Digital twin (DT) can achieve real-time information fusion and interactive feedback between virtual space and physical space. This technology involves a digital model, real-time information management, comprehensive intelligent perception networks, etc., and it can drive the rapid conceptual development of intelligent construction (IC) such as smart factories, smart cities, and smart medical care. Nevertheless, the actual use of DT in IC is partially pending, with numerous scientific factors still not clarified. An overall survey on pending issues and unsolved scientific factors is needed for the development of DT-driven IC. To this end, this study aims to provide a comprehensive review of the state of the art and state of the use of DT-driven IC. The use of DT in planning, design, manufacturing, operation, and maintenance management of IC is demonstrated and analyzed, following which the driving functions of DT in IC are detailed from four aspects: information perception and analysis, data mining and modeling, state assessment and prediction, intelligent optimization and decision-making. Furthermore, the future direction of research, using DT in IC, is presented with some comments and suggestions. This work will help researchers gain in-depth and systematic understanding of the use of DT, and help practitioners to better promote its implementation in IC.
The concept of IC has been described by the Global Engineering Frontier 2018 [
With increasing advances in the construction industry, Industry 4.0 proposed a new construction technology to achieve high adaptability, rapid design changes using IT, and a more flexible technical workforce training. The advanced construction technologies employed include CPS [
The concept of DT was described as an information-mirror model by Grieves [
Number | Definition | Refs. |
---|---|---|
1 | DT is digital copies of biological or non-biological physical entities. By bridging the physical and virtual worlds, data are seamlessly transferred, allowing virtual entities to exist simultaneously with physical entities. | Abdulmotaleb [ |
2 | DT is an integrated multi-physics, multi-scale, probabilistic simulation of completed vehicles or systems that use the best physical models, sensor updates, fleet histories, etc., to reflect the life of their corresponding flying twin. | Glaessgen et al. [ |
3 | A coupled model of real machines running on a cloud platform that uses a combination of data-driven analysis algorithms and other available physics knowledge to simulate health conditions. | Lee et al. [ |
4 | Real-time optimization using digital copies of physical systems. | Soderberg et al. [ |
5 | The dynamic virtual representation of a physical object or system throughout its life-cycle, using real-time data to achieve understanding, learning, and reasoning. | Bolton et al. [ |
6 | DT shares physical, virtual and interactive data between them to map all components in the product life-cycle. | Tao et al. [ |
Previously, it was a challenge to present the state of an IC system in real time. However, the emergence of DT has solved this problem [
This study summarizes the IC driven by DT from the perspective of sustainable development. First, we introduce the application of DT, based on five stages: planning, design, manufacture, installation, operation, and maintenance. Second, we introduce the DT-driven IC from four aspects: information perception and interaction, data mining and modeling, state assessment and prediction, intelligent optimization and decision.
IC includes the relevant equipment, processes, and systems that support each other, as shown in
Professor Michael Grieves of the University of Michigan proposed the concept of DT in 2003. This technology was first used by NASA to simulate and analyze the flying spacecraft in the Apollo project [
The use of DT in planning and design is mainly based on the research on DT planning and design methods, which make it more efficient. This includes digital planning, design, and simulation.
Wang et al. [
In the field of construction engineering, there have been rapid advances in technology from the development of three-dimensional (3D) computer-aided design to BIM. BIM-related technology is the core technology for DT [
In the design stage, the adaptability and superiority of the simulation can be simulated using a DT model. Haag et al. [
The use of DT in construction engineering is different from that in the manufacturing industry. After establishing a virtual simulation model entity, based on a DT, it is not only applied to the processing and assembly of prefabricated components, but also used to optimize and manage the entire construction process to achieve accurate process control. It includes process simulation, digital production line, condition monitoring, and so on.
(1) Processing of assembly component
Before the processing of assembly components, the virtual simulation method can be used to carry out an in-depth design and process simulation of components, and to comprehensively analyze the rationality of assembly component design, processing rate, and efficiency. A new method was proposed for resource supply and demand-matching manufacturing, based on complex networks and the IoT, to realize intellectual perception and access to manufacturing resources [
(2) Simulation of assembly process
With the development of IT, the influence of traditional construction technology has become increasingly profound, especially with the deepening of complex construction technology. Through the CPS simulation of the DT model, the construction assembly scheme is determined and optimized to be safe, more efficient, and reasonable. Ellinger et al. [
All the elements of the production stage are integrated into a closely coordinated process through digital methods to achieve an automated production process. By means of a 3D virtual simulation model of an industrial robot, a DT model was established to control the robot for the automatic assembly of large-scale spacecraft components [
The production process can be monitored visually by collecting real-time operation data of the production equipment. Abnormal equipment must be dealt with and adjusted in time to optimize the production process. Botkina et al. [
The use of DT in maintenance management is based on the virtual simulation model of DT. The real state of the physical model is fed back to the virtual simulation model through the signal data of the IoT devices. When the physical model fails, the virtual simulation model will synchronously produce faults. Therefore, the virtual simulation model can judge whether the physical model will produce faults according to the real-time state data, which can effectively reduce the failure rate. This part includes fault warning, maintenance, and management-scheme optimization.
By reading the real-time parameters of the sensors or control systems on the building components, a visual remote monitoring model is established to analyze the state of the building components with AI. A maintenance strategy to reduce the loss is proposed, and an early warning is given. Based on DT, fault detection and health management of different parts or areas can be realized. According to the state of the virtual model’s data, the operating mode of the building components is optimized, the safety risk is reduced, while the stability of the building structure is improved.
Combined with DT and virtual reality, a new application mode based on a DT is formed. In the manufacturing industry, Wang et al. [
Research progress in DT has led to continuous development in the medical, sports, and other fields. In the medical field, Groth et al. [
Compared with other industries, the development of IT in the construction industry has been relatively slow because of the complexity of buildings. However, with the concept of digital earth, proposed by IBM, the smart city system has been incorporated into all aspects of daily city management. From the investigation of the use of DT-concept roads, the deployment of a DT box, and the feedback of real-world, real-time data flow through the IoT, equipment has become the key component for solving the problem of automatic driving vehicles (intelligent mobile devices) [
(1) Optimization of the production line
By analyzing the status data of the production line, the configuration parameters of manufacturing instruments are modified to ensure the performance and quality of products, as well as to optimize the production index. Tao et al. [
(2) Optimization of management decision
In the context of Industry 4.0, simulation-based decision-support tools, commonly called DT [
In recent years, the construction industry has been experiencing low production, high accident rates, labor shortage, and rising costs every year. With the rapid development of digital and intelligent technology, IC is considered the mainstream direction of development in the industry. However, owing to the uniqueness of construction products, the non-repeatability and fragmentation of the construction process, and the complexity of the site environment, IC still faces many difficulties in being implemented. Therefore, research on DT-driven IC has become a popular trend, and has achieved good results in intelligent design [
IC is not a technology oriented to a single production stage, but a highly integrated multi-link construction system (as shown in
The construction industry’s development has evolved in stages: from human force to mechanical power construction 1.0 (prefabricated building), standardized modular assembly-line construction 2.0 (industrial building), and CAD + CNC/CAM + BIM informatization construction 3.0 (digital building). It has come to the fourth Industrial Revolution led by intelligent manufacturing in the construction industry. Therefore, the world’s traditional and emerging construction powers have proposed the development of IC planning.
Country/organization | Intelligent-construction development plan | Refs. |
---|---|---|
Germany |
Industry 4.0 | [ |
South Korea |
Manufacturing Industry Innovation 3.0 Strategy | [ |
European Union |
Horizon 2020 | [ |
Japan |
Industrial Value Chain Initiative | [ |
United States |
Advanced Manufacturing Partnership | [ |
Unites Kingdom |
Modern Industrial Strategy | [ |
France |
New France Industrial | [ |
China |
Made in China 2025 | [ |
In this study, the system of IC is divided into five intelligent layers: connection, analysis, network, cognition, and execution layers. The main function of the connection layer is intelligent perception—that is, the ability to obtain information and data from different parts of the construction industry at different stages through various sensors. It also includes the ability to recognize the environment, object categories, and attributes through memory, learning, judgment, reasoning, and other processes. The main function of the intelligent analysis layer and the intelligent network layer is data mining, which transforms the perceptual information into a 3D information model and uses big-data technology to mine the data needed by the collaborators. The main function of the intelligent cognitive layer is to evaluate and make decisions, that is, using the real-time state data of buildings to evaluate and make optimal decisions. The main function of the intelligent execution layer is to provide feedback-entity control, that is, to provide a feedback optimization scheme for construction equipment through an intelligent device, so as to realize automatic construction with the machinery. This section focuses on the IC system based on CPS, and introduces the development of intelligent construction from four aspects.
An effective AI system is based on its perception, memory, and thinking ability, as well as learning, adaptive, and autonomous behavior abilities. With the ability of dynamic intelligent perception in complex scenes, we need to use multisource information-fusion technology to collect and fuse both similar and different types of sensor information across time and space. Only through memory, learning, judgment, and reasoning can we achieve the purpose of recognizing the environment, object categories, and attributes. On this basis, we can enable experience-based judgment and the intelligent processing of decision-making.
In the process of using information, the first step is to obtain accurate and reliable information; sensors are the main means of obtaining such information in the fields of science and production. A sensor can feel the measured information and transform it into an electrical signal or other required information output according to certain rules to meet the requirements of information transmission, processing, storage, display, recording, and control. It has the characteristics of miniaturization, digitization, intellectualization, multi-functionality, systematization, and networking [
In basic subject research, the sensing has a more prominent position. The progress of modern science and technology has led to the development of many new fields. In addition, various kinds of extreme-technology research play an important role in deepening the understanding of materials, and in developing new energy and materials such as ultra-high and ultra-low temperatures, ultra-high pressure and vacuum, and super-strong and ultra-weak magnetic fields. Without appropriate sensors, it would be impossible to obtain significant information that cannot be directly obtained by the human senses. The first obstacle in many basic scientific studies lies in the difficulty of obtaining object information, and the emergence of new mechanisms and high-sensitivity detection sensors certainly leads to breakthroughs in this field. The development of sensing is often the pioneer of some frontier disciplines.
Sensors have already been used in many fields such as industrial production [
Sensors are also widely used in the construction industry. Traditional applications have mainly been for structural safety monitoring [
The automation and intelligence of construction are inseparable from the application of industrial robots, which are widely used in all aspects of construction. As the construction industry has developed, the types and functions of industrial robots have become more diverse. With the shift from the process of traditional construction to IC, industrial robots have ushered in new and greater development.
In recent years, more and more researchers have focused on industrial robots to make them more competitive. With higher precision, fewer errors, more reasonable structure, more convenient programming, and friendlier human-computer interaction, industrial robots have become increasingly important in industrial applications. Liu et al. [
The important technical means of modern intelligent-sensing systems is to obtain sufficient sensing information and feature information generated by the systems. The information from various sensors has different characteristics, and one of the important tasks of intelligent sensing is to extract the various characteristics of objects from various types of sensor information. The urban driving environment-collection method was developed using a road side unit camera and vehicle Global Positioning System with two sensors, and also the Dempster Shafer theory to map the credibility root to the credibility map [
Data mining has attracted significant attention in the information industry. The main reason is the explosion of data; there is an urgent need to transform these data into useful information and knowledge that can then be used in various applications, including business management, production control, market assessment, engineering design, and scientific exploration. Data mining uses ideas from the following areas: (1) sampling, estimation, and hypothesis-testing from statistics; (2) search algorithm, modeling technology, and learning theory of AI, pattern recognition, and machine learning. Data mining has also rapidly accepted ideas from other fields, including optimization, evolutionary computing, information theory, signal processing, visualization, and information retrieval. Some other areas, especially in database systems, need to provide effective storage, index, and query-processing support, which also play an important role.
Data mining is a popular topic in the field of AI and databases. It refers to the process of revealing hidden, previously unknown, and potentially valuable information from a large amount of data in a database. Data mining is a type of decision-making process. It is mainly based on AI, machine learning, pattern recognition, statistics, databases, visualization technology, etc. It can analyze the data of enterprises automatically, perform inductive reasoning, and extract potential patterns to help decision-makers adjust market strategies, reduce risks, and make correct decisions. The common methods of data mining include classification, regression, clustering, association rules, features, change and deviation, and web-page mining. The difference among these methods indicates the mining of data from different angles.
Classification is used to determine the common characteristics of a group of data objects in the database and to divide them into different classes according to the classification mode. Its purpose is to map the data items in the database to a given class through the classification model, which is used to predict the discrete class of data objects. Classification technology has been applied in many areas including the classification of building components, the evaluation of material properties and characteristics, structural safety assessment, and health prediction. The main classification methods are the decision tree [
Regression is a statistical prediction model used to describe and evaluate the relationship between dependent variables and one or more independent variables. It reflects the temporal characteristics of the attribute’s values in the transaction database, generates a function that maps data items to a real-value predictive variable, and determines the dependency relationship between the variables or attributes. The main research issues include the trend characteristics of the data series, the prediction of the data series, and the correlation among the data. The regression method has been widely used in construction engineering [
Clustering divides a group of objects into several categories according to their similarity and differences. The similarity between the objects in each category is high. However, the similarity between objects in different categories is low, which means that the difference is obvious. The difference in classification is that clustering does not rely on a given category to divide objects. The algorithm is divided into the following: (1) partition method, (2) hierarchical method, (3) density-based method, (4) grid-based method, and (5) model-based method. It can be applied to the classification of smart city management, management-background big-data decision, and management-scheme trend prediction [
Association rules describe the relationships between two data items in the database. We can infer the occurrence of another thing from the occurrence of one thing, which is the association or mutual relationship hidden in the data, so as to better understand the development law of things. In the construction industry, it can be applied to association rules for anomaly detection and reliability evaluation of intelligent devices [
Feature extracts feature expressions from a set of data in a database, which express the overall characteristics of the dataset. The purpose of feature selection is to extract useful information from massive amounts of data to improve the efficiency of data use. The selection and evaluation of feature validity include probability theory, mathematical statistics, and information theory. For example, in structural damage detection, prediction can be carried out according to the features [
Deviations are small objects in a dataset. Generally, deviation objects are called exceptions or outliers. The purpose of deviation is to identify objects that differ from most other objects. For example, by analyzing abnormal cases in the classification, abnormal mode, observation results, expected deviation, and other information [
Through web mining, we can use the web to collect information about politics, the economy, policy, science and technology, finance, various markets, competitors, supply-and-demand information, customers, and so on. This study focuses on analyzing and processing information about external environment and internal operation that has significant or potentially significant impacts on an enterprise, and identifies various problems and possibilities in the process of enterprise management, according to the results. In order to identify, analyze, evaluate, and manage crises, we need to analyze and deal with the information that web mining gathers.
The construction industry has the largest amount of data at the largest scale. With the development and popularization of big data and BIM, building industrialization involves transforming the original “design-site construction” into “design-factory manufacturing-site assembly”. The construction of a building is also a product-manufacturing process. The purpose of industrialization is to make the building-construction process the manufacturing process of the manufacturing industry [
BIM technology provides a good technical platform for the early construction of industrialization projects, and later management and maintenance of them. Using BIM technology to establish a library of layout, the assembly component library of the industrialization building can standardize property types and components, reduce design errors, and improve drawing efficiency, especially in the processing and on-site installation of prefabricated components [
With the impact of big data, cloud computing, IoT, GIS, mobile internet, and other advances in IT the integration of social resources will be the main problem the construction industry must overcome. BIM enables the integration of cross-border resources in the construction industry, and provides a valuable tool to find the optimal resource integration. The application of the IoT, based on BIM, will give every brick and piece of rubble their own identity, and an unprecedented value in their own posts [
State assessment is a method for estimating the internal state of a dynamic system, based on available measurement data. The data obtained by measuring a system’s input and output can only reflect its external characteristics; the dynamic law of the system needs to be described by the internal state variables (usually not measured directly). Therefore, state estimation is of great significance for understanding and controlling an IC system. Taking the 5-1 District of Shizong City as an example, a digital dual model for simulating soil resistance was established using GIS technology, and a prediction system for earthquake liquefaction and foundation-damage amplification was established, while liquefaction risk-assessment was also carried out [
Decision theory aims to achieve an optimal decision by assuming an ideal decision-maker has complete information, accurate calculation, and complete rationality. The practical application of this normative model is called decision, and its goal is to help people make better decisions with its tools and methodologies. Bayram [
The intelligent optimization method is a very active research field, developed in recent years. Many scholars and students of systems engineering, automation, computer, management engineering, mining, machinery, and other majors have widely used intelligent optimization methods. For example, the genetic, tabu search, simulated annealing (SA), ant colony, and particle-swarm optimization have been widely used in various industries of the national economy.
In the field of construction engineering, Buitrago et al. [
The development of IC technology will phase out traditional construction technology with its heavy environmental burden. With the increasing attention to the concept of sustainability, the sustainable development of the construction industry has been clearly positioned. This study proposes a framework of DT-driven IC, as shown in
Sustainability has become an important topic [
Traditional modeling and simulation technology is an indispensable tool and method for the construction industry. However, because it is only an independent unit, modeling and simulation are not suitable for an IC system. DT contains not only model simulation, but also dynamic simulation. The data interaction between digital and physical objects must be realized, so that the DT operation can add value by continuously improving industrial applications. Compared with the business environment, the interaction between uplink and downlink data needs to consider periodicity, data interface, and information modeling in order to improve efficiency.
DT, based on engineering, construction, and maintenance, includes the three main components of an IC system, covering the whole building life-cycle business from product design and construction to service. From the virtual engineering design to the real construction site, and then to building operation, DT always serves every stage of the building life-cycle.
Through blockchain, we can reduce logistics costs, trace the production and delivery processes of goods, and improve the efficiency of supply-chain management. Through the hierarchical structure of a scattered network connected by nodes, blockchain can realize the comprehensive transmission of information in the whole network and test the accuracy of information, which improves the convenience and intelligence of IoT transactions, to a certain extent [
Significant greenhouse gas emissions have led to global warming. Environmental protection is a major concern in the global construction industry [
Humans and machines can interact more collaboratively in IC, which enables a human to make efficient and effective decisions.
With the implementation of IC, the accuracy of construction, component quality, and processing efficiency will continue to improve. With the help of DT and big data technology, a virtual simulation model of an entity model can be established to produce feedback for the entity state in real time. It can also be used for real-time health detection, safety diagnosis, and maintenance of structures.
IT has developed rapidly since 2000. In recent years, AI has also rapidly developed in the fields of medicine, monitoring, and interaction, thereby greatly changing people’s lifestyles. In the future, IC processes will inevitably reduce human factors. Therefore, the application of AI will be the future research direction for IC, gradually replacing the role of humans with robots.
IoT enables all objects that can perform independent functions to access the network, to achieve both interconnection and the effect of interconnection of all things. The combination of IoT and IC realizes the interconnection between prefabricated components and construction parts.
With the transformation and upgrading of the construction industry, sustainable IC has become increasingly important. The combination of IC and DT, which have intelligent sensing and simulation functions, makes the building-production process more efficient and intelligent. At the same time, it can monitor the status of the structure and prefabricated components in real time, as well as predict potential safety hazards. After introducing a DT and its application, this paper introduced four aspects of sustainable IC driven by DT: information perception and interaction, data mining and modeling, condition evaluation and prediction, and intelligent optimization and decision-making. A framework for sustainable IC, driven by a DT, was proposed. The development direction of digital dual-drive sustainable intelligent construction was also discussed.