AI collaborative review whole process traceable supervision service information management system

Abstract

This invention discloses an AI-powered collaborative drawing review system with full-process traceability for engineering supervision services. It relates to the field of engineering supervision information management technology and includes: a project information archiving module that establishes unique identification codes, associates basic data, and categorizes and archives it; an intelligent drawing upload and parsing module that supports multi-format uploads, extracts core elements, and stores them in a structured manner; an AI-powered collaborative drawing review module that automatically identifies drawing problems and generates rectification suggestions through a trained model; a manual review and annotation module that provides annotation tools and creates review records; a drawing review process traceability module that records all process data and assigns unique traceability codes; an opinion collaborative circulation module that enables closed-loop online problem circulation; a supervision results management module that generates standardized reports and statistical indicators; and a system security and maintenance module that assigns permissions to ensure data security and system stability. This invention improves efficiency and accuracy through AI-powered collaborative drawing review, ensures service quality from multiple dimensions, and provides efficient and reliable support for engineering supervision.

Description

Technical Field

[0001] This invention relates to the field of information management technology for engineering supervision, and in particular to an information management system for supervision services that allows for full-process traceability of AI-assisted drawing review. Background Technology

[0002] Construction project supervision is a core component in ensuring project quality, safety, and compliance. Construction drawing review, as the first step in supervision, directly impacts the scientific rigor and safety of the project. With the increasing scale and interdisciplinary nature of construction projects, the complexity of construction drawings has significantly increased, making traditional supervision and review methods inadequate for modern project management needs. Traditional review relies on manual, sheet-by-sheet verification, which is not only time-consuming and labor-intensive but also susceptible to subjective factors such as the reviewers' experience and energy levels, leading to oversights of issues like dimensional discrepancies and specification conflicts, thus compromising review efficiency and accuracy. Furthermore, the lack of efficient collaborative mechanisms in multi-disciplinary review results, with fragmented management of review outcomes across disciplines, makes problem aggregation and cross-verification difficult, easily resulting in gaps in interdisciplinary coordination.

[0003] Existing supervision information systems often focus on single-stage functionality, lacking integrated management capabilities across the entire process. Some systems only support drawing uploads and simple annotations, failing to incorporate AI-powered intelligent drawing review technology and still requiring manual review. Collaborative communication relies on traditional methods such as offline meetings and email, resulting in low efficiency in the flow of review issues and rectification suggestions, unclear responsibility allocation, and potential communication gaps leading to delayed rectification. The process traceability mechanism is inadequate; data such as review operation trajectories and issue handling records are stored in a fragmented manner, failing to form a complete traceability chain and lacking effective evidence for subsequent verification and responsibility determination. Furthermore, the management of supervision deliverables lacks standardized processes; report generation relies on manual compilation, data statistical analysis capabilities are weak, and it is difficult to quickly output core indicators such as review pass rates and rectification closure rates, failing to provide accurate data support for project management decisions.

[0004] Outdated standards and regulations, and an unscientific quality assessment system are also prominent problems with the existing system. National and industry standards related to construction engineering are constantly being revised and improved, but the existing system struggles to keep pace with these updates in real time. This leads to a lag in the basis for drawing review and increases the likelihood of biased compliance judgments. Furthermore, the quality assessment of supervision services relies heavily on subjective evaluations, lacking a multi-dimensional quantitative indicator system. This fails to objectively reflect core service quality aspects such as the accuracy of drawing review and the timeliness of deliverables, hindering the continuous optimization and performance evaluation of supervision services. Simultaneously, drawings and supervision data contain a large amount of sensitive information, and the existing system's data security measures are inadequate, posing risks of data leakage and tampering, impacting the compliance and credibility of supervision services. These problems result in inefficient, inconsistent, and chaotic supervision services, hindering the digital and intelligent development of the construction engineering supervision industry. Summary of the Invention

[0005] The present invention proposes an AI-based collaborative drawing review process traceability information management system for supervision services, which aims to solve the problems mentioned in the prior art.

[0006] To achieve the above objectives, the present invention adopts the following technical solution: an AI-assisted collaborative drawing review process traceability information management system for supervision services, comprising the following modules: The project information filing module receives relevant information about the supervision project entered by the user, establishes a unique identification code for the project, links it to project establishment documents, contract documents, and technical specifications and standards, and automatically classifies and archives the project by type to form a standardized project information database. The intelligent drawing upload and parsing module supports batch upload and online preview of CAD, PDF and BIM format construction drawings. It extracts the core elements of the drawings through the drawing parsing algorithm and automatically identifies the drawing version number and modification mark. The AI ​​collaborative drawing review module uses deep learning to build a drawing review model. It takes the parsed drawing elements and technical specifications as input, automatically identifies drawing problems, marks related content, and generates rectification suggestions. The manual review and annotation module provides an online annotation tool, allowing reviewers to view the AI ​​review results, confirm, reject, or supplement the explanations for issues, upload review comments and signed documents, and create a manual review record. The drawing review process traceability module records data from the entire drawing review process in real time, assigns a unique traceability code to each issue and associates it with relevant information; The feedback and feedback module establishes a collaborative channel among participating units, pushing review issues and rectification suggestions to the responsible parties, and supporting online responses to rectification plans and uploading of modified drawings; The project supervision results management module integrates relevant drawing review materials, generates standardized supervision reports according to the supervision service specifications, and automatically calculates the core indicators of project drawing review. The system security and operation module sets up a multi-level user permission system, uses data encryption technology to ensure security, and records user logs and alarms for abnormal behavior.

[0007] Furthermore, it also includes a project progress control module, which establishes a progress plan model based on key project time nodes and supervision workflow, links the progress of drawing review, problem rectification and project construction, monitors the progress of each stage in real time, and automatically sends reminder notifications for overdue tasks.

[0008] Furthermore, it also includes a supervision quality assessment module, which constructs a multi-dimensional supervision service quality assessment system. It conducts quantitative assessments from the dimensions of drawing review accuracy, problem rectification closure rate, timeliness of deliverables, and satisfaction of participating units, collects feedback and evaluation scores from participating units, and automatically generates supervision quality assessment reports.

[0009] Furthermore, the AI ​​collaborative drawing review module introduces a dynamic priority allocation model for drawing review tasks, which scientifically sorts the review order of different drawings using the following formula: in Indicates the priority value of the drawing review task. This indicates the importance coefficient of the related engineering parts in the drawing. This indicates the risk impact coefficient that may be caused by problems with the drawings. This indicates the complexity coefficient of modifying the drawing. This indicates the current congestion coefficient of the drawing review queue. This indicates the urgency level of the project schedule. This indicates the professional correlation coefficient of the drawing.

[0010] Furthermore, the intelligent drawing upload and parsing module adopts a unified parsing engine for multi-format drawings. For CAD drawings, it extracts layer information, block references, and structured data of dimension annotations. For PDF drawings, it uses OCR technology to recognize text and graphic elements. For BIM drawings, it extracts three-dimensional model parameters and component relationships. The integrity of the drawings is automatically verified during the parsing process.

[0011] Furthermore, the supervision quality assessment module introduces a comprehensive scoring model for supervision service quality, which quantifies quality using the following formula: in This indicates the overall score for the quality of the supervision services. Indicates the basic quality coefficient. Indicates the first The weight of each evaluation indicator Indicates the first The actual score of each evaluation indicator Indicates the first The deviation coefficient of the evaluation indicators Indicates the first The impact attenuation coefficient of the evaluation index This indicates the total number of evaluation indicators.

[0012] Furthermore, in the drawing review process traceability module, blockchain technology is used to record key traceability data to form a distributed traceability ledger. Each traceability code corresponds to a unique blockchain hash value, supporting anti-counterfeiting verification and full traceability of traceability data.

[0013] Furthermore, the opinion collaboration and circulation module establishes an intelligent reminder mechanism, sets the reminder frequency according to the urgency of the problem and the rectification deadline, and pushes notifications through system messages, SMS, and emails. Problems that have not been rectified within the deadline are automatically upgraded with reminders and the overdue status is recorded.

[0014] Furthermore, it also includes a standard update module, which establishes a technical standard database covering current national, industry, and local building engineering-related standards and specifications, and supports online querying, retrieval, and related citation of standards and specifications.

[0015] Furthermore, the supervision results management module integrates data visualization and analysis functions, which intuitively display the review data in the form of bar charts, line charts, and heat maps, supports custom chart dimensions and query conditions, and automatically generates data statistical analysis reports.

[0016] Compared with existing technologies, the beneficial effects of this invention are: In terms of efficiency and accuracy in drawing review, the system integrates an AI collaborative drawing review module, which automatically identifies dimensional deviations, specification conflicts, and structural defects in drawings based on a deep learning model. This covers the review needs of multiple disciplines, significantly reducing the workload of manual drawing review and improving efficiency. The manual review and annotation module serves as a supplement, accurately capturing hidden problems that the AI ​​fails to identify, forming a dual guarantee mechanism of "AI initial review + manual review." This significantly improves the accuracy of drawing review, avoids omissions due to human negligence, and builds a solid first line of defense for project quality.

[0017] In terms of collaborative management and process traceability, the feedback collaboration module establishes an online collaborative channel for participating units, enabling the entire process of raising, rectifying, reviewing, and closing loops of review issues online. It automatically records feedback and timelines from all participants, eliminating communication barriers and improving rectification efficiency. The review process traceability module records all process data in real time, assigning a unique traceability code to each issue, linking responsible personnel to the processing results, and using blockchain technology to ensure the immutability of traceability data. This ensures that the entire review process is verifiable, traceable, and accountable, providing a solid basis for liability determination and compliance verification.

[0018] In terms of deliverables management and decision support, the supervision deliverables management module generates standardized supervision reports according to specifications, automatically calculates core indicators, and supports online previewing, exporting, and archiving of deliverables, significantly reducing manual processing workload and improving the timeliness and standardization of deliverables. The data visualization and analysis function presents data such as the distribution of review issues and rectification progress intuitively, providing data support for supervision decisions; the specification and standard update module synchronizes with the latest specifications in real time, ensuring the accuracy and timeliness of the review basis.

[0019] In terms of quality assessment and schedule control, the supervision quality assessment module constructs a multi-dimensional quantitative assessment system to objectively reflect the quality of supervision services and provide a scientific basis for service optimization and performance evaluation. The project schedule control module links the review progress with the project construction progress, monitors the progress in real time, and automatically sends overdue reminders to ensure that the project progresses as planned. The system security and operation and maintenance module comprehensively protects the security of drawings and supervision data through multi-level permission division, encrypted data storage and transmission, and abnormal behavior alarms, preventing the risk of data leakage and tampering, and improving the compliance and credibility of supervision services.

[0020] Overall, this invention significantly improves the digitalization and intelligence of supervision services, enabling efficient collaborative review of drawings, full traceability of the process, standardized management of results, scientific quality assessment, and controllable data security. It effectively reduces the workload of supervision, improves service quality and management efficiency, provides strong guarantees for the quality, safety and compliance of construction projects, and promotes the digital and intelligent transformation of the supervision industry. It has important practical value and industry-driving significance. Attached Figure Description

[0021] Figure 1 This is a schematic block diagram of the AI-based collaborative drawing review and full-process traceability information management system for supervision services proposed in this invention. Figure 2 A bar chart comparing the efficiency of multi-disciplinary drawing review under different review modes; Figure 3 A line graph showing the change in the recognition accuracy of the AI ​​image analysis model as a function of the number of training samples. Figure 4 A pie chart showing the distribution of problem types in municipal engineering drawing review; Figure 5 A scatter plot showing the correlation between the supervision service quality score and various evaluation indicators. Detailed Implementation

[0022] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0023] In the description of this invention, it should be understood that the terms "center," "longitudinal," "lateral," "length," "width," "thickness," "upper," "lower," "front," "rear," "left," "right," "vertical," "horizontal," "top," "bottom," "inner," "outer," "clockwise," and "counterclockwise," etc., indicate the orientation or positional relationship based on the orientation or positional relationship shown in the accompanying drawings. They are only for the convenience of describing this invention and simplifying the description, and do not indicate or imply that the device or element referred to must have a specific orientation, or be constructed and operated in a specific orientation. Therefore, they should not be construed as limitations on this invention.

[0024] Furthermore, the terms "first" and "second" are used for descriptive purposes only and should not be construed as indicating or implying relative importance or implicitly specifying the number of indicated technical features. Thus, features defined with "first" and "second" may explicitly or implicitly include one or more of the stated features. In the description of this invention, "a plurality of" means two or more, unless otherwise explicitly specified. Furthermore, the terms "installed," "connected," and "linked" should be interpreted broadly; for example, they may refer to a fixed connection, a detachable connection, or an integral connection; they may refer to a mechanical connection or an electrical connection; they may refer to a direct connection or an indirect connection through an intermediate medium; and they may refer to the internal connection of two components. Those skilled in the art can understand the specific meaning of the above terms in this invention based on the specific circumstances. The invention will now be described in further detail with reference to the accompanying drawings.

[0025] Reference Figures 1 to 5 A supervisory service information management system with AI-assisted collaborative drawing review and full-process traceability, comprising the following modules: The project information filing module receives basic information about the supervision project, information of participating units, project construction scale, scope of supervision services and key time nodes entered by users, establishes a unique identification code for the project, associates it with project establishment documents, contract documents and technical specifications and standards, supports batch import, modification and query of project information, automatically classifies and archives it according to project type, and forms a standardized project information database. The intelligent drawing upload and parsing module supports batch uploading and online preview of construction drawings in CAD, PDF, and BIM formats. It extracts core elements such as component types, size parameters, material information, and node construction from the drawings through the drawing parsing algorithm, automatically identifies the drawing version number and modification mark, establishes the association between the drawings and the project, and realizes the structured storage and fast retrieval of drawings. The AI ​​collaborative drawing review module is based on deep learning to build a drawing review model. The model is trained on a large number of compliant drawing samples, covering building structure, water supply and drainage, electrical, and HVAC disciplines. Input the parsed drawing elements and corresponding technical specifications and standards, and automatically identify dimensional deviations, specification conflicts, and structural defects in the drawings. Mark the location of the problem, associate the relevant specification clauses, and generate preliminary rectification suggestions. It supports simultaneous drawing review and problem summary by multiple disciplines. The manual review and annotation module provides online annotation tools, including text annotation, graphic drawing, and voice annotation functions. Reviewers can view the AI ​​review results, confirm, reject, or supplement the identified issues, manually mark hidden issues not identified by AI, upload review opinions and signed confirmation documents, and form a manual review record. The drawing review process traceability module records the entire drawing review process data in real time, including drawing upload time, AI drawing review start and end time, manual review operation trajectory, problem modification records, and opinion flow nodes. It assigns a unique traceability code to each problem, links the problem discoverer, handler, handling time, and rectification results, and supports querying traceability records by project, profession, and time range, and generates process traceability reports. The feedback collaboration module establishes a collaborative channel among participating units, automatically pushing review issues and rectification suggestions to the relevant responsible parties such as design and construction units. It supports online responses to rectification plans and uploading of revised drawings, realizing the entire process of problem identification, rectification, review, and closure online. It automatically records the feedback content and time nodes of each participating party, achieving an efficient closed loop for feedback. The project supervision results management module integrates materials such as the review issue list, rectification response documents, review records, and process traceability reports. It generates standardized supervision reports according to the supervision service specifications, including review summary reports, issue rectification tracking reports, and phased supervision evaluation reports. It supports online preview, export, printing, and archiving of deliverables and automatically calculates key indicators such as project review pass rate and issue rectification closure rate. The system security and operation and maintenance module sets up a multi-level user permission system, dividing users into roles such as project administrators, drawing review engineers, participating units, and system operation and maintenance personnel, and assigning different data operation and viewing permissions. It adopts data encryption technology to ensure the security of the storage and transmission of drawings and supervision data, records user login logs, operation logs and abnormal behavior alarms, and provides system parameter configuration, data backup, automatic fault diagnosis and recovery functions to ensure stable system operation.

[0026] This invention also includes a project progress control module, which establishes a progress plan model based on key project time nodes and supervision workflow, links the progress of drawing review, problem rectification and project construction, monitors the progress of each stage in real time, automatically sends reminder notifications for overdue tasks, supports the adjustment and change records of the progress plan, generates progress deviation analysis reports, and provides data support for supervision progress control.

[0027] This invention also includes a supervision quality assessment module, which constructs a multi-dimensional supervision service quality assessment system. It conducts quantitative assessments from the dimensions of drawing review accuracy, problem rectification closure rate, timeliness of deliverables, and satisfaction of participating units. It collects feedback and evaluation scores from participating units and automatically generates a supervision quality assessment report, providing a basis for optimizing supervision services and performance evaluation.

[0028] In this invention, the AI ​​collaborative drawing review module introduces a dynamic allocation model for drawing review task priorities, which scientifically sorts the review order of different drawings through the following formula: in This indicates the priority of the drawing review task; the larger the value, the higher the priority. This indicates the importance coefficient of the engineering parts related to the drawings, and is determined based on the critical node level of the project. This indicates the risk impact coefficient that may be caused by problems with the drawings, calculated based on the degree of impact on safety, quality, and cost. This represents the complexity coefficient of drawing modifications, determined by the number of components in the drawing and the scope of the modifications. This indicates the current congestion coefficient of the drawing review queue, calculated based on the number of drawings awaiting review and the workload of the reviewers. This indicates the project's urgency level, determined by considering the remaining construction period and drawing review deadlines. This represents the professional correlation coefficient of the drawings, which is determined based on the degree of cross-relationship among multiple disciplines. By dynamically calculating each coefficient, key drawings are given priority for review, thereby improving the efficiency of review resource allocation.

[0029] In this invention, the intelligent drawing upload and parsing module adopts a unified parsing engine for multiple drawing formats. For CAD drawings, it extracts layer information, block references, and structured data of dimension annotations. For PDF drawings, it uses OCR technology to recognize text and graphic elements. For BIM drawings, it extracts three-dimensional model parameters and component relationships. During the parsing process, it automatically verifies the integrity of the drawings and sends a reminder to re-upload drawings with missing key elements. The parsing results support linked viewing with the original drawings.

[0030] In this invention, the supervision quality assessment module introduces a comprehensive scoring model for supervision service quality, which achieves precise quantification of quality through the following formula: in This represents the overall score for the quality of the supervision services, ranging from 0 to 100 points. A higher score indicates better quality. This represents the basic quality coefficient, determined based on the supervision unit's qualification level and historical service reputation. Indicates the first The weights of the evaluation indicators were determined using the analytic hierarchy process (AHP) to ascertain the importance of each indicator. Indicates the first The actual score of each evaluation indicator is determined based on quantitative standards and actual performance. Indicates the first The deviation coefficient of each evaluation indicator is calculated by the difference between the actual score and the standard score. Indicates the first The attenuation coefficient of the impact of each evaluation indicator is determined based on the persistence of the indicator's impact on overall quality. This indicates the total number of evaluation indicators, and a scientific evaluation of the quality of supervision services is achieved through the comprehensive calculation of multi-dimensional indicators.

[0031] In this invention, the drawing review process traceability module uses blockchain technology to record key traceability data, including drawing review results, signed confirmation documents, rectification closed-loop records, and other tamper-proof information, forming a distributed traceability ledger. Each traceability code corresponds to a unique blockchain hash value, supporting anti-counterfeiting verification and full traceability of traceability data, ensuring the compliance and credibility of supervision services.

[0032] In this invention, the opinion collaboration and circulation module establishes an intelligent reminder mechanism, sets the reminder frequency according to the urgency of the problem and the rectification period, and pushes notifications through system messages, SMS, and emails. Problems that have not been rectified within the time limit are automatically upgraded with reminders and the overdue status is recorded. It supports online review of rectification plans and multi-party signature confirmation, and realizes an efficient closed loop for opinion circulation.

[0033] This invention also includes a standard update module, which establishes a technical standard database covering current national, industry, and local building engineering-related standards and specifications. It supports online querying, retrieval, and association of standards and specifications, automatically monitors the update dynamics of standards and specifications, promptly pushes update notifications, and provides new version downloads, thereby achieving real-time updates and accurate matching of the basis for drawing review.

[0034] In this invention, the supervision results management module integrates data visualization and analysis functions, which intuitively displays data such as the distribution of review issues, rectification progress, and quality assessment results in the form of bar charts, line charts, and heat maps. It supports custom chart dimensions and query conditions, and automatically generates data statistical analysis reports to provide data support for supervision decision-making and project management.

[0035] The following two examples further illustrate the specific implementation of this system: Example 1: Application of Construction Supervision Services for Large Commercial Complexes This embodiment targets a large-scale commercial complex project with a total construction area of ​​250,000 square meters, covering four disciplines: building structure, water supply and drainage, electrical, and HVAC. It applies the AI ​​collaborative drawing review and full-process traceability supervision service information management system of this invention to realize multi-disciplinary collaborative drawing review, full-process traceability and standardized management of supervision results, and fully implement all functional modules.

[0036] 1. Multi-module collaborative operation The project information filing module receives basic project information entered by the user, including project name, construction location, total investment, and structural type. Information on participating units includes the names, contact persons, and contact information of the construction unit, design unit, construction unit, and supervision unit. The project construction scale is clearly defined as 20 floors above ground and 3 floors underground. The scope of supervision services covers construction drawing review, construction process supervision, and final acceptance supervision. Key time nodes include the completion time of drawing review, commencement time, main structure topping-out time, and completion time. The system generates an 18-digit unique project identification code, with the coding rule being: the first 6 digits are the regional code, the middle 6 digits are the project type code, and the last 6 digits are the project approval sequence code. It links to basic data such as project approval documents, supervision contracts, and unified standards for construction quality acceptance. The system supports batch import and online modification of project information and automatically archives it to the project database according to commercial building type.

[0037] The intelligent drawing upload and parsing module supports design units in batch uploading CAD format structural drawings, PDF format water supply and drainage drawings, and BIM format HVAC and electrical drawings. The system provides online preview functionality, allowing users to zoom in, zoom out, and pan to view drawing details. The unified multi-format drawing parsing engine extracts structured data such as layer information, beam and column block references, and rebar dimensions from CAD drawings; it uses OCR technology to recognize text descriptions and pipe diameter markings from PDF drawings; and it extracts parameters such as component dimensions, material types, and pipeline routes from BIM drawings. During parsing, the system automatically identifies drawing version numbers, marks modification traces, establishes the association between drawings and projects, and stores the data in a structured drawing database, supporting quick retrieval by specialty and drawing number.

[0038] The AI-powered collaborative drawing review module employs a deep learning model trained on 100,000 compliant commercial building drawing samples. Inputs include the parsed drawing elements and corresponding technical specifications such as building structural load codes and water supply and drainage design standards. A dynamic priority allocation model for drawing review tasks is also introduced. set up The importance coefficient of the engineering parts associated with the drawings, core tube area The value is 0.9, for ordinary floors. The value is 0.6; The risk impact coefficient is related to structural safety. Value is 0.9, related to function optimization. The value is 0.5; To modify the complexity coefficient, for components with more than 50 components... Values ​​of 0.8 and fewer than 30 The value is 0.3; The congestion factor for the drawing review queue is the number of drawings awaiting review (more than 20). Values ​​of 0.7 and less than 10 portions The value is 0.2; The project's urgency factor indicates that the remaining construction period is less than 30 days. Value is 0.9, over 60 days The value is 0.5; This refers to the professional relevance coefficient, which represents the intersection of multiple disciplines. Value is 1.2, single major The value is set to 1.0. Taking the core tube structure drawing as an example, =0.9、 =0.9、 =0.8、 =0.7、 =0.9、 =1.2, calculated as follows: The first term = 0.9 × 0.9 × 0.57 ≈ 0.46, and the second term = 0.9 / 1.2 = 0.75. =0.46+0.75=1.21, the highest priority, the system prioritizes allocating drawing review resources. The model automatically identifies problems such as core tube beam reinforcement dimension deviations and conflicts between water supply and drainage pipes and electrical wiring specifications, marks the specific location coordinates of the problems, associates them with the corresponding specification clause numbers, and generates preliminary rectification suggestions.

[0039] The manual review and annotation module provides tools for text annotation, graphic drawing, and voice annotation. The review engineer views the AI ​​review results, confirms all 12 identified dimensional deviation issues, rejects 3 misjudged code conflict issues, adds 2 beam-column joint structural defects that the AI ​​did not recognize, and uploads a review opinion document with an electronic signature to form a complete manual review record.

[0040] The drawing review process traceability module records in real time the drawing upload time as 9:00 AM on March 10, 2024, the AI ​​review start time as 9:05 AM and the end time as 9:20 AM, and the manual review operation trajectory includes annotation time and modification content. The problem modification record covers the design unit's feedback time and modification plan. Each problem is assigned a unique traceability code, which is associated with the discoverer, the handler, and the rectification result. Blockchain technology is used to record tamper-proof information such as review results and signed documents. Each traceability code corresponds to a unique hash value, and it supports querying traceability records by profession and time range, generating process traceability reports.

[0041] The feedback collaboration module pushes review issues to the design unit via system messages and emails. The design unit responds online with rectification plans and uploads the revised drawings. Once the supervision unit reviews and approves, a closed loop is formed. The intelligent reminder mechanism sets up to remind once a day based on the urgency of the issue. If the issue is not rectified within 3 days, the reminder is automatically escalated. The system records the feedback content and time nodes of each participating party.

[0042] The project supervision results management module integrates the review issue list, rectification documents, and review records to generate review summary reports and issue rectification tracking reports. It automatically calculates indicators such as review pass rate and rectification closure rate, and visualizes the issue distribution and rectification progress through bar charts and line charts. It also supports exporting results files to PDF format.

[0043] The system security and operations module is divided into user roles: project administrator, drawing review engineer, and participating unit users. Administrators have full permissions, drawing review engineers can only operate the drawing review and verification functions, and participating unit users can only view related issues and feedback. Encryption technology is used to ensure the security of data storage and transmission, and user login logs and operation logs are recorded and retained for one year, providing data backup and fault recovery functions.

[0044] The standards update module detects updates to building water supply and drainage design standards, promptly pushes notifications, and provides new version downloads. The system automatically associates the updated standard clauses for drawing review.

[0045] The project progress control module establishes a progress plan model, links the review progress with the construction progress, monitors the core tube drawing review in real time if it is overdue by 1 day, automatically sends a reminder notification, supports progress plan adjustments and records the reasons for changes, and generates a progress deviation analysis report.

[0046] 2. Application effect data Table 1 is a comparison table of the application performance of the commercial complex project supervision service system: Performance indicators Existing supervision methods This invention system Multi-disciplinary drawing review efficiency 48 hours / batch 8 hours / batch Drawing problem recognition rate 75% 96% Problem rectification closed-loop cycle 72 hours 24 hours Process traceability integrity 60% 100% Supervision report generation time 12 hours 1 hour Table 1 shows that existing supervision methods rely on manual review of drawings and offline communication, resulting in low efficiency in multi-disciplinary review, insufficient problem identification rate, long rectification cycle, fragmented process traceability data, and time-consuming report generation. This invention's system significantly improves review efficiency through AI-assisted collaborative drawing review. The AI-based initial review combined with manual verification mode significantly improves the problem identification rate. Online collaborative workflow shortens the rectification cycle, and the blockchain-enabled traceability module enables full traceability. Automated results management quickly generates standardized reports. The system integrates full-process functions, solving the pain points of low efficiency, poor collaboration, and difficult traceability in traditional supervision services, providing efficient and accurate supervision support for large-scale commercial complex projects.

[0047] Example 2: Application of Supervision Services for Urban Main Road Municipal Engineering This embodiment focuses on an 8-kilometer-long urban main road reconstruction and expansion project, involving four disciplines: road engineering, water supply and drainage pipeline engineering, power pipeline engineering, and communication pipeline engineering. The system of this invention is applied to realize intelligent drawing review, collaborative management, and quality assessment of municipal engineering supervision, and fully implements all functional modules.

[0048] 1. Multi-module collaborative operation The project information filing module receives basic project information entered by the user, including road length, width, design speed, information of participating units, and the scope of supervision services covering drawing review, construction supervision, and quality acceptance. Key time nodes include the completion time of drawing review, road closure for construction, and opening to traffic. It generates a unique project identification code, links it to project approval documents, supervision contracts, urban road engineering construction and quality acceptance specifications, and other materials, and archives them in the project database according to municipal road type. Online querying and modification of project information are supported.

[0049] The intelligent drawing upload and parsing module supports construction units in uploading CAD format road structure drawings, PDF format pipeline layout drawings, and BIM format integrated pipeline models. The system's online preview function allows users to view drawing details. The parsing engine extracts data such as road cross-sectional dimensions and base layer thickness from CAD drawings; it uses OCR technology to identify pipeline diameters and burial depth markings from PDF drawings; and it extracts parameters such as pipeline 3D coordinates and material types from BIM drawings. It automatically identifies drawing versions and modification marks, establishes a link between drawings and projects, stores them in a structured manner, and supports rapid retrieval.

[0050] The AI-powered collaborative drawing review module employs a deep learning model trained on 80,000 compliant municipal engineering drawings. It takes the parsed drawing elements and corresponding municipal engineering technical specifications as input, and introduces a dynamic priority allocation model to calculate the priority of water supply and drainage pipeline drawings. =0.8 indicates that the key parts of the main pipeline are associated with this value. A value of 0.9 indicates a high safety risk. =0.7 means there are many components. =0.6, which means there are 18 drawings pending review. =0.8 indicates an urgent construction period. =1.3, which represents the intersection of multiple disciplines, calculated as follows: The first term = 0.8 × 0.9 × 0.66 ≈ 0.47, and the second term = 0.8 / 1.3 ≈ 0.62. =0.47+0.62=1.09, indicating a high priority. The model automatically identifies issues such as insufficient road base thickness and pipeline intersection conflicts, marks the locations and associated specifications, and generates rectification suggestions.

[0051] In the manual review and annotation module, the drawing review engineer confirmed 15 issues identified by AI, added 3 hidden defects in pipeline interfaces, uploaded the signed review document, and formed a manual review record.

[0052] The drawing review process traceability module records all process data in real time, including drawing upload time, AI review duration, manual review operations, and problem rectification records. It assigns a unique traceability code to each problem, associates it with the responsible personnel, uses blockchain technology to store key data, supports traceability queries and anti-counterfeiting verification, and generates process traceability reports.

[0053] The feedback collaboration module pushes review issues to design and construction units via system messages and SMS. It supports online responses to rectification plans and uploading of modified drawings. The intelligent reminder mechanism sets the reminder frequency according to the rectification period. If the rectification is not completed within the deadline, the reminder will be automatically upgraded. The module also records the feedback content and time nodes.

[0054] The supervision results management module integrates various data, generates review summary reports and phased supervision evaluation reports, automatically calculates core indicators, displays the distribution of pipeline problems through heat maps, supports result export and archiving, and data visualization functions assist in supervision decision-making.

[0055] The system security and operation and maintenance module divides permissions into multiple levels, uses encryption technology to ensure data security, records operation logs and anomaly alarms, and provides data backup and fault recovery functions to ensure stable system operation.

[0056] The standards and specifications update module automatically monitors updates to relevant municipal engineering standards, sends notifications, and provides downloads to ensure that the basis for drawing review is real-time and accurate.

[0057] The supervision quality assessment module constructs a multi-dimensional assessment system and introduces a comprehensive scoring model for supervision service quality: set up =0.9 indicates that the supervision unit has a Class A qualification and a good historical reputation. =4 evaluation indicators, =0.3, which represents the weighting for drawing accuracy. =95 points =0.05 means the deviation is 5 points, =0.2; =0.25, which is the weight of the rectification closure rate. =98 points =0.02, =0.3; =0.25, which is the weight for timely delivery of results. =96 points, =0.04, =0.2; =0.2 represents the satisfaction weight of the participating units. =94 points, =0.06, =0.1. Calculation yields: , =0.9×95.46≈85.91 points, generating a supervision quality assessment report.

[0058] 2. Application effect data Table 2 is a performance comparison table of the urban main road project supervision service system: Performance indicators Existing supervision methods This invention system Supervision service quality rating 72 points 86 points Standardized adaptation accuracy 78% 98% Multi-disciplinary collaboration efficiency 36 hours / batch 6 hours / batch Objectivity of quality assessment Subjective evaluation Quantitative scoring Data security level generally high Table 2 shows that existing supervision methods rely on subjective judgment for quality assessment, suffer from lagging standard adaptation, low efficiency in multi-disciplinary collaboration, and insufficient data security. This invention's system achieves objective scoring through a multi-dimensional quantitative evaluation model, improves adaptation accuracy through real-time updates of standards, significantly enhances efficiency through online collaboration mechanisms, and strengthens data security through encryption and blockchain technology. Addressing the characteristics of municipal engineering projects—numerous pipelines and multiple professional overlaps—the system achieves intelligent drawing review and full-process management, resolving the problems of uneven quality, poor collaboration, and outdated standards in traditional supervision services, providing reliable supervision support for urban main road reconstruction and expansion projects.

[0059] Reference Figure 2 This diagram visually illustrates the efficiency differences between different review modes, highlighting the significant advantages of the system in multi-disciplinary review efficiency. Purely manual review relies on manual verification by each specialty, which is affected by personnel energy and professional barriers, taking up to 48 hours per batch. Single-specialty information-based review only solves the digitization problem of that single specialty; cross-specialty collaboration still requires manual coordination, taking 36 hours per batch. Traditional multi-specialty collaborative review, while achieving basic online workflow, still takes 24 hours per batch without AI-powered intelligent review support. AI-assisted review only completes preliminary identification and lacks a full-process collaborative mechanism, taking 12 hours per batch. The system of this invention, through its AI collaborative review module automatically identifying problems and its opinion collaboration module achieving efficient cross-party communication, reduces time to 8 hours per batch, improving efficiency by 83% compared to purely manual review. This fully demonstrates the value of the system in integrating multi-disciplinary review processes and applying intelligent technology, adapting to the high-efficiency needs of multi-disciplinary cross-review in large-scale projects.

[0060] Reference Figure 3This figure clearly demonstrates the impact of sample size on the accuracy of different AI drawing review models, highlighting the technical advantages of the model in this invention. Traditional CNN models, limited by their feature extraction capabilities, only see their accuracy increase from 70% to 82% when the sample size increases from 20,000 to 100,000. This limited increase and overall low accuracy make it difficult to meet the demands of high-precision drawing review. The improved deep learning model in this invention optimizes the network structure for engineering drawing features, enhancing its ability to extract elements from multiple disciplines. With a sample size of 20,000, the accuracy reaches 82%, continuously improving with increasing sample size, reaching 96% with 100,000 samples, and the growth rate is consistently higher than that of traditional models. This indicates that the model in this invention possesses stronger learning and generalization capabilities, maintaining high accuracy even with a limited sample size, adapting to the drawing review needs of projects of different scales, and addressing the pain points of traditional AI models' excessive reliance on sample size and insufficient accuracy.

[0061] Reference Figure 4 This diagram accurately presents the distribution characteristics of core issues in municipal engineering drawing review, providing data support for system function design and application. Pipeline intersection conflicts account for 35%, making it the primary issue. Due to the diverse types and complex layouts of pipelines in municipal engineering, traditional drawing review methods easily overlook these conflicts. Structural dimensional deviations account for 25%, a critical issue affecting project safety. Conflicts in regulatory clauses account for 20%, stemming from rapid regulatory updates and the difficulty of manual verification. Incompatible material selection accounts for 15%, and other hidden defects account for 5%. This invention addresses the most prevalent issue of pipeline intersection conflicts by extracting 3D coordinates from BIM drawings and automatically identifying conflicts using an AI drawing review model. A regulatory standard update module is set up to synchronize with the latest regulations in real time for conflicts in regulatory clauses, significantly reducing the omission rate of such issues. The visual presentation of the pie chart allows supervisors to focus on high-frequency issues, optimize the key points of drawing review, and improve the accuracy and effectiveness of supervision services.

[0062] Reference Figure 5 This graph clearly reflects the positive correlation between each evaluation indicator and the overall score of the supervision service quality, verifying the scientific nature of the quality evaluation system of this invention. A drawing review accuracy rate of 90 points corresponds to an overall score of 85 points; a rectification closure rate of 95 points corresponds to 90 points; timely delivery of results of 88 points corresponds to 83 points; participant satisfaction of 92 points corresponds to 87 points; and standard adaptation accuracy of 96 points corresponds to 91 points. The positive slope of the linear trend line indicates that the higher the indicator score, the higher the overall score. This invention's supervision quality evaluation module introduces a multi-dimensional quantitative model, incorporating these core indicators into the evaluation system, rather than relying on a single subjective evaluation. The scatter plot visually shows that the standard adaptation accuracy rate has the most significant impact on the overall score, also confirming the important value of the system's standard update module. This graph provides data support for supervision units to optimize service priorities, allowing for targeted improvement of weak indicator scores and comprehensive improvement of service quality.

[0063] The above are merely preferred embodiments of the present invention, but the scope of protection of the present invention is not limited thereto. Any equivalent substitutions or modifications made by those skilled in the art within the scope of the technology disclosed in the present invention, based on the technical solution and inventive concept of the present invention, should be covered within the scope of protection of the present invention.

Claims

1. An AI-powered collaborative drawing review and full-process traceability information management system for supervision services, characterized in that: Includes the following modules: The project information filing module receives relevant information about the supervision project entered by the user, establishes a unique identification code for the project, links it to project establishment documents, contract documents, and technical specifications and standards, and automatically classifies and archives the project by type to form a standardized project information database. The intelligent drawing upload and parsing module supports batch upload and online preview of CAD, PDF and BIM format construction drawings. It extracts the core elements of the drawings through the drawing parsing algorithm and automatically identifies the drawing version number and modification mark. The AI ​​collaborative drawing review module uses deep learning to build a drawing review model. It takes the parsed drawing elements and technical specifications as input, automatically identifies drawing problems, marks related content, and generates rectification suggestions. The manual review and annotation module provides an online annotation tool, allowing reviewers to view the AI ​​review results, confirm, reject, or supplementary explanations for issues, upload review comments and signed documents, and create a manual review record. The drawing review process traceability module records data from the entire drawing review process in real time, assigns a unique traceability code to each issue and associates it with relevant information; The feedback and feedback module establishes a collaborative channel among participating units, pushing review issues and rectification suggestions to the responsible parties, and supporting online responses to rectification plans and uploading of modified drawings; The project supervision results management module integrates relevant drawing review materials, generates standardized supervision reports according to the supervision service specifications, and automatically calculates the core indicators of project drawing review. The system security and operation module sets up a multi-level user permission system, uses data encryption technology to ensure security, and records user logs and alarms for abnormal behavior.

2. The AI-based collaborative drawing review process traceability information management system for supervision services, as described in claim 1, is characterized in that: It also includes a project progress control module, which establishes a progress plan model based on key project time nodes and supervision workflow, links the progress of drawing review, problem rectification and project construction, monitors the progress of each stage in real time, and automatically sends reminder notifications for overdue tasks.

3. The AI-based collaborative drawing review process traceability information management system for supervision services, as described in claim 1, is characterized in that... It also includes a supervision quality assessment module, which constructs a multi-dimensional supervision service quality assessment system. It conducts quantitative assessments from the dimensions of drawing review accuracy, problem rectification closure rate, timeliness of deliverables, and satisfaction of participating units, collects feedback and evaluation scores from participating units, and automatically generates supervision quality assessment reports.

4. The AI-based collaborative drawing review process traceability information management system for supervision services, as described in claim 1, is characterized in that... The AI ​​collaborative drawing review module introduces a dynamic priority allocation model for drawing review tasks, which scientifically sorts the review order of different drawings using the following formula: in Indicates the priority value of the drawing review task. This indicates the importance coefficient of the related engineering parts in the drawing. This indicates the risk impact coefficient that may be caused by problems with the drawings. This indicates the complexity coefficient of modifying the drawing. This indicates the current congestion coefficient of the drawing review queue. This indicates the urgency level of the project schedule. This indicates the professional correlation coefficient of the drawing.

5. The AI-based collaborative drawing review process traceability information management system for supervision services according to claim 1, characterized in that, The intelligent drawing upload and parsing module uses a unified parsing engine for multiple drawing formats. It extracts layer information, block references, and structured data such as dimensions from CAD drawings, recognizes text and graphic elements from PDF drawings using OCR technology, and extracts 3D model parameters and component relationships from BIM drawings. The integrity of the drawings is automatically verified during the parsing process.

6. The AI-based collaborative drawing review process traceability information management system for supervision services, as described in claim 3, is characterized in that... The supervision quality assessment module introduces a comprehensive scoring model for supervision service quality, which quantifies quality using the following formula: in This indicates the overall score for the quality of the supervision services. Indicates the basic quality coefficient. Indicates the first The weight of each evaluation indicator Indicates the first The actual score of each evaluation indicator Indicates the first The deviation coefficient of the evaluation indicators Indicates the first The impact attenuation coefficient of the evaluation index This indicates the total number of evaluation indicators.

7. The AI-based collaborative drawing review process traceability information management system for supervision services, as described in claim 1, is characterized in that... In the drawing review process traceability module, blockchain technology is used to record key traceability data to form a distributed traceability ledger, with each traceability code corresponding to a unique blockchain hash value.

8. The AI-based collaborative drawing review process traceability information management system for supervision services, as described in claim 1, is characterized in that... In the aforementioned opinion collaboration and circulation module, an intelligent reminder mechanism is established. The reminder frequency is set according to the urgency of the problem and the rectification period. Notifications are pushed through system messages, SMS, and emails. Problems that have not been rectified within the time limit are automatically upgraded with reminders and the overdue status is recorded.

9. The AI-based collaborative drawing review process traceability information management system for supervision services according to claim 1, characterized in that, It also includes a standard update module, which establishes a technical standard database covering current national, industry, and local building engineering-related standards and specifications, and supports online querying, retrieval, and association of standards and specifications.

10. The AI-based collaborative drawing review process traceability information management system for supervision services according to claim 1, characterized in that, The supervision results management module integrates data visualization and analysis functions, which intuitively display the review data in the form of bar charts, line charts, and heat maps. It supports custom chart dimensions and query conditions, and automatically generates data statistical analysis reports.

Images