Modelling industrial construction operations using a multi-agent resource allocation framework

Purpose -Modelling construction resources and their dynamic interactions and constraints are a challenging problem. The allocation of these resources to competing activities is usually a function required in any scheduling process. Performing such allocation under a dynamic and diverse set of constraints adds more complexity to the problem. This study seeks a structured approach for representing resources and their allocation to different activities through the use of an agent-oriented modelling framework. Design/methodology/approach -A model is developed for a real case of assembly operations of industrial construction modules. The model follows a multi-agent resource allocation structure and is implemented within an agent-based simulation environment. The model is used to evaluate the effects of different optimization algorithms and modelling parameters on the generation of a construction schedule. Different experiments run through the model and their results are analyzed and discussed. Findings -The model showed sensitivity only under large and continuous workloads. Overall the structured approach followed in developing the model provided a flexible medium for experimenting with different elements of the resource allocation problem. Research limitations/implications -The work is limited to the studied case and the results cannot be generalized beyond similar cases. The modelling approach used in the study provides a platform that can facilitate future research in construction resource allocation strategies. Originality/value -The presented work demonstrates a new approach for modelling construction resource allocation problems that enables structured experimentation with alternative allocation algorithms. It also presents a novel way for modelling modular industrial construction operations.