Simulating and vizualising emergent production in construction (Epic)using agents and BIM

We present an agent-based simulation model developed for studying and improving production control in construction processes. The model represents individuals' decision making, knowledge and uncertainty. Simulation methods are particularly useful for assessing the impacts of different production control methods and information flows on production on site because field experiments in building projects suffer difficulties with isolating cause and effect. Existing methods such as Discrete Event Simulation (DES) are limited to model decision-making by individuals with distinct behaviour, context and knowledge representation. Agent-Based Simulation (ABS) may offer a better solution. The simulation developed exhibits the interdependence of individual crews as they interact with each other and share resources, reflecting the influence of crew leaders' perception of the project state on their workflow decisions. The model uniquely distinguishes between reality and perceived reality. Significantly, this allows experimentation with uncertainty as agents function within the context of what they know. Different management policies, such as the LPS, can be tested, as can the impact of different site information-flow systems. Unlike the few existing agent-based simulation models for construction, the simulation is situated in a realistic virtual environment modelled using BIM, allowing future experimental setups that can incorporate human subjects and real buildings.