Dynamic Simulation of Enterprise-Level Strategic Choices in Intelligent Construction: Integration of Evolutionary Game Theory and System Dynamics

Abstract

The decision-making regarding the development of intelligent construction in construction enterprises is crucial for the transformation and upgrading of the construction industry. This paper constructs an evolutionary game model among construction enterprises and applies system dynamics for simulation analysis of the game model. It explores the impact of key factors on the strategy choices of the game participants. The research findings indicate that the initial state of construction enterprises’ willingness to transition to intelligent construction in the evolutionary game model influences the final stable strategy. Direct benefits, the strength of government incentives, penalty intensity, and reduced costs through joint transition positively affect the probability of construction enterprises implementing intelligent construction, while incremental transition costs and positive spillover effect are negatively correlated. When the direct benefit rate exceeds 5%, costs are jointly reduced by more than 2%, and transition costs are below 35 CNY/m², it can significantly motivate enterprises to adopt intelligent construction. A certain level of government incentives (at least greater than 5 CNY/m²) has a positive effect on the transformation process; however, once the incentives exceed 10 CNY/m², their impact stabilizes. Penalties only affect the speed at which the system evolves toward a stable point. Current policy incentives do not require further enhancement. Meanwhile, reducing incremental transition costs is more effective than increasing the intensity of government incentives. The research conclusions contribute to the quantitative analysis of how changes in different key factors affect the dynamic evolution of strategy adjustments by construction enterprises over time, thereby providing corresponding recommendations for transformation and upgrading.

1. Introduction

The construction industry is one of the important pillars of China’s economy. However, due to rising labor costs and changes in the international market environment, the traditional development model, which relies on rapid expansion through scale, has become unsustainable. Common challenges in the industry include low resource utilization, lagging informatization and intelligent development, and insufficient innovation vitality [1,2]. At the same time, emerging technologies represented by artificial intelligence, big data, industrial internet, internet of things, and digital twins continue to break through application boundaries, empowering various fields to transition and upgrade towards digitalization, intelligence, and sustainability [3]. The development of these information technologies is gradually changing the organizational structure and production methods across industries, leading to an increasing demand for informatization in the construction industry. This has given rise to intelligent construction, which has become a new productive force and an important development trend in the global construction industry’s transformation and upgrade [4,5]. Intelligent construction is a new construction model centered on informatization technologies such as cloud computing, BIM, internet of things, big data, and artificial intelligence. It aims to achieve information integration and efficient collaboration throughout the entire industrial chain of engineering projects, spanning the entire lifecycle of design, production, construction, and operation [6,7,8,9]. Developing intelligent construction is an important pathway to drive transformation in the construction industry, becoming a crucial lever to enhance building quality, efficiency, and sustainability capabilities [3,10,11,12,13].

Although the Chinese government has issued specialized strategic plans and guidelines, positioning the development of intelligent construction as a key element for advancing new urban construction and enhancing urban quality at the national level. However, due to the cost sensitivity and profit orientation of construction enterprises in the market [14], their willingness to transform towards intelligent construction is easily influenced by economic interests. The most significant factors are direct benefits and incremental transformation costs [15,16]. Currently, due to significant technological investments and weak platform foundations, the incremental costs during the initial promotion and practice process are high, leading to insufficient intrinsic motivation for transformation. Based on policy experiences in recent years in areas such as corporate carbon reduction, electric vehicle promotion, green building, and prefabricated construction, the establishment of government reward and punishment mechanisms has proven to be very effective [17]. As the advocate and promoter of intelligent construction transformation, the government primarily employs two types of policies: incentives and penalties. To stimulate the interest of construction enterprises in transformation, pilot programs and subsidies have been established as part of the current policy framework. Additionally, the transformation process is also affected by other factors, which indicates that the intelligent construction transition is a repeated, long-term, and dynamic evolutionary process. Ultimately, this leads to the formation of an intelligent construction technology innovation system that is led by enterprises, guided by the government, and involves the coordination of upstream, midstream, and downstream players, as well as collaboration among large, medium, and small enterprises [18]. As key stakeholders in the transformation of the construction industry towards intelligent construction, it is essential to explore the strategic choices of construction enterprises influenced by critical factors such as direct benefits, incremental transformation costs, and government policies during the implementation of intelligent construction models. This exploration can effectively address the main contradictions associated with the transformation process. This paper constructs an evolutionary game model among construction enterprises to visually illustrate the dynamic evolution of strategic choices and uses Vensim simulation to study the impact of various key factors on construction enterprises, ultimately proposing recommendations for advancing intelligent construction development in China based on the simulation results. As the development stages progress, this research approach can provide targeted recommendations for promoting the transformation to intelligent construction at different stages of development.

Previous research has focused on the more macro and broad applications and significance of intelligent construction technology, examining stakeholders such as the government, construction enterprises, and suppliers, without considering the construction enterprise as the fundamental unit of the industry. In the early stages of technological development, enterprise decisions are often based more on internal evaluations rather than external environments, necessitating a focus on the proactive engagement of core entities. Therefore, this paper will separate the government from the stakeholder group and include it as an external policy factor in the model. Additionally, past static analysis methods may overlook the dynamic processes of decision-making in the transformation of construction enterprises and the interaction mechanisms among them. The evolutionary game method is particularly suitable for characterizing issues of bounded rationality and long-term dynamic evolution, making it a favorable theoretical tool for studying the transformation of construction enterprises towards intelligent construction. Thus, building on previous research, this paper will establish a horizontal evolutionary game model among construction enterprises from their perspective to study the evolutionary process of implementing intelligent construction models. Furthermore, simulation will be used to visually represent the stability process of the model. Finally, the paper will propose policy recommendations suitable for the transformation of construction enterprises towards intelligent construction in China.

This paper’s main contribution is to shift the government from being a stakeholder to an external policy factor within the model, focusing on the strategic evolution process among construction enterprises. Additionally, it quantitatively analyzes the impact of different key factors on the willingness of construction enterprises to implement intelligent construction, determining the threshold values of key indicators, which can help in setting work objectives for both construction enterprises and the government.

The remainder of this study is organized as follows: Section 2 reviews the existing literature and highlights the contributions of this paper. Section 3 outlines the primary research methods employed, presents the assumptions of evolutionary game theory, and constructs the model. Section 4 utilizes numerical simulations to analyze and compare results based on varying parameters. Section 5 discusses the insights gained from the model and simulations. Finally, Section 6 concludes the study and provides recommendations for future actions.

2. Literature Review

2.1. Research on the Development of Intelligent Construction

As a new technology and management approach for the transformation and upgrading of traditional construction models, the concept of intelligent construction originates from smart cities (SC). It aims to achieve sustainable urban development through the application of new technologies, thereby enhancing the quality of life for residents [19,20,21]. Inspired by the concepts of smart earth and smart cities, the notion and application of intelligent construction have gained attention in academic research, government policies, and construction practices [22]. By integrating advanced computer technologies such as Building Information Modeling (BIM), cloud computing, big data, the Internet of Things (IoT), blockchain, and Geographic Information Systems (GIS), intelligent construction connects various fields with smart equipment and devices (such as intelligent production lines for steel structures, construction robots, and smart tower cranes). It applies digital construction organization management and practical application theories to provide development directions and ideas for the transformation and upgrading of traditional job sites [11,12,23]. Research on intelligent construction mainly focuses on technology applications and development strategies.

The application of intelligent construction technologies is primarily reflected in three levels. First, there is the integration of new-generation information and communication technologies with construction management systems. Building Information Modeling (BIM) is widely used in project management, and the integrated application of BIM, parametric design, generative artificial intelligence, multi-objective optimization, and visual collaborative platforms is regarded by developed countries as the technological entry point and foundation of intelligent construction [24,25,26]. By combining VR, sensors, the Internet of Things (IoT), Geographic Information Systems (GIS), and BIM, applications have emerged in areas such as quality safety monitoring, construction safety management, building disaster prediction, and asset management [27,28,29]. A common framework of “BIM + GIS + IoT” has been established to achieve a three-dimensional spatial positioning and data-driven construction management model, resulting in safe, efficient, and high-quality construction [30,31,32,33,34]. Germany has developed an intelligent construction site system (SCS) centered on “perception-control-feedback”. It utilizes sensors, RFID, and environmental monitoring equipment to collect real-time data on temperature, humidity, noise, and other environmental factors, combined with edge computing and cloud platforms for dynamic monitoring and intelligent interaction. At the same time, German companies use BIM collaborative platforms to achieve schedule simulation, path optimization, and resource conflict prediction. The United States, on the other hand, emphasizes multi-technology integration and platform-based construction management, linking information on schedule, safety, personnel, equipment, and costs through BIM and platforms like Procore and PlanGrid into a visual interface to support multi-party collaboration [3]. Second, there is the intelligent upgrading of construction equipment and processes, utilizing construction robots and intelligent machinery to achieve human–machine collaborative operations. Japanese companies, such as Shimizu Corporation, have built standardized and modular component industrial systems, deploying automated welding robots and digital management platforms for production in factories. Construction robots are widely used in steel structure welding, panel installation, and material transportation. Third, there is the transformation of construction methods. The collaborative development of intelligent construction and industrialized building is a key pathway for the transformation and upgrading of the construction industry [35].

Intelligent Construction Technology (ICT) breaks through traditional construction models, and its foresight and uncertainty often lead companies to face the decision-making dilemma of “high expectations-low adoption-difficult sustainability” [36]. Due to the inability of various stakeholders to obtain short-term benefits in the early stages of development, investments are likely to be rejected [37]. Consequently, scholars have also studied the main stakeholders in intelligent construction and their roles, identifying stakeholders as core elements in promoting the development of intelligent construction. In the process of engineering construction, differing interests among various stakeholders can create both cooperative and adversarial relationships. High levels of coordination among stakeholders are necessary to facilitate project advancement, and differentiated strategies are particularly important during this process [38]. Feng et al. [39] argue that Smart Construction Sites (SCS), as complex systems involving multiple stakeholders, rely heavily on the participation strategies of the government, enterprises, and project parties. Therefore, a promotion system for SCS should be established, led by enterprises, guided by the government, and involving project parties. Wang [40], Tang [41], and Ye [42] analyze the game evolution process from different perspectives, focusing on the interactions between “government-developer”, “developer-integrated service provider”, and “construction enterprise-technology supplier”. They argue that the incentive policies of the government and developers can significantly enhance the enthusiasm of organizational members. However, a stable equilibrium strategy is only possible if developers pursue maximum profit. Goerzig D et al. [43] found that the higher the demand for digitalization among small and medium-sized enterprises, the greater their dependence on service providers; conversely, this also holds true. Miao et al. [44] argue that in the context of weak digital foundations within companies, the construction of a digital ecosystem relies on support from third-party platform services. Lu analyzes the behaviors of participants in promoting the development of prefabricated construction, viewing the government, developers, and consumers as players in the game. He suggests that consumers, as the ultimate owners of prefabricated buildings, are primarily influenced by the incremental costs of purchasing homes and policy subsidies. The government needs to formulate dynamic subsidy policies based on the development status of prefabricated construction [16]. Existing research demonstrates that construction companies, as direct practitioners of intelligent construction, play a key role in influencing the transformation to intelligent construction. Support from technology service providers, developers, and the government can facilitate this transformation, while consumer behavior is mainly influenced by other stakeholders. From this, it is evident that the development of intelligent construction reflects the proactive engagement of developers, construction enterprises, and technology service providers. The government plays a regulatory role in the construction market and occupies a dominant position in promoting the formulation of policies related to the construction industry. It is particularly important not to overlook the significant endogenous driving characteristics presented by construction enterprises as the main body of the transformation in the implementation and promotion of intelligent construction [45]. Wei et al. [46], based on the integration of the Unified Theory of Acceptance and Use of Technology (UTAUT) and the improved Task-Technology Fit (TTF) theory, constructed a hypothetical model of the factors influencing the adoption and continued use of intelligent construction technology. They found that performance expectancy, effort expectancy, and facilitating conditions positively influence technology adoption behavior, while social influence did not reach significance. This indicates that the industry is currently in the early stages of technology diffusion, where policy advocacy or peer pressure has not yet formed effective external driving forces, and corporate decisions are more based on internal evaluations rather than the external environment. Therefore, the focus should primarily be on the strategic choices of construction enterprises, based on government guidance and external influences.

Currently, intelligent construction in developed countries has evolved from localized technology applications to an integrated system covering the entire lifecycle of construction, having established a relatively mature development system that encompasses policy support, technological innovation, industrial collaboration, and standard system construction [3]. In comparison, although China has accumulated certain experiences across 24 pilot cities and 758 pilot projects, with progress in supportive policies, innovation platform construction, quota standards formulation, and talent cultivation, it remains largely in a fragmented trial phase. Therefore, it is particularly important to study the influencing factors of intelligent construction transformation and the strategies for promoting the transformation of construction enterprises. Chen Ke et al. [47] argue that China still faces challenges in market environment, enterprise deployment, and core resources. They emphasize the need for comprehensive empowerment across different stages of the engineering supply chain, production systems and organizational methods, as well as cooperation between enterprises and industries. Ejidike et al. [48,49,50] suggest that in economically underdeveloped countries, the transformation of the construction industry is hindered by macroeconomic barriers and limited financing channels. Gao et al. [51] conduct research on the obstacles to the development of industrialized construction, identifying the main factors as insufficient policy and industry support, an imperfect coordination system, high construction costs, and low market recognition. Zheng et al. [52] analyzed the barriers to the development of intelligent construction technology based on Bayesian networks, ultimately identifying key and sensitive factors hindering its progress. Shao et al. [53] employed the ISM method to explore the key driving factors for the development of intelligent construction and proposed targeted countermeasures. Wang [54] used Social Network Analysis (SNA) to investigate the key influencing factors in the promotion process of intelligent construction, focusing on core technology applications, government policies, and collaborative management capabilities, and put forward specific strategies. Duan [55] found, based on the ISM model, that cultivating innovative talents is the strongest driver among the factors influencing the collaborative development of intelligent construction and industrialization. Yan et al. [56] identified 38 factors affecting the promotion of intelligent construction and constructed a PLS-SEM model based on the UTAUT2 theory, proposing different key pathways for the promotion of intelligent construction technology at the initiation, rapid development, and maturity stages. This article will focus on analyzing the degree of influence of the main factors on the transformation of construction enterprises towards intelligent construction.

2.2. Research on Evolutionary Game Theory and Its Application in Intelligent Construction

Evolutionary game theory, with “bounded rationality” as its core premise, breaks through the traditional assumption of “complete rationality” in game theory. It focuses on the dynamic process in which multiple agents adjust their strategies through learning and imitation in repeated interactions, ultimately converging to an Evolutionarily Stable Strategy (ESS). This approach better analyzes how different individuals make strategic choices in cooperative and competitive situations to achieve their own interests [57]. Evolutionary game theory integrates the study of game theory with dynamic evolutionary processes, drawing from behavioral ecology and biological evolution [58]. Due to the assumption of bounded rationality, it posits that participants can continuously adjust their strategies through learning and imitation. Additionally, a limited number of interactions among individuals can lead to the formation of specific equilibrium strategies [59,60]. Its advantages in the construction field are significant: first, it aligns with the bounded rationality characteristics of various stakeholders in the construction industry (such as government, enterprises, and consumers), making it more applicable to real decision-making scenarios; second, it can dynamically reveal the evolutionary paths of strategies rather than only presenting static equilibrium results, which fits the long cycles and multiple phases of construction projects; third, it can integrate multiple variables such as returns, costs, policy rewards and punishments, and risk perception, thereby quantifying the core driving factors behind stakeholders’ strategic choices. Consequently, it has been widely applied in recent years in areas such as smart construction sites, zero-carbon buildings, green buildings, prefabricated construction, and digital transformation in the industry. Wu [61] believes that evolutionary game theory is an important method for analyzing the strategic choices of multiple agents.

Feng [39] constructed a tripartite model involving the government, enterprises, and projects. He identified three types of stable equilibrium strategies: “all active”, “government intervention-enterprise promotion-project non-participation”, and “government intervention-enterprise and project negativity”, pointing out that the balance of government punishment and the interests of enterprises and projects is a core driver. Hong [62] explored the behavioral decision logic of the government, developers, and consumers in a zero-energy market based on the tripartite interest distribution mechanism, identifying and optimizing key factors that hinder the development of zero-energy buildings at various stages, and clarifying the pattern of initial government dominance followed by market withdrawal after maturity. Yang [63] introduced environmental organizations into a tripartite model under the traditional government regulatory framework, finding that moderate rewards and punishments from the government can promote low-carbon construction by enterprises, while excessive rewards and punishments are detrimental to system stability. Some scholars [64,65] have constructed a three-party evolutionary game model involving the “government-developer-consumer” to formulate strategies for promoting green buildings, focusing on government incentives and regulatory efforts, as well as the mechanisms of low-carbon trading markets. Yang [66] proposed through government-enterprise game theory that raising carbon trading prices, expanding tax rate differences, promoting technological upgrades, enhancing enterprise enthusiasm, and increasing social awareness can reduce the cost–benefit ratio for enterprises applying BIM, thereby effectively promoting the application of new technologies. Additionally, reasonable administrative measures, policy subsidies, and penalties will have a positive impact on the government incentive mechanisms and enterprise BIM application strategies. Zhang [67] constructed an evolutionary game model for the digital transformation involving the government, service providers, and construction enterprises, verifying that government tax rebates and subsidies within thresholds can enhance enterprise enthusiasm, while the comprehensive digital transformation of enterprises depends on the benefits of system solution strategies. Zhu et al. [68] explore the key factors influencing decision-making from the perspective of adopting intelligent construction technologies. They identify government subsidies, taxes, and penalties as significant influences and propose an incentive mechanism dominated by government macro incentives, which coexists with adoption and development incentives.

Evolutionary game theory provides an effective analytical tool for resolving multi-party interest conflicts and complex strategic choices in the construction field. It is a rational method for studying the strategic choices of various stakeholders in the transformation of intelligent construction. In the game process, government policies, cost–benefit considerations, and external influences among entities are core influencing factors.

3. Methodology

The analytical framework of this study consists of three main steps, as illustrated in Figure 1. First, a literature review is conducted to analyze the internal and external factors affecting whether construction enterprises implement intelligent construction transformation. The focus is on the enterprises themselves, with the government considered as an external policy factor rather than a stakeholder. Next, based on evolutionary game theory, an evolutionary game model is constructed to perform stability analysis, identifying equilibrium points and influencing variables. Finally, system dynamics theory is applied to simulate the evolutionary game model developed in the second part, allowing for a quantitative analysis of the dynamic impacts of varying variables on the strategic choices of construction enterprises, leading to the formulation of targeted development strategies.

3.1. Model Assumptions and Construction

3.1.1. Problem Description

China’s construction industry generally faces the challenge of low profit margins. At the same time, with the increasing severity of the aging workforce issue, transitioning from a traditional extensive development model to a refined and intelligent approach has become an inevitable choice for construction enterprises. However, the transformation process is hindered by high initial investments, long payback periods, and significant uncertainties in returns, preventing most enterprises from making substantial progress [69]. To expedite the advancement of new productive capabilities and facilitate the implementation of intelligent construction, the government aims to create a favorable market environment and reduce the pressure on enterprises to transform through appropriate incentive and penalty policies. Therefore, this paper considers the government’s incentives and penalties as representations of external governmental influence in constructing an evolutionary game model for the transformation among construction enterprises. By assigning different values to the incentive and penalty coefficients, the study investigates how governmental influence affects the strategic decisions of construction enterprises.

The revenues generated by construction enterprises during the construction process often determine their development strategies. Following the first round of the game, construction enterprises will evaluate their revenue outcomes before determining their strategy for the second round [70]. For construction enterprises themselves, the key factors in the game include direct benefits, transformation costs, and synergy effects (costs reduced through joint efforts and indirect benefits obtained by one party’s initiatives alone).

Firstly, regarding transformation costs, intelligent construction is an innovative mode of engineering built on the integration of new information technologies and engineering practices. In the initial development phase, significant costs must be invested in technology development and equipment procurement, while also requiring the cultivation of new industrial technology talent [71,72]. In the short term, this is likely to result in profit losses or even deficits, with a long return cycle, greatly reducing the willingness of construction enterprises to undergo transformation.

Secondly, direct benefits arise from digitalizing construction entities, informatizing production factors, and smartening project management processes. This reduces fragmentation in the construction industry, improves operational and construction efficiency, and lowers high costs associated with insufficient operability [73]. By smartening construction methods, production modes can be transformed, quality and efficiency can be improved, resources can be allocated more rationally, and material waste can be significantly reduced, while also eliminating the possibility of design errors and rework, along with the related costs and time waste [74].

Finally, the synergy effect is another factor. When different construction enterprises simultaneously engage in intelligent construction practices within a project, synergy benefits can arise. Enterprises can collaboratively solve technical problems, fully utilize shared resources to achieve excess returns, and jointly reduce rework and design errors, leading to cost savings. Even if only one party undergoes transformation, the other party can benefit as a ‘free rider’ to achieve improvements in efficiency or receive credit rewards.

3.1.2. Assumptions and Parameters

Assumption 1.

Construction enterprises, as the players in the game, operate under the premise of limited rationality and strive to achieve the highest possible profits. Under conditions of incomplete information for both parties, they adjust their strategies based on changes in the internal and external environments to achieve a stable state.

Assumption 2.

Two construction enterprises are randomly selected from a large number of firms to act as players in the game. During the game, there are two strategies available, and they can freely choose whether to undergo intelligent construction transformation. The probability that Enterprise 1 chooses to undergo intelligent construction transformation is x (where 0 ≤ x ≤ 1), and the probability of not implementing it is 1 − x. Similarly, Construction Enterprise 2 has probabilities of y (where 0 ≤ y ≤ 1) for implementing and 1 − y for not implementing intelligent construction methods. During the promotion period of intelligent construction, the internal entities involved in the transformation possess the following characteristics: they have the capability to carry out innovations and practical activities related to intelligent construction; they can execute relevant transformation decisions; the entities engaged in technological collaborative innovation must bear the risks and responsibilities that may arise during the intelligent construction process; at the same time, these innovative entities are also the beneficiaries of the intelligent construction transformation [75]. According to Sun’s viewpoint, the assumptions regarding the interaction relationships of micro-level individuals are overly rigorous, leading to research results that may deviate from reality. By utilizing the evolutionary game theory within the collaborative innovation network, it can be observed that the internal entities of intelligent construction technology exhibit homogeneous characteristics [76].

Assumption 3.

The revenue for construction companies using traditional methods is denoted as Ei. The costs incurred by these companies to transition to intelligent construction are denoted as Ci, which include training personnel, purchasing hardware and software, and investing in technology. After adopting intelligent construction methods, the direct benefit coefficient is a (where 0 ≤ a ≤ 1). This includes enhancements in production management efficiency, improvements in the quality of on-site construction and management, and the increased potential for the enterprise to secure contracts. Intelligent construction represents an integration of innovative technologies in the construction industry. Given the current intense competitive environment and a lack of innovative resources, multiple construction companies are required to collaborate during project implementation for better development. If both companies actively cooperate and jointly engage in key technological innovations, they can effectively reduce transformation costs and promote complementary advantages in resources, with a cost reduction coefficient of b (where 0 ≤ b ≤ 1). Moreover, the synergy generated from their cooperation can achieve goals that would be difficult for a single company to accomplish, effectively enhancing their market competitiveness and yielding excess benefits denoted as Ii. However, due to the complexity of the transformation process and the constraints of resource capabilities that result in high initial transformation costs, along with uncertainties in technological development, there is a possibility of opportunistic behavior leading to passive cooperation aimed at maximizing individual interests. When one party undergoes transformation unilaterally, while the other party engages in “free-riding” behaviors, the transforming party still generates indirect benefits denoted as Pi. Meanwhile, the cooperating party, despite its passive cooperation, can obtain additional benefits (positive spillover effect) from the “free-riding” behavior, with a coefficient of c (where 0 ≤ c ≤ 1) [77]. For example, sharing a smart construction site management system for online quality and safety inspections can yield labor benefits, or different units within the same project can use BIM for design review to reduce rework costs [70]. However, the indirect benefits are far less than the excess benefits that would result from active cooperation, significantly affecting the effectiveness of cooperative transformation and the stability of the collaborative alliance, leading to a fragile synergy between companies and even potential breakdowns in cooperation [78].

Assumption 4.

To accelerate the transformation process of ordinary construction companies, the government provides policy incentives denoted as Si to actively participating enterprises, such as tax and fee reductions, special funds, training and guidance, land use support, and platform development to promote the transformation of construction companies. On the other hand, as the industry regulator, the government may impose penalties denoted as Li for management flaws in traditional construction methods, particularly concerning green construction and quality safety supervision. Relevant parameter settings are summarized in

Table 1. Notation Description.

Parameters	Definition
x	Probability of Construction Enterprise 1 adopting intelligent construction
y	Probability of Construction Enterprise 2 adopting intelligent construction
E1, E2	The revenue for construction enterprises using traditional methods
C1, C2	Incremental costs for construction enterprises using intelligent construction methods
a	The coefficient for direct benefit obtained by construction enterprises using intelligent construction methods
b	The coefficient for reducing costs through active cooperation by both parties using intelligent construction methods.
I1, I2	Excess benefits through active cooperation by both parties using intelligent construction methods.
P1, P2	Indirect benefits when only one party uses intelligent construction methods
c	The coefficient for the positive spillover effect obtained by one party due to opportunism when only one party uses intelligent construction methods.
S1, S2	Incentives for the implementation of intelligent construction promoted by the government.
L1, L2	Possible penalties for construction enterprises opting traditional construction

.

3.1.3. Payoff Matrix and Replication Dynamic Equation

Based on the above assumptions, the evolutionary game payoff matrix between construction enterprises can be obtained under the active promotion of the government, as shown in

Table 2. Payoff matrix for evolutionary game between Construction Enterprise 1 and Construction Enterprise 2.

Construction Enterprise 1	Construction Enterprise 2
	Implement Intelligent Construction (y)	Implement Traditional Construction (1 − y)
Implement intelligent construction (x)	E1 + aE1 + I1 + S1 − (1 − b) C1	E1 + aE1 + S1 + P1 − C1
	E2 + aE2 + I2 + S2 − (1 − b) C2	E2 − L2 + cE2
Implement traditional construction (1 − x)	E1 − L1 + cE1	E1
	E2 + aE2 + S2 + P2 − C2	E2

.

3.2. Stability Analysis of the Evolutionary Game

The expected returns for Construction Enterprise 1 opting for intelligent construction methods can be represented as U₁. Expected returns for Construction Enterprise 1 when not opting for intelligent construction methods can be denoted as U₂. Average expected returns can be expressed as Ux. They can be formulated as follows:

$$U_1=y\left[E1+aE1+I1+S1-(1-b)C1\right]+(1-y)(E1+aE1+S1+P1-C1),$$

(1)

$$U_2=y(E1-L1+cE1)+(1-y)E1$$

(2)

$$U_x=x{U}_1+(1-x)U_2$$

(3)

The replication dynamics equation for Construction Enterprise 1, denoted as F(x). This can be formulated using the following general form:

$$\mathrm{F}(\mathrm{x})={\displaystyle \frac{dx}{dt}}=\mathrm{x}(U_1-U_x)=\mathrm{x}(1-\mathrm{x})\left[aE1+S1+P1-C1+y\left(bC1+L1+I1-P1-cE1\right)\right]$$

(4)

Similarly, the expected returns for Construction Enterprise 2 opting for intelligent construction methods can be represented as V₁. The expected returns for Construction Enterprise 2 when not opting for intelligent construction methods are V₂. The average expected returns can be expressed as Vy. These can be formulated as follows:

$$V_1=x\left[E2+aE2+I2+S2-(1-b)C2\right]+(1-x)(E2+aE2+S2+P2-C2)$$

(5)

$$V_2=x(E2-L2+cE2)+(1-x)E2$$

(6)

$$V_y=y{V}_1+(1-y)V_2$$

(7)

The replication dynamics equation for Construction Enterprise 2 can be represented as F(y). This can be formulated using the following general form:

$$\mathrm{F}(\mathrm{y})={\displaystyle \frac{dy}{dt}}=\mathrm{y}(V_1-V_y)=\mathrm{y}(1-\mathrm{y})\left[aE2+S2+P2-C2+x\left(bC2+L2+I2-P2-cE2\right)\right].$$

(8)

By setting F(x) = F(y) = 0, we can determine five local equilibrium points within the evolutionary game dynamics, which are: (0, 0), (0, 1), (1, 0), (1, 1), (x*, y*), where $x^{*}={\displaystyle \frac{C2-P2-S2-aE2}{bC2+L2+I2-P2-cE2}}, y^{*}={\displaystyle \frac{C1-P1-S1-aE1}{bC1+L1+I1-P1-cE1}}$.

By analyzing the local stability of the system equilibrium points through the Jacobian matrix [79], the Jacobian matrix of the evolutionary game system can be obtained as follows:

$$\begin{array}{c}\begin{array}{l}\mathrm{J}=\left(\begin{array}{cc}{\displaystyle \frac{\partial F\left(x\right)}{\partial x}}& {\displaystyle \frac{\partial F\left(x\right)}{\partial y}}\\ {\displaystyle \frac{\partial F\left(y\right)}{\partial x}}& {\displaystyle \frac{\partial F\left(y\right)}{\partial y}}\end{array}\right)\\ =\left(\begin{array}{cc}(1-2x)\left[aE1+S1+P1-C1+y(bC1+L1+I1-P1-cE1)\right]& x(1-x)(bC1+L1+I1-P1-cE1)\\ y(1-y)(bC2+L2+I2-P2-cE2)& (1-2y)\left[aE2+S2+P2-C2+x(bC2+L2+I2-P2-cE2)\right]\end{array}\right)\end{array}\end{array}$$

Using the Friedman matrix local analysis method, we can assess whether the equilibrium points are in a locally stable state by analyzing if they satisfy the conditions of Det(J) > 0, Tr(J) < 0. For detailed information, please refer to

Table 3. Stability analysis results for the evolutionary game between Construction Enterprise 1 and Construction Enterprise 2.

Equilibriums	Case 1			Case 2			Case 3
	Det(J)	Tr(J)	Results	Det(J)	Tr(J)	Results	Det(J)	Tr(J)	Results
O (0, 0)	+	−	ESS	+	+	unstable point	−	±	saddle point
A (0, 1)	+	+	unstable point	+	−	ESS	+/−	±	saddle point
B (1, 0)	+	+	unstable point	+	−	ESS	+/−	±	saddle point
C (1, 1)	+	−	ESS	+	+	unstable point	−	±	saddle point
D (x*, y*)	−	0	saddle point		0	saddle point	−	0	saddle point

.

Based on the conditions of the equilibrium points, the discussion can be divided into the following three scenarios:

Scenario 1: When C1 − P1 − S1 − aE1 > 0 and C2 − P2 − S2 − aE2 > 0, both (0, 0) and (1, 1) are stable strategies. This means that when the transformation costs for construction enterprises are higher than the direct benefits, government incentives, and unilateral transformation indirect benefits, there are two evolutionary outcomes: both enterprises ultimately choose to implement intelligent construction methods, promoting cooperative results in intelligent construction and thereby obtaining excess returns; or both enterprises do not implement intelligent construction methods.

Scenario 2: When C1 − P1 − S1 − aE1 < 0 and C2 − P2 − S2 − aE2 < 0, both (0, 1) and (1, 0) are stable strategies. This means that when the transformation costs for construction enterprises are lower than the direct benefits, government incentives, and unilateral transformation indirect benefits, only one enterprise is likely to be willing to implement intelligent construction methods. Since both enterprises are rational economic agents, motivated by the goal of maximizing their own interests, they will engage in opportunistic behaviors such as “free-riding” to reduce their own costs and achieve higher returns, leading to the overall failure of intelligent construction implementation.

Scenario 3: When C1 − P1 − S1 − aE1 < 0 and C2 − P2 − S2 − aE2 > 0; or C1 − P1 − S1 − aE1 > 0 and C2 − P2 − S2 − aE2 < 0, there are no stable strategies.

Given that we are currently in the early stages of promoting intelligent construction and the transformation costs are relatively high, we will primarily discuss the situation in Scenario 1. Figure 2 illustrates the evolutionary phase diagram of construction enterprises implementing intelligent construction methods, reflecting the dynamic evolution process of the transformation. Points O and C are stable points, representing stable strategies (0, 0) and (1, 1), where both parties either choose to implement or not implement intelligent construction methods. Points A (0, 1) and B (1, 0) are unstable points, indicating that construction enterprises 1 or 2 choose different strategies. Point D (x*, y*) is saddle point of the game system. Each region converges from unstable points to stable points, such as region ADC evolving from unstable point A (0, 1) to stable point C (1, 1).

3.3. Analysis of Influencing Factors

From Figure 2, it can be inferred that the size of ${\mathrm{S}}_{\mathrm{O}\mathrm{A}\mathrm{D}\mathrm{B}}={\displaystyle \frac{1}{2}}\left(\mathrm{x}\mathrm{*}+\mathrm{y}\mathrm{*}\right)={\displaystyle \frac{1}{2}}\left({\displaystyle \frac{\mathrm{C}2-\mathrm{P}2-\mathrm{S}2-\mathrm{a}\mathrm{E}2}{\mathrm{b}\mathrm{C}2+\mathrm{L}2+\mathrm{I}2-\mathrm{P}2-\mathrm{c}\mathrm{E}2}}+{\displaystyle \frac{\mathrm{C}1-\mathrm{P}1-\mathrm{S}1-\mathrm{a}\mathrm{E}1}{\mathrm{b}\mathrm{C}1+\mathrm{L}1+\mathrm{I}1-\mathrm{P}1-\mathrm{c}\mathrm{E}1}}\right)$ affects the evolutionary direction of the system. If it is smaller, then ${\mathrm{S}}_{\mathrm{A}\mathrm{D}\mathrm{B}\mathrm{C}}$ becomes larger, and the system will ultimately evolve towards C (1, 1), meaning that either construction enterprise 1 or 2 will converge on implementing intelligent construction methods. Conversely, if it is larger, the system will evolve towards O (0, 0), indicating that both construction enterprises will choose traditional construction methods. Analyzing the factors that influence the area, we can determine the direction of system evolution by taking partial derivatives of the influencing elements (see

Table 4. Effects of Influencing Factors on Intelligent Construction Transformation.

Influencing Factors	Partial Derivative Results	$\mathbf{Impact}\mathbf{on}{\mathbf{S}}_{\mathbf{O}\mathbf{A}\mathbf{D}\mathbf{B}}$
a	<0	↓
b	<0	↓
c	>0	↑
C1, C2	>0	↑
S1, S2	<0	↓
L1, L2	<0	↓

for details). Here, “↑” indicates a positive correlation and “↓” indicates a negative correlation. Based on the analysis results of the influencing factors in Table 4, the following conclusions can be drawn.

The intelligent construction methods can significantly enhance construction efficiency and safety. The higher the direct benefit coefficient a, the more favorable it is for the transformation of construction methods. Taking the partial derivative of a, we have: ${\displaystyle \frac{\partial S_{OADB}}{\partial a}}=-{\displaystyle \frac{1}{2}}\left({\displaystyle \frac{E2}{bC2+L2+I2-P2-cE2}}+{\displaystyle \frac{E1}{bC1+L1+I1-P1-cE1}}\right)$. From the previous derivation, we have 0 ≤ x* ≤ 1 and 0 ≤ y* ≤ 1, thus $\mathrm{b}\mathrm{C}2+\mathrm{L}2+\mathrm{I}2-\mathrm{P}2-\mathrm{c}\mathrm{E}2>0$ and $\mathrm{b}\mathrm{C}1+\mathrm{L}1+\mathrm{I}1-\mathrm{P}1-\mathrm{c}\mathrm{E}1>0$. Therefore, ${\displaystyle \frac{\partial S_{OADB}}{\partial a}}$ < 0, indicating that S is a monotonically decreasing function of a. This means that as the direct benefit coefficient a increases, S decreases, leading construction enterprises 1 or 2 to ultimately converge on implementing intelligent construction methods.

Due to the uniqueness of developing innovative technologies for intelligent construction and the degree of integration among participants, it is necessary to reduce costs through complementary advantages between enterprises. The higher the cost reduction coefficient b, the more favorable it is for the transformation of construction methods. Taking the partial derivative of b, we have ${\displaystyle \frac{\partial S_{OADB}}{\partial b}}=-{\displaystyle \frac{1}{2}}\left[{\displaystyle \frac{C2\left(C2-P2-S2-aE2\right)}{{\left(bC2+L2+I2-P2-cE2\right)}^2}}+{\displaystyle \frac{C1\left(C1-P1-S1-aE1\right)}{{\left(bC1+L1+I1-P1-cE1\right)}^2}}\right]$. From previous derivations, we can obtain ${\displaystyle \frac{\partial S_{OADB}}{\partial b}}$ < 0. Thus, S is a monotonically decreasing function of b. This means that as the cost reduction coefficient b increases, b decreases, leading to a higher willingness for construction enterprises 1 or 2 to actively cooperate in transformation, ultimately converging on implementing intelligent construction methods. The promotion of intelligent construction often accelerates the integration of various production factors within the industry. For example, by sharing information through smart construction sites and making joint decisions, it effectively breaks down existing technological barriers, collaboratively resolving technical issues and achieving greater benefits with lower costs. The more costs that are jointly reduced, the more favorable it is for construction enterprises to implement intelligent construction methods.

Since enterprises are rational economic agents, it is difficult to avoid the situation of free riding in the early stages of applying new technologies. Therefore, when one party actively implements transformation, the other party can also gain positive spillover effect. The higher the benefit coefficient c, the less favorable it is for the transformation of construction methods. Taking the partial derivative of c, we have ${\displaystyle \frac{\partial S_{OADB}}{\partial c}}={\displaystyle \frac{1}{2}}\left[{\displaystyle \frac{E2\left(C2-P2-S2-aE2\right)}{{\left(bC2+L2+I2-P2-cE2\right)}^2}}+{\displaystyle \frac{E1\left(C1-P1-S1-aE1\right)}{{\left(bC1+L1+I1-P1-cE1\right)}^2}}\right]$. From previous derivations, we can obtain ${\displaystyle \frac{\partial S_{OADB}}{\partial c}}$ > 0. Thus, S is a monotonically increasing function of c. This means that as the positive spillover effect coefficient c increases, S increases, leading construction enterprises 1 or 2 to ultimately converge on not implementing intelligent construction methods.

In the process of promoting intelligent construction, the higher the transformation costs for construction enterprises 1 or 2, the less favorable it is for both to implement intelligent construction methods. Since C1 and C2 have the same impact on the evolutionary direction of the game system, taking the partial derivative of C1 and C2, we have ${\displaystyle \frac{\partial S_{OADB}}{\partial C1}}={\displaystyle \frac{1}{2}}·{\displaystyle \frac{\mathrm{L}1+\mathrm{I}1+\mathrm{b}\mathrm{S}1+(\mathrm{b}-1)\mathrm{P}1+(\mathrm{a}\mathrm{b}-\mathrm{c})\mathrm{E}2}{{\left(bC1+L1+I1-P1-cE1\right)}^2}}$. From previous derivations, we can obtain ${\displaystyle \frac{\partial S_{OADB}}{\partial C1}}$ > 0. Thus, S is a monotonically increasing function of C1. This means that as the transformation costs increase, S increases, leading construction enterprises 1 or 2 to ultimately converge on O (0, 0). Intelligent construction, as a new technology integration, requires significant production factors such as labor, capital, materials, and machinery, which demand substantial funding and have a long investment return cycle. For example, investing in fixed production lines incurs high costs, which greatly reduces the willingness of most small and medium-sized construction enterprises or private enterprises to promote intelligent construction.

The government plays an extremely important role in the promotion of intelligent construction, acting not only as a supervisor but also as a promoter. The more policy incentives construction enterprises 1 or 2 receive and the greater the penalties for traditional methods, the more favorable it is for the transformation of construction methods. Taking the partial derivative of L1, we have ${\displaystyle \frac{\partial S_{OADB}}{\partial L1}}=-{\displaystyle \frac{1}{2}}·{\displaystyle \frac{\mathrm{C}1-\mathrm{P}1-\mathrm{S}1-\mathrm{a}\mathrm{E}1}{{\left(bC1+L1+I1-P1-cE1\right)}^2}}$. From previous derivations, we can obtain ${\displaystyle \frac{\partial S_{OADB}}{\partial L1}}$ < 0. Thus, S is a monotonically decreasing function of L1. Similarly, taking the partial derivative of S1, we have ${\displaystyle \frac{\partial S_{OADB}}{\partial S1}}=-{\displaystyle \frac{1}{2}}·{\displaystyle \frac{1}{\mathrm{b}\mathrm{C}1+\mathrm{L}1+\mathrm{I}1-\mathrm{P}1-\mathrm{c}\mathrm{E}1}}$. From previous derivations, we can obtain ${\displaystyle \frac{\partial S_{OADB}}{\partial S1}}$ < 0, indicating that S is a monotonically decreasing function of S1. This means that the higher the penalties for traditional methods and the greater the government incentives, the smaller S becomes, leading construction enterprises 1 or 2 to ultimately converge on C (1, 1). Appropriate government incentive policies can help construction enterprises alleviate the burdens of transformation and break through the developmental bottleneck caused by low levels of early-stage intelligent construction. The increased government oversight resulting from traditional construction methods can provide external motivation for enterprise transformation.

In summary, the final stable strategy of the system depends on the magnitude of parameters such as the direct benefits of implementing intelligent construction, the joint efforts of enterprises to reduce costs, the transformation costs for construction enterprises, and the intensity of government incentives and penalties.

4. Numerical Simulation Analysis

4.1. Construction of the System Dynamics Model and Parameter Settings

System Dynamics (SD) is a methodology based on feedback control theory and systems thinking that studies the dynamic evolution of complex systems over time. It reveals the nonlinear relationships, delay effects, and causal feedback loops among various elements within the system. Evolutionary game theory, grounded in the premise of “bounded rationality”, examines the process by which participants in a group achieve dynamic evolution of strategy frequencies through adjustments such as imitation, learning, and mutation. The methodological characteristics of system dynamics align closely with the research needs of evolutionary game theory. The advantages of simulating evolutionary games using system dynamics are significant: first, it breaks through static analysis by depicting the temporal paths and phase characteristics of strategy evolution; second, it integrates multiple strategies and external factors to restore realistic game scenarios; third, it presents feedback mechanisms and nonlinear relationships, clarifying evolutionary logic; and fourth, it supports multi-scenario simulations, aiding policy pre-execution and scientific decision-making. Therefore, utilizing system dynamics models can effectively assist decision-makers in predicting the outcomes of evolutionary games under different environments and conditions.

Using Vensim for system dynamics research is both convenient and widely adopted. A stock-and-flow diagram has been created through Vensim PLE 7.3.5, as shown in Figure 3. The model has two stocks, representing the probabilities x and y of Construction Enterprises 1 and 2 adopting intelligent construction methods, respectively. There are 15 external variables, as detailed in Table 1. The model includes two flows, which are the implementation change rates F(x) and F(y). These are set according to Equations (1)–(8) in the game model. The arrows in the stock-flow diagram convey the influence relationships, indicating the direction of causality.

Due to the early stage of promoting intelligent construction in China, many cities lack substantial practical application cases and publicly available survey data, making it challenging to determine accurate parameter data. However, system dynamics (SD) simulation and evolutionary games do not require precise empirical data [80]. They do not demand extremely accurate results based on the precision of parameter settings and real-world fidelity. The core value lies in explaining the patterns of strategy change through the rational design of the model, utilizing the benefit functions of various games to reveal the relationships between auxiliary variables and external variables [81,82]. This approach can still effectively showcase the internal evolutionary dynamics of the system. As long as the parameter settings satisfy the theoretical analysis’s prerequisite conditions, proportions may vary, and modeling results do not have to be constant [39].

Xiamen, as one of the first pilot cities for intelligent construction, received recognition from the Ministry of Housing and Urban-Rural Development for its 2024 annual work (one of only eight pilot cities). It is highly representative in terms of industrial systems, pilot projects, and technological applications across the country. Therefore, by referencing existing literature on the transformation of other technological fields in the industry for the setting of simulation values [68,78,83], alongside statistical analyses of the construction industry’s development, recent local policies, and the survey results of Xiamen’s first to fourth batches of intelligent construction pilot projects for 2024–2025, relevant data for the model can be calculated.

According to data from the Xiamen Price Information Network, the average construction investment for projects is 3500 CNY/m². Considering an industry average profit margin of 5%, the return from traditional construction methods is approximately 180 CNY/m². Based on the investigation of 15 pilot projects for intelligent construction in Xiamen, the incremental cost for the transformation to intelligent construction is about 30 CNY/m². The direct benefits from obtaining rewards for shortening construction periods and reducing rework losses are approximately 5%. If only one side undergoes transformation, that side can reduce some rework losses by about 1%. Although the actively promoting side may not achieve excess benefits, it still gains indirect benefits such as quality safety evaluations and excellence awards, amounting to about 5 CNY/m². According to the incentive policies for pilot or demonstration projects announced by major pilot cities (such as Hefei and Shenzhen), subsidies are generally provided on a per-square-meter basis, amounting to 10 CNY/m², with a maximum limit not exceeding 2 million yuan. Set the initial parameters for the game system as shown in

Table 5. Initial parameter values.

Parameter Name	E1, E2	C1, C2	a	b	I1, I2	P1, P2	c	S1, S2	L1, L2
Initial Value	180	30	5%	40%	8	5	1%	10	10

. The simulation will start at time 0 and end at time 5, with a step size of 0.015625, measured in years. The initial probabilities for the strategy choices of construction enterprises are set as x = 0.5 and y = 0.5. It is important to clarify that when studying a specific parameter, the values of all other parameters remain constant. The internal subjects of intelligent construction technology exhibit homogeneous characteristics; therefore, the transformation probabilities of Construction Enterprise 1 and 2 can be viewed as changing in the same direction [76].

4.2. Analysis of the Impact of Transformation Elements on System Evolution in Construction Enterprises

4.2.1. Impact of Initial Conditions on the Strategic Evolution of Construction Enterprises

As shown in Figure 4, the final outcomes of the game between the two enterprises vary depending on different initial conditions. When the initial state of the enterprise is in areas OAD and ODB, the probability of the enterprise transitioning to intelligent construction is at a low level, leading to a final stable strategy of O (0, 0), meaning the construction enterprise does not transform. When the initial state is in areas ADB and BDC, the enterprise shows a higher willingness to transition to intelligent construction, resulting in a final stable strategy of C (1, 1), indicating that the construction enterprise chooses to implement intelligent construction methods. Furthermore, the higher the initial willingness to transform, the faster the system stabilizes. The simulation results of the SD model are consistent with the strategy scenario in Scenario 1 of the game model, confirming the validity of the SD model.

4.2.2. Impact of Direct Benefit Coefficient a on the Evolution of Strategies in Construction Enterprises

When other influencing factors remain constant, the direct benefits brought by the transition to intelligent construction affect the implementation probability for construction enterprises, as shown in Figure 5. When the direct benefit coefficient a is 2%, the probability curve for construction enterprises to implement intelligent construction rises slowly and approaches C (1, 1). However, when the direct benefit coefficients are 5%, 8%, and 10%, the probability curves for implementing the transition to intelligent construction have a steeper slope and converge to 1 at a rapid pace, with only slight differences among the three values. As the direct benefits increase, it becomes more favorable for construction enterprises to implement the transition to intelligent construction.

4.2.3. Impact of Cost Reduction Coefficient b on the Evolution of Strategies in Construction Enterprises

When other influencing factors remain constant, the proportion of cost reduction brought by the transition to intelligent construction affects its implementation probability for construction enterprises, as shown in Figure 6. When the cost reduction coefficient b is 1%, the probability curve for construction enterprises to implement intelligent construction rises slowly and approaches C (1, 1). However, when the cost reduction coefficients are 2%, 3%, 4%, 5% and 6%, especially when exceeding 2%, the probability curves for implementing the transition to intelligent construction have a steeper slope and converge to 1 at a rapid pace.

4.2.4. Impact of Positive Spillover Effect Coefficient c on the Evolution of Strategies in Construction Enterprises

When other influencing factors remain constant, the impact of one party adopting intelligent construction while the other party gains positive external benefits due to opportunism on its implementation probability is illustrated in Figure 7. Regardless of the presence of external benefits, the probability for construction enterprises to implement intelligent construction converges to 1 at a rapid pace.

4.2.5. Impact of Transition Incremental Cost C on the Evolution of Strategies in Construction Enterprises

When other influencing factors remain constant, the impact of the incremental costs incurred by construction enterprises when implementing intelligent construction on their transition probability is illustrated in Figure 8. It can be seen that when the incremental transition cost exceeds 40 CNY/m², the probability of construction enterprises adopting intelligent construction shows a downward trend, ultimately approaching 0. When the incremental cost is below 35 CNY/m², the transition probability rises slowly and converges to 1 (indicating a willingness to implement intelligent construction), especially when the incremental transition cost is below 30 CNY/m², where the slope of the probability curve increases significantly.

4.2.6. Impact of Government Incentives S on the Evolution of Strategies in Construction Enterprises

When other influencing factors remain constant, the adjustment of government incentive levels S from 0, 2, 5, 8, 10, to 15 CNY/m² shows the impact of government incentives on the transformation probabilities, as illustrated in Figure 9. It is evident that without government incentives, construction companies find it difficult to spontaneously implement intelligent construction methods. When the subsidy is very low (2 CNY/m²), the probability of construction companies adopting intelligent construction shows a downward trend, ultimately approaching zero. At a slightly higher subsidy level (5 CNY/m²), the growth of the transformation probability curve is relatively slow. As the incentive subsidies continue to increase, when the incentives exceed 10 CNY/m², the probability of construction companies implementing intelligent construction rises rapidly and converges quickly towards 1 with little difference in the slope of the curves for incentives of 10 CNY/m² and 15 CNY/m². Figure 10 also indicates that the impact of government incentives on transformation probabilities exhibits phased characteristics. Particularly during periods of weak government incentives, the willingness of construction companies to transform is not significant. However, as more companies become pilot projects and receive actual incentives, the transformation probability increases rapidly. Once the market matures, construction companies tend to favor intelligent construction methods.

4.2.7. Impact of Government Penalties L on the Evolution of Strategies in Construction Enterprises

As shown in Figure 10, with the penalty amounts set at 5, 10, and 15 CNY/m², as the penalty values are introduced, the evolutionary outcome for construction enterprises tends toward 1, indicating both parties choose the intelligent construction method. When there is no penalty for the traditional construction method, the probability of choosing intelligent construction ultimately also converges to 1, but it requires a longer time for negotiation. This indicates that the intensity of government penalties does not affect the evolutionary equilibrium point of the system, but rather influences the speed at which the system evolves to a stable point.

4.3. Sensitivity Analysis

Taking the process of changing strategies from (Traditional Construction, Traditional Construction) to (Intelligent Construction, Intelligent Construction) as an example, we discuss the sensitivity of the proportion x of construction companies choosing to adopt intelligent construction methods to external variables. Given that some external variables are difficult to change in the short term under current market conditions, we primarily analyze the impact of the incremental costs of intelligent construction C1 or C2, government incentives S1 or S2, and government fines L1 or L2 on the model. Considering a long-term implementation process, while keeping the original model data unchanged, we set both x and y to 0.2 (representing the early stage of intelligent construction promotion as described in Section 4.2.1). We examine the effects of a 20% and 50% change in each individual variable on the transformation ratios x or y of construction companies adopting intelligent construction. The parameter values are shown in

Table 6. Parameter values for sensitivity analysis.

Parameter Name	Value
	Control Group	Experimental Group 1-1	Experimental Group 1-2	Experimental Group 2-1	Experimental Group 2-2	Experimental Group 3-1
C1, C2	30	24	15	30	30	30
S1, S2	10	10	10	12	15	10
L1, L2	10	10	10	10	10	15
Description	Initial State	Incremental Cost Decrease by 20%	Incremental Cost Decrease by 50%	Government Incentive Increase by 20%	Government Incentive Increase by 50%	Government Fine Increase by 50%

.

As shown in Figure 11, based on comparisons between the control group and experimental groups 1-2, 2-2, and 3-1, simulations reveal that under a 50% change in all variables, the promotion effect of implementing intelligent construction for construction companies is ranked as follows: Experimental Group 1-2 > Experimental Group 2-2 > Experimental Group 3-1. The sole use of penalty measures does not promote construction companies to converge towards the implementation of intelligent construction when initial willingness is low. Reducing the incremental costs of the intelligent construction transition is more effective than increasing government incentive policies. Comparisons among Experimental Groups 1-1, 1-2, 2-1, and 2-2 indicate that the evolutionary paths of a 20% reduction in incremental costs and a 50% increase in government incentive levels are essentially overlapping.

By adjusting the initial state transformation probability, we can reflect the phase of intelligent construction promotion to some extent. As shown in Figure 12, during the emergence phase, the transformation cost must be less than 20 CNY/m² for construction companies to ultimately converge the probability of choosing intelligent construction to 1. In the market promotion phase, the transformation cost must be less than 35 CNY/m² for the transformation probability to rise slowly and converge to 1. After reaching the stabilization phase, when construction companies have sufficient awareness and willingness to transform, a transformation cost exceeding 50 CNY/m² will cause hesitation among them.

For incentive policies, during the emergence phase, only strong incentives can lead construction companies to converge the probability of choosing intelligent construction to 1, while weak incentives have a negligible effect on the transformation probability, as shown in Figure 13. In the market promotion phase, when the incentive exceeds 5 CNY/m², the transformation probability rises slowly and converges to 1. After reaching the stabilization phase, providing incentives can further enhance the willingness of construction companies to transform. However, regardless of government subsidies, they will choose the intelligent construction method.

5. Discussion and Limitations

5.1. The Impact of Initial Conditions on Strategy Selection in Construction Enterprises

From Section 4.2.1, it can be seen that the initial state of construction enterprises’ willingness to transition to intelligent construction affects the final stable strategy of the game. The higher the initial transition probability, the faster the system tends to stabilize. This is because, in the initial state, a low probability of both parties implementing intelligent construction indicates that enterprises perceive the benefits of adopting intelligent construction as uncertain, leading to a pessimistic attitude towards the overall promotion of intelligent construction transformation in society. From a game theory perspective, both parties may believe that the other will not adopt intelligent construction technology, resulting in a situation where unilateral attempts lead to higher costs for one party without guaranteed benefits. Moreover, the other party, not actively cooperating in implementation, can still gain some benefits from the project. In this context, enterprises tend to avoid high-risk, high-cost investments and prefer to maintain the status quo by sticking to traditional construction methods. This aligns with the findings of Liu [70] and Chang [84] regarding the impact of initial conditions on game strategies in digital transformation and the industrialization of passive buildings. Liu [69] suggests that in the early stages of digital transformation, construction companies lack the necessary resources, technology, and environment for transformation, requiring support from government policies and digital solution providers to overcome these challenges. Of course, there are distinctions between intelligent construction practices and digital transformation; some construction techniques require support from service providers. On the other hand, construction companies with strong technical research and development capabilities or those that have mature technologies like BIM and prefabricated construction often have their own teams capable of independent execution.

Therefore, it is essential to enhance companies’ awareness of the intelligent construction transformation, leverage the guiding role of government policies, and, along with support from technology suppliers, harness market forces to drive construction companies towards intelligent construction practices.

5.2. Analysis of Factors Influencing the Transition to Intelligent Construction in Construction Enterprises

Section 4.2 simulates the evolutionary game analysis results, validating the conclusions drawn in Section 3.3. The direct benefit coefficient, cost reduction coefficient, government incentive strength, and penalty strength have a positive impact on the probability of construction enterprises implementing intelligent construction, while transition costs and positive spillover effect are negatively correlated.

Construction enterprises can empower production and construction through intelligent construction technology, deeply integrating digital technology into communication and coordination between enterprises. This significantly enhances construction efficiency and the success rate on the first attempt, reduces rework costs, and consequently leads to a substantial increase in return on investment. When the direct benefit rate exceeds 5%, it greatly motivates enterprises to adopt intelligent construction. This is significantly influenced by the regional environment. In mature pilot areas, a direct benefit rate of 5% can be easily achieved. For instance, the Changquan public housing project in Shenzhen, one of the first intelligent construction pilot projects approved by the Ministry of Housing and Urban-Rural Development, utilized 49 key technological achievements, resulting in a cost saving of approximately 75 million CNY, a reduction of about 10% in the overall construction period, and creating direct economic benefits of CNY 650 million (approximately an 11% direct benefit rate). However, in non-pilot cities or smaller construction markets where the penetration rate of BIM is low, achieving the benefits of reduced rework through collaborative review becomes challenging [85].

The transition process involves multiple stakeholders, and collaboration among participants is crucial, particularly in reducing transition costs. Cost savings have a significant positive impact on enterprises’ transition decisions, as reduced costs imply lower expected risks. If costs can be collectively reduced by more than 2%, the probability of construction enterprises implementing intelligent construction can be greatly enhanced. When one party actively promotes the transition to intelligent construction, it can lead to benefits and cost reductions for the other party involved in the project. Even if excess benefits are not realized, it can encourage the other party to attempt the transition. This also validates the current governmental focus on pilot projects in policy promotion; only by fully leveraging the demonstrative role of pilot projects and ensuring that all parties involved benefit can initial reluctance gradually be transformed into enthusiasm.

Transition incremental costs are a critical factor affecting construction enterprises’ decisions to implement intelligent construction. As costs increase, the implementation of intelligent construction does not yield more benefits for enterprises; instead, the complexity of cost expenditures and uncertain effectiveness can lead to a gradual decline in the willingness to transition, ultimately resulting in a preference for maintaining the status quo. When transition costs are below 35 CNY/m², there is still a realistic demand for intelligent construction. According to survey results, during the early promotion phase, the breadth and depth of application of intelligent construction technologies among enterprises are limited, and an average transition cost of 30 CNY/m² aligns with the model’s requirements for decision-making toward transition. This indicates that the initially high costs need to be reduced through technological innovation. However, transformation costs are also regionally dependent, with significant price differences for the same technology applied in different areas. High-end technologies often face more pronounced information barriers. In contrast, the unit prices for more mature technologies like BIM show less variation across provinces and cities (approximately 1–2 CNY/m²). This suggests that there is no need to pursue comprehensive application immediately; instead, it is advisable to start with cost-effective technology applications and gradually expand the scope from point to area. Additionally, expanding the range of participating enterprises and projects can achieve economies of scale and further reduce transition costs.

A certain level of government incentives (at least greater than 5 CNY/m²) has a positive effect on the transition, with even better results when exceeding 10 CNY/m². This can significantly stimulate hesitant construction enterprises and motivate them to transition to intelligent construction, thereby accelerating industry transformation. This aligns with Chen’s viewpoint [86], which suggests that only high-intensity special fund support (with a minimum subsidy of 8 CNY/m² per project) can promote the transition and upgrade of construction enterprises. It is also noted that, in practice, to achieve effective strategies, very few implementers use a single strategy tool, and the combined use of multiple strategy tools is considered more effective. Simulation results indicate that after exceeding a subsidy of 10 CNY/m², the convergence speed of transition probability does not significantly increase, which may instead increase the government’s burden. This underscores the critical role of policy support methods and intensity in the selection of game strategies. Existing research suggests that government subsidies and special funds are important incentive mechanisms for promoting the application of new technologies and transitions in enterprises. However, inappropriate use can lead to counterproductive effects [87]. For example, excessively high subsidies can create fiscal pressure on the government, reducing its willingness to guide, thus hindering transitions [88]. Similarly, high government subsidies may lead to a “free-riding” phenomenon, potentially weakening the vitality of enterprises in technology research and application [89]. Government subsidies have a U-shaped impact on enterprises’ digital innovation, inhibiting innovation beyond a critical threshold [90]. Currently, in various pilot cities in China, such as Zhengzhou, Shenzhen, and Hefei, incentive policies for intelligent construction are set with subsidies per square meter, with a maximum subsidy of 2 million CNY, equivalent to 10 CNY/m². In Xiamen, different levels of pilot cities receive rewards ranging from 500,000 to 1 million CNY, equivalent to 5–10 CNY/m². Suzhou has included “intelligent construction” in its city-wide science and technology development plan, with each project eligible for up to 500,000 CNY in financial support. The current level of incentives has reached the conditions necessary to promote transformation, and there is no need to continuously increase these incentives; otherwise, it may lead to negative effects. However, the major cities are still primarily concentrated in the 24 national pilot cities, and other provinces and municipalities have not followed suit with similar policies.

Penalties have a certain degree of influence on enterprises’ strategic choices during the transition process but only affect the speed at which the system evolves to a stable point. This conclusion aligns with the findings of Shi [91] and Zhu [68] in similar new technology fields (digital transformation, intelligent construction technology adoption). However, Jiang [92] noted in the evolutionary game of low-carbon development of green building materials that lower penalty levels lead building material suppliers to prefer producing traditional materials, while increased penalties shift strategy choices toward producing green materials, primarily because green materials have a larger share in construction applications and higher incremental costs. Only by increasing penalty levels can the risks of opportunistic behavior be heightened. Therefore, in the early stages of technology promotion, the government should focus more on incentivizing and guiding, strengthening market regulation, and, when necessary, using penalties to accelerate the promotion of intelligent construction implementation.

5.3. Sensitivity Analysis of Key Influencing Factors

The sensitivity analysis of key factors in Section 4.3 reveals that the promotion of intelligent construction requires the joint realization of technology, market, motivation, and support, similar to the viewpoint proposed by Feng [39]. In the early stages of development, the high costs associated with transformation are the primary barrier to implementing intelligent construction. Reducing transformation costs can be approached in two steps.

Firstly, technology application should not aim for comprehensiveness but should focus on solving practical problems, such as using BIM for collision detection to reduce rework or employing prefabricated construction to improve efficiency. According to research by McGraw Hill, the adoption rate of BIM technology in the U.S. construction industry was 28% in 2007 and reached 71% by 2012, maintaining a high level since then. Although the application of BIM technology in Singapore began in 2010, since 2015, the Singapore Building and Construction Authority has required all projects over 5000 square meters to submit BIM models, achieving an application rate of over 80% in construction projects. The industrialization rate of the construction sector in developed countries is generally over 70%, with the U.S., Japan, and France exceeding 85%. On the other hand, since the 1980s, developed countries have made many positive attempts regarding construction robotics; however, due to the complexity of processes and high costs, most construction robots remain in the research and development phase, with only a few entering pilot production, and large-scale market application has not yet been realized [11]. In terms of the strategy for intelligent construction development, it is essential to focus on low-cost, widely applicable technologies to enhance the level of intelligence and digitization in projects.

Secondly, when the application promotion reaches a certain scale, the economies of scale can gradually reduce incremental costs. As market demand and awareness improve, the collaboration between industry, academia, and research can drive innovation and overcome technical bottlenecks. Furthermore, when construction companies enter a stable transformation phase, even a 50% increase in incremental costs can still maintain their enthusiasm for transformation, allowing for the practical application of robotics and large intelligent equipment. Japan and South Korea have explored models for collaborative research and development among companies. For example, three major Japanese construction firms—Kajima Corporation, Shimizu Corporation, and Takenaka Corporation—signed a preliminary memorandum of understanding for technical cooperation in the field of construction robotics, forming an alliance to jointly develop new types of construction robots while strengthening the mutual use of existing robots. This collaborative R&D model will lower research and development costs and production expenses, expand the market impact of products, and accelerate the widespread application of construction robots.

From the perspective of motivation and support, government policy guidance, incentive measures, and training are important external drivers and essential guarantees for enhancing the confidence of construction companies in their transformation. Many industrialized countries have already elevated intelligent manufacturing and industrial development to a national strategic level. In countries like the U.S., Germany, and France, the focus is mainly on supporting advanced equipment manufacturing, while the UK and Japan have specifically proposed concepts similar to intelligent construction as important directions for reform and development in the construction industry. By building a digitally driven new ecosystem for the industry through the industrial internet, three development models are worth referencing. First, a government-led model with active participation from enterprises, such as the “Industry 4.0 Platform” jointly established by the German Mechanical Engineering Industry Association, the Electrical and Electronics Industry Association, and Siemens, which has become a national project included in Germany’s “High-Tech Strategy 2020”. Second, a model led by large enterprises with government encouragement and support. For instance, General Electric, along with AT&T, Cisco, and Intel, established the Industrial Internet Consortium to encourage more companies to join the industrial internet, and through federal government funding, a Digital Manufacturing and Innovation Design Research Center was established to encourage SMEs and makers to actively participate in industrial internet R&D. Third, a model led by industry organizations with enterprises “cooking together”. For example, a Japanese industry organization launched the “Industry 4.1J” experimental program, attracting large manufacturing companies such as Sun Electronics and Fuji Electric, connecting factories dispersed around the world to achieve the goal of extending industrial intelligence from individual enterprises to the overall value chain of the industry [11]. In recent years, China has elevated intelligent construction to a strategic level, gradually identifying development priorities and phased tasks. However, it still needs to learn from developed countries the principle of “standards first” to improve standards and fully leverage the roles of government and industry organizations.

For construction companies, internal motivation for transformation can be driven by self-initiated technological innovation, professional talent development, and improvements in operational management. Currently, the level of information technology application among companies in China is still low, and the enthusiasm for independent R&D is not strong enough. Stimulating innovation vitality for technological R&D and cultivating multi-level, cross-disciplinary professionals in intelligent construction can effectively enhance the ability to creatively solve engineering problems using new technologies and methods, providing a means to reduce costs and increase efficiency. The conclusions from Section 4.3 indicate that providing incentives after a stabilization period can further enhance construction companies’ willingness to transform. Regardless of government subsidies, companies will still choose intelligent construction methods, fully leveraging the intrinsic motivation of construction enterprises, which also provides a basis for the exit mechanism of policy incentives.

5.4. Limitation and Future Work

Given the complexity and uncertainty of promoting intelligent construction, which involves not only a single technical field but is also characterized by a situation where “there is much sound but little action” in the early stages, and considering the limited professional capacity and data sample availability, this study has several limitations and will focus on exploratory research in the future. Firstly, the scope of stakeholders is limited, as this paper only considers a two-party game between construction enterprises. In reality, many stakeholders are involved in promoting intelligent construction, including technology consultants, equipment suppliers, and other participants in the upstream and downstream industrial chain. As the promotion process accelerates, it will involve various management relationships, such as government-association-enterprise and government-enterprise-project relationships. Constructing a multi-party game model would provide a more comprehensive understanding of the strategy choices of each participant and their overall impact on the system. Secondly, the current model relies on static parameter assumptions, with values reflecting the current stage of development, which may not capture the dynamic market changes. As market dynamics evolve, the model and results will differ; future research could consider conducting evolutionary impact simulations based on changes in development stages. Thirdly, the limitations of simulation data pose challenges, as there are currently few pilot projects, making data collection difficult. This may lead to discrepancies between parameter definitions, assumptions, and real-world conditions during the simulation process. The data focuses mainly on Xiamen, a national pilot city, which may limit its representativeness. Future efforts should aim to enhance professional capabilities to ensure accurate model construction, clear parameter definitions, and collaboration with relevant authorities to obtain more data sources. Additionally, the accuracy of the simulation results should be further validated through enterprise surveys and actual case studies. Moreover, the internal collaborative relationships among various influencing factors have not been explored, such as the changing relationships between incentive policies, joint cost reductions, transformation costs, and direct benefits.

6. Conclusions

This paper utilizes evolutionary game theory to analyze the dynamic evolutionary process of strategy selection among construction enterprises concerning the implementation of intelligent construction methods, employing system dynamics for simulation analysis of the game model. Firstly, the research concludes that the initial state of construction enterprises’ willingness to transition to intelligent construction significantly affects the final stable strategy of the game; specifically, a higher initial transition probability leads to a faster stabilization of the system. Secondly, several key factors positively influence the probability of construction enterprises adopting intelligent construction, including direct benefits, the strength of government incentives, penalty intensity, and reduced costs associated with a joint transition. In contrast, transition costs and positive spillover effects are negatively correlated. From the enterprises’ perspective, integrating intelligent construction technology with the construction industry yields direct benefits. Furthermore, reducing construction costs through technology sharing and platform collaboration helps mitigate the drawbacks of traditional construction methods. The development costs linked to emerging technologies such as the Internet of Things, artificial intelligence, cloud computing, and 5G also play a critical role in influencing whether construction enterprises adopt intelligent construction models. Notably, when the direct benefit rate exceeds 5%, and costs can be jointly reduced by more than 2% while transition costs remain below 35 CNY/m², this can significantly motivate enterprises to adopt intelligent construction. From the government’s external perspective, both incentive and penalty measures can accelerate the transition to intelligent construction. A certain level of government incentives (at least greater than 5 CNY/m²) positively impacts the transition, effectively stimulating hesitant construction enterprises. However, once the incentives exceed 10 CNY/m², their effectiveness in sustaining stable transition levels diminishes, indicating that currently established policy incentives do not require further enhancement. While penalties do influence enterprises’ strategy choices during the transition process, they primarily affect the speed at which the system evolves to a stable point. Moreover, relying solely on punitive measures when initial willingness is low does not effectively promote construction companies towards the implementation of intelligent construction. Among the external factors influencing transformation, reducing the incremental costs of intelligent construction is more effective than increasing government incentive policies. When construction companies reach a stable level of awareness, they can still be encouraged to transition towards intelligent construction even without government incentives.

Thus, to better encourage construction companies in their transition to intelligent construction, several key measures and recommendations are suggested. Firstly, it is crucial to enhance awareness of intelligent construction. Both construction enterprises and the government must collaborate to dispel concerns regarding this transition, ensuring a full recognition of the benefits intelligent construction offers over traditional models, as well as the importance of technical collaboration among enterprises to achieve these advantages. Secondly, supporting policy formulation should be strengthened. Policies must be designed to effectively guide and serve the sector by enhancing resource coordination and optimizing the technical incubation environment. Concerted efforts should be made to advance the construction of new information infrastructure, such as 5G, artificial intelligence, and big data, which will establish a solid foundation for the transformation centered on information technology. The government should also timely adjust its reward and punishment mechanisms, ensuring that appropriate levels of subsidies and incentives maintain the game’s stability. A “subsidy tapering” policy is recommended, where the government initially provides a certain level of subsidies and incentives, gradually reducing them as the number of pilot projects increases until they are phased out. Additionally, the government can play a significant regulatory role, determining whether to increase penalties based on the pace of transition to guide enterprises in implementing intelligent construction practices. Finally, empowering enterprises and reducing incremental costs are essential. The transformation and upgrading of intelligent construction are primarily driven by market economic forces. During the early development stages, when the market is not yet mature, regulation through policy incentives can be effective. As the market expands, the focus should shift to strengthening the role of enterprises as key players. By applying intelligent construction models, companies can enhance project management and improve resource allocation efficiency, thereby increasing core competitiveness and expanding overall returns. Moreover, enhancing technological innovation capabilities can shorten the development cycle for new processes and technologies, cultivate new industrial workers, and steadily accelerate the application of new technologies, ultimately minimizing the incremental costs of intelligent construction.