SYSTEM DYNAMICS MODELING FOR CONSTRUCTION MANAGEMENT RESEARCH: CRITICAL REVIEW AND FUTURE TRENDS

. As a result of growing complexities in the construction industry, system dynamics modeling (SDM) has been increasingly used in construction management (CM) research to explore complicated causal relationships at the various levels of construction and management processes. Given the rapid growth of SDM applications over the past two decades, a systematic review is needed to ascertain the state of the art and further trends in the area. This paper provides the results of a systematic analysis of 103 papers from 41 selected peer-reviewed journals from 1997 to 2016. The contributions of the papers are first analyzed, structured and formulated in terms of the year of publication, software involved, the combined use with other methods, and research design. With the assistance of the a keyword co-occurrence network analysis, eight research topics involving different internal and external complexities are identified, including: (1) sustainability, (2) project planning and control, (3) performance and effectiveness, (4) strategic management, (5) site and resource management, (6) risk analysis and management, (7) knowledge management, and (8) organization and stakeholder management. The analysis results reveal the pivotal role of SDM in streamlining different complicated casual relationships at the activity, project, and industry levels across the eight topics and its significant potential in uncovering the impact of complicated contextual conditions on project planning and control, effectiveness and performance, strategic management, and sustainability at the project and industry levels. Lastly, trends and recommendations for SDM applications are provided for future CM research. This paper provides a state of the art of SDM in CM applications and insights into opportunities and useful references for the future.


Introduction
Over the past two decades, system dynamics modeling (SDM) has been increasingly applied in construction management (CM) research to explore the feedback and interaction of factors in construction activities, as shown in Figure 2, reported later. A similar challenge is faced by construction engineering and management as increasing construction scales and ever-changing construction environments have increased the complexity and process dynamics of construction activities and projects (Ogunlana, Li, & Sukhera, 2003). Theoretically, these construction systems can be further conceptualized as a collection of complex and dynamic interdependent components, including multiple feedback processes and nonlinear re-lationships (Ozcan-Deniz & Zhu, 2016). SDM provides a powerful way to model the relationships in complicated construction systems, and explore feedback processes and their causal relationships. Such method is pivotal to secure construction efficiency and performance.
From a systems perspective, the construction system involves an accumulating and evolving behavioral process of on-site progress and resources that need to provide response feedback to maintain required performance levels (Lyneis & Ford, 2007). SDM has an extremely strong ability to deal with behavioral factors of construction systems and their interactions with construction processes (Rodrigues & Bowers, 1996), especially for those involved in a Review article megaproject delivery. This advantage cannot be obtained from most traditional modeling methods (e.g.-work breakdown structures, Gantt charts, PERT/CPM networks, project crashing analysis and trade-off analysis) owing to their limited functions (Love, Holt, Shen, Li, & Irani, 2002).
Despite the significance of SDM in addressing construction-related problems and the care needed in the correct use of this technique, no critical review has yet been undertaken to provide a systemic understanding of its use in CM research. Accordingly, this study conducts a comprehensive review of SDM applications through a robust content analysis of peer-reviewed journal papers. Through a comprehensive literature review, this study aims to: (1) ascertain the annual productivity of SDM research published in leading CM journals between 1997 and 2016, (2) uncover the topic coverage of SDM research regarding different complexity issues in CM, (3) explore the combined use of SDM with other methods and the main research design involved, and (4) identify future directions of SDM applications in CM.

SDM applications in CM research
SDM is a modeling method used to explore and understand a complex system in a holistic manner by ascertaining its feedback structures and resultant behavior (Sterman, 2000). The method was first proposed in 1958 by Forrester at the Massachusetts Institute of Technology (MIT) to explore the dynamics of industrial systems. In the past two decades, SDM has evolved into a pivotal approach to modeling the behavior and characteristics of complex systems in terms of internal feedback (Sterman, 1992(Sterman, , 2000Yuan & Wang, 2014). The method can solve macro-level problems while avoiding micro-level fragmented details (Sterman, 2000;Ko & Chung, 2014). Thus, it is suitable for handling multi-level complex systems (e.g., modern corporations and social organizations) (Ko & Chung, 2014).
SDM has developed into an important area in the construction engineering and related fields in preference to quantitative system modeling methods due to its combination of systems theory and computer simulation (Lyneis & Ford, 2007;Lee, Han, & Peña-Mora, 2009;Moradi, Nasirzadeh, & Golkhoo, 2015). It can incorporate the technical, organizational, human, and environmental factors involved in dynamic system processes while simulating the behavior of major outputs of a system over time (Lyneis & Ford, 2007). Enabled by the rapid development of SDMrelated software in the last two decades, SDM has been increasingly applied in construction engineering areas, such as transportation engineering (Shepherd, 2014), mining engineering (Yang, Love, Brown, & Spickett, 2012), and gas engineering (Hu, Zhang, Ma, & Liang, 2010).
SDM is suited to CM research because of its ability to analyze the interrelationships and feedbacks that exist within complex systems (Sterman, 2000(Sterman, , 2001Thomas, Menassa, & Kamat, 2016). It can deal with issues in systems with highly dynamic complexity, derived from inter-actions between interrelated components that evolve over time, such as stocks and flows, time delays, nonlinearities, and feedback loop structures (Terouhid & Ries, 2016). In addition, the project organization can be viewed as a complex system (Sheng & Lin, 2018), and SDM can make sub-systems interrelated to pursue and reach project goals (Love et al., 2002). Thus, SDM can be widely used in CM research owing to the characteristics of the construction process and its organization. With the help of SDM modeling, construction project managers can react appropriately to changes, and understand how they influence the behavior of the entire project system (Love et al., 2002). An increasing number of studies have suggested the use of SDM in current CM research, such as planning project activities, managing construction risks, and identifying the causes of rework (Rodrigues & Bowers, 1996;Love, Mandal, & Li, 1999;Love, Mandal, Smith, & Li, 2000;Love et al., 2002;Nasirzadeh, Afshar, & Khanzadi, 2008a).

Data collection
CM research is a combination of multiple disciplines covering technical and managerial topics (Xiong, Skitmore, & Xia, 2015). Thus, a comprehensive search method for SDM journal papers is necessary. Meanwhile, the identified papers were necessarily selected through a rigorous process as the foundation of research.
In stage 1, to ensure search comprehensiveness, the Scopus database is used as the search source because it includes all the leading CM journals and other journals publishing CM papers (Ke, Wang, Chan, & Cheung, 2009;Xiong et al., 2015;Zhang, Chan, Feng, Duan, & Ke, 2016b). In this study, only journal papers were selected for analysis, whereas book reviews, editorials and conference papers were excluded. Journal papers usually provide more comprehensive and higher-quality information than other types of publications, and most reviews of CM studies solely focus on journal papers (Zhao, 2017). Adopting the same process as Xiong et al. (2015), the key search words of "system dynamics" and "construction" were used. The full search code is as follows: (TITLE-ABS-KEY ("system dynamics") AND TITLE-ABS-KEY (construction) AND LANGUAGE (english)) AND DOCTYPE (ar OR re) AND SUBJAREA (mult OR ceng OR chem OR comp OR eart OR ener OR engi OR envi OR mate OR math OR phys OR mult OR arts OR busi OR deci OR econ OR psyc OR soci) (time: 2017-12-21). The above search identified 373 potential SDM papers in CM from 1997 to 2016. Then, another review of these papers was carried out in terms of the following criterion: (a) all papers are peer-reviewed journal papers related to CM issues, and (2) the SDM is a main research method used. This review reduced the number of papers to a final number of 103. These were closely examined to reveal the trends and topic coverage of SDM-based papers to obtain a holistic picture of the SDM applications involved.

Data analysis
A hybrid research methodology was adopted to achieve four purposes, respectively. First, annual productivity analysis was conducted to understand the basic trend and changes of SDM research between 2007 and 2016. This method has been commonly advocated used in previous reviews regarding CM issues (Zheng, Le, Chan, Hu, & Li, 2016;Hu, Xia, Skitmore, & Chen, 2016). Second, a topic coverage analysis was undertaken to determine the major topics involved, which may reveal the application potential of SDM. CiteSpace software, one of the most frequently used scientometric technique, was used to assistant in analyzing the keyword co-occurrence network of the papers (Hosseini et al. 2018), which has been increasingly advocated in recent reviews (e.g. Zheng et al., 2016;Zhao, 2017;Hosseini et al., 2018). Keywords are direct expression of the core content of papers and provide a clue to the topic covered (Hosseini et al., 2018). Third, the combined use of SDM with other advanced methods were evaluated, which may reflect latest trend in the application of SDM to strong and reliable findings. Last, an in-depth analysis was undertaken in the research design of the identified 103 papers.
The overall research process and procedure is illustrated in Figure 1.

Annual productivity of the SDM research
As shown in Figure 2, although fluctuating in the process, the number of SDM-based journal papers increased since 1997. This finding is particularly apparent over the last decade, showing that the SDM is becoming more popular and has wide application in CM. Table 1 shows the number of papers published each year between 1997 and 2016. The majority are contained in leading CM journals (see Chau, 1997 Table 1 further shows the article distribution and percentage of identified papers in these journals.

Complexity-related topic coverage
CiteSpace 5.1 R8 was used to visualize the keyword co-occurrence network as shown in Figure 3. The results involve 104 nodes and 110 edges. The network's density is 0.0211, and thus it is regarded as a sparse network. The node size refers to the frequency of keyword co-occurrence and the edge refers to the two different keywords that together reflect a main theme. The 10 most frequently occurring keywords are "simulation", "project management", "performance", "Hong Kong", "waste management", "policy", "feedback", "demolition waste", "rework", and "sustainability". The high frequency keywords also include the United States and China, which indicates the popularity of SDM in CM research in these countries.   Each of the identified papers was thoroughly examined to identify its main research topic by using the scientometric technology. In case that a paper involved more than one topic, only one topic with the best fit was selected (Hu, Chan, Le, & Jin, 2015;Zheng et al., 2016) in combination with the keywords identified through the co-occurrence network. Considering SDM has a distinct advantage in dealing with CM complexity, especially with risk management and organizational issues (Luo, He, Jaselskis, & Xie, 2017). Accordingly, this study made an evaluation of the topic coverage of selected SDM topics in terms of project complexity classification For the definition and classification of complexity (internal and external), see Hu et al. (2015). Finally, on the basis of the results of the keyword co-occurrence network analysis, a total of eight topics were identified, namely (1) sustainability; (2) project planning and control; (3) performance and effectiveness; (4) strategic management; (5) site and resource management; (6) risk analysis and management; (7) knowledge management; and (8) organization and stakeholder management. Figure 4 shows the trend in topic coverage in the four time-slices from 1997 to 2016. Table 2 shows the complexity analysis results of SDM in CM.
Sustainability attracts the most research attention with 26 papers involved, which has significantly increased from 2012 to 2016. Based on the research content analysis, from 2007 to 2011, sustainability mainly focused on waste and demolition management. For example, Yuan, Shen, Hao, and Lu (2011) used SDM to streamline internal relationships in evaluating the costs and benefits of construction and demolition waste management. From 2012 to 2016, in addition to the constant themes of the internal complexities of waste management from material deterioration (Thomas et al., 2016), the availability of resources (Ozcan-Deniz & Zhu, 2016), and waste generation and recycling (Yuan & Wang, 2014), SDM was expanded to carbon emission reduction (J. Sim & J. Sim, 2016), nuclear power development (X. Guo & X. Guo, 2016), prefabrication (Li, Shen, & Alshawi, 2014), and the interactions between management complexity and macro-policy complexity were examined by SDM.
Project planning and control, mainly relating to cost/ schedule control and change management, is the second most popular topic. The number of papers on this topic has been relatively stable over the four 5-year periods. These studies often use SDM to explore the complicated relationships between internal activities and their feedback in solving problems (e.g., Sing, Love, Edwards, & Liu, 2016;Parvan, Rahmandad, & Haghani, 2015) as shown in Table 2. Chapman's (1998) study piloted the application of SDM in this area by examining the decrease in design productivity because of staff changes. From 1997 to 2001, the main SDM application focus in this area was on dynamic planning and construction rework. For example, Pena-Mora and Park (2001) and Pena-Mora and Li (2001) developed SDM-based dynamic planning methods for fast-tracking building construction projects, whereas Love et al. (1999) explored the causal structure of construction rework influences. SDM was gradually extended to web-enabled SDM (Lee, Pena-Mora, & Park, 2006), early-warning and forecasting systems (Huang & Wang, 2005), cash flow management strategies (Cui, Hastak, & Halpin, 2010), decision making from an integrated system   (Motawa, Anumba, Lee, & Peña-Mora, 2007), and learning (Love, Edwards, & Irani, 2008). From 2012 to 2016, SDM was widely used in addressing new concepts (e.g., lean production) and with an emphasis on CM issues at the macro level. For example, a new design workflow based on SDM was proposed, using lean concepts to smoothen design work, reduce unnecessary design errors, and increase design reliability (Ko & Chung, 2014). Schedule delays and cost overruns in design and construction projects at the macro level were systematically evaluated with the assistance of a SDM-based model for analysis (Han, Lee, & Pena-Mora, 2012). Performance and effectiveness is ranked third, with 19 papers involved, with SDM helping to optimize management plan changes and policy making to enhance project performance in a real-world setting (Moonseo & Peña-Mora, 2003). Chasey, Garza, and Drew (1997) first used SDM to evaluate the capacity of infrastructure maintenance systems in addressing their complex nature involving multiple feedback loops. Subsequently, an increasing number of studies were devoted to this topic, especially those related to performance at the organizational level. For example, Ogunlana (2003a, 2003b), for example, examined how to develop and implement improvement policies to enhance organizational performance from a senior manager's perspective based on a SDMbased case study of a publicly listed construction organization. Ogunlana et al. (2003), on the other hand, used SDM to integrate engineering processes and local influencing factors to simulate the performance of construction firms. SDM was also used for performance evaluation at the project level, such as the performance of construction enterprise resource planning systems (Tatari, Castro-Lacouture, & Skibniewski, 2008), the effectiveness of contractors' bidding management (Lo, Lin, & Yan, 2007), construction quality management (Nasirzadeh, Khanzadi, Afshar, & Howick, 2013), labor productivity (Nasirzadeh & Nojedehi, 2013), the impact of design rework on project performance (Li & Taylor, 2014), and the effectiveness of contractors' green innovation (Hsueh & Yan, 2013). Based on these studies, both the impacts of internal and external complexities on organizational/project performance can be streamlined using SDM.
Strategic management is ranked fourth, with 13 papers involved. As construction projects are a man-made goaloriented open system, they tend to be unpredictable and changeable. The complexities of construction projects and their environments trigger the disruptive effect of subjective human factors, which cannot be addressed solely by the experience of individuals, and SDM provides a systematic understanding of the strategic issues involved (Rodrigues & Bowers, 1996). SDM applications on this topic are grouped under the macro and micro levels. At the macro level, prior research focuses on government and company value-engineering policies (Park, Ahn, Lee, & Yoon, 2012), urban-land use policy changes (Wu, Zhang, & Shen, 2011), strategies for the design-build industry (Park, Ji, Lee, & Kim, 2009), and strategic-operational CM (Peña-Mora, Han, Lee, & Park, 2008), which emphasizes on the exploration of the external complexities of project management. At the micro level, SDM is mainly used to analyze internal complexities, such as design strategies for construction waste minimization (Wang, Li, & Tam, 2015), managing the consequences of cost overruns and schedule delays (Peña-Mora et al., 2008), and the complex interactions of construction operations . Therefore the SDM can be widely in this topic, whether in the macro or micro levels, to analyze the strategic goals of construction projects by connecting the external and internal complexities.
Site and resource management, which is related to safety and resource management, is ranked fifth with 10 papers involved. SDM applications in this area mainly involved dynamic resource management (Park, 2005), equipment management (Prasertrungruang & Hadikusumo, 2009), construction operation management , and safety management (Han, Saba, Lee, Mohamed, & Peña-Mora, 2014). In particular, the topic of safety management is related to an integrated method for safety performance in construction operations , workers' safety attitudes and behaviors (Shin, Lee, Park, Moon, & Han, 2014), and construction safety behavior patterns (Guo, Yiu, & González, 2015) from a systems perspective. "Systems thinking" is widely accepted as an effective tool to conceptualize a group of complexity factors and their dynamics of safety management (Guo et al., 2015). The papers involved in this topic are more focused on the internal complexities (8 of 10 papers) of construction activities, especially for safety management (e.g., Guo et al., 2015;Shin et al., 2014;Han et al., 2014).
Risk analysis and management is an important area for SDM applications. As a large number of risk factors usually involve complex casual-loop relationships, SDM is very suited to analyze such relationships, with a number of risk analysis treatments, such as tunnel construction risk analysis (Wang, Ding, Love, & Edwards, 2016), and risk allocation between owners and their contractors (Nasirzadeh, Khanzadi, & Rezaie, 2014). As shown in Table 2, the SDM applications in this topic mainly focus on the risks factors related to internal complexities.
Knowledge management received only minor attention with four papers involved. In an early application, SDM was used in a learning organization to represent singleloop learning and double-loop learning (Senge, 1991). From 2000, knowledge management has become a major topic in SDM papers, such as in strategic learning in a dynamic environment (Chen & Fong, 2013) and experience transfer in complex learning systems (Lê & Low, 2009), especially combined with organization behavior (Bajracharya, Ogunlana, & Bach, 2000).
The topic of organization and stakeholder management involves only two SDM-based papers. These are concerned with organizational capabilities and excellence (Terouhid & Ries, 2016) and employee work-family conflict management (Wu, Duan, Zuo, Yang, & Wen, 2016), both involving internal complicated causal relationships. Organization adaptability is the ability to adapt to changes in contextual complexity (Luo et al., 2017), and thus future SDM applications may be well directed at this issue.
The discussions above indicate that the SDM papers involve internal and external complexity issues across all eight topics. Table 2 reveals a distinct advantage of SDM in handling CM complexity issues. Two topics of project planning and control, and performance and effectiveness were maintained across the four time periods of 1997-2016 as shown in Figure 4. This finding indicates that dominating methodological role of SDM accepted by researchers in these areas. In addition, 71% of the identified 103 papers involved internal complexity issues in CM, especially with site and resource management, risk analysis and management, knowledge management, project planning and control, performance and effectiveness, and organization and stakeholder management. SDM application has an emphasis on external complexity issues involving industrial policy, regarding sustainability and strategic management.

Combined use of SDM with other methods
Some studies have attempted to combine SDM with other methods such as networks, fuzzy logic, discrete event simulation (DES), and agent-based simulation to provide complementary support by solving the complex problems involved (Rodrigues & Williams, 1998).
The imprecise and uncertain nature of many CM factors means that the traditional deterministic SDM may not always be an appropriate modeling tool (Nasirzadeh et al., 2008a;Nasirzadeh, Afshar, Khanzadi, & Howick, 2008b;Nasirzadeh et al., , 2014. Fuzzy logic is sometimes integrated into the SDM structure to account for this, forming a fuzzy-SDM model to accommodate the uncertainties involved . The application of Za-deh's extension principle and interval arithmetic has been proposed for the SDM to enable the system outcomes to take into account uncertainties in the input variables (Nasirzadeh et al., 2014). The fuzzy-SDM model is widely used for risk management (Nasirzadeh et al., 2008a(Nasirzadeh et al., , 2008b owing to the imprecise and uncertain nature of risks involved. Data processing methods are often combined with SDM as collected data cannot be used directly. Several methods, such as factor analysis (Zhang et al., 2016a), time-series forecasting models, regression techniques, judgement techniques (Sing et al., 2016), and AHP (Perng, Hsia, & Lu, 2007), have been used to formulate raw data to provide the necessary data for the SDM.

Categories of research designs of the SDM papers
Identifying complexity factor and its constructs is the basis of SDM applications in CM research. Meanwhile, research process objectives and it's the level of analysis is another research design factor. CM research using SDM faces the following dilemma (Sterman, 1992): if a simple model is built, then it is criticized for ignoring real-world complex relationships; if a complex model is built instead, then it is criticized for being a 'black box' that no one can understand or check it's working. Thus, the model building process, especially research process objectives, depends on the necessity and practicality of the research (Richardson, 1986). Therefore, balancing complex factors (complexity) and practicality (suitable level of analysis) is significant for SDM applications in CM. An in-depth analysis was conducted to evaluate the design of the identified SDM papers in Table 3. This finding shows that most selected SDM papers focus on the activity (62%) and project levels (18.5%).The industry-level papers only involve three topics: sustainability, performance and effectiveness, and strategic management. These topics do not only need SDM to explore micro-level but also emphasize its macro research objectives. Of the eight topics involved, only strategic management is more focused on the industry-level. Although SDM is a quantitative method for dealing with complex problems in CM, it can also be used to provide an analysis and explanatory framework to explore the complex CM relationships. SDM is used as a qualitative method to explain behavioral logic in CM for 16 of the selected papers (e.g., Park et al., 2012).
Computer simulation is an indispensable part in SDM research and can overcome the limitations of mental models as it: (i) is explicit and their assumptions are open to review; (ii) is able to interrelate many factors simultaneously; and (iii) can be carried out under controlled conditions, allowing analysts to conduct experiments that are not feasible or ethical in the real system (Love et al., 2000). The early computer languages for SDM were SIMPLE and DYNAMO from the 1950s. These languages were transformed into software owing to the development of the Windows operating system, the most famous software still being STELLA/iThink, Vensim, Powersim, and AnyLogic from the 1990s. Vensim (52 papers, 50.49%) and iThink (30 papers, 29.13%) are the most popular in our sample, with Vensim being especially widely used because of its friendly interface and ease of use (Wang et al., 2015).

Evaluation of current research and future directions
On the basis of the earlier discussions, a dual-dimensional framework involving the types of complexity and unit of analysis is proposed to assess previous SDM research and identify future directions in CM. SDM is commonly used to tackle internal complexity issues in CM at the activity and project levels (Boxes I and III), which are evidenced by the majority of the papers (68/103) as shown in Figure 5. Early SDM applications were in the area of project rework and then spread to other topics, such as project planning and control, site and resource management, risk analysis and management, and knowledge management.
Exploring internal complexity issues at the activity level does not only involves traditional topics (e.g., project planning and control, and performance and effectiveness), but also several emerging topics, such as sustainability, as shown in Box (I). These applications indicate the strong ability of SDM in ascertaining the complicated causal relationships of activities at the micro level.
Boxes (II) and (IV) indicate that another major application of SDM, as evidenced by 14 papers involved, is in uncovering the impact of contextual complexity (e.g. policy changes, industrial norms, and law requirements) on project activities and arrangements, especially with the topics on project sustainability, performance, and effectiveness (e.g., Onat, Egilmez, & Tatari, 2014;Tam, Li, & Cai, 2014). In particular, SDM has been regarded as a leading method for investigating the impact of external policies on sustainability issues (9/14) (Levitt, 2007). Recognizing the dynamic and changeable nature of built environments faced by current CM practice, SDM will have a significant potential to make more contributions in these areas.
Box (III) refers to SDM applications exploring internal complexities at the project level, especially those different project types involved, such as infrastructure (Sing et al., 2016), construction , and power (Ford, 2001). Although the number of papers (13/103) in this domain is not large, these studies are quite dispersed with five topics involved. This number revealed the ability of SDM to conduct cross-type analysis Box (VI) refers to the 16 SDM papers that explored external complexities at the industry level. Compared with the SDM studies in Box (I), these papers mainly focus on the influence of external factors (e.g., new policies and technology) on the industry (e.g., Zhang et al., 2016a;J. Sim & J. Sim, 2016;X. Guo & X. Guo, 2016;Tang & Ogunlana, 2003a). The number of papers is small. However it can be an important direction due to such rapid changes at the Figure 5. A dual-dimensional framework for evaluating SDM-based research in CM industry level, such as new technology (BIM, artificial intelligence and big data) and procurement policies (publicprivate partnerships).
Direction (V) refers to SDM applications in addressing internal complexities at the industry level with five papers involved. These papers mainly look at the impact of macro industrial policies on such micro construction activities as sustainable building developments (Yan, 2015), the design-build model on strategies (Park et al., 2009), and environmental policies on waste reduction management (Ding, Yi, Tam, & Huang, 2016). These findings reveal the ability of SDM to uncover the interactions and relationships between contextual conditions and micro construction activities. Based on the results above, SDM is widely accepted because of its essential ability in streamlining complicated casual relationships at the activity, project and industry levels and its significant potential in uncovering the impact of contextual complicated conditions on project planning and control, effectiveness and performance, strategic management, and sustainability at the project and industry levels. Recognizing the increasing complexity of the economic, natural, and social environments faced by construction projects and the relatively limited applications of SDM, these areas will represent a significant potential for future SDM applications.

Conclusions
This systematic review indicates that SDM has been increasingly advocated by researchers over the past two decades to explore nonlinear and dynamic complexity issues involved in CM. The sampled 103 papers indicate that prior efforts can be grouped into eight topics, of: (1) sustainability, (2) project planning and control, (3) performance and effectiveness, (4) strategic management, (5) site and resource management, (6) risk analysis and management, (7) knowledge management, and (8) organization and stakeholder management. In addition, the review results show that these SDM applications involve the use of a mixed method, combining network analysis, fuzzy logics analysis, discrete event simulation, and agent-based simulation. SDM is identified as not only an important approach for streamlining complicated causal relationships across topics at the activity, project, and industry levels, but also for its significant potential in uncovering the impact of contextual complicated conditions in project planning and control, effectiveness and performance, strategic management and sustainability. The research findings present both a holistic knowledge map of SDM applications in CM and insights into opportunities and useful references for future applications of SDM in CM research.