A single individual or organization cannot have complete knowledge and abilities to handle all matters in complex projects; therefore, effective communication, coordination, and information-handling processes are necessary to accomplish project objectives. How well organizational knowledge creation processes and organizational development processes are executed determines the quality of information created; poor organizational processes could cause unplanned events, such as design defects and construction errors, which would result in schedule delays and additional expenses. Nevertheless, current planning methods do not explicitly consider organizational and information-handling issues and thereby have resulted in planning fallacy.

Based on this premise, this research introduces an agent-based simulation of "evolution of collaboration" within inter-organizational networks of construction project teams from a complex system perspective. The agent-based simulation approach allowed this research to achieve things that might not be able to do with traditional approaches. First and foremost, a theoretical framework in this research was expressed with an explicit and intuitively understandable computational language. Also, adaptive and boundedly rational behaviors of heterogeneous individuals were reflected in the model. It endowed the model with greater conviction, usefulness, and validity. Besides, both individual behavior at micro level and emergent network behavior at macro level were obtained.

As a result, the simulation shows that the fewer individuals are familiar with others in the network and thereby aware of networks they are involved in, the more likely it is to take a longer time for networks to reach stable states. It was also found that the stronger tendency of cohesion arises as the effort that individuals have to make for forming a relation with outside partner increases. Furthermore, the results indicate that the more efforts individuals need to make to form relations with those from other organizations, the less efficient networks are formed. The model of organizational processes developed in this research which explicitly incorporates a realistic representation of individual decision making and interaction among them from an active and dynamic standpoint is expected to contribute not only for enhancing understanding of the impact of organizational aspects on project management but also for facilitating the development of tools for investigating organizations in construction.