VIET NAM NATIONAL UNIVERSITY HO CHI MINH CITY HO CHI MINH CITY UNIVERSITY OF TECHNOLOGY HUNG VIET NGUYEN DEVELOP A DATA-DRIVEN APPROACH UNDER THE INTEGRATION OF 4D VISUALIZATION AND PROCESS MINING TO SIMULATE, DIAGNOSE AND PREDICT REAL- WOLRD CONSTRUCTION EXECUTION Major: Construction Management Major code: 8580302 MASTER’S THESIS HO CHI MINH CITY, June 2024 THIS THESIS IS COMPLETED AT HO CHI MINH CITY UNIVERSITY OF TECHNOLOGY – VNU-HCM Supervisor 1: Dr. Cuong Viet CHU Supervisor 2: Assoc. Son Vu Hong PHAM Examiner 1: Assoc. Sy Tien DO Examiner 2: Dr.
Viet Thanh NGUYEN This master’s thesis is defended at HCM City University of Technology, VNU-HCM on June 20, 2024. Master’s Thesis Committee: (Please write down full name and academic rank of each member of the Master’s Thesis Committee) 1. Long Hoai LE - Chairman 2. Minh Nhat HUYNH - Member, Secretary 3.
Sy Tien DO - Reviewer 1 4. Viet Thanh NGUYEN - Reviewer 2 5. Beng Tiang QUEK - Member Approval of the Chairman of Master’s Thesis Committee and Dean of Faculty of Civil Engineering after the thesis being corrected (If any). CHAIRMAN OF THESIS COMMITTEE DEAN OF FACULTY OF CIVIL ENGINEERING Dr.
Long Hoai LE Assoc. Tuan Anh LE VIETNAM NATIONAL UNIVERSITY-HO CHI MINH CITY SOCIALIST REPUBLIC OF VIETNAM HO CHI MINH CITY UNIVERSITY OF TECHNOLOGY Independence – Freedom – Happiness THE TASK SHEET OF MASTER’S THESIS Full name: Hung Viet NGUYEN Student code: 2270380 Date of birth: February 07, 1999 Place of birth: HCMC, Vietnam Major: Construction Management Major code: 858032 I. THESIS TITLE: Develop a data-driven approach under the integration of 4D visualization and process mining to simulate, diagnose and predict real-world construction execution. Phát triển cách tiếp cận dựa trên dữ liệu dưới sự tích hợp của trực quan hóa 4D và khai thác quy trình để mô phỏng, chẩn đoán và dự đoán việc thực hiện xây dựng trong thế giới thực.
TASKS AND CONTENTS: Optimization in construction project by proposing an approach as follows: ▪ Plan: Utilizing Synchro 4D to generate as-planned IFC from as-designed IFC and construction timetable. ▪ Monitor: Utilizing LIDAR technology to collect as-happened point cloud on site during execution phase then mapping with as-planned IFC to get the list of missing elements thanks to the RAAMAC service on BIMserver. ▪ Diagnose: Utilizing BIMserver Eventlog Service to parse information from as-planned IFC then manually mapping with the list of missing elements to scrutinize the punctuality of each task. After that, using Disco by Fluxicon to exploit the as-happened log to evaluate the bottlenecks in the current state of the construction progress.
▪ Improve: Utilizing Statgraphics and Matlab to build a predictive model (ARIMA) to forecast the amount of activities could be completed in the future. THESIS START DAY: January 15, 2024 IV. THESIS COMPLETION DAY: May 20, 2024 V. SUPERVISOR: Supervisor 1: Dr.
Cuong Viet CHU Supervisor 2: Assoc. Son Vu Hong PHAM Ho Chi Minh City, May 20, 2024 SUPERVISOR 1 SUPERVISOR 2 Dr. Cuong Viet CHU Assoc. Son Vu Hong PHAM HEAD OF DEPARTMENT Dr.
Long Hoai LE DEAN OF FACULTY OF CIVIL ENGINEERING Assoc. Tuan Anh LE i ACKNOWLEDGEMENT Embarking on the journey of completing my master's thesis in Construction Management as part of the International Master Programs at the Department of Civil Engineering, Ho Chi Minh City University of Technology, has been an enriching and formidable experience, made possible through the collective support and guidance of many. At the heart of this journey have been Dr. Chu Viet Cuong and Assoc.
Pham Vu Hong Son, my advisors, whose invaluable mentorship, deep expertise, and constant encouragement have been pivotal in shaping the high quality of my work. Their patience and thoughtful guidance have not only advanced my thesis but also profoundly impacted my personal and professional growth. I'm equally grateful to the faculty of the Department of Civil Engineering, whose wisdom and support have deepened my understanding of the field, enriching my academic and research capabilities. This journey was further enhanced by the camaraderie and insight of my peers, whose shared experiences and discussions added invaluable dimensions to my learning process.
The supportive staff and librarians at the University have provided essential resources and assistance, facilitated my research and contributed significantly to my thesis. Above all, the unwavering love and support of my family have been the cornerstone of my achievements, providing me with the motivation and strength to pursue my academic goals. Their endless encouragement has been my anchor throughout this academic endeavor. To all who have been part of this journey, my heartfelt appreciation for your indispensable role in my achievements and for fostering an environment of growth, learning, and success.
Hung Viet NGUYEN ii ABSTRACT In a dynamic shift toward the digitalization of the construction industry, this research heralds a novel data-centric methodology that merges the temporal capabilities of 4D simulation with the analytical power of process mining to enhance the management and execution of construction projects. The primary objective of this research is to develop an integrated framework that combines real-world data from Internet-of-Things (IoT) devices and Building Information Modeling (BIM) with advanced data mining techniques, aiming to improve both the simulation and analysis of project workflows. The study’s findings reveal that this approach significantly refines the prediction and management of construction risks, enabling the accurate forecasting of number of completed activities per day through the use of ARIMA model. Additionally, the research highlights the establishment of a robust diagnostic mechanism designed to identify bottlenecks and deviations from planning (BIMserver and Disco by Fluxicon).
By applying this novel framework to a practical case study in Søborg, Denmark, the research not only validates the effectiveness and adaptability of the proposed methodology but also contributes valuable insights that bridge the gap between temporal simulation and process analysis. This innovative approach offers a substantial advancement in both academic research and industry practices, paving the way for enhanced efficiency and performance in construction project management. Keywords: 4D Simulation, Process Mining, Construction Project Management, Data- centric Methodology, Predictive Models, Risk Mitigation iii TÓM TẮT LUẬN VĂN THẠC SĨ Trong bước chuyển mình năng động hướng tới sự số hóa ngành xây dựng, nghiên cứu này giới thiệu một phương pháp mới tập trung vào dữ liệu, kết hợp tiềm năng của mô phỏng bốn chiều và sức mạnh phân tích của khai thác quy trình để nâng cao việc quản lý trong giai đoạn thực hiện thi công của dự án xây dựng. Mục tiêu chính của nghiên cứu này là phát triển một khung tích hợp dữ liệu trong thế giới thực từ các thiết bị IoT và mô hình thông tin tòa nhà (BIM) với kỹ thuật khai thác dữ liệu tiên tiến, nhằm cải thiện đồng thời sự mô phỏng và phân tích luồng công việc của dự án.
Kết quả nghiên cứu cho thấy phương pháp này cải tiến đáng kể việc dự đoán và quản lý rủi ro trong xây dựng, cho phép dự báo chính xác số lượng công tác hoàn thành mỗi ngày thông qua việc sử dụng mô hình dự đoán ARIMA. Ngoài ra, nghiên cứu nhấn mạnh việc thiết lập một cơ chế chuẩn đoán mạnh mẽ được thiết kế để xác định các điểm nghẽn và sai lệch so với kế hoạch (BIMserver và Disco by Fluxicon). Bằng cách áp dụng phương pháp mới này vào một dự án thực tế ở Søborg, Đan Mạch, nghiên cứu không chỉ xác nhận tính hiệu quả và khả năng thích ứng của phương pháp đề xuất mà còn đóng góp những hiểu biết có giá trị giúp thu hẹp khoảng cách giữa mô phỏng thời gian và phân tích quy trình. Cách tiếp cận sáng tạo này mang lại sự tiến bộ đáng kể trong cả nghiên cứu học thuật và thực tiễn công nghiệp, mở đường cho việc nâng cao hiệu quả và hiệu suất trong quản lý dự án xây dựng.
Từ khóa: Mô phỏng 4D, Khai thác Quy trình, Quản lý Dự án Xây dựng, Phương pháp Tập trung vào Dữ liệu, Mô hình Dự đoán, Giảm thiểu Rủi ro iv AUTHOR’S COMMITMENT The undersigned below: Student full name: Hung Viet NGUYEN Student ID: 2270380 Place and date of born: Ho Chi Minh City, Vietnam, February 07, 1999 Address: District 03, Ho Chi Minh City With this declaration, the author finishes his master’s thesis entitled “DEVELOP A DATA-DRIVEN APPROACH UNDER THE INTEGRATION OF 4D VISUALIZATION AND PROCESS MINING TO SIMULATE, DIAGNOSE AND PREDICT REAL-WORLD CONSTRUCTION EXECUTION” under the advisor's supervision. All works, ideas, and materials that were gained from other references have been cited correctly. Ho Chi Minh City, May 20, 2024 Hung Viet NGUYEN v TABLE OF CONTENTS ACKNOWLEDGEMENT. ii TÓM TẮT LUẬN VĂN THẠC SĨ .iii AUTHOR’S COMMITMENT.
iv TABLE OF FIGURES. viii TABLE OF TABLES. x LIST OF ABBREVIATIONS. xi CHAPTER 1: INTRODUCTION.
Structure of the study. 6 CHAPTER 2: LITERATURE REVIEW. Definitions and concepts. Construction planning and monitoring.
Implementation of BIM and IT systems in construction management 14 2. Industry Foundation Classes (IFC). Internet of Things (IoT). Optimization in Construction.
Time series prediction. 36 CHAPTER 3: ARCHITECTURE OF PROPOSED APPROACH. Difference between proposed approach and Deming cycle. 39 CHAPTER 4: MODEL IMPLEMENTATION AND VALIDATION.
Case study – Vindspor Højde Residence. As-planned IFC generation. As-happened monitoring. Diagnosis of current state of construction progress.
Prediction of future state of construction progress. 52 CHAPTER 5: DISCUSSION AND RECOMMENDATION. ANNEX A: Construction schedule of case study. ANNEX B: Workflow of event-log-generation.
ANNEX C: Detailed illustration of process map. 74 viii TABLE OF FIGURES Figure 2. Example of a Gantt chart (Tory et al. Example of a PERT network (Soroush, 1994).
Flowline illustrating four tasks and demonstrating the impact of delays (Kenley and Seppanen, 2009). Activity on Arrow (Tarigan, 2021). Activity on Node (Tarigan, 2021). The history of digital twin (Lucchi, 2023).
Direct Cost, Indirect Cost, and Total Cost in Construction (Hegazy, 2002). Optimization in Exact Method (Tarigan, 2021). Deming cycle (Sokovic et al. Architecture of event logs (Aalst and Mining, 2011).
Proposed approach inspired by Deming cycle. Real-world project captured in February 2024. As-designed 3D model captured in Enscape. As-designed architectural drawing of ground floor plan.
As-deigned architectural drawing of typical floor plan. As-designed architectural drawing of elevation front view and side view. As-designed architectural drawing of main cross-section. Workflow demonstration in planning phase.
As-planned IFC models from week 22 to week 24 in 2019 (demonstrated by Synchro 4D). Workflow demonstration in monitoring phase. Comparison of as-planned IFC and as-happened point cloud. A portion of task-centered-workflow diagram demonstrated by Disco by Fluxicon.
A portion of worker-centered-workflow diagram demonstrated by Disco by Fluxicon. Illustration of the original data. Illustration of stationary data after the first-order. ACF plot for stationary data after the first-order difference.
PACF plot for stationary data after the first-order difference. Plot of the predicted value and true value in entire dataset. Residual errors in testing set. Residual errors in training set.
Illustration of residual distribution in training set. Illustration of residual distribution in testing set. 58 x TABLE OF TABLES Table 2. Performance of progress monitoring solutions (Kopsida et al.
Illustration of an event log created from an as-planned IFC (Author’s elaboration). Collection of missing entities in accordance with comparison of as-planned IFC and as-happened point cloud in week 22. A portion of an as-scheduled log generated from as-scheduled IFC. A portion of an as-happened log with additional attribute of punctuation.
Summary of data for analysis. Goodness-of-fit analysis for five ARIMA models. Coefficient of ARIMA (1, 1, 1) model. Evaluation of prediction from developed algorithm.