Engineering projects are the main content of investment for heavy asset enterprises. For power supply companies primarily responsible for power grid construction, their engineering projects are numerous and wide-ranging. Engineering project audits are a focal point of high concern for both internal and external stakeholders. Traditional methods of auditing power grid engineering projects often suffer from low quality and efficiency due to various constraints. With the development of big data and artificial intelligence technologies, intelligent auditing has increasingly become a practical necessity for auditing power grid engineering projects. This article discusses how to implement intelligent auditing using the example of audits in power grid enterprises.
1. Content and Current Status of Power Grid Engineering Project Audits
(1) Content of Power Grid Engineering Project Audits
The scope of power grid engineering project audits covers the entire process of internal control and construction management, primarily including pre-project management (including investment approval, surveying, and design), project implementation management (including project cost, procurement management, contract management, material management, and project implementation), as well as project completion and subsequent management (including completion acceptance, final settlement, and post-evaluation of the project, along with economic benefit assessments).
When conducting audits of power grid engineering projects, auditors need to comprehensively verify the following materials to determine whether the engineering projects meet technical standards and specifications, and whether they are compliant, reasonable, and effective. The materials include but are not limited to: contracts, agreements, and other documents related to the power grid engineering project; planning designs, construction drawings, budget proposals, and acceptance materials; approval, acceptance, and verification documents from relevant departments and units involved in the power grid engineering project; technical materials, technical summaries, design calculations, and on-site testing results related to the engineering project; financial materials related to the power grid engineering project, including financial statements, accounting vouchers, total investment plans, and summaries of investment situations; investigation documents and records related to the implementation of the power grid engineering project, including project approvals, acceptance reports, safety inspection records, project visas, etc.
(2) Current Status of Power Grid Engineering Project Audits
Taking the municipal power supply enterprises under the State Grid Corporation as an example, the sources of audit materials for power grid engineering projects are diverse, including both online system data and offline provided site drawings, resulting in complex audit scenarios. Currently, the main difficulties in the auditing process of power grid engineering projects include challenges in data acquisition, data processing, and data verification.
2. Concepts and Methods of Intelligent Auditing
(1) Concept of Intelligent Auditing
So far, there is no clear academic definition of intelligent auditing. The intelligent auditing discussed in this article refers to the use of modern information technology tools such as big data, artificial intelligence, and machine learning to conduct comprehensive and in-depth analysis and mining of a company’s financial and business data, aimed at improving the quality and efficiency of audit work. Compared to traditional auditing, intelligent auditing has advantages such as high efficiency, accuracy, objectivity, comprehensiveness, and early warning, making it an inevitable trend for future audit work.
(2) Methods of Intelligent Auditing
Optical Character Recognition (OCR) is a technology that converts printed text in paper documents or images into digital text. OCR technology can automatically parse the text in scanned paper documents or image files, converting it into a computer-readable format for subsequent processing and storage. In audit work, OCR technology can help auditors quickly and accurately identify text in a large number of paper documents and image files, reducing manual input errors and omissions, thus improving audit efficiency and accuracy.
Robotic Process Automation (RPA) is a technology that uses software robots (or “bots”) to automatically perform repetitive and monotonous tasks in business processes. RPA bots can simulate human actions, accessing, interacting with, and controlling various applications and systems on computers to complete predefined tasks and processes. In audit work, RPA bots can help auditors quickly and accurately complete repetitive data analysis and processing tasks, such as data extraction, data verification, and anomaly detection.
3. Exploration of Intelligent Auditing in Power Grid Engineering Project Audits
(1) Automatic Recognition of Power Grid Engineering Project Document Content Based on OCR Technology
During the implementation of power grid engineering projects, there is a large amount of unstructured data. These project documents may sometimes be in paper form and other times saved as photos, making verification work for auditors quite inconvenient. During the verification process, auditors need to flip through relevant documents and manually enter the information from these documents, converting the content into structured data before conducting statistical analysis, which affects work quality and efficiency.
By using the open-source Tesseract-OCR engine from HP, images and PDF files can be batch converted into editable formats, exploring the realization of automatic recognition of power grid engineering project document content based on OCR technology.
1. Image Correction.
For documents created through scanning or photographing, image tilt often occurs due to manual operations, and tilted images severely affect text recognition accuracy. Therefore, before text recognition, it is necessary to correct and check tilted images.
2. Content Recognition.
(1) Content recognition of common image files such as JPG and PNG.
For JPG, PNG, and other common image formats, Tesseract-OCR can be used directly for text recognition in the images.
(2) Content recognition of TIF format files.
TIF files generally contain multiple scanned images. During content recognition, each image’s content needs to be recognized separately before merging the recognized content.
(3) Content recognition of PDF format files.
PDF files can use the third-party library PyMuPDF to convert PDF files page by page into images, then recognize the content of each image.
(2) Automatic Review of Power Grid Engineering Project Audit Materials Based on RPA Technology
1. Review of Completeness of Audit Materials.
Conducting a preliminary review of the audit materials for engineering projects is an important aspect of the power grid engineering project audit process. The documents for each stage of the power grid engineering project have relatively fixed formats, which can narrow the scope of information recognition and reduce the difficulty of data capture, providing convenience for automated review by bots.
The automatic review of power grid engineering project audit materials based on RPA technology involves pre-configuring various types of audit material databases for power grid engineering projects. The construction unit of the power grid engineering project can customize the standardized audit material package required for the project audit based on the project type. This approach meets common management needs while also accommodating differentiated management requirements.
For example, for a common 0.4kV distribution network engineering project, the audit materials for settlement review include but are not limited to: design drawings, budget documents, completion drawings, settlement documents, construction (completion) reports, acceptance reports, actual consumption tables for materials supplied by the owner, engineering visa documents, concealed works acceptance records, design change orders, construction contracts, etc.
The reviewer places the audit documents into the corresponding folders, and the RPA bot can automatically check their completeness. If any audit documents are found to be incomplete, feedback will be provided to the relevant auditors in a timely manner, allowing them to promptly identify and collect missing documents.
Auditors only need to select the folder containing the audit materials for the power grid engineering project, and they can use the program to review the audit materials item by item according to the customized audit material package list, analyzing and determining whether the submitted materials meet the audit requirements. The review process is automatically recorded, and the results can be generated instantly, including but not limited to: the names and quantities of documents that have passed the audit, the names and quantities of documents that have not passed the audit, and the final audit conclusions.
2. Review of Accuracy of Audit Materials.
Most of the data in the audit materials have certain logical relationships in terms of time or amounts. By presetting certain audit rules, the RPA bot can clarify the correct relationships between various data and automatically review for logical errors in the data based on the recognized structured information.
(1) Correspondence of data within the audit material package.
Reviewing the correspondence of time sequence, amounts, etc., within the audit material package.
(2) Correspondence of audit materials with system data.
Based on the completion of the correspondence review of the data within the audit material package, the RPA bot will also review the correspondence of the audit materials with system data.
4. Effectiveness of Intelligent Auditing in Power Grid Engineering Projects
(1) Multi-source Data Input
Through OCR technology, images, scanned documents, and other unstructured materials in the audit documents of power grid engineering projects can be continuously and automatically recognized, ensuring that multi-source data can be input and utilized under intelligent conditions, providing strong support for subsequent audit analysis phases, big data comparisons, and auditing issue identification.
(2) Rapid Automatic Operation
In the past, auditors could only complete a limited amount of preliminary review work on power grid engineering project audit materials each day and often struggled to ensure that all content was adequately addressed. RPA bots have made automatic auditing of power grid engineering project data possible. After auditors configure the review logic rules, the RPA bot runs automatically and can quickly provide feedback on inspection results, significantly improving the quality and efficiency of audit work while reducing the workload of auditors.
(3) Quality of Service
Intelligent auditing of power grid engineering projects is a result of continuous exploration and attempts in a rapidly developing intelligent environment, targeting the characteristics of large, numerous, complex, and scattered investments in power grid engineering. Through intelligent auditing of power grid engineering projects, it supports the digital transformation required by power grid enterprises and safeguards the lean construction and management of power grid engineering projects.
5. Outlook for Intelligent Auditing in Power Grid Engineering Projects
In the future, with the continuous development and application of information technology, intelligent auditing of power grid engineering projects will present new and more possibilities. The goals of full-process, full-coverage, and high-efficiency audits will gradually become a reality. At the same time, with the continuous improvement of data quality and scale, intelligent auditing of power grid engineering projects will also achieve data sharing, risk control, and visualization of audit results, which will greatly promote the modernization process of auditing power grid engineering projects and continuously enhance the quality and efficiency of intelligent auditing work.
This article is excerpted from the July 2023 issue of China Internal Audit.
Authors: Zheng Hongyi, Liu Xia, Zhao Weiming, Nie Tongyun, Wang Meili
Affiliation: State Grid Shandong Electric Power Company, Binzhou Power Supply Company
Editor: Sun Zhe
Currently, over 160,000 people have followed us; what are you waiting for?
Click “Read the Original” to charge up together.