Construction engineering safety and quality inspection system and method

Abstract

The invention discloses a constructional engineering safety and quality inspection system and method.The constructional engineering safety and quality inspection system comprises an inspection robot, a BIM model database, a digital twin engine, an AI analysis module and a cloud management platform which are in data connection with one another. According to the method, the engineering safety and quality are ensured, synchronization of virtuality and reality is realized through a digital twinning technology and integration of a BIM model, analysis and rectification of problems are facilitated through an AI analysis model, industry big data are accumulated, rectification closed-loop management is supported, it is ensured that the problems are thoroughly solved, the rectification closed-loop period is shortened to be within 24 hours, and the method has high popularization value.

Description

Technical Field

[0001] The present invention relates to the technical field of construction engineering safety and quality inspection, and in particular to a construction engineering safety and quality inspection system and method. Background Art

[0002] Safety and quality are crucial during construction. Traditional inspection methods rely primarily on manual labor, which is inefficient due to the scattered locations of construction sites. Furthermore, in some work scenarios, when project supervisors or supervisors need to review on-site safety and quality control, they can only do so by visiting the construction site, which remains inefficient.

[0003] In addition, although some existing projects have established BIM models, there is a disconnect between the BIM models and the on-site data, and dynamic mapping cannot be achieved; the discovery of quality problems is delayed, and the rectification tracking lacks accurate temporal and spatial positioning correlation evidence, making closed-loop management difficult; and the amount of hidden engineering in construction projects is large, and there is currently no panoramic traceability mechanism for on-site management of construction projects; for measurement issues, such as vertical steel bar spacing, support frame standardization and other issues, they can only be measured on-site. Summary of the Invention

[0004] To overcome the defects in the prior art, the present invention provides a construction engineering safety and quality inspection system and method to improve inspection efficiency and accuracy and ensure project safety and quality.

[0005] In order to achieve the above objectives, the present invention provides a construction engineering safety and quality inspection system, comprising:

[0006] Inspection robots, BIM model databases, digital twin engines, AI analysis modules, and cloud management platforms that are interconnected with each other;

[0007] The inspection robot is equipped with a walking mechanism, and is also equipped with a laser radar, a panoramic camera, and a GPS positioning mechanism, and the laser radar, the panoramic camera, and the GPS positioning mechanism are used to collect raw data of the inspection robot in real time during its walking process;

[0008] The BIM model database is pre-installed inside the inspection robot. The BIM model database pre-stores BIM model data that matches the construction site, and the BIM model database is respectively connected to the laser radar, the panoramic camera, and the GPS positioning mechanism data;

[0009] The digital twin engine is used to dynamically map the raw data collected by the inspection robot during its walking process to the BIM model database based on digital twin technology, thereby achieving synchronization between the raw data and the virtual model;

[0010] The AI ​​analysis module has a built-in question library of different types and levels. The AI ​​analysis module is used to match and analyze the raw data collected by the inspection robot during the walking process based on the question library, identify the problem type and level corresponding to the abnormal data, and mark the points corresponding to the abnormal data;

[0011] The cloud management platform is equipped with a client, and the cloud management platform is used to generate corresponding inspection tasks and issue corresponding task instructions to the corresponding responsible personnel through the client terminal based on the identified problem type and level and the marked points, and finally perform closed-loop feedback on the system by returning the on-site processing results to the digital twin engine.

[0012] A construction project safety and quality inspection method, which implements construction project safety and quality inspection through the construction project safety and quality inspection system, and the method includes the following steps:

[0013] Build a BIM model database that pre-stores BIM model data that matches the construction site. Workers create new task instructions and set inspection routes in the BIM model space.

[0014] The inspection robot receives task instructions, walks along the inspection route, and uses lidar, panoramic camera, and GPS positioning components to collect raw data in real time during the inspection robot's walking process;

[0015] Build a digital twin engine and dynamically map the collected raw data to the BIM model database, thereby achieving synchronization between the raw data and the virtual model;

[0016] Through the AI ​​analysis module, and based on the built-in problem library types and levels, the original data collected by the inspection robot during its movement is matched and analyzed, thereby identifying the problem type and level corresponding to the abnormal data and marking the points corresponding to the abnormal data;

[0017] Based on the identified problem type and level and the marked points, corresponding inspection tasks are generated for the corresponding responsible personnel through the corresponding client terminal in the cloud management platform and corresponding task instructions are issued at the same time.

[0018] Preferably, before the staff creates a new task instruction and sets the inspection path in the BIM model space, the initial position information origin of the inspection robot is mapped to the BIM model database, thereby calibrating the initial position of the inspection robot.

[0019] Preferably, when collecting the original data of the inspection robot during its walking process in real time through the laser radar, panoramic camera and GPS positioning component, the point cloud data of the inspection robot is collected in real time through the laser radar, the panoramic image of the inspection robot is collected in real time through the panoramic camera, and the real-time coordinates of the inspection robot are positioned through the GPS positioning component.

[0020] Preferably, when building a digital twin engine and dynamically mapping the collected raw data to the BIM model database to achieve synchronization between the raw data and the virtual model, the coordinate systems of the lidar, the inspection robot, and the holographic image are consistent;

[0021] By building a rigid body transformation model and optimizing the rotation matrix and translation vector, the LiDAR coordinate system and the BIM model coordinate system are unified.

[0022] Based on the precise point selection algorithm, the corresponding pixel area in the holographic image is selected for spherical coordinate conversion, and the axis-pair bounding box of the selected area is generated based on the point cloud data, thereby realizing the spatial mapping between the holographic image and the BIM model coordinate system.

[0023] Preferably, the AI ​​analysis module is used to match and analyze the raw data collected by the inspection robot during its walking process based on the types and levels of the built-in question library, so as to identify the problem types and levels corresponding to the abnormal data and mark the points corresponding to the abnormal data. For data measurement problems, the AI ​​measurement algorithm is used and automatic analysis is performed based on the point cloud data to generate quantitative indicators.

[0024] Preferably, based on the identified problem type and level and the marked points, after a corresponding inspection task is generated for the corresponding responsible personnel through the corresponding client terminal in the cloud management platform and the corresponding task instruction is issued at the same time, closed-loop management of the problem is performed, including the following steps:

[0025] The responsible personnel assign inspection issues to the rectification personnel according to the corresponding task instructions and inspection tasks;

[0026] After receiving the inspection issues, the remediation personnel will carry out on-site rectification and record the rectification process to form rectification evidence;

[0027] The person making the rectification will upload the recorded rectification evidence to the cloud management platform, and the responsible personnel will verify and provide feedback on the uploaded content, thereby achieving closed-loop management of the problem.

[0028] Preferably, when recording the rectification process to form rectification evidence, the rectification evidence meets the following conditions:

[0029]

[0030] in, Indicates coordinate information, Indicates the three-dimensional coordinates of the points corresponding to the abnormal data marked by the AI ​​analysis module in the BIM model coordinate system, Interchangeable Image File Format metadata representing an image file, Indicates the timestamp when the rectification person uploaded the rectification evidence to the cloud management platform, Indicates the timestamp when the cloud management platform sends the inspection task to the responsible person. Indicates the maximum allowed period for problem rectification.

[0031] Due to the adoption of the above technical solution, the present invention has the following beneficial effects:

[0032] 1) Improve inspection efficiency and accuracy: Inspection robots conduct autonomous inspections along pre-set routes, collecting data every 5 meters. Combined with lidar obstacle avoidance and dynamic route adjustment, this reduces the risks of manual inspections and increases inspection efficiency by over 60%. This also strengthens collaboration among project management stakeholders. This enables rapid discovery and precise location of safety and quality issues, with a spatial error of ≤5cm for problem marking and over 90% accuracy in problem identification.

[0033] 2) Accurately locate and describe problems: By integrating digital twins, BIM, and AI technologies, the spatial error of problem marking is reduced to ≤5cm through precise coordinate mapping algorithms. AI is combined with the problem database to achieve fuzzy search and problem classification, with a problem identification accuracy rate of over 90%, which can accurately locate and describe problems.

[0034] 3) Synchronize virtual and real life: Through digital twin technology, the data collected by the inspection robot is linked to the BIM model in real time, and a dynamic digital twin engine is built to synchronize virtual and real life. Managers can intuitively compare and analyze virtual and real scenes to quickly identify and solve problems.

[0035] 4) Support closed-loop management of rectification: By establishing a closed-loop management mechanism for problems, the person responsible for rectification uploads rectification images containing 3D location information, the system verifies the location consistency, and the initiator confirms the closed loop or requests re-inspection to ensure that the problem is rectified in place, and the rectification closed-loop cycle is shortened to within 24 hours.

[0036] 5) Accumulate industry big data: By collecting a large amount of engineering data during the inspection process and archiving it, industry big data can be formed. This data provides a reference for subsequent projects, helps analyze common problems, optimize construction processes, and improve safety and quality management standards.

[0037] 6) Strengthen multi-party collaboration: Stakeholders involved in construction project safety and quality management, such as supervision units and headquarters safety and quality management personnel, can manage online simultaneously through mobile terminals or web terminals. All parties can view inspection information, mark issues, and track rectification in real time, breaking down information barriers and improving collaborative work efficiency. BRIEF DESCRIPTION OF THE DRAWINGS

[0038] In order to more clearly illustrate the technical solutions in the embodiments of the present invention, the following briefly introduces the drawings required for use in the description of the embodiments. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without creative work.

[0039] Figure 1 It is a system architecture diagram of the construction engineering safety and quality inspection system in an embodiment of the present invention.

[0040] Figure 2 This is a flowchart of the inspection process in an embodiment of the present invention.

[0041] Figure 3 This is a flowchart of marking problems using an AI analysis module using the spacing between steel bars as an example in an embodiment of the present invention.

[0042] Figure 4 It is a flowchart of closed-loop problem management in an embodiment of the present invention. DETAILED DESCRIPTION

[0043] The following is a further description of specific embodiments of the present invention in conjunction with the accompanying drawings. It should be noted that the description of these embodiments is intended to facilitate understanding of the present invention and does not constitute a limitation of the present invention. In addition, the technical features involved in the various embodiments of the present invention described below may be combined with each other as long as they do not conflict with each other.

[0044] See also Figures 1 to 4 As shown, an embodiment of the present invention provides a construction engineering safety and quality inspection system, including an inspection robot, a BIM model database, a digital twin engine, an AI analysis module, and a cloud management platform that are data-connected to each other; wherein,

[0045] The inspection robot is equipped with a walking mechanism, a laser radar, a panoramic camera, and a GPS positioning mechanism. The laser radar, the panoramic camera, and the GPS positioning mechanism are used to collect raw data of the inspection robot in real time during its walking process.

[0046] The BIM model database is pre-installed inside the inspection robot. The BIM model database contains BIM model data that matches the construction site, and the BIM model database is connected to the laser radar, panoramic camera, and GPS positioning device data respectively;

[0047] The digital twin engine is used to dynamically map the raw data collected by the inspection robot during its movement to the BIM model database based on digital twin technology, thereby achieving synchronization between the raw data and the virtual model;

[0048] The AI ​​analysis module has a built-in question library of different types and levels. Based on the question library, the AI ​​analysis module is used to match and analyze the raw data collected by the inspection robot during its walking process, identify the problem type and level corresponding to the abnormal data, and mark the points corresponding to the abnormal data;

[0049] The cloud management platform is equipped with a client, and the cloud management platform is used to generate corresponding inspection tasks and issue corresponding task instructions to the corresponding responsible personnel through the client terminal based on the identified problem type and level and the marked points. Finally, the system is closed-loop fed back by transmitting the on-site processing results to the digital twin engine.

[0050] See also Figure 1 As shown, the system architecture in this embodiment mainly includes:

[0051] S1 physical layer (data collection and execution)

[0052] Core equipment: inspection robot.

[0053] Main hardware configuration: LiDAR, high-precision three-dimensional environment scanning, generating point cloud data; panoramic camera, 360° image acquisition, generating holographic images; GPS positioning mechanism.

[0054] Core functions: Real-time collection of point cloud data (the basis of spatial modeling), panoramic images (visual monitoring), and real-time coordinates (dynamic positioning); through multi-sensor fusion, providing accurate physical world data input to the upper layer.

[0055] S2 IoT layer (data transmission and connection)

[0056] The core function is wireless transmission, which uploads the raw data (point cloud, image, coordinates) collected at the physical layer to the cloud in real time; low-latency communication ensures efficient synchronization of data between the robot and the cloud.

[0057] Technical value: Serving as a "bridge" between the physical layer and the cloud layer, it ensures the smooth flow of the entire data chain.

[0058] S3 cloud layer (data processing and twin mapping)

[0059] BIM model database: Builds a high-precision building information model based on 1:1 modeling of the construction project site; provides a static basic framework for digital twins.

[0060] Digital Twin Engine: Dynamically maps the inspection robot's real-time coordinates to the BIM model to synchronize virtual and real spaces; supports accurate overlay analysis of physical data and virtual models.

[0061] AI analysis module: Problem library matching, comparing the types and levels of preset problem libraries (such as safety problem library S and quality problem library Q) to quickly identify anomalies; measurement and calculation, automatically analyzing point cloud data to generate quantitative indicators such as size and deformation.

[0062] Data storage: archives original data, analysis results, and emergency storage information, supporting backtracking and retrieval.

[0063] S4 application layer (business interaction and closed-loop management)

[0064] Terminal coverage: WEB (desktop office) and mobile (on-site operation).

[0065] Core features:

[0066] a. Problem marking: marking abnormal points in the digital twin model;

[0067] b. Track rectification and record the progress of problem handling to form a traceable task chain;

[0068] c. Issue issues and push task instructions from the cloud to the human end;

[0069] d. Corrective feedback: on-site processing results are transmitted back to the cloud to update the system status.

[0070] The embodiment of the present invention further discloses a construction project safety and quality inspection method, which implements construction project safety and quality inspection through the construction project safety and quality inspection system. The method includes the following steps:

[0071] Build a BIM model database that pre-stores BIM model data that matches the construction site. Workers create new task instructions and set inspection routes in the BIM model space.

[0072] The inspection robot receives task instructions, walks along the inspection route, and uses lidar, panoramic camera, and GPS positioning components to collect raw data in real time during the inspection robot's walking process;

[0073] Build a digital twin engine and dynamically map the collected raw data to the BIM model database, thereby achieving synchronization between the raw data and the virtual model;

[0074] Through the AI ​​analysis module, and based on the built-in problem library types and levels, the original data collected by the inspection robot during its movement is matched and analyzed, thereby identifying the problem type and level corresponding to the abnormal data and marking the points corresponding to the abnormal data;

[0075] Based on the identified problem type and level and the marked points, corresponding inspection tasks are generated for the corresponding responsible personnel through the corresponding client terminal in the cloud management platform and corresponding task instructions are issued at the same time.

[0076] See also Figure 2 As shown, before the staff creates a new task instruction and sets the inspection path in the BIM model space, the initial position information origin of the inspection robot is mapped to the BIM model database (the robot's initial position is aligned with the BIM model coordinate system), thereby calibrating the initial position of the inspection robot.

[0077] See also Figure 1 As shown, in this embodiment, when the original data of the inspection robot during its walking process is collected in real time through the laser radar, panoramic camera, and GPS positioning component, the point cloud data of the inspection robot is collected in real time through the laser radar, the panoramic image of the inspection robot is collected in real time through the panoramic camera, and the real-time coordinates of the inspection robot are positioned through the GPS positioning component.

[0078] Furthermore, when building a digital twin engine and dynamically mapping the collected raw data to the BIM model database to achieve synchronization between the raw data and the virtual model, the coordinate systems of the laser radar, inspection robot and holographic image are consistent, and the BIM model coordinate system is consistent. The default coordinate system is WGS-84 (WGS-84 is the standard coordinate system that GPS positioning relies on). When the digital twin engine is working, it models at intervals of 5 meters and stores data in the cloud in real time. The lidar scans for obstacles in real time and triggers dynamic path adjustments.

[0079] By building a rigid body transformation model, optimizing the rotation matrix and translation vector, the laser radar coordinate system and the BIM model coordinate system are unified.

[0080]

[0081] in, 、 、 is the rotation matrix, which is calculated by registering the BIM model feature points with the LiDAR point cloud;

[0082] 、 、 is the translation vector, with an error of ≤ 2 cm (optimized by the least squares method), 、 、 They represent the three-dimensional coordinates (x-axis, y-axis, and z-axis components) of the point cloud data collected by the lidar in the lidar local coordinate system, and are used to convert them to the BIM model coordinate system through rigid body transformation.

[0083] Registration optimization function:

[0084]

[0085] Based on the precise point selection algorithm, the user selects the corresponding pixel area [u1, v1; u2, v2] in the holographic image for spherical coordinate conversion, and generates an axis-pair bounding box for the selected area based on the point cloud data, thereby realizing the spatial mapping between the holographic image and the BIM model coordinate system.

[0086] Preferably, the spherical coordinate conversion method is:

[0087]

[0088]

[0089]

[0090]

[0091]

[0092] Among them, r is the farthest effective distance of the point cloud in the viewing direction, W is the pixel width of the holographic image collected by the panoramic camera (unit: pixel), H is the pixel height of the holographic image collected by the panoramic camera (unit: pixel), is the azimuth angle in the spherical coordinate system (rotation angle around the z axis), ranging from 0 to 2π, and is calculated from the horizontal coordinates of the image pixels. It is the pitch angle (the angle with the z axis) in the spherical coordinate system, ranging from 0 to π, and is calculated from the vertical coordinate of the image pixel.

[0093] It should be noted that when AI-assisted problem marking is performed, for data measurement problems, the AI ​​measurement algorithm is used to automatically analyze the point cloud data to generate quantitative indicators. In this embodiment, the AI ​​measurement algorithm is analyzed in specific steps using the steel bar spacing as an example:

[0094] •Point cloud segmentation: Use RandLA-Net network to extract steel point cloud clusters;

[0095] • Centerline fitting: Perform PCA principal component analysis on the point cloud of a single steel bar to extract the axis equation:

[0096]

[0097] in, 、 、 They represent the three-dimensional coordinates (x-axis, y-axis, and z-axis components) of a reference point on the steel bar axis in the BIM model coordinate system. 、 、 Respectively represent the components of the steel bar axis direction vector on the x-axis, y-axis, and z-axis, and are used to describe the spatial direction of the steel bar;

[0098] • Spacing calculation:

[0099]

[0100] in, is the direction vector of the adjacent steel bar axis, 、 are the coordinates of the corresponding points, The total number of sampling points when calculating the spacing (5 to 10 evenly distributed sampling points are used by default to reduce errors).

[0101] • Single point problem marking (BIM space location, specific problem, problem level, rectification deadline, etc.);

[0102] • Lassoing the panoramic image space (3D bounding box calculation) for a single problem point;

[0103] •The system automatically generates an AI inspection task list (to be dispatched);

[0104] •A record is created in the cloud database, and supervisors / safety / quality personnel review AI issues (to decide whether to dispatch them).

[0105] See also Figure 3 and Figure 4 As shown, based on the identified problem type and level and the marked points, the corresponding inspection tasks are generated for the corresponding responsible personnel through the corresponding client terminal in the cloud management platform and the corresponding task instructions are issued at the same time. Then, closed-loop management of the problem is carried out, including the following steps:

[0106] Responsible personnel (supervisors / safety / quality personnel) assign inspection issues to the correctors according to the corresponding task instructions and inspection tasks. One is discovered by themselves in the digital twin model, and the other is automatically generated by the system. AI inspection task list (to be distributed);

[0107] After receiving the inspection issues, the remediator will carry out on-site rectification and record the rectification process to form rectification evidence (verification can be done by taking photos with a mobile phone app. The phone must have GPS enabled and the photos must include coordinate information);

[0108] The person making the rectification uploads the recorded rectification evidence to the cloud management platform. The responsible personnel will verify and provide feedback on the uploaded content through the rectification evidence verification logic algorithm, thereby achieving closed-loop management of the problem.

[0109] Furthermore, when recording the rectification process to form rectification evidence, the rectification evidence must meet the following conditions:

[0110]

[0111] in, Indicates coordinate information, Indicates the three-dimensional coordinates of the corresponding points of abnormal data marked by the AI ​​analysis module in the BIM model coordinate system (i.e. the original positioning coordinates where the problem occurred), which are used for spatial consistency verification with the coordinates uploaded by the rectification evidence. Represents the Exchangeable Image File Format (EIF) metadata of an image file. This refers specifically to the metadata carried by photos of the rectification site taken by the remediation personnel through a client (such as a mobile phone), including information such as the shooting time, GPS location (latitude, longitude, and elevation), Indicates the timestamp when the person who made the rectification uploaded the rectification evidence (such as photos, text records) to the cloud management platform. Indicates the timestamp when the cloud management platform issues the inspection task to the responsible person (i.e. the problem is first released). Indicates the maximum allowed period for problem rectification (the default value is 7 days). That is, the time interval from the publication of the problem to the upload of rectification evidence must not exceed this threshold to ensure the timeliness of rectification.

[0112] It should be noted that, in this embodiment The coordinate system is consistent with the coordinate system of the mobile phone GPS. The elevation data is called but no mandatory matching calculation is performed. When the verification fails, a rejection report is automatically generated and the responsible person is notified.

[0113] The embodiments of the present invention are described in detail above with reference to the accompanying drawings, but the present invention is not limited to the described embodiments. It is apparent to those skilled in the art that various changes, modifications, substitutions, and variations to these embodiments may be made without departing from the principles and spirit of the present invention, and these changes and modifications still fall within the scope of protection of the present invention.

Claims

1. A construction engineering safety and quality inspection system, characterized in that: It includes inspection robots, BIM model database, digital twin engine, AI analysis module and cloud management platform with data connection between each other; The inspection robot is equipped with a walking mechanism, and is also equipped with a laser radar, a panoramic camera, and a GPS positioning mechanism, and the laser radar, the panoramic camera, and the GPS positioning mechanism are used to collect raw data of the inspection robot in real time during its walking process; The BIM model database is pre-installed inside the inspection robot. The BIM model database pre-stores BIM model data that matches the construction site, and the BIM model database is respectively connected to the laser radar, the panoramic camera, and the GPS positioning mechanism data; The digital twin engine is used to dynamically map the raw data collected by the inspection robot during its walking process to the BIM model database based on digital twin technology, thereby achieving synchronization between the raw data and the virtual model; The AI ​​analysis module has a built-in question library of different types and levels. The AI ​​analysis module is used to match and analyze the raw data collected by the inspection robot during the walking process based on the question library, identify the problem type and level corresponding to the abnormal data, and mark the points corresponding to the abnormal data; The cloud management platform is equipped with a client, and the cloud management platform is used to generate corresponding inspection tasks and issue corresponding task instructions to the corresponding responsible personnel through the client terminal based on the identified problem type and level and the marked points, and finally perform closed-loop feedback on the system by returning the on-site processing results to the digital twin engine.

2. A construction engineering safety and quality inspection method, characterized in that: The construction project safety and quality inspection system of claim 1 is used to implement the construction project safety and quality inspection, and the method includes the following steps: Build a BIM model database that pre-stores BIM model data that matches the construction site. Workers create new task instructions and set inspection routes in the BIM model space. The inspection robot receives task instructions, walks along the inspection route, and uses lidar, panoramic camera, and GPS positioning components to collect raw data in real time during the inspection robot's walking process; Build a digital twin engine and dynamically map the collected raw data to the BIM model database, thereby achieving synchronization between the raw data and the virtual model; Through the AI ​​analysis module, and based on the built-in problem library types and levels, the original data collected by the inspection robot during its movement is matched and analyzed, thereby identifying the problem type and level corresponding to the abnormal data and marking the points corresponding to the abnormal data; Based on the identified problem type and level and the marked points, corresponding inspection tasks are generated for the corresponding responsible personnel through the corresponding client terminal in the cloud management platform and corresponding task instructions are issued at the same time.

3. The construction engineering safety and quality inspection method according to claim 2, characterized in that: Before the staff creates a new task instruction and sets the inspection path in the BIM model space, the initial position information origin of the inspection robot is mapped to the BIM model database to calibrate the initial position of the inspection robot.

4. The construction engineering safety and quality inspection method according to claim 2, characterized in that: When the original data of the inspection robot during its walking process is collected in real time through the laser radar, panoramic camera and GPS positioning component, the point cloud data of the inspection robot is collected in real time through the laser radar, the panoramic image of the inspection robot is collected in real time through the panoramic camera, and the real-time coordinates of the inspection robot are located through the GPS positioning component.

5. The construction engineering safety and quality inspection method according to claim 4, characterized in that: When building a digital twin engine and dynamically mapping the collected raw data to the BIM model database to synchronize the raw data with the virtual model, the coordinate systems of the lidar, inspection robot, and holographic image are consistent; By building a rigid body transformation model and optimizing the rotation matrix and translation vector, the LiDAR coordinate system and the BIM model coordinate system are unified. Based on the precise point selection algorithm, the corresponding pixel area in the holographic image is selected for spherical coordinate conversion, and the axis-pair bounding box of the selected area is generated based on the point cloud data, thereby realizing the spatial mapping between the holographic image and the BIM model coordinate system.

6. The construction engineering safety and quality inspection method according to claim 4, characterized in that: The AI ​​analysis module matches and analyzes the raw data collected by the inspection robot during its walking process based on the types and levels of the built-in question library, thereby identifying the problem types and levels corresponding to the abnormal data and marking the points corresponding to the abnormal data. For data measurement problems, the AI ​​measurement algorithm is used and automatic analysis is performed based on point cloud data to generate quantitative indicators.

7. The construction engineering safety and quality inspection method according to claim 2, characterized in that: Based on the identified problem type and level and the marked points, the corresponding inspection tasks are generated and the corresponding task instructions are issued to the corresponding responsible personnel through the corresponding client terminal in the cloud management platform. The closed-loop management of the problem includes the following steps: The responsible personnel assign inspection issues to the rectification personnel according to the corresponding task instructions and inspection tasks; After receiving the inspection issues, the remediation personnel will carry out on-site rectification and record the rectification process to form rectification evidence; The person making the rectification will upload the recorded rectification evidence to the cloud management platform, and the responsible personnel will verify and provide feedback on the uploaded content, thereby achieving closed-loop management of the problem.

8. The construction engineering safety and quality inspection method according to claim 7, characterized in that: When recording during the rectification process to form rectification evidence, the rectification evidence must meet the following conditions: ； in, Indicates coordinate information, Indicates the three-dimensional coordinates of the points corresponding to the abnormal data marked by the AI ​​analysis module in the BIM model coordinate system, Interchangeable Image File Format metadata representing an image file, Indicates the timestamp when the rectification person uploaded the rectification evidence to the cloud management platform, Indicates the timestamp when the cloud management platform sends the inspection task to the responsible person. Indicates the maximum allowed period for problem rectification.

Images