The processes in construction are more likely than others to breed unsafe behaviors, and the consequences of these actions can be serious. This paper first reviews the research status on unsafe behavior in construction teams. It then analyzes the complex mechanisms that lead to unsafe behavior and constructs a three-layer structural model based on agent-based modeling (ABM) technology. This modeling deals with complexity and elaborates on key points and potential research ideas in the study of unsafe behavior in construction teams. Using the ABM method, the effects of different incentive strategies on the safe behavior of construction teams under different management scenarios were studied. The results showed that when members have a fair perception of the situation, the effect of the excess performance reward distribution, according to the member’s safety awareness level, is better than the average distribution effect. This is the case whether the member’s safety behavior level is positively or negatively related to the member’s safety awareness level. This study proves the feasibility, validity, and universality of the three-layer structural model. It also reaches certain management conclusions and ideas for further development. The purpose of this paper is to provide a reference for research on the containment and prevention of unsafe behavior in construction teams.
According to statistics, about 60,000 construction workers die each year worldwide, which is equivalent to an accident every 9 minutes [
In the frequency statistics of fatal construction safety accidents published by the U.S. Occupational Safety and Health Administration (OSHA), the top four accidents are directly related to the unsafe behavior of workers. Many theories on causes of accidents also regard the unsafe behavior of workers as a direct cause of safety incidents [
In the process of construction, the project faces enormous challenges such as environmental complexity, technical complexity, and management complexity. As the establishment of key equipment requires the coordination of multiple teams to achieve the optimization of multiple goals such as project progress, cost, and quality, its operational processes are generally based on team cooperation. Haslam et al. pointed out that 70% of construction safety accidents are closely related to the construction team [
Due to diversity and difference of unsafe behavior in construction teams, there are multiple sources and complexities of motivation. Motivation can be nonlinear and may follow uncertain evolutionary paths. Project conditions such as regulatory cost constraints, incomplete security management systems, and information asymmetry also play a key role in determining behaviors. At an individual level, cognition and decision-making heterogeneity make it difficult to supervise and control the unsafe behavior of construction teams. The practical effect of preventing unsafe behavior of construction teams is thus not ideal. Therefore, analyzing and summarizing the research status of unsafe behavior in the construction team and examining the complex mechanisms that lead to unsafe behavior in construction teams are vital. In addition, seeking research ideas to deal with the complexity of unsafe behavior in construction teams that are realistic is urgently required in the field of engineering management.
To address these issues, this paper first summarizes and analyzes the research status of unsafe behavior in construction teams and divides the existing research into three categories covering three different aspects: subjects’ attributes and decision making, subject interaction and environment, and system safety performance. The complexity of unsafe behavior in construction teams and the proposed ABM framework provide the theoretical basis for this study. Secondly, it analyzes the complex and diverse mechanisms governing the subjects’ unsafe behavior in the construction team, the interaction between the subjects and the complex environment, and the emergence and dynamic evolution of system performance. On this basis, the method, based on ABM, is applied to the research on unsafe behavior in construction teams to deal with the challenges of their complexity. Based on the research status and the characteristics of the ABM, a multiagent model is constructed. The three-layer structural model of the modular method illustrates the thoughts and technical route of the model in the study of unsafe behavior in construction teams. Finally, the ABM method, based on the three-layer structural model, is applied to study a real case with a realistic background. The evolution mechanism of unsafe behavior in construction teams under different incentive mechanisms proves that the model is feasible, effective, and universal.
Human behavior is the product of cognition. If people’s behavior is insecure, it must be a failure in the cognitive process that generates the behavior [
Cognitive and decision-making studies on unsafe behavior in construction teams.
Factors | Subfactors | References |
---|---|---|
Psychological | Emotion | People who are in a positive emotional state can better assess the consequences of their behavior than those who are in a negative emotional state [ |
Attention | The unsafe behavior of construction workers is mainly due to the incorrect estimation of potential risks and a lack of attention [ | |
Work pressure | Hofmann and Stetzer believe that complex and overloaded tasks can lead to work pressure and thus affect safety behavior [ | |
The results of the study indicate that work stress is negatively correlated with safety behavior [ | ||
Cognition and assessment | Satisfaction | The results of the case study showed that the workers in the accident group were dissatisfied with the job compared with the workers in the nonaccident group (control group) [ |
Attitude | Attitude plays an important role in the behavior of the decision makers [ | |
Workers perceive risk by collecting various kinds of information, and workers establish safety attitudes based on perceived risks. The workers decided to take action based on the established attitude [ | ||
Motivation | Motivational factors (risk/benefit tradeoffs) play an important role in the process of violations, and cognitive factors may influence the results of violations. Errors may require the interpretation of individual cognitive processing capabilities [ | |
Risk perceptions | Risk perception plays a crucial role in eliminating work-related hazards [ | |
Physiological | Fatigue | When the miners feel tired, they will be negligent and more vulnerable to unsafe behavior [ |
Age | Occupational injury is related to the age-based curve, with injuries at first increasing with age, then decreasing. The two safety attitude scales were related to age, and the elderly are more positive about safety [ | |
Young people under the age of 26 have low scores for safety performance, poor safety knowledge, and an aversion to safety management [ | ||
Individual attributes | Personality | Studying the relationship between the five dimensions of personality and work-related accidents also found a close correlation between personality traits and worker accident trends [ |
Risk preference | When emphasizing decision-making options for avoiding losses, most people adopt risk-taking strategies [ | |
Knowledge | The degree of professional knowledge will directly affect the workers in dealing with professional projects [ | |
Employees’ safety risk tolerance will be affected by work knowledge and work experience [ | ||
Work experience | Work experience affects the safety behavior of workers [ | |
Construction workers generally lack objective and rational safety knowledge, and the judgment of the degree of danger is mainly based on personal intuitive experience and past experience [ | ||
Experienced workers clearly recognize that it is very important to incorporate disaster reduction measures into building technology from the beginning of the project [ | ||
Historical behavior | Past behaviors | Goles et al. believe that because of the positive emotional experience brought about by past behaviors, individuals will have a more positive attitude toward such behaviors, which in turn will increase their willingness to implement the behavior again [ |
Habitual behaviors | When experiencing problems, workers often choose a habit as a center of consciousness from experience. Bad habits can lead to injury accidents [ | |
Customary unsafe behavior is the “burner” for unsafe production behaviors, and there are seven personal factors that affect habitual violations [ |
With the deepening of research on security incidents, people are gradually realizing that individual safe behavior depends not only on individual cognition but also on other members of the construction team and the operating environment. For example, when insecurity of the operating environment coexists with unsafe behavior of individuals, accidents may occur [
Summary of subject interactions and environment on unsafe behavior in construction teams.
Factors | Subfactors | References |
---|---|---|
Hardware environment | Construction equipment | Equipment maintenance has an important effect on the occurrence and severity of accidents [ |
Weather | Hot weather conditions will increase the health and safety risks of construction workers and the possibility of workers suffering from heat-related diseases [ | |
Work condition | The characteristics of miners’ workplaces have a major impact on the occurrence of accidents [ | |
Chi et al. studied the impact of unsafe working conditions on unsafe behavior of workers by analyzing the specific types of accidents and determining the degree of damage [ | ||
Technology | Adeleke et al. showed that there is a significant positive correlation between technical factors and construction risk management [ | |
Interaction | Imitation | Imitation and learning are the main methods for replicating and spreading unsafe behaviors. It can lead to or catalyze some new unsafe behaviors [ |
Workers tend to imitate others who appear to have adopted successful strategies [ | ||
Social conformity | Rozin shows how herd characteristics among group members can affect personal judgment and attitude [ | |
Safety training | Safety training plays a crucial role in risk prevention. It can help builders master the knowledge of construction technology [ | |
Training is an indispensable part of a successful scientific and technological management system. In the design of a technical training program for a specific position, we should pay more attention to the characteristics of the staff and the awareness of risk [ | ||
Safety training has a positive effect on construction workers’ safe behavior [ | ||
Incentive mechanisms | Vredenburgh showed that incentive can effectively reduce the injury rate in industry [ | |
Li et al. explained that the recognition of safety behavior through bonuses, penalties, awards, advancement, and so on is the most important factor in motivating us to work safely [ | ||
Leadership | Managers’ safety management decisions and attitudes have a significant impact on miners’ safety motivation and safety behavior [ | |
Teamwork behavior | Dov found that the group leader directly influences the employee’s operational behavior through daily task decisions [ | |
The teamwork climate has a significant positive effect on workgroup members’ in-role, extra role, and deference behavior [ | ||
Workers’ safety behavior is influenced by the perceived teamwork climate [ | ||
Supervisory behavior | The supervisory behavior affects the safe climate of the group and the safety behavior of the employees [ | |
Soft environment | Regulations | Safety regulation has significant effects on safety performance [ |
Safety management systems | After using the safety management system, the accident rate of the project dropped significantly [ | |
Organizational structure | The characteristics of organizational structures in construction make the effects of social norms on safety more complex [ | |
Work schedule | Work plans have a significant effect on the risk behavior of construction workers [ | |
Safety standards | National safety standards set out various mechanical safety measures and procedures to ensure safe work [ | |
Safety climate | The safety climate has a significant direct effect on safety behavior [ | |
The safety climate of the construction team is becoming more and more important in the process of construction safety management, because the casualty rate of the construction team is very high [ | ||
Safety culture | The safety culture has an indirect effect on safety behavior [ | |
Macro environment | Political | Political factors help construction companies to reduce risks in construction activities [ |
Economic | Israelsson and Hansson affirmed a significant relationship between economic factors and construction risk management [ | |
Insurance | Insurance companies can minimize losses, and contractors can be motivated to invest in safety [ |
System safety performance research, as an important part of safety management, helps to clarify the influencing factors and preventive measures of unsafe behavior in construction teams. Firstly, some scholars have analyzed the definition of system safety performance, where the difference is relatively large and can be roughly divided into three categories: one is to directly define the safety performance with the occurrence of safety production accidents and their consequences; the second is to use the actual performance of the company to consider the operational effects of safety work; and the third is to define the safety performance by using accidents and the actual performance of the company [
System safety performance indicators in unsafe behavior in construction teams.
Author | Indicators |
---|---|
Hidley[ |
Leadership qualities, workplace culture, workplace pressure |
Chen et al. [ |
Work pressure, safety climate, management commitment to safety, supervisor safety perception |
Skeepers and Mbohwa [ |
Safety culture, safety leadership, attitudes, behaviors of individuals |
Chinda and Mohamed [ |
Safety attitude, safety behavior, safety culture, safety climate |
Wehbe et al. [ |
Individual resilience, supervisor safety perception, safety climate, communication, organizational factors, work pressure |
Griffin and Neal [ |
Safety knowledge, safety skill, motivation |
Feng [ |
Safety investments (investments in basic safety measures and investments in voluntary safety measures), safety culture |
Dearmond et al. [ |
Safety compliance, safety participation |
In general, the study of unsafe behavior in construction teams is a complex, scenario-dependent problem, because it involves many system elements, heterogeneous entities (e.g., owners, designers, supervisors, contractors, subcontractors, and team members), multiple goals [
Due to the universality of the construction team’s unsafe behavior, the diversity of its manifestations, the complexity of the impact mechanisms, and the severity of the consequences, it has gradually attracted the attention of researchers in many fields and has become an important topic for research. This paper asserts that the evolutionary process of unsafe behavior in construction teams can be regarded as a multiagent, multilevel dynamic interaction process, and it may face complex challenges in the process of its research. These challenges are summarized in the following section.
Firstly, there are multisource causes for unsafe behavior in the construction team. There are many factors influencing the unsafe behavior of the subject in the system including the subject’s complex relationship and interaction mechanisms with the construction team that are unsafe for the construction team. There is a difference between the mode of action and the effectiveness of the impact, which may have a nonlinear effect on the unsafe behavior of the construction team. The subject of unsafe behavior is often the result of a combined effect of many factors, which leads to the multisource characteristics of the unsafe behavior of the subject. For example, Bohm and Harris [
Secondly, the manifestation of unsafe behaviors of construction team workers are diverse. Members of the construction team are heterogeneous in their external environment, goals, abilities, preferences, and so on. The decision making of subjects is influenced by game choice and active feedback under the conditions of an integrated external environment, self-factors, and other conditions. In addition, the subject’s complex psychological and behavioral characteristics can have an important influence on decision making, such as their attitude to trust, concern for reputation, attention to fairness, and risk appetite. Therefore, the subject’s behavior and decision making will show diverse characteristics. For different scenarios in the project construction phase, such as schedule control [
Firstly, the scenarios contained in the construction phase are relatively rich and complex. Many elements in the scenarios may affect unsafe behavior. The influencing factors are mainly the external environment and institutional conditions, the team’s own attributes, and the multiagent dynamic interaction process. In addition, there are differences in the acting pathway, impact effectiveness, and transmission mechanisms of various factors on unsafe behavior. For example, the safety behavior of the construction team is affected by differences in factors such as team leader safety awareness, concern for social reputation, and staff quality, knowledge, and skill levels.
Secondly, the generation and development of unsafe behavior also depend on certain objective and realistic foundations. Due to the complex and open operating environment of the project, the numerous stakeholders with heterogeneous characteristics, the interaction of the subject’s behavior, and other complicating factors, there may be defects and loopholes in the institutional guarantees and supervision of the team members’ behavior. For example, it is difficult for owners to know the safety behavior of the construction team during on-site operations, and there is information asymmetry between the subjects, the standards and norms of safety behavior are subjectively ambiguous to some extent, and they are often difficult to define clearly in contractual terms.
In addition, conflicts, interactions, and games between the construction teams and other stakeholders may also affect safety behavior. The scenario of the construction team is a dynamic collection of multiple stakeholders, numerous distributed resources and information, and other elements. Various elements are intertwined and linked to form a certain level or network structure, which changes continuously with the evolution of the system. Deep indefinite features are also generally present, so the dynamic effects and mutation of scenarios can induce the complexity that leads to insecure behaviors, such as the owner’s time limits and cost constraints, the supervision of the unit’s safety, and the engineering supply chain member’s interactions. In the context of the owner’s shortcuts, the construction team may be subject to the owners’ administrative orders and controls, so they will emphasize the construction period but ignore safety.
Construction team members are both learning and adaptable in the process of a dynamic interaction. Subjects’ decision making is itself an interaction with other individuals and the environment, and through learning, imitation, conformity, and other means, it changes their behavior to adapt to environmental changes. Adaptability creates complexity, and the scene dependence of unsafe behavior in construction teams is a manifestation of their adaptability.
Firstly, there is a close relationship between the unsafe behavior of the construction team and the main decision making at different stages of the construction project. The unsafe behavior of the construction team may be affected by the results of the previous stage of behavioral decision making and may also affect the behavior of the next stage of the project. For example, under the influence of economic interests or performance goals, the owner’s actions to shorten the construction period may cause the construction unit to take on long overtime operations to avoid delays in the construction period, which may cause unsafe behaviors such as fatigue.
Secondly, the unsafe behavior of the construction team is often accompanied by the opportunistic behavior of the system’s subjects. For example, in the “Shanghai downstairs incident,” various violations by the clients, general contractors, construction workers, supervisors, and management were involved. As long as any one of them seriously implements the rules, security accidents will not happen, but the result has become a tragedy of “mutual cooperation and joint directorship.” On the one hand, when there is supervisory dereliction of duties, failure to properly perform or passively perform their own duties and obligations, and other opportunistic behaviors, the behaviors driven by the noted interests may induce construction teams to generate unsafe behavior. On the other hand, when the construction team is motivated by unsafe behavior, it may also try to induce other stakeholders to generate opportunistic behaviors (such as “rent seeking” in the supervision process) and then create objective conditions for its unsafe behavior. It can be seen that there may be interdependence, mutual causation, emergence mechanisms, and interactional paths between the opportunistic behavior of stakeholders and the unsafe behavior of construction teams.
In addition, unsafe behaviors of construction teams often have dynamic evolutionary characteristics. In the environment of information asymmetry, inadequate supervision, incomplete safety management systems, and uncertain environments, the construction teams will realize the spread and evolution of security behavior through interaction, coordination, organization, imitation, learning, and conformity, which is based on the team members’ own attributes and the external scenarios.
In summary, the above factors lead to multisource, diverse causes and forms of unsafe behavior as well as nonlinearity and uncertainty in the evolutionary path of unsafe behavior in construction teams. In addition, insecure construction environments, inadequate management departments, regulatory cost constraints, incomplete safety management systems, unclear operating standards and specifications, and other objective conditions, as well as inadequate supervision by government regulatory agencies and imperfect insurance systems, have made the supervision and control of unsafe behaviors in construction teams extremely complicated and difficult to understand.
The complexity of unsafe behavior in construction teams has been widely recognized by researchers, and it poses challenges for us to study the field in more depth. The ABM method, as an important tool in the research of complex systems, has achieved remarkable results in its application in social science-related fields [
This study introduces the ABM method into the study of unsafe behaviors in construction teams and aims to lay a methodological foundation for the construction of unsafe behavior prevention mechanisms. The study of this issue should follow the spiral of “practice-theory-practice.” The research ideas and solutions for the situational adaptation of unsafe behavior in construction teams are shown in Figure
Research ideas and design of strategies for prevention of unsafe behavior in construction teams.
Based on the background of construction practice, a systematic analysis of unsafe behaviors in construction teams was conducted, and the key scientific management issues were condensed. Using scale analysis, case analysis, empirical research, and in-depth interviews to lay the foundation for building an agent-based modeling to improve accuracy, then based on the engineering practice and empirical evidence, the model assumptions, subject decision rules, and interaction rules were continuously revised to construct a more realistic ABM. In addition, the design of the prevention mechanism of unsafe behavior and theoretical results needed to be repeatedly placed in different scenarios to simulate and test the effects. Then, through in-depth comparison, analysis, and verification of the effectiveness of the implementation for preventing unsafe behavior in construction teams and their long-term evolution, the rational selection and creation of specific strategies were implemented to ensure their robustness and effectiveness. These were analyzed and tested to achieve a reasonable selection and creation of specific strategies to ensure their robustness and effectiveness.
Constructing a multiagent experimental model is an important part of the ABM research on unsafe behavior in construction teams. Based on the ABM research ideas on unsafe behavior in construction teams, this paper proposes a three-layer structural model of ABM. As shown in Figure The system performance layer belongs to the macrolevel of the system and describes the system macro features of the unsafe behaviors in the construction teams that need to be studied, such as system macro performance indicators like safety atmosphere, safety awareness, accident rate, number of accidents, and other system macro performance indicators. The agent interaction layer is mainly used to describe the behavioral characteristics and interaction mechanisms of intelligent agents in the system. It involves the interactive relationship between subject and subject such as learning, imitation, rewards and punishments, and other interactions; it also includes the interaction between subject and environment. The agent attribute layer is mainly used to describe the basic characteristics of the subject such as individual characteristics, behavioral preferences, and psychological cognition. It is the most basic level in constructing the evolution of artificial social systems and research systems. Based on memory, cognition, behavior, learning, and preference of the subject, this level focuses on the dynamic process of the individual’s psychology and behavior in the system, and this layer can abstractly reflect the self-evolutionary mechanism of the intelligent subject.
Three-layer structural model framework for studying unsafe behavior in construction teams.
In the process of system self-evolution, “emergence” is expressed as the output of cognitive decisions formed under the internal interaction of the subject’s attribute layer (microlevel), which in turn affects the interactions between the various entities within the interaction layer of the subject and the interaction between the subject and the environment. The results of the interactions serve as input to the system’s performance layer and reflect the macrolevel behavioral characteristics and performance in the system performance layer. During the construction process, it exists in the microlevel upward transmission and affects the macrolevel phenomenon. It also has an influence on the impact and feedback of the macrolevel on the microlevel. This is embodied in the setting of system performance indicators, the construction safety climate, and safety culture. Other factors will affect many other factors such as the interaction of the agent within the system, the establishment of safety regulations and penalties, and the layout of the construction environment. The interaction of agents and the situation of the construction environment further feedback and influence the attribute indicators such as cognitive, psychological, and risk perception of each subject. Therefore, in the process of system self-evolution, the three-layer structural model not only has bottom-up “emerging” but also top-down feedback and correction under interaction.
Using the three-layer structural model, we can study the behavior in decision making, interactive games, and the self-organization and self-adaptive evolution of the system in the construction process and then establish the relationship between the subjective microbehavior and the macro emergence of the system. In the process of model building, it is necessary to focus on the modeling of learning and adaptability in the decision-making process of the agent. In the process of multistage construction practice, each agent generally adjusts and optimizes its behavioral decisions based on learning, imitation, trial and error, and so on, and the agent can dynamically adjust its behavior according to its own attributes and external environment.
In order to further verify the universality and effectiveness of the three-layer structural model proposed in Section
A project construction team consists of individuals such as project managers, technicians, professional foremen, quality inspectors, material managers, and safety managers. Team members’ safety performance awards consist of basic performance income
The basic performance income is positively correlated with the team members’ safety awareness level, and the basic performance income is the normal distribution
This paper considers two types of correlations between
Four management scenarios for safety performance incentives.
Management scenario settings | Excessive performance bonus distribution plan | ||
---|---|---|---|
According to the level of safety awareness | Equally distributed | ||
Level of distribution of member safety behavior | Positive correlation with members’ safety awareness level | Scenario 1 | Scenario 2 |
Negative correlation with members’ safety awareness level | Scenario 3 | Scenario 4 |
Assuming that the team members’ safety behavior level is
The overall safety performance output of the construction team
Assuming
Assuming that team members are risk-neutral, the total income received by each team member based on their safety performance output is the sum of basic performance income and excessive performance incentives, as shown in
Assuming that the team members all have fairness perception. Fairness perception is a key factor in team members forming cooperative relationships, and good partnerships can help improve project management performance. Love learned from interviews that if the contractors feel that in terms of distribution, interaction, and information openness, it is unfair, then this can lead to negative cooperation, and it will in turn affect the contractor’s violations and unsafe behavior [
If we adjust the level of safety behavior according to fairness perception, then the specific adjustment rules are shown in
Here,
Assuming that the number of members is 20 (
Member security cognition level and initial setting of safety behavior level.
1 | 2 | 3 | 4 | 5 | 6 | 7 | 8 | 9 | 10 | 11 | 12 | 13 | 14 | 15 | 16 | 17 | 18 | 19 | 20 | |
---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
4 | 7 | 9 | 9 | 10 | 11 | 13 | 14 | 14 | 15 | 17 | 19 | 20 | 21 | 22 | 28 | 30 | 33 | 34 | 40 | |
8 | 9 | 9 | 11 | 12 | 12 | 12 | 12 | 13 | 14 | 17 | 17 | 19 | 26 | 26 | 26 | 29 | 29 | 33 | 40 | |
40 | 33 | 29 | 29 | 26 | 26 | 26 | 19 | 17 | 17 | 14 | 13 | 12 | 12 | 12 | 12 | 11 | 9 | 9 | 8 |
First of all, we select the variance of basic performance income
Evolution of safety behavior of construction team members under different variances.
From Figure When the level of safety behavior of members is positively or negatively correlated with the level of safety perception of members, the effect of the distribution of excessive performance awards based on the level of member safety awareness is better than the effect of equal distribution. When the variance of basic performance income is small, although the effect of reward distribution based on members’ safety awareness levels is better than the effect of average distribution, the difference is not very obvious. In addition, when the variance of basic performance income increases, the effect of the distribution of excessive performance rewards based on members’ safety awareness levels will be greatly enhanced compared with the average distribution effect. It can be seen that the variance of basic performance income will increase the incentive effect of safety behavior to some extent. The variance of basic performance income is different. The safety behavior of team members in scenario 3 is better than in other scenarios. It can be seen that when the member’s safety behavior level is negatively correlated with their safety cognition level, it shows that although the safety norms show good understanding, strong security experience, and high perception of risk, it is not possible to observe and practice safety regulations during the operation process. For such members, the use of rewards based on their level of security awareness will result in better incentive effects. The safety behavior of team members in scenario 2 is worse than in other scenarios. When the level of safety behavior of members is positively correlated with their level of safety awareness, this type of member has a good perception of safety and is willing to follow safety norms and requirements in construction operations. For such members, if they adopt the method of equal distribution, they will not have a positive incentive effect under the effect of fair perception.
During the construction phase of the project, team members are more likely to breed unsafe behavior, which may have serious consequences. First of all, this paper reviewed the current research status on unsafe behavior in construction teams from three perspectives: agent cognition and decision making, the interaction among subjects in specific environments, and system safety performance. Secondly, this paper analyzed the multisource factors and diversity of manifestations of unsafe behavior in construction teams, the adaptability and environmental dependence of the agent behavior, and the multistage correlation and dynamic evolutionary characteristics of the unsafe behavior. In addition, ABM was applied to study the unsafe behaviors of construction teams and to deal with the challenges of complexity. Finally, a three-layer structural model based on agent-based modeling was constructed to explain key points and research ideas on unsafe behavior in construction teams.
Using the above research, this paper studied the incentive effects of different incentive strategies for the safety behavior of construction teams under different management scenarios based on the multiagent modeling method and proved the feasibility, effectiveness, and universality of the model. It also offers some management conclusions and inspirations. It aims to provide a reference for research on the containment and prevention of unsafe behaviors in engineering construction teams. The results show that when the level of safety behavior of members is positively or negatively related to the level of safety perception of members, the effect of the distribution of excess performance rewards based on the level of member safety perception is better than the effect of average distribution. Accepting that fair perception exists among team members, when the two are negatively correlated, adopting a reward based on their safety awareness level will achieve better incentive effects, while when the two are positively related, the average incentive is less effective. The variance of basic performance income has an influence on the incentive effect of team members’ safety behavior.
The three-layer structural model constructed in this paper has the following limitations. This paper mainly considers the system performance indicators at the macrolevel of the system; however, different risks emerge from a bottom-up system, such as management risk, financial risk, environmental risk, and other risk factors, so at the macrolevel of the system, the design and description of risk indicators and risk assessment methods should be considered. In summary, future research based on the model constructed in this paper should focus on the following: (1) through the identification, feature extraction, and classification of the typical unsafe behavior of members of a construction team, the effects of the heterogeneity of the team members, the complex psychology and behavior preference of the members, the external environment, and the institutional conditions on the unsafe behavior of the members of the team can be further analyzed; (2) the influence of the interaction mechanism (resource allocation, information communication, and other mechanisms) between the members of the construction team and the mechanism of benefit coordination (such as mechanism of reward and punishment and risk sharing) on the evolution of unsafe behavior of team members can be studied; and (3) considering the complex behavior of the members of the construction team, such as psychology, physiology, trust, and imitation learning, the traditional prevention strategy is improved or innovative, and the strategy of preventing unsafe behavior among team members can be further emphasized and built on.
The data used to support the findings of this study are available from the corresponding author upon request.
The authors declare that there is no conflict of interest regarding the publication of this paper.
This work was supported by the National Natural Science Foundation of China (nos. 71671078, 71501084, and 71390521), Qing-Lan Project of Jiangsu Province, Youth Backbone Teacher Training Project of Jiangsu University, The Key Project of Philosophy and Social Science Research in Colleges and Universities in Jiangsu Province (no. 2018SJZDI052), Senior Personnel Scientific Research Foundation of Jiangsu University (15JDG108), and The National Social Science Fund of China (17AGL010).