A location based approach to study the impact of construction logistics and production control on crew performance.

Abstract

Construction logistics and production control can enhance project performance. Several techniques can be applied to these systems, each having different effects on project performance. Push planning and pull planning have been studied and implemented to these systems; push planning mainly uses end dates to control the project, whereas pull planning pulls from milestones and the state of the system. Researchers have been studying and contrasting the mentioned approaches from scheduling and control perspectives trying to improve project performance, mainly considering tasks and project durations. However, zooming into the level of crews and locations within a project, it is noticed that the effect of the mentioned production control approaches on crew performance including labor productivity, crew allocation to areas, idle time, and other crew performance metrics still need to be studied. This research reviews onsite construction logistics and production control techniques, studies them at the level of locations, and proposes hypotheses relating push and pull production control techniques to several project and crew performance metrics. The significance of this study is materialized in two main contributions. First, the study contrasted push and pull planning techniques at the level of locations, and second, it exposed the effects of push and pull planning on crew performance including labour productivity and other project performance metrics. This enabled a convergence to generalized conclusions regarding optimum methods to deal with different production control issues. Agent-based modelling is used to develop a simulation model that describes how each of push and pull production control techniques affects crew performance. The model depicts crews’ interactions and allocation to tasks within the project locations. During simulation, several metrics indicating the performance of crews, activities, and the project were measured and stored. After conducting several runs to accumulate enough data, the collect