses the matching monitoring device that does not have an associated inspection risk point as a strongly associated device, and determines whether the collaborative inspection risk point has a strongly associated device. If yes, proceed to step S33; otherwise, determine the inspection control scheme of the collaborative inspection risk point as the second preset control scheme.

[0170] Strongly correlated devices refer to cameras in the matching monitoring devices for a target risk point that serve only the current target risk point and not any other collaboratively inspected risk points. In other words, the list of associated inspection risk points for that camera is empty. "Strongly correlated devices" are dedicated, exclusive monitoring resources for the target risk point and are the cornerstone of monitoring reliability. If even one such dedicated resource is missing, it means that the monitoring of that point relies entirely on cameras shared with other points, making its monitoring status extremely unstable and subject to adjustment at any time due to the needs of other points. In this case, a more complex second-preset control scheme specifically designed for highly shared resources must be adopted to identify the key weaknesses in the monitoring resource structure. If strongly correlated devices are lacking, it is directly determined that an advanced dynamic sharing scheduling scheme must be adopted; this is a strategy to ensure the bottom line of monitoring.

[0171] S33 determines the inspection and control plan for the collaborative inspection risk points based on the impact monitoring device data of the collaborative risk points and the associated inspection risk points in different matching monitoring devices.

[0172] When a risk point has dedicated resources (strongly correlated devices) and faces situations where these resources are partially occupied (affecting monitoring devices) and partially shared, a quantitative method is needed to comprehensively assess the "reliability" or "independence" of its overall monitoring resources. By calculating a comprehensive "monitoring reliability coefficient," it is possible to objectively determine whether its monitoring capability is more inclined towards stability and reliability (suitable for the preset scheme) or more inclined towards reliance on complex scheduling (suitable for the second preset scheme).

[0173] A decision-making model for scientific scheme selection under a mixed resource model is provided. This model makes an optimal trade-off between "stability" and "scheduling complexity" through quantitative evaluation, making scheme selection more objective and adaptable.

[0174] Specific example: Calculate a coefficient by combining the number of strongly correlated devices and the proportion of controllable resources. For example, a high coefficient indicates that the point mainly relies on dedicated resources, has high monitoring stability, and can adopt the preset scheme; a low coefficient indicates a high dependence on shared resources, requiring a second preset scheme for dynamic management.

[0175] It is understood that the monitoring reliability coefficient of the collaborative inspection risk point is determined by the number of strongly correlated devices and the proportion of the matching monitoring devices of the collaborative inspection risk point excluding the devices that affect the monitoring. It is then determined whether the monitoring reliability coefficient of the collaborative inspection risk point is greater than the preset reliability coefficient threshold. If it is, the inspection control scheme of the collaborative inspection risk point is determined to be the preset control scheme. If not, the inspection control scheme of the collaborative inspection risk point is determined to be the second preset control scheme.

[0176] It should be noted that the first preset control scheme only requires setting the strongly correlated device at an angle that can monitor the construction quality risk of the collaborative inspection risk point. The second preset control scheme, in addition to setting the strongly correlated device at the angle that can monitor the largest number of construction quality risk points of the collaborative inspection risk point, sets a matching monitoring device with a number of associated inspection risk points less than a preset threshold for the number of associated risk points at the collaborative inspection risk point. The matching monitoring device is set at the collaborative inspection risk point with the smallest number of matching monitoring devices at the angle that can monitor the largest number of construction quality risk points of the collaborative inspection risk point. That is, it is set at the angle with the largest number of construction quality risk points of the smallest number of collaborative inspection risk points.

[0177] Furthermore, apart from the strongly correlated devices and the matching monitoring devices where the number of correlated inspection risk points is less than the preset threshold for the number of correlated risk points, the remaining matching monitoring devices will, if any correlated risk point has a construction quality problem, set it at the angle with the highest number of construction quality risk points among the correlated risk points with construction quality problems. Otherwise, they will switch between the angles with the highest number of construction quality risk points among different correlated risk points according to a preset time period.

[0178] Example: Development of a collaborative inspection and control plan for risk points in the core tube construction area of ​​a high-rise building:

[0179] Scenario and settings (strictly consistent with all the aforementioned embodiments):

[0180] The identified control point is risk point E (the opening and closing status of the elevator shaft safety door inside the core tube). Its matching monitoring devices [CAM_2, CAM_7] have been fixed at the optimal monitoring angle.

[0181] Risk points for collaborative inspections to be formulated: A (excessive spacing of vertical reinforcement bars in shear walls), B (tightness of formwork joints), C (positioning accuracy of reserved sleeves).

[0182] Global data association of monitoring devices:

[0183] Risk Point A: Matching devices [CAM_1, CAM_3, CAM_5]. (Related: CAM_3 is shared with C)

[0184] Risk Point B: Matching devices [CAM_2, CAM_4, CAM_5, CAM_6]. (Related: CAM_2 is occupied by E, CAM_4 is shared with C, and CAM_5 is shared with A.)

[0185] Risk point C: Matching devices [CAM_3, CAM_4]. (Related: CAM_3 is shared with A, CAM_4 is shared with B)

[0186] Determination of risk point A in collaborative inspection:

[0187] S31: Among the matching devices [CAM_1, CAM_3, CAM_5] for risk point A, none of them are matching devices for identifying control point E, therefore the list of affected monitoring devices is empty.

[0188] S311: After removing the monitoring devices that affect the system (0), the number of remaining matching devices is 3, which is greater than the preset threshold of 2. Therefore, the inspection and control plan for risk point A is determined to be the preset control plan.

[0189] Execution of preset control plan:

[0190] Strongly correlated device identification: CAM_1 serves only A and is a strongly correlated device.

[0191] Solution: Fix the strongly correlated device CAM_1 at the angle that best monitors the risk of "vertical reinforcement spacing in shear walls". Although CAM_3 and CAM_5 are shared with C and B, since A has sufficient resources, they can be used for auxiliary monitoring according to default or simple rules.

[0192] Determination of risk point B in collaborative inspection:

[0193] S31: CAM_2 is identified as serving the control point E, therefore the monitoring device affected is [CAM_2].

[0194] S311: After removing CAM_2, the remaining matching devices are [CAM_4, CAM_5, CAM_6], totaling 3, which is greater than the preset threshold of 2. Therefore, the inspection and control plan for risk point B is determined to be the preset control plan.

[0195] Execution of preset control plan:

[0196] Strongly correlated device identification: CAM_6 serves only B and is a strongly correlated device.

[0197] Solution: The strongly correlated device CAM_6 is fixed at an angle that best monitors the risk of "template joint tightness". CAM_4 and CAM_5 participate in shared scheduling, that is, they switch between different collaboratively monitored risk points on a 30-minute cycle.

[0198] Scenario: Assume there is a risk point F, and its matching devices are [CAM_8, CAM_9, CAM_10]. Among them, CAM_8 is an impact monitoring device (occupied by a certain identification and control point), CAM_9 is a strongly associated device (serving only D), and CAM_10 is associated with another risk point.

[0199] Procedure: After judgment by S311 / S312, proceed to S32. S321 determines that there is a strongly correlated device CAM_9, and proceeds to S33.

[0200] S33 calculation: Number of strongly correlated devices: 1, Total number of matched devices: 3, Excluding the proportion of devices affecting monitoring: (3-1) / 3 = 2 / 3 ≈ 66.7%, which is less than 0.8.

[0201] Monitoring reliability coefficient calculation (example formula): Coefficient = (Number of strongly correlated devices / Total number of matched devices) * 0.6 + (Proportion excluding influencing devices) * 0.4 = (1 / 3) * 0.6 + (2 / 3) * 0.4 = 0.2 + 0.2667 = 0.4667.

[0202] Judgment: Coefficient 0.4667 < preset reliability coefficient threshold 0.7, therefore the inspection and control scheme for risk point F is determined to be the second preset control scheme. Although there are strongly correlated devices, the monitoring reliability is insufficient after comprehensive evaluation, and a dynamic sharing strategy is still required.

[0203] Implementation of the second pre-set control plan:

[0204] Low-association device processing: Determine whether the number of associated inspection risk points is less than the preset threshold of 2. The number of associated points for CAM_10 is 1, which is less than 2, so both are "low-association devices".

[0205] Rule: Set these low-association devices at the angle that has the highest number of construction quality risk points at point C and the lowest number of matching monitoring devices at the collaborative inspection risk points.

[0206] Analysis: For CAM_10, it can monitor both F and G. We need to choose which one is "more needed" between F and G. Let's examine which cameras can monitor point F: CAM_8, CAM_9, and CAM_10. Point F is covered by 3 cameras (CAM_8, CAM_9, CAM_10), and point G is covered by 2 cameras (CAM_10, CAM_11). Therefore, for CAM_10, we should assign it to the point with the fewest cameras covering the risk of point G being compromised, which is point G itself (since point G only has 2 cameras).

[0207] This embodiment strictly follows all the determination steps from S31 to S33. Through coherent scenarios and data, it fully demonstrates the triggering conditions, decision logic, and specific execution content of the two strategies, the preset control scheme and the second preset control scheme, ensuring the comprehensiveness and operability of the scheme formulation.

[0208] The collaborative inspection risk point control scheme determination method provided by this invention is the final strategy output layer based on the prior risk point classification and identification. Its core value lies in realizing a closed loop of intelligent construction quality monitoring from "perception and analysis" to "decision execution." Under hard resource constraints (identification and control point occupancy) and soft resource constraints (camera sharing network), this method generates customized monitoring strategies for each collaborative inspection risk point. Through scientific decision-making logic, it ensures that while prioritizing the protection of the most critical areas, all other high-priority risk points can obtain optimal monitoring arrangements matching their resource conditions, thereby maximizing the overall effectiveness of the entire monitoring network.

[0209] S3 uses the aforementioned inspection and control scheme to identify and process construction quality risks at collaborative inspection risk points. Based on the identification and processing results of construction quality risks at collaborative inspection risk points and the identification and processing results of associated control points of the monitoring device, the risk control method for the collaborative inspection risk points is determined.

[0210] Furthermore, the method for determining the risk control method for the collaborative inspection risk points is as follows:

[0211] In this embodiment, continuous effectiveness evaluation and dynamic optimization are performed on collaborative inspection risk points for which inspection and control schemes (pre-set control schemes or second pre-set control schemes) have been implemented. The core logic is to evaluate the actual monitoring effectiveness of the existing scheme by analyzing newly generated risk event data after the scheme's implementation. Based on the evaluation results—including the number of new problems, the reliability of monitoring findings, and the status of associated control points—a dynamic decision is made to either maintain the existing scheme or initiate a more proactive resource-sharing strategy (e.g., including its impact on monitoring devices in dynamic scheduling). This maximizes the ability to detect quality problems at collaborative inspection risk points while ensuring that the highest priority monitoring task (control point identification) is not compromised.

[0212] S41 Based on the identification and processing results of the construction quality risks of the collaborative inspection risk points, determine the risk events monitored after the collaborative inspection of the collaborative inspection risk points, and treat them as monitoring risk events.

[0213] It should be noted that in the above steps, it is determined whether the number of monitored risk events at the collaborative inspection risk point is less than the preset risk event number threshold. If so, the risk control method for the collaborative inspection risk point is determined to still be carried out according to the original collaborative inspection plan. If not, proceed to step S42.

[0214] The results of the identification and handling of construction quality risks refer to the reports of construction quality problems that the system automatically identifies through camera monitoring during the operation of the inspection and control plan.

[0215] Monitoring risk events: Specifically refers to construction quality issues that occur after the implementation of the inspection and control plan and are successfully monitored and recorded by the system, targeting the risk point of the collaborative inspection.

[0216] This forms the basis for effectiveness evaluation. Direct evidence of the program's effectiveness comes from whether new quality issues emerge after the program's implementation, and how these issues were discovered, establishing the timeline benchmark and data sources for the evaluation. It clearly distinguishes between the "risk events" defined in the historical analysis phase and the "monitored risk events" generated during the program's implementation phase, ensuring that the evaluation is based on the latest actual performance under the program's influence.

[0217] Specific example: In the case of taking over risk point F, after the "second preset control plan" was deployed for it, the system recorded two new quality problem alarms for risk point F within one observation period. These two events are the "monitoring risk events" for point F.

[0218] A preset threshold for the number of risk events is used as a benchmark to measure whether the frequency of new problems during the observation period is at an acceptablely low level. If the occurrence rate of new problems at this risk point is very low after the implementation of the solution, it indicates that the existing solution can effectively control the risk or detect problems in a timely manner, the solution is operating well, and there is no need for adjustment. This provides a fast and stable optimization criterion. Prioritizing the maintenance of a well-functioning system and avoiding unnecessary adjustments that introduce new uncertainties is in line with the principle of robustness management.

[0219] S42 Based on the monitoring risk events, determine the monitoring devices that can detect construction quality problems in different monitoring risk events, and regard the monitoring devices that can detect construction quality problems in the monitoring risk events as reliable monitoring devices.

[0220] Reliable monitoring device: refers to a camera that has successfully identified and reported a quality problem in a specific monitoring risk event.

[0221] When a large number of new problems arise, a thorough diagnosis of the "effectiveness" of existing surveillance resources is necessary. Analyzing which cameras were truly effective when the problem occurred can reveal blind spots or untimely responses in the current camera configuration or scheduling strategy, moving beyond "whether it was detected" to "who detected it" and "how it was detected." This provides crucial factual evidence for subsequent assessments of the "redundancy" and "reliability" of surveillance.

[0222] Specific example: For two monitoring risk events occurring at risk point F, analysis reveals that: Event 1 was captured simultaneously by CAM_9 and CAM_10; Event 2 was captured only by CAM_9. Therefore, in these two events, CAM_9 and CAM_10 acted as "reliable monitoring devices" in different events.

[0223] It is understandable that the above steps include the following:

[0224] S421 Obtain the number of reliable detection devices for the collaborative inspection risk point in different monitoring risk events, and determine whether the number of reliable detection devices for the collaborative inspection risk point in different monitoring risk events is less than a preset reliable device number threshold. If so, determine that the risk control method for the collaborative inspection risk point is to remove the risk control processing according to the original collaborative inspection plan, and for the impact monitoring device of the collaborative inspection risk point, switch processing according to the preset time period from the angle with the largest number of construction quality risk points in the collaborative inspection risk point and the identification control point. If not, proceed to step S422.

[0225] Number of reliable detection devices: The number of cameras used as "reliable detection devices" in a single monitoring risk event. Preset threshold for the number of reliable devices: A benchmark used to determine whether the "intensity" or "redundancy" of a single event successfully captured by the monitoring system meets the standard.

[0226] If only a very small number (e.g., one) of the cameras can capture each new incident, it indicates that the monitoring capability at that point is extremely "fragile" and "unstable." This could mean that the fixed camera angles are inadequate, or that the dynamic scheduling strategy has failed to effectively cover the risks. In this case, the most radical optimization is needed—releasing the "impact monitoring devices" originally fixed to the identification and control point to participate in dynamic scheduling. This significantly increases the monitoring resources allocated to that risk point, representing a deep diagnosis of monitoring "stability" and the most powerful remedial measure. It identifies risk points with weak monitoring capabilities and triggers the highest level of resource sharing to fundamentally improve monitoring reliability.

[0227] S422 defines a monitoring risk event where the number of reliable monitoring devices is not less than a preset threshold as a reliable monitoring event. It then determines whether all of the monitoring risk events are reliable monitoring events. If so, the risk control method for the collaborative inspection risk point is to continue to handle the risk according to the original collaborative inspection plan. If not, the process proceeds to step S43.

[0228] Reliable monitoring events: These refer to events where the number of cameras successfully capturing the problem reaches or exceeds a preset standard. If all new problem events can be detected simultaneously by a sufficient number of cameras (meeting the redundancy standard), it indicates that the existing monitoring scheme is effective, and the monitoring reliability is high. Therefore, the existing monitoring strategy should be maintained, and the management focus should shift to improving the construction process and accurately identifying the root causes of problems. This avoids erroneous adjustments to a well-functioning monitoring scheme due to problems in the construction phase, enabling more scientific decision-making.

[0229] S43 utilizes the monitored risk events and reliable monitoring devices in different monitored risk events, and combines the monitored risk events of the identification control points associated with the impact monitoring devices of the collaborative inspection risk points to determine the risk control method for the collaborative inspection risk points.

[0230] Furthermore, it is determined whether there are any monitoring risk events at the identification and control points associated with the monitoring device. If so, no dynamic switching is required. If not, the monitoring device for the impact of the collaborative inspection risk points is switched according to the preset time period, based on the angle with the highest number of construction quality risk points at the collaborative inspection risk points and identification and control points.

[0231] When optimization is required (i.e., the judgment direction of S421 or S422 needs to be adjusted), the overall security of resource scheduling must be considered. If the identification and control point associated with the "impact monitoring device" of the target risk point is also in a period of high risk, then the camera must be prioritized for the highest priority task and cannot be scheduled away. Conversely, if the identification and control point is in a stable state, it can be safely allowed to participate in dynamic scheduling to support the current risk point.

[0232] It enables secure and flexible scheduling of global monitoring resources. It ensures that any optimization action is taken without compromising the highest priority monitoring safeguards, reflecting the system's hierarchical security concept.

[0233] Specific example: For risk point F, if it is decided to dynamically affect the monitoring device CAM_8, the system will check whether the identification and control points served by CAM_8 have monitored any risk events during the same period. If not, the system will optimize by adding CAM_8 to the polling process; if so, the system will prohibit scheduling of CAM_8 and maintain its fixed monitoring of the identification and control points.

[0234] Example: Dynamic optimization of risk management methods for collaborative inspection risk point F:

[0235] Scenario and settings (strictly consistent with the implementation of the inspection and control plan):

[0236] Current status: Risk point F (matching devices [CAM_8, CAM_9, CAM_10]) has been determined to adopt the second preset control scheme because its monitoring reliability coefficient (0.4667) is lower than the threshold (0.7). According to this scheme:

[0237] The strongly correlated device CAM_9 is fixed at the optimal angle for monitoring point F, while the weakly correlated device CAM_10 is fixedly assigned to the associated risk point G, which has less camera coverage (assuming there are only 2 cameras at point G).

[0238] The monitoring device CAM_8 remains fixed in serving its associated identification and control point (let's say point H).

[0239] Observation period: 15 working days after the implementation of the second pre-set control plan.

[0240] Data on newly occurring monitored risk events:

[0241] Risk point F: Three quality incidents occur.

[0242] Event F1: The reliable monitoring device is [CAM_9].

[0243] Event F2: The reliable monitoring device is [CAM_9].

[0244] Event F3: The reliable monitoring device is [CAM_9].

[0245] Control point H: 0 monitoring risk events occurred.

[0246] S41: The number of monitored risk events for risk point F is 3, which is not less than the preset threshold of 2. Proceed to S42.

[0247] S42: Determine that the reliable monitoring device for each event is [CAM_9].

[0248] S421: Determine whether the number of reliable devices for the three events is "all less than" the threshold 2.

[0249] The number of events F1 is 1, which is less than 2; the number of events F2 is 1, which is less than 2; and the number of events F3 is 1, which is less than 2.

[0250] Since all events are below the threshold, the risk management method for risk point F is as follows: While continuing to implement the original second preset control plan, the monitoring device CAM_8, which affects the risk point, will also switch between risk point F and the angle with the highest number of construction quality risk points identified at control point H, according to a preset time cycle. This means that CAM_8 is released to participate in dynamic patrols, working with CAM_9 to strengthen the monitoring coverage of point F.

[0251] This embodiment takes risk point F as the object and fully covers the judgment steps from S41 to S43. It demonstrates how to trigger the dynamic upgrade and optimization of the scheduling strategy for "impact monitoring device" based on the monitoring data after the implementation of the scheme, thus realizing a management closed loop.

[0252] The collaborative inspection risk point risk control method provided by this invention serves as the "adaptive optimization engine" and "safety scheduling hub" of the entire intelligent monitoring system. Its core value lies in achieving closed-loop management from static plan execution to dynamic effect optimization. This method establishes a complete closed loop of "plan deployment → operation monitoring → performance evaluation → strategy adjustment." It uses real data generated after plan execution as the basis for decision-making, enabling the monitoring system to self-diagnose and self-optimize based on actual results. This automates and automates the management process, achieving a leap from "monitoring problems" to "monitoring performance management." Moving beyond the traditional focus on merely "whether there is a problem," this method proactively assesses the reliability, redundancy, and stability of the monitoring system itself by analyzing in-depth indicators such as the "number of reliable monitoring devices," extending the management focus to proactively ensuring the health of the monitoring system.

[0253] By verifying the status of associated identification and control points through step S43, this method strictly adheres to the principle of "highest priority task assurance" in any optimization decision. It only allows for deeper resource sharing after ensuring the safety of the identification and control points, achieving a precise balance between safety baseline and resource flexibility.

[0254] Example 2

[0255] Secondly, such as Figure 4 As shown, this invention provides an AI-integrated project risk management workbench, employing the aforementioned AI-integrated project risk management method, specifically including:

[0256] Collaborative identification module, inspection and control module, risk control module;

[0257] The collaborative identification module is responsible for determining the collaborative inspection risk points among the construction management risk points;

[0258] The inspection and control module is responsible for determining the inspection and control plan for the collaborative inspection risk points;

[0259] The risk management module is responsible for determining the risk management methods for the collaborative inspection risk points.

[0260] The various embodiments in this specification are described in a progressive manner. Similar or identical parts between embodiments can be referred to mutually. Each embodiment focuses on describing the differences from other embodiments. In particular, the embodiments of apparatus, devices, and non-volatile computer storage media are basically similar to the method embodiments, so the descriptions are relatively simple; relevant parts can be referred to the descriptions of the method embodiments.

[0261] The foregoing has described specific embodiments of this specification. Other embodiments are within the scope of the appended claims. In some cases, the actions or steps recited in the claims may be performed in a different order than that shown in the embodiments and may still achieve the desired result. Furthermore, the processes depicted in the drawings do not necessarily require the specific or sequential order shown to achieve the desired result. In some embodiments, multitasking and parallel processing are possible or may be advantageous.

[0262] The above description is merely one or more embodiments of this specification and is not intended to limit this specification. Various modifications and variations can be made to the one or more embodiments of this specification by those skilled in the art. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principle of one or more embodiments of this specification should be included within the scope of the claims of this specification.

Claims

1. A project risk management method integrating AI, characterized in that, Specifically, it includes: Based on the detection data of construction quality risk points of the construction project by the monitoring device based on the integrated AI model, the historical identification results of the construction quality risk points in different monitoring devices are determined. Based on the identification data of different monitoring devices in the risk event, the collaborative inspection risk points in the construction quality risk points are determined. The collaborative inspection risk points are construction quality risk points that require the activation of a multi-camera collaborative monitoring mechanism; Based on the monitoring devices in the collaborative inspection risk points, and in combination with the association between the monitoring devices in the collaborative inspection risk points and the monitoring devices in other collaborative inspection risk points, the identification and control points in the collaborative inspection risk points are determined. Based on the association between the monitoring devices of the identification and control points and the monitoring devices of the identification and control points, the inspection and control plan for the collaborative inspection risk points is determined. The inspection and control scheme is used to identify and process the construction quality risks of collaborative inspection risk points. Based on the identification and processing results of the construction quality risks of the collaborative inspection risk points and the identification and processing results of the identification and control points associated with the monitoring devices in the collaborative inspection risk points, the risk control method of the collaborative inspection risk points is determined. The method for determining the identification and control points in the collaborative inspection risk points is as follows: Based on the monitoring device data of the collaborative inspection risk points, determine the matching monitoring device for the collaborative inspection risk points; Based on the association between the matching monitoring device and other matching monitoring devices of collaborative inspection risk points, the matching monitoring device is determined to be a collaborative inspection risk point of the matching monitoring device and is regarded as an associated inspection risk point. The aforementioned correlation refers to the situation where the same camera is recorded as a matching monitoring device for two or more different collaborative inspection risk points; Based on the matching monitoring device for the collaborative inspection risk points and the associated inspection risk points of different matching monitoring devices, determine whether the collaborative inspection risk points are identification and control points. The method for determining the risk control method for the collaborative inspection risk points is as follows: Based on the identification and processing results of the construction quality risks of the collaborative inspection risk points, the risk events monitored after the collaborative inspection risk points are determined and used as monitoring risk events. Based on the aforementioned monitoring risk events, a monitoring device capable of detecting construction quality problems in different monitoring risk events is identified, and the monitoring device capable of detecting construction quality problems in the aforementioned monitoring risk events is designated as a reliable monitoring device. By utilizing the aforementioned monitoring risk events and reliable monitoring devices in different monitoring risk events, and combining the monitoring risk events of the identification and control points associated with the impact monitoring devices of the collaborative inspection risk points, the risk control method for the collaborative inspection risk points is determined. The aforementioned impact monitoring device refers to the matching monitoring device that belongs to the matching monitoring device for identification and control points among the matching monitoring devices for collaborative inspection risk points.

2. The project risk management method integrating AI as described in claim 1, characterized in that, The construction quality risk points are determined based on the construction locations in the construction project where there are construction quality risks.

3. The project risk management method integrating AI as described in claim 1, characterized in that, The detection data for the construction quality risk points are determined based on the detection results of the construction quality problems at the construction quality risk points.

4. The project risk management method integrating AI as described in claim 1, characterized in that, The risk event is any event in which a construction quality problem is identified by any monitoring device.

5. The project risk management method integrating AI as described in claim 1, characterized in that, The method for determining the collaborative inspection risk points among the construction quality risk points is as follows: Based on the identification data of different monitoring devices in the risk event, identify the monitoring devices that identified construction quality problems in the risk event and use them as matching monitoring devices. The number of matching monitoring devices in the risk event is determined based on the data from the matching monitoring devices in the risk event. Based on the number of matching monitoring devices in the risk events, and the risk event data of different monitoring devices belonging to matching monitoring devices, it is determined whether the construction quality risk point belongs to the collaborative inspection risk point. The risk event data for the different monitoring devices mentioned above is determined based on the number or frequency with which each camera is listed as a matching monitoring device in all historical risk events at the construction quality risk point.

6. The project risk management method integrating AI as described in claim 5, characterized in that, If the number of risk events at the construction quality risk point is less than a preset risk event number threshold, then the construction quality risk point is determined not to be a collaborative inspection risk point.

7. The project risk management method integrating AI as described in claim 1, characterized in that, If the collaborative inspection risk point is an identification and control point, then the matching monitoring device for the collaborative inspection risk point will be set at the angle that can monitor the largest number of construction quality risk points of the collaborative inspection risk point.

8. A project risk management workbench integrating AI, employing the project risk management method integrating AI as described in any one of claims 1-7, characterized in that, Specifically, it includes: Collaborative identification module, inspection and control module, risk control module; The collaborative identification module is responsible for determining the collaborative inspection risk points among the construction quality risk points. The inspection and control module is responsible for determining the inspection and control plan for the collaborative inspection risk points; The risk management module is responsible for determining the risk management methods for the collaborative inspection risk points.

Images